Guidelines For Intensified Tuberculosis Case-Finding and Isoniazid Preventive Therapy For People Living With HIV in Resource-Constrained Settings

Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resourceconstrained
settings
Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resourceconstrained settings
Department of HIV/AIDS Stop TB Department World Health Organization, Geneva, Switzerland
Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resource-constrained settings
WHO Library Cataloguing-in-Publication Data Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resource-constrained settings. 1.Tuberculosis - prevention and control. 2.Tuberculosis - diagnosis. 3.HIV infections - complications. 4.Isoniazid therapeutic use. 5.Predictive value of tests. 6.Developing countries. 7.Guidelines. I.World Health Organization. Stop TB Dept. II.World Health Organization. Dept of HIV/AIDS. ISBN 978 92 4 150070 8
World Health Organization 2011
(NLM classification: WF 220)
All rights reserved. Publications of the World Health Organization can be obtained from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: bookorders@who.int). Requests for permission to reproduce or translate WHO publications whether for sale or for noncommercial distribution should be addressed to WHO Press, at the above address (fax: +41 22 791 4806; e-mail: permissions@who.int). The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. Edited and proof read by Bandana Malhotra Printed in Italy Layout design: www.blossoming.it www.who.int/hiv/topics/tb/en/index.html www.who.int/tb/challenges/hiv/en/index.html
WHO Steering Group Siobhan Crowley (HIV/AIDS Department, WHO), Andrew Doupe (HIV/AIDS Department, WHO), Haileyesus Getahun (Stop TB Department, WHO), Reuben Granich (HIV/AIDS Department, WHO), Lulu Muhe (Child and Adolescent Health Department, WHO), Delphine Sculier (Stop TB Department, WHO), Caoimhe Smyth (HIV/AIDS Department, WHO) WHO Consultants for Systematic Review Christopher Akolo (USA), Anand Date (USA), Martina Penazzato (Italy), Georgina Russell (UK), Abhishek Sharma (Australia) Co-chairs of the WHO Guidelines Group Kevin De Cock (Centers for Disease Control and Prevention, Kenya), Holger Schnemann (McMaster University Health Sciences Centrer, Canada), Suzanne Hill (Essential Medicines and Pharmaceutical Policies Department WHO) WHO Guidelines Group The following group represents experts from the fields of HIV, TB, HIV/TB, sexually transmitted infections, child health, infectious and tropical diseases, clinical research, maternal health, infectious disease research, clinical epidemiology and biostatistics (additional information on request): Helen Ayles (ZAMBART Project, Zambia), Draurio Barreira (National TB Program, Brazil), Franois-Xavier Blanc (Agence Nationale de Recherche sur le SIDA et les Hpatites virales, France), Charlene Brown, [US Agency for International Development, (USAID), USA], Kevin Cain [Centers for Disease Control and Prevention (CDC), USA], Rolando Cedillos (Proyecto Regional VIH SIDA para Centroamerica, El Salvador), Richard Chaisson (Johns Hopkins University, USA), Mean Chhivun (National AIDS Programme, Cambodia), Anupong Chitwarakorn (Ministry of Public Health, Thailand), Gavin Churchyard (Aurum Institute for Health Research, Republic of South Africa), Mark Cotton (Stellenbosch University, Republic of South Africa), Anand Date (CDC, USA), Dmytro Donchuk (State Medical University, Ukraine), Wafaa El-Sadr [International Center for AIDS Programs (ICAP), Columbia University, USA], Peter Godfrey-Faussett (London School of Hygiene and Tropical Medicine, UK), Olga Petrovna Frolova (Ministry of Health and Social Development, Russian Federation), Paula Fujiwara [International Union Against Tuberculosis and Lung Disease (The Union), France], Alison Grant (London School of Hygiene and Tropical Medicine, UK), Mark Harrington (Treatment Action Group, USA), Catherine Hewison [Medecins sans Frontieres (MSF), France], Maureen Kamene Kimenye (Ministry of Public Health, Kenya), Michael Kimerling (Bill and Melinda Gates Foundation, USA), Stephen D. Lawn, (University of Cape Town, South Africa), Gary Maartens (University of Cape Town, Republic of South Africa), Barbara Jean Marston, (CDC, USA), Thombile Mbengashe (National Department of Health, Republic of South Africa), Zenebe Melaku (ICAP, Ethiopia), Peter Mgosha (Ministry of Health and Social Welfare, Tanzania), Muhamed Mulongo (Tropical Medical and Maternity Centre, Uganda), Sharon Nachman (Stony Brook University Medical Center, USA), Alasdair Reid [Joint United Nations Programme on HIV/ AIDS (UNAIDS), Geneva], Stewart Reid [Centre for Infectious Disease Research in Zambia (CIDRZ), Zambia], Taraz Samandari (CDC, USA), Paula Isabel Samo Gudo (Ministry of Public Health, Mozambique), Mauro Schechter (AIDS Research Laboratory, Brazil), Wim Vandevelde (European Community Advisory Board, European AIDS Treatment Group, Belgium), Eric van Praag (Family Health International, Tanzania), Jay K. Varma (CDC, USA), Fujie Zhang (National Center for AIDS/STD Control and Prevention, Peoples Republic of China) WHO Headquarters and Regional offices Lopold Blanc (Stop TB Department, WHO), Colleen Daniels (Stop TB Department, WHO), Puneet Dewan (WHO SEARO), Massimo Ghidinelli (WHO WPRO), Sandra Gove (HIV/AIDS Department, WHO), Malgorzata Grzemska (Stop TB Department, WHO), Teguest Guerma (HIV/AIDS Department, WHO), Christian Gunneberg (Stop TB Department, WHO), Rafael Lopez Olarte (WHO AMRO), Frank Lule (WHO AFRO), Eyerusalem Negussie (HIV/AIDS Department, WHO), Rose Pray (Stop TB Department, WHO), Mario Raviglione (Stop TB Department, WHO) Peer reviewers Jesus Maria Garcia Calleja (HIV/AIDS Department, WHO), Jacob Creswell (Stop TB Department, WHO), Irina Eramova (WHO EURO), Robert Gie (University of Stellenbosch, South Africa), Steve Graham (The Union, Australia), Prakash Kudur Hanumaiah (Karnataka Health Promotion Trust, India), Charles Mwansambo (Kamuzu Central Hospital, Malawi), Nguyen Viet Nhung (National TB programme, Viet Nam), Carla Obermeyer (HIV/AIDS Department, WHO), Ikushi Onozaki (Stop TB Department, WHO), Cyril Pervilhac (HIV/AIDS Department, WHO), Renee Ridzon (Bill and Melinda Gates Foundation, USA), Rifiloe Matji (Ministry of Health, South Africa), Quaid Saeed (WHO EMRO), Fabio Scano (WHO WPRO), Sahu Suvanand (Stop TB Partnership), Richard Zaleskis (WHO EURO) Overall coordination Haileyesus Getahun and Reuben Granich
Summary of declarations of interest All members of the Guidelines Group were asked to complete a World Health Organization (WHO) declaration of interest form and only two declared a conflict of interest. These were discussed within the WHO steering group and then with the Guidelines Group before the deliberations. Alison Grant declared receiving financial support of GB 200 from Roche to attend the International AIDS Conference, Sydney, 2007 when the aeroplane she was flying in broke down and the GB 200 was paid for a flight deviation. Helen Ayles declared receiving financial support amounting to US$ 15 000 from the Bill and Melinda Gates Foundation, US$ 50 000 from Senter Novum and 150 000 from Delft Imaging Systems to conduct research on intensified TB case-finding and isoniazid preventive therapy for TB, and the development of computer-aided diagnostics for TB/HIV. The Guidelines Group discussed these and concluded that there was no conflict of interest. Declarations of interest were collected from all non-WHO peer reviewers. No peer reviewer declared a conflict of interest. All declarations of interest are on electronic file with the Department of HIV/AIDS of WHO. Acknowledgements The development of these guidelines was financially supported by the Joint United Nations Programme on HIV/AIDS Unified Budget and Workplan (UNAIDS UBW) and the US Presidents Emergency Plan for AIDS Relief (PEPFAR) through the Centers for Disease Control and Prevention (CDC) and United States Agency for International Development (USAID).
ii
Contents
Contents
Abbreviations and acronyms Executive summary 1. Background and process 1.1 Background 1.2 Target audience 1.3 Scope 1.4 Process of formulating the guidelines 1.5 Strength of recommendations 1.6 Adaptation of the guidelines 2. Intensified case-finding for and prevention of tuberculosis in adults and adolescents living with HIV 2.1 Screening for TB 2.2 Efficacy, regimen and duration 2.2.1 Efficacy 2.2.2 Regimen and duration 2.2.2.1 Table 1: Comparison of the efficacy of different drug regimens 2.2.3 Immune status and concomitant use of IPT with ART 2.2.4 Pregnant women 2.2.5. Patients previously treated for TB (secondary prophylaxis) 2.2.6 Special populations 2.2.7 Figure 1. Algorithm for TB screening in adults and adolescents living with HIV in HIV-prevalent and resource-constrained settings 2.3 Detecting latent TB infection in resource-constrained settings 2.3.1. Tuberculin skin test (TST) and IPT 2.3.2. Interferon-gamma release assays (IGRA) 2.4 Issues to consider for implementation of IPT 2.4.1 Primary ownership by HIV service providers 2.4.2 IPT and drug-resistant TB 2.4.3 Adherence and clinical follow up 2.4.4 Cost-effectiveness of IPT 3. Intensified tuberculosis case-finding and prevention in children living with HIV 3.1 Screening for TB 3.2 Regimen and duration 3.3 Secondary prophylaxis and IPT with ART in children 3.3.1 Secondary prophylaxis 3.3.2 IPT with ART in children 3.4 The role of TST and IGRA in evaluating children for IPT 3.5 Figure 2: Algorithm for TB screening in children more than one year of age and living with HIV 4. Research gaps 4.1 Screening for TB 4.2 Preventive treatment for TB 4.3 Operational research 5. References 6. Selected GRADE profiles IV V 1 1 1 2 2 3 4 5 5 6 6 6 7 7 7 8 8 8 9 9 9 10 10 10 10 11 12 12 13 14 14 14 14 15 16 16 16 17 18 20
iii
Abbreviations and acronyms

AIDS ART ARV CDC CPT GRC GRADE HCV HIV ICF IGRA INH IPT LTBI MDR M&E PCR PEPFAR PMTCT TB The Union TST UNAIDS USAID WHO XDR acquired immunodeficiency syndrome antiretroviral therapy antiretroviral (drug) Centers for Disease Control and Prevention co-trimoxazole preventive therapy Guideline Review Committee Grading of Recommendations Assessment, Development and Evaluation hepatitis C virus human immunodeficiency virus intensified case-finding interferon-gamma release assay isonicotinic acid hydrazide/isoniazid isoniazid preventive therapy latent tuberculosis infection multidrug-resistant (TB, resistant to at least isoniazid and rifampicin) monitoring and evaluation polymerase chain reaction US Presidents Emergency Plan for AIDS Relief prevention of mother-to-child transmission (of HIV) tuberculosis International Union Against Tuberculosis and Lung Disease tuberculin skin test The Joint United Nations Programme on HIV/AIDS United States Agency for International Development World Health Organization extensively drug-resistant TB (defined as resistance to at least rifampicin and isoniazid from among the first-line anti-TB drugs, in addition to resistance to any fluoroquinolone, and to at least one of three injectable second-line anti-TB drugs used in TB treatment [capreomycin, kanamicin and amikacin])
iv
Executive summary
Executive Summary
IV is the strongest risk factor for developing tuberculosis (TB) disease in those with latent or new Mycobacterium tuberculosis infection. The risk of developing TB is between 20 and 37 times greater in people living with HIV than among those who do not have HIV infection. TB is responsible for more than a quarter of deaths in people living with HIV. Relatively more women than men were detected to have TB in countries with a prevalence of HIV infection of more than 1%. In response to the dual epidemics of HIV and TB, the World Health Organization (WHO) has recommended 12 collaborative TB/HIV activities as part of core HIV and TB prevention, care and treatment services. They include interventions that reduce the morbidity and mortality from TB in people living with HIV, such as the provision of antiretroviral therapy (ART) and the Three Is for HIV/TB: intensified case-finding of TB (ICF), isoniazid preventive therapy (IPT), and infection control for TB. On 2527 January 2010, WHO conducted a global policy meeting to review the evidence regarding ICF and IPT, and to reconceptualize the 1998 WHO/ Joint United Nations Programme on HIV/AIDS (UNAIDS) Policy on TB prevention. Key questions were identified and a comprehensive review of the available scientific evidence was conducted to formulate the recommendations. The evidence was evaluated using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria. The quality of the evidence was categorized as high (when further research is very unlikely to change our confidence in the estimate of effect), moderate (further research is likely to have an important impact on our confidence in the effect) and low (further research is very likely to have an estimate of effect and is likely to change the estimate). Reports were also commissioned from people living with HIV and affected communities regarding the key questions and the summary of the evidence. After the initial draft was reviewed by the
Guidelines Group, the comments were incorporated into a draft that was then sent to over 200 people for peer review. Comments from around 30 internal and external peer reviewers were used to finalize the recommendations. The final recommendations take into consideration the quality of evidence, cost, feasibility, and values and preferences of the community and health-care workers. The recommendations were classified as strong when the guidelines group was confident that the desirable effects of adherence to a recommendation outweigh the undesirable effects, and as conditional (weak) when the desirable effects of adherence to a recommendation probably outweigh the effects, but the panel was not confident about these tradeoffs. These new guidelines recommend the use of a simplified screening algorithm that relies on four clinical symptoms to identify those eligible for either IPT or further diagnostic work-up for TB and other conditions. Chest radiography is no longer a mandatory investigation before starting IPT. In contrast to the 1998 Policy, the new guidelines strongly recommend at least six months of IPT for children and adults including pregnant women, people living with HIV, those receiving ART, and those who have successfully completed TB treatment. IPT for a duration of 36 months is conditionally recommended in settings with a high transmission of TB among people living with HIV. The revised guidelines also emphasize that a tuberculin skin test (TST) is not a requirement for initiating IPT in people living with HIV. However, in some settings where it is feasible, it can help to identify those who would benefit most from IPT. The guidelines also emphasize that IPT is a core component of HIV prevention and care, and should be the primary responsibility of AIDS programmes and HIV service providers. In addition, the provision of IPT should not be viewed as an isolated intervention for people living with HIV. Rather, it should be part of a TB prevention package along with infection control for TB, ICF and provision of ART.
Key recommendations 1
Adults and adolescents living with HIV should be screened for TB with a clinical algorithm and those who do not report any one of the symptoms of current cough, fever, weight loss or night sweats are unlikely to have active TB and should be offered IPT.
Strong recommendation, moderate quality of evidence1
Adults and adolescents living with HIV and screened with a clinical algorithm for TB, and who report any one of the symptoms of current cough, fever, weight loss or night sweats may have active TB and should be evaluated for TB and other diseases.
Strong recommendation, moderate quality of evidence
Adults and adolescents living with HIV who have an unknown or positive TST status and are unlikely to have active TB should receive at least six months of IPT as part of a comprehensive package of HIV care. IPT should be given to such individuals irrespective of the degree of immunosuppression, and also to those on ART, those who have previously been treated for TB and pregnant women.
Strong recommendation, high quality of evidence
Adults and adolescents living with HIV who have an unknown or positive TST status and who are unlikely to have active TB should receive at least 36 months of IPT.2 IPT should be given to such individuals irrespective of the degree of immunosuppression, and also to those on ART, those who have previously been treated for TB and pregnant women.
Conditional recommendation, moderate quality of evidence3
5 6
1
TST is not a requirement for initiating IPT in people living with HIV.
People living with HIV who have a positive TST benefit more from IPT; TST can be used where feasible to identify such individuals.
A strong recommendation is one for which the panel is confident that the desirable effects of adherence to a recommendation outweigh the undesirable effects. 2 The considerations for implementation should include the local context such as the epidemiology of TB and HIV, and settings with the highest rates of prevalence and transmission of TB among people living with HIV. 3 A conditional recommendation is one for which the panel concludes that the desirable effects of adherence to a recommendation probably outweigh the undesirable effects and data to support the recommendation are scant. Therefore, the recommendation is only applicable to a specific group, population or setting, or new evidence may result in changing the balance of risk to benefit, or the benefits may not warrant the cost or resource requirements in all settings.
vi
Key recommendations
Providing IPT to people living with HIV does not increase the risk of developing isoniazid (INH)-resistant TB. Therefore, concerns regarding the development of INH resistance should not be a barrier to providing IPT.
8 9
Children living with HIV who do not have poor weight gain,4 fever or current cough are unlikely to have active TB.
Strong recommendation, low quality of evidence
Children living with HIV who have any one of the following symptoms poor weight gain, fever, current cough or contact history with a TB case may have TB and should be evaluated for TB and other conditions. If the evaluation shows no TB, such children should be offered IPT regardless of their age.
10
Children living with HIV who are more than 12 months of age and who are unlikely to have active TB on symptom-based screening, and have no contact with a TB case should receive six months of IPT (10 mg/kg/ day) as part of a comprehensive package of HIV prevention and care services.
11
In children living with HIV who are less than 12 months of age, only those children who have contact with a TB case and who are evaluated for TB (using investigations) should receive six months of IPT if the evaluation shows no TB disease.
12
All children living with HIV who have successfully completed treatment for TB disease should receive INH for an additional six months.
Conditional recommendation, low quality of evidence
Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weightfor-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening,
vii
Background and process
1. Background and process
1.1 Background
has been recommended since 1998 by WHO and the Joint United Nations Programme on HIV/AIDS (UNAIDS) as part of a comprehensive HIV and AIDS care strategy.[11] It has subsequently been included in a number of WHO guidelines and recommendations. [3,12] However, its implementation has been very slow and has been impeded by several barriers including lack of an accepted approach to exclude active TB disease and restricted access to isoniazid for fear of developing drug resistance. By the end of 2009, globally only 85 000 people living with HIV received IPT.[1] It is not known what proportion of these were children. In April 2008, WHO convened the Three Is for HIV/ TB Meeting, which called for a re-conceptualization of the existing WHO/UNAIDS Policy on IPT to reflect new scientific evidence and thinking about HIV and TB prevention, care and treatment, and expedite the implementation of this important intervention in tandem with ICF.[4] Therefore, the objective of these guidelines is to provide guidance for national TB and AIDS programmes by updating existing WHO recommendations with new evidence, taking into consideration the changing context of HIV and TB prevention, treatment and care. The new guidelines focus on facilitating the implementation of IPT and ICF. The guidelines are also intended to highlight and strengthen the leadership role of national AIDS programmes and HIV stakeholders to scale up the implementation of TB screening and provision of IPT among people living with HIV.
IV is the strongest risk factor for developing tuberculosis (TB) disease in those with latent or new Mycobacterium tuberculosis infection. The risk of developing TB is between 20 and 37 times greater in people living with HIV than among those who do not have HIV infection.[1] TB is responsible for more than a quarter of deaths among people living with HIV.[2] Relatively more women than men were detected to have TB in countries with a prevalence of HIV infection of more than 1%.[1] In response to the dual epidemics of HIV and TB, the World Health Organization (WHO) has recommended 12 collaborative TB/HIV activities as part of core HIV and TB prevention, care and treatment services.[3] These include interventions that reduce the morbidity and mortality from TB in people living with HIV, such as the provision of antiretroviral therapy (ART) and the Three Is for HIV/TB: intensified case-finding of TB (ICF), isoniazid preventive therapy (IPT) and infection control for TB.[4] A high rate of previously undiagnosed TB is common among people living with HIV.[5,6] ICF and treatment of TB among people living with HIV interrupts disease transmission by infectious cases,[7,8] reduces morbidity and delays mortality.[9] Most importantly, active screening for TB offers the opportunity to provide preventive therapy for those who do not have symptoms and signs of TB.[10] IPT is a key public health intervention for the prevention of TB among people living with HIV and
1.2 Target audience

he guidelines are aimed at health-care workers providing care for people living with HIV, policy-makers and health programme managers working in the field of HIV/AIDS and
TB. These guidelines are also intended for governments, nongovernmental organizations, donors and patient support groups that address HIV and TB.
1.3 Scope
The guidelines present a set of recommendations that will help reduce TB disease in people living with HIV, their families and communities through a combination of screening for TB and provision of IPT. The following eight questions were used to guide the review of the evidence for developing the guidelines. 1. What is the best combination of signs, symptoms and diagnostic procedures (e.g. smear microscopy, radiography, serum-based tests such as interferon-gamma release assays [IGRA]) that can be used as screening tools to determine the eligibility for treatment of latent TB infection (LTBI)? 2. What is the optimal duration and drug regimen (e.g. INH, rifampicin, etc.) for treatment of LTBI to reduce the risk of developing TB among people living with HIV? 3. What is the optimal time to start considering initiation of IPT (i.e. should immune status be considered and should IPT be started with ART)? 4. Should secondary treatment of LTBI be provided for people living with HIV to prevent reinfection or recurrence of TB after successful completion of TB treatment? 5. Does treatment for LTBI among people living with HIV lead to significant development of monoresistance against the drug(s) used for LTBI treatment? 6. Will low adherence rates to treatment for LTBI be a barrier to the implementation of LTBI treatment among people living with HIV? 7. Is the provision of treatment for LTBI costeffective? 8. Is the use of tuberculin skin test (TST) feasible in resource-limited settings? The guidelines include evidence-based recommendations for adults, children and infants, the summary and grading of evidence, implementation issues and key research gaps. In contrast to the 1998 WHO/UNAIDS Policy, these new guidelines reconceptualize ICF and the provision of IPT as integral and interlinked components of quality care for people living with HIV. The revised guidelines recommend the use of an evidence-based, simplified TB screening algorithm that relies on four clinical symptoms to identify those eligible for either IPT or further diagnostic work-up for TB or other diseases. Although a subject of another set of WHO guidelines, screening for TB also allows for improved infection control measures to prevent nosocomial transmission. These guidelines also include recommendations for people living with HIV who are pregnant, on ART and have completed TB treatment. The guidelines will be reviewed and updated in five years according to WHO procedure.[13]
1.4 Process of formulating the guidelines

s part of the Guideline Review Committee (GRC)-recommended process, the WHO HIV/ AIDS and Stop TB Departments conducted a global policy meeting on 2527 January 2010 to review the evidence regarding ICF and IPT, and to reconceptualize the 1998 WHO/UNAIDS Policy on TB prevention (Annexes 13). Key questions were identified and a comprehensive review of the available scientific evidence was conducted to formulate the recommendations. A WHO Guidelines Group to review the evidence and formulate the recommendations was established and a comprehensive review of the available scientific evidence for eight key questions (see above) was prepared. Systematic literature reviews of studies related to the eight questions among people living with HIV were conducted using PubMed, and various combinations of keywords were used to search for studies related to each question. A search was also conducted for abstracts presented at conferences on TB and lung disease organized by the International Union Against TB and Lung Disease (The Union) and the International AIDS Society between 2000 and 2008. All retrieved titles and abstracts were reviewed for their relevance to the topic in the question. The reference lists of the retrieved studies were also reviewed to identify further studies that met the eligibility criteria. In addition, recognized experts in the field were contacted to identify studies that were not available (e.g. unpublished) in the initial electronic search for each question. The quality of evidence and strength of recommendation was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology.[14] In the GRADE assessment process, the quality of a body of evidence is defined as the extent to which one can be confident that the reported estimates of effect (desirable or undesirable) available
Background and process
from the evidence are close to the actual effects of interest. The usefulness of an estimate of the effect (of the intervention) depends on the level of confidence in that estimate. The higher the quality of evidence, the more likely a strong recommendation can be made; however, the decision regarding the strength of the evidence also depends on other factors. Although the GRADE evidence assessment process was used for all of the questions, it was not always possible
to calculate GRADE profiles for all the questions because there was a lack of data and information to calculate the necessary risk ratios. The initial ranking of the evidence for each question was collectively done by the consultants of the systematic review and members of the WHO Steering Group, which was later presented and discussed by the Guidelines Group. In the GRADE profiles, the following levels of assessment of the evidence were used:
Evidence level High Moderate Low Very low
Rationale
Further research is very unlikely to change our confidence in the estimate of effect. Further research is likely to have an important impact on our confidence in the effect. Further research is very likely to have an estimate of effect and is likely to change the estimate. Any estimate of effect is very uncertain. strength of the evidence. The Group used open voting and discussion to arrive at a consensus for each of the recommendations. After the initial draft was reviewed by the Guidelines Group, the comments were incorporated into a draft that was then sent to over 200 people for peer review. The Coordinators of the process, representing the two technical units of WHO (HIV/AIDS and Stop TB Departments), incorporated comments from around 30 internal and external peer reviewers to finalize the recommendations.
Reports were also commissioned from people living with HIV and affected communities regarding the key questions and the summary of the evidence (Annex 4). The final recommendations also take into consideration the quality of the evidence, cost, feasibility, and values and preferences of the community, and health-care workers. The Guidelines Group, which included two GRADE methodologists, assessed the evidence along with the risks and benefits of each recommendation, and determined their recommendations and the
he strength of the recommendations reflects the degree of confidence of the Guidelines Group that the desirable effects of adherence to the recommendations outweigh the undesirable effects. Desirable effects considered include beneficial health outcomes (e.g. prevention and early diagnosis of TB, reduced TB-related morbidity and mortality), less burden and savings, whereas undesirable effects can include harms, more burden and costs. Burdens considered include the demands of adhering to the recommendations that programmes, patients or caregivers (e.g. family) may have to bear, such as having to undergo more frequent tests, taking additional medications or opting for a treatment that has a risk for toxicity. The recommendations in these guidelines were graded into two categories as follows:
1.5 Strength of recommendations

A STRONG RECOMMENDATION is one for which the Guidelines Group is confident that the desirable effects of adherence to the recommendation outweigh the undesirable effects. This can be either in favour of or against an intervention. A CONDITIONAL (WEAK) RECOMMENDATION is one for which the panel concludes that the desirable effects of adherence to the recommendation probably outweigh the undesirable effects, but the panel is not confident about these trade-offs. Reasons for not being confident can include: absence of highquality evidence, presence of imprecise estimates of benefits or harms, uncertainty or variation regarding how different individuals value the outcomes, small benefits, and benefits that may not be worth the costs (including the costs of implementing the recommendation).
Strength of recommendation Strong Conditional (weak)
Rationale
The panel is confident that the desirable effects of adherence to the recommendation outweigh the undesirable effects. The panel concludes that the desirable effects of adherence to the recommendation probably outweigh the undesirable effects. However: Data to support the recommendation are scant; or The recommendation is only applicable to a specific group, population or setting; or New evidence may result in changing the balance of risk to benefit; or The benefits may not warrant the cost or resource requirements in all settings.
1.6 Adaptation of the guidelines
he guidelines have been developed for a global audience and it is expected that regions and countries will adapt the recommendations to suit their own circumstances. These include consideration of the epidemiology of TB and HIV, and defining settings with the highest rates of prevalence and transmission of TB among people living with HIV (for example, to implement IPT lifelong or for 36 months). The ultimate goal of these adaptations should be to scale up implementation of services for TB screening, prevention and treatment as core functions of HIV prevention, treatment and care services. Depending on the situation of the country, a national consultation process involving all the stakeholders should help ensure the creation of a policy and programme environment that is conducive to implementation. Critical factors that need to be addressed during the national adaptation
process include incorporation of TB screening and IPT as core interventions in the treatment and care package for people living with HIV. Other critical functions include the development of standardized operating procedures, access to INH (preferably 300 mg tablets) for HIV service providers and implementers, and establishment of an effective and standardized monitoring and evaluation (M&E) system. The evaluation of the efficacy of the guidelines will be done through the global TB and HIV/AIDS reporting system, which will monitor country and global implementation of IPT and ICF. In addition, WHO and ministries of health, along with key stakeholders, will participate in countrylevel programme reviews to monitor adaptation and implementation of the guidelines. Feedback from the community and other stakeholders will be used to revise the next edition of the guidelines.
Intensified case-finding for and prevention of tuberculosis in adults and adolescents living with HIV
2. Intensified case-finding for and prevention of tuberculosis in adults and adolescents living with HIV
2.1 Screening for TB
Adults and adolescents living with HIV should be screened for TB with a clinical algorithm and those who do not report any one of the symptoms of current cough, fever, weight loss or night sweats are unlikely to have active TB and should be offered IPT.
Adults and adolescents living with HIV and screened for TB with a clinical algorithm and who report any one of the symptoms of current cough, fever, weight loss or night sweats may have active TB and should be evaluated for TB and other diseases.
Strong recommendation, moderate quality of evidence All people living with HIV, wherever they receive care, should be regularly screened for TB using a clinical algorithm at every visit to a health facility or contact with a health worker. Screening for TB is important, regardless of whether they have received or are receiving IPT or ART. As part of the guidelines development process, a comprehensive systematic primary patient data meta-analysis, including 12 observational studies involving over 8000 people living with HIV, was used to develop the best screening rule to identify adults and adolescents living with HIV who are unlikely to have active TB disease (Annex 5).[15] The analysis found that the absence of all the symptoms of current cough, night sweats, fever or weight loss can identify a subset of people living with HIV who have a very low probability of having TB disease. This best screening rule has a sensitivity of 79% and a specificity of 50%. At 5% TB prevalence among people living with HIV, the negative predictive value was 97.7% (95%CI 97.498.0). This high negative predictive value ensures that those who are negative on screening are unlikely to have TB and hence can reliably start IPT. Therefore, the Guidelines Group recommends that adults and adolescents living with HIV should be screened for TB using a clinical algorithm at every visit to a health facility or contact with a health worker. Those who do not have current cough, fever, weight loss or night sweats are unlikely to have active TB
and should be offered IPT. This recommendation is applicable for those living with HIV irrespective of the degree of immunosuppression, and for those on ART, those who have previously been treated for TB and pregnant women (Figure 1). Furthermore, the GRADE assessment of the evidence showed that the addition of abnormal findings on chest radiography to the four-symptom-based rule increases the sensitivity from 79% to 91% with a drop in specificity from 50% to 39%. At a 5% TB prevalence rate among people living with HIV, augmenting the symptom-based rule with abnormal findings on chest radiography increases the negative predictive value by a margin of only 1% (98.7% versus 97.8%). On the other hand, the addition of abnormal chest radiographic findings to the symptom-based rule at a TB prevalence of 20% among people living with HIV increases the negative predictive value by almost 4% (94.3% versus 90.4%). This suggests that chest radiography could be considered to augment the utility of symptom-based screening in settings with high TB prevalence rates among people living with HIV. However, the Guidelines Group recognized that the desire for increased sensitivity and negative predictive value is often accompanied by significant feasibility concerns such as cost, workload, infrastructure and qualified staff. Therefore, the Guidelines Group recommends that in most settings, the symptom-based rule should be implemented,
regardless of the availability of radiography, consistent with the recommended algorithm (Figure 1). Adults and adolescents living with HIV who have any one of the four symptoms (current cough, fever, weight
loss or night sweats) may have active TB and should be evaluated for TB and other diseases. The diagnostic work-up for TB should be done in accordance with national guidelines and sound clinical practice to identify either active TB or an alternative diagnosis.
2.2 Efficacy, regimen and duration

Adults and adolescents living with HIV who have an unknown or positive TST status and who are unlikely to have active TB should receive at least six months of IPT as part of a comprehensive package of HIV care. IPT should be given to such individuals irrespective of the degree of immunosuppression, and also to those on ART, those who have previously been treated for TB and pregnant women.
Adults and adolescents living with HIV who have an unknown or positive TST status and are unlikely to have active TB should receive at least 36 months of IPT. IPT should be given to such individuals irrespective of the degree of immunosuppression, and also to those on ART, those who have previously been treated for TB and pregnant women.
Conditional recommendation, moderate quality of evidence
T T
6
2.2.1 Efficacy
0.87). For those who were TST positive, the reduction in confirmed, probable or possible TB increased to 64% (RR [95% CI] 0.36 [0.220.61]). Although not statistically significant, the reduction among TSTnegative persons was 14% (RR [95% CI] 0.86 [0.59, 1.26]) and in those with unknown TST status it fell by 14% (RR [95% CI] 0.86 [0.48, 1.52]).[16] The Guidelines Group concluded that there is benefit in providing TB preventive therapy to people living with HIV regardless of the TST status, with greater protective benefit seen in those with a positive TST.
he Guidelines Group reviewed the available evidence regarding the benefit of chemotherapy preventing TB disease (latent disease, reactivation or reinfection) in adults living with HIV (Annex 6). A GRADE assessment was used to examine the existing evidence on drug regimens including the 12 randomized controlled trials used in the Cochrane review of preventive therapy.[16] For those with confirmed, probable or possible TB disease, preventive chemotherapy reduces the overall risk of developing TB by 33% (relative effect 0.67; CI 0.51
2.2.2 Regimen and duration

were associated with higher rates of toxicity (Table 1). The Guidelines Group concluded that INH at 300 mg/ day remains the drug of choice for chemotherapy to prevent TB in adults living with HIV. The Guidelines Group also reviewed the evidence on the duration and durability of effect of IPT in people living with HIV. The critical outcomes of interest considered were the efficacy of IPT in preventing active TB, relapse, reinfection and toxicity.
he Guidelines Group reviewed the evidence on a wide range of regimens used for TB prevention and their duration among people living with HIV, including results from three unpublished trials (Annex 6). The Group reviewed studies of the drug combinations used for prevention including INH, rifampicin, pyrazinamide and rifapentine. A total of eight studies compared INH alone with other regimens, and found that regimens that included pyrazinamide, rifampicin and rifapentine were as efficacious as INH alone, but
2.2.2.1 Table 1: Comparison of the efficacy of different drug regimens

Intervention INH INH INH INH and rifampicin INH and rifapentine Comparator
Rifampicin and pyrazinamide INH and rifampicin
RR (95% CI)
1.03 (0.751.4) 0.97 (0.521.83)
Quality of evidence
Moderate Moderate Low Moderate Moderate
INH, rifampicin and pyrazinamide 0.69 (0.231.57) INH, rifampicin and pyrazinamide 0.75 (0.211.82) INH 1.05 (0.561.97)
The Guidelines Group considered the existing evidence on the optimal duration of IPT including for six, nine, 12 and 36 months (Annex 6). The evidence primarily focused on the comparison of a six- and 12-month duration of IPT, and found no significant difference in efficacy.[16] Although nine months of IPT is supported by evidence and recommended in some guidelines, there are no studies that have directly compared IPT for six and nine months. This led the Guidelines Group to strongly recommend the six-month duration. The protective effect of IPT decreases with time and the durability ranges for up to five years. The Guidelines Group reviewed emerging unpublished evidence from two clinical trials that suggest increased benefit with a 36-month
or longer duration of IPT, particularly in people who are TST positive.[17,18] Given that longer trials are expensive and unlikely to be done, the Guidelines Group considered a duration of at least 36 months as a surrogate for lifelong treatment. It also emphasized the potential benefit of extended IPT for people living with HIV in settings with a background of high HIV and TB prevalence and transmission. Given the preliminary and scanty nature of the evidence, feasibility concerns and potential adverse events, the Guidelines Group conditionally recommends 36 months duration of IPT for people living with HIV in settings with high TB prevalence and transmission, as determined by the local context and national guidelines.
2.2.3 Immune status and concomitant use of IPT with ART

potential benefit of concomitant use of IPT with ART, the Guidelines Group strongly recommends that IPT be given irrespective of immune status and whether or not a person is on ART. IPT initiation or completion should not be the cause for a delay in starting ART for eligible people living with HIV.[21] However, the Guidelines Group recognizes the absence of evidence on whether concomitant initiation of IPT with ART or delayed initiation of IPT is better in terms of efficacy, toxicity or the development of immune reconstitution.
he Guidelines Group reviewed available data regarding the initiation of IPT and immune status, including concomitant use with ART. Six studies were examined which showed contrasting results regarding the reduction of TB risk by immune status (Annex 6). Additional protective benefits of concomitant use of IPT with ART were demonstrated in two observational studies from Brazil [19] and South Africa,[20] and a sub-analysis of data from an unpublished randomized clinical trial from Botswana.[17] Based on this evidence and the
regnant women living with HIV are at risk for TB, which can impact on maternal and perinatal outcomes.[22] These could range from death of the mother and the newborn, to prematurity and low birth weight of the newborn. [23] The Guidelines Group stressed the importance of screening pregnant women living with HIV for active TB using the clinical algorithm as mentioned above. This implies the introduction of the clinical algorithm into maternal HIV services in order
2.2.4 Pregnant women

to prevent, diagnose and treat TB. The Group concluded that evidence and experience from the pre-HIV and HIV era suggest that IPT is safe in pregnant women. Therefore, the Guidelines Group strongly recommends that pregnancy should not exclude women living with HIV from symptombased TB screening and receiving IPT. However, sound clinical judgement is required for decisions such as the best time to provide IPT to pregnant women.
he Guidelines Group reviewed the evidence and discussed IPT as secondary prophylaxis for people who have previously been successfully treated for TB. GRADE assessment of the evidence from four studies including three randomized controlled trials [2426] and one observational study [27] showed the value of providing IPT immediately after successful completion of TB treatment (Annex 7). The Guidelines Group strongly recommends that adults and adolescents living with HIV who successfully complete their TB
2.2.5 Patients previously treated for TB (secondary prophylaxis)

treatment should continue receiving INH for another six months and should conditionally receive it for 36 months based on the local situation (e.g. high rates of TB prevalence and transmission) and existing national guidelines. There was no evidence on the potential role of IPT for those who had successfully completed treatment for multidrug-resistant (MDR) or extensively drug-resistant (XDR) TB. Therefore, the Guidelines Group did not make any recommendation on the use of IPT after successful treatment for MDR or XDR TB.
eople living with HIV in congregate settings, such as prisons and centres for refugees or internally displaced persons, have a higher risk for and incidence of TB, HIV infection and drug use.[2] Special attention has to be paid to ensure screening for TB and provision of IPT for these groups. Injecting drug users have a higher risk of coinfections with HIV, TB and hepatitis-causing viruses. Screening
2.2.6 Special populations

for TB and providing IPT for injecting drug users should be combined with harm reduction measures, including the provision of testing for hepatitis B and hepatitis C infection, and referral for positive cases. [28] Sound clinical judgement is required to weigh the benefits of IPT among injecting drug users with hepatitis coinfection. IPT should not be provided in the presence of active hepatitis.
2.2.7 Figure 1. Algorithm for TB screening in adults and adolescents living with HIV in HIV-prevalent and resource-constrained settings
Adults and adolescents living with HIV* Screen for TB with any one of the following symptoms: Current cough Fever Weight loss Night sweats
No
Yes
Assess for contraindications to IPT No Give IPT Yes Defer IPT
Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Not TB Follow up and consider IPT TB Treat for TB
Screen for TB regularly at each encounter with a health worker or visit to a health facility
* Every adult and adolescent should be evaluated for eligibility to receive ART. Infection control measures should be prioritized to reduce M. tuberculosis transmission in all settings that provide care. Chest radiography can be done if available, but is not required to classify patients into TB and non-TB groups. In high HIVprevalence settings with a high TB prevalence among people living with HIV (e.g. greater than 10%), strong consideration must be given to adding other sensitive investigations. Contraindications include: active hepatitis (acute or chronic), regular and heavy alcohol consumption, and symptoms of peripheral neuropathy. Past history of TB and current pregnancy should not be contraindications for starting IPT. Although not a requirement for initiating IPT, TST may be done as a part of eligibility screening in some settings. Investigations for TB should be done in accordance with existing national guidelines.
FOOTNOTES TO ALGORITHM FOR ADULTS
2.3 Detecting latent TB infection in resource-constrained settings

2.3.1 Tuberculin skin test (TST) and IPT
TST is not a requirement for initiating IPT in people living with HIV.
People living with HIV who have a positive TST benefit more from IPT; TST can be used where feasible to identify such individuals.
Strong recommendation, high quality of evidence TST relies on a competent immune response to identify people with latent Mycobacterium tuberculosis infection. Multiple studies in people living with HIV demonstrate that IPT is more effective in people with a positive TST than in those with a negative test.[16] In addition, the use of TST could reduce the number of patients receiving IPT and the numbers needed to treat to prevent one case of active TB. However, in resource-constrained settings, operational challenges to the implementation of TST are significant impediments for access to IPT. Such challenges include the costs of procuring tuberculin and administering the test, maintaining an effective supply chain, training staff in administering and accurately reading the test, and the need for the patient to attend the clinic at least twice over 4872 hours with its associated inconvenience and cost.[29] In addition, the immunological status of the patient and the negative results in anergic patients or those with a long lapse between infection and the TST may affect its interpretation.[30,31] Although some studies suggest that using TST is cost-effective, there is a limited supply of tuberculin worldwide, it is costly to ship and requires an adequate cold chain to ensure accurate test performance (see Annex 8). The Guidelines Group strongly recommends that in resource-constrained settings, TST should not be a requirement for initiating IPT for people living with HIV. People living with HIV whose TST status is unknown should be started on IPT after symptom-based screening for TB. However, given that TST-positive patients benefit more from IPT than those who are TST negative, the test can be used where feasible. People living with HIV who are TST negative should be assessed on a case-bycase basis for their individual risk of TB exposure and the added advantage of the provision of IPT (e.g. health-care workers, prisoners, miners and others who live in a high TB transmission setting). In settings where TST is not available, the Guidelines Group encourages national programmes to explore its expanded use as a potential adjunct to enhancing IPT implementation.
he Guidelines Group discussed the GRADE assessment of the evidence on the use of IGRA as a screening tool to identify patients with latent TB infection (Annex 9). Two types of IGRA were considered: Quantiferon Gold in tube assay and T-Spot assay. Two studies considered the ability of IGRA to predict development of TB over time. [32,33] Eight studies evaluated the performance of Quantiferon Gold in tube assay among HIV-infected adults with confirmed TB, and one study evaluated its sensitivity among children living with HIV diagnosed with TB. Similarly, five studies reported the sensitivity of a T-Spot assay among adults living with HIV and TB, and two studies reported its sensitivity among children living with HIV with confirmed TB.
2.3.2 Interferon-gamma release assays (IGRA)

However, IGRA cannot generally distinguish between active TB disease and latent infection [34] and their performance is compromised among people living with HIV compared to those without HIV. Significantly higher rates of indeterminate test results were found with Quantiferon Gold in tube test in persons with HIV compared to persons without HIV, and in persons with low CD4 cell counts compared to persons with higher CD4 cell counts. Its sensitivity was also markedly reduced among patients with low CD4 counts. Similarly, while most studies found no impact of low CD4 cell count on the sensitivity of T-Spot assay, at least one study found that sensitivity was significantly reduced among patients with low CD4 counts.
The Guidelines Group noted that the data suggesting the use of IGRA to identify latent TB in persons living with HIV are restricted to studies conducted in low TB-prevalence settings and there is no evidence that
IGRA will determine who will benefit most from IPT. Based on the available data, the Guidelines Group concluded that IGRA are not recommended to screen people living with HIV for eligibility to receive IPT.
2.4 Issues to consider for implementation of IPT

IV treatment and care services should include a comprehensive approach to preventing, diagnosing and treating TB with an emphasis on the Three Is for HIV/TB. The Guidelines Group concluded that providing IPT as a core component of HIV preventive care should be the responsibility of national AIDS programmes and HIV service providers. In addition, IPT should not be viewed as an isolated intervention and should be part of a TB prevention package along with infection control for TB, ICF and the provision of early ART to those with CD4 counts <350 cells/mm3 (people with TB should receive ART irrespective of CD4 count). National AIDS programmes and providers
2.4.1 Primary ownership by HIV service providers

of HIV services should ensure the meaningful engagement of people living with HIV, persons with TB and their communities in both the planning and implementation of these interventions.[35] The implementation of TB screening and IPT needs to be monitored and evaluated through established and recommended patient M&E systems [36] that should use internationally recommended indicators. [37] HIV stakeholders implementing TB screening and IPT in resource-limited settings outside of the facilities run by the government should ensure that a reporting mechanism is established so that their data are captured in one national M&E system.
2.4.2 IPT and drug-resistant TB

Providing IPT to people living with HIV does not increase the risk of developing INH-resistant TB. Therefore, concerns regarding the development of INH resistance should not be a barrier to providing IPT.
Strong recommendation, moderate quality of evidence One of the reasons commonly cited for not offering IPT to people living with HIV is the fear of developing drug-resistant TB. The Guidelines Group reviewed the evidence on the provision of IPT and drug-resistant TB, which was presented after GRADE assessment of the evidence (Annex 10). This included eight studies and the results of a meta-analysis which concluded that INH resistance is not significantly associated with the provision of IPT.[38] The GRADE assessment of the evidence examined the relative risk of developing INH-resistant TB among all of those receiving isoniazid and found no statistically significant increased risk of resistance (RR 95% CI= 1.87 [0.655.38]). In addition, the results of a study that was under publication and showed no risk of development of drug resistance after provision of IPT to gold miners were also presented and discussed.[39] The Guidelines Group also noted that regular TB screening for those taking IPT will help identify those who could develop TB as early as possible. This early identification will allow for prompt diagnosis and treatment, which should also help to prevent the development of drug-resistant TB. The Guidelines Group noted that in settings with high INH resistance, fewer patients are likely to benefit from IPT, and the decision to provide access to IPT for people living with HIV should thus be based on the local context. Programmes implementing IPT are encouraged to introduce international and national TB drug-resistance surveillance systems that also include HIV testing as an integral component.
10
2.4.3 Adherence and clinical follow up

he Guidelines Group reviewed the evidence regarding the importance of adherence to IPT (Annex 11). The available data were observational and did not directly address whether poor adherence adversely affects individual or programme outcomes. Adherence rates for IPT
varied widely from 34% to 98%, and a number of factors were identified to improve adherence.[40 44] The Guidelines Group noted that, although treatment completion is important for good individual and programme outcomes, the primary objective should be to ensure that people do not continue to take IPT in the rare instance of active TB or development of toxicity. People living with HIV and receiving IPT should have regular clinical
follow up based on the national, local and clinical context. This includes regular screening using the TB symptom-based rule during every contact with a health-care provider. The Guidelines Group noted that the co-formulation of INH with other drugs (e.g. ART or CPT) could reduce the pill burden and enhance adherence, and called for expedited development of such co-formulations. The Guidelines Group strongly recommends that concerns regarding adherence should not be a barrier to implementing IPT.
he Guidelines Group concluded that the available data on the cost-effectiveness of IPT is of low quality with significant variability between outcome measures, assumptions and analytical procedures (Annex 12). It was recognized that this is an area that requires additional research to better inform programmatic decision-making. However,
2.4.4 Cost-effectiveness of IPT

after a review of the evidence, the Guidelines Group strongly recommends that the provision of IPT is likely to be cost-effective. This supports the overall recommendation for the wide use of IPT within comprehensive HIV prevention, care and treatment services, both as a measure of good clinical practice and as a likely cost-effective measure.
11
3. Intensified tuberculosis case-finding and prevention of tuberculosis in children living with HIV
Children living with HIV who do not have poor weight gain,* fever or current cough are unlikely to have active TB. Children living with HIV who have any one of the following symptoms poor weight gain*, fever, current cough or contact history with a TB case may have TB and should be evaluated for TB and other conditions. If the evaluation shows no TB, they should be offered IPT regardless of their age.
* Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weight-for-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening.
Strong recommendations, low quality of evidence Encouraging efforts have been made to expand access to early diagnosis of HIV in children as part of HIV prevention, care and treatment. TB screening, prevention and treatment should be an integral part of these services. This section of the guidelines is specifically targeted at children living with HIV. However, in circumstances where HIV-exposed infants and children are receiving HIV care pending a result of a virological or serological test, they should be considered as children living with HIV and get the appropriate services until their results are known. For infants less than 6 weeks of age and unknown HIV exposure, and in settings where the HIV epidemic is generalized (i.e. >1% prevalence in the population attending antenatal care services), programmes are strongly recommended to provide HIV serological testing to mothers or their infants in order to establish exposure status. Virological testing should be conducted at 46 weeks of age for infants known to be exposed to HIV, or at the earliest possible opportunity for those seen after 46 weeks of birth. For children 12 18 months of age, diagnosis using virological testing is recommended. However, in resource-constrained settings where access to virological testing is limited, it is recommended that, for this age group, virological tests be performed only after positive serological testing. A definitive diagnosis of HIV in children aged 18 months or more (with known or unknown HIV exposure) can be made with HIV serological tests, including rapid serological tests following standard testing algorithms used for adults.[45] The Guidelines Group stressed that infants and children living with HIV should routinely be screened for TB as a part of standard clinical care, whether they are receiving TB prophylaxis or ART. However, the diagnosis of TB in children, with or without HIV, is difficult and clinicians need a high index of suspicion at all times and should follow national guidelines. A history of contact of the infant or child with someone with TB (regardless of the type of TB disease) within the home is particularly important and should motivate the health-care worker to screen for TB in the child and among the other family members. Based on this analysis and the relative lack of good studies, the Guidelines Group concluded that the quality of evidence is low and available data are limited regarding the best approach to screening infants and children for TB. The range of evidence assessed using GRADE included a number of scoring systems for children who are not infected with HIV. However, such scoring systems were not found to be as effective in children living with HIV (Annex 13).[46] The evidence also included one unpublished study that investigated a combination of signs and symptoms to reliably exclude active TB in a child with HIV. The
12
Intensified tuberculosis case-finding and prevention in children living with HIV
study showed that the absence of cough of more than two weeks duration, fever and failure to thrive could identify children unlikely to have active TB with a 99% negative predictive value. Such children would therefore be eligible for IPT. Similarly, the presence of cough for more than two weeks, or failure to thrive or fever has a sensitivity of 90% and specificity of 65%, and is therefore useful for identifying children in need of further screening for TB or alternative diagnoses. [47] Another study among 1024 children suggested that weight loss and cough for more than two weeks and fatigue had a sensitivity of only 56% and specificity of 62%.[48] In order to facilitate programmatic implementation and increase the likelihood of identifying children without active TB for IPT, the Guidelines Group recommends that the duration of cough as a screening rule should be reduced to the presence of any current cough, in line with the recommendation for adolescents
and adults. Unlike the screening rule for adults and adolescents, this recommendation is based on expert opinion and clinicians need to broaden the differential diagnosis to include other diseases that may cause children with HIV to present with current cough, fever and poor weight gain. Similarly, contact history with a known TB case should raise the clinical suspicion of TB in children living with HIV. The Guidelines Group recommends that children living with HIV without poor weight gain, fever and current cough are unlikely to have active TB and should be offered IPT (see below for age-specific recommendations). Similarly, children living with HIV with any one of the following symptoms poor weight gain, fever, current cough and contact with a TB case may have TB and should be evaluated for TB and other diseases. If the evaluation shows no TB, such children should be offered IPT regardless of their age (Figure 2).
3.2 Regimen and duration

Children living with HIV who are more than 12 months of age and who are unlikely to have active TB on symptom-based screening, and have no contact with a TB case should receive six months of IPT (10 mg/kg/ day) as part of a comprehensive package of HIV prevention and care services. In children living with HIV who are less than 12 months of age, only those who have contact with a TB case and who are evaluated for TB (using investigations) should receive six months of IPT if the evaluation shows no TB disease.
Strong recommendations, moderate quality of evidence
All children living with HIV who have successfully completed treatment for TB disease should receive INH for an additional six months.
Conditional recommendation, low quality of evidence Two studies were considered for the GRADE assessment of the evidence (Annex 14). One study suggested considerable benefits for children receiving INH for six months, in particular, with regard to significant reductions in mortality.[49] However, findings from a randomized control trial conducted in South Africa showed that when HIV-infected infants with no known exposure to a TB source case are identified in the first three to four months of life, given rapid access to ART and carefully monitored for new TB exposure or disease on a monthly basis, there is no benefit from IPT (Madhi 2008, unpublished). Therefore, based on this, the Guidelines Group recommends that all children living with HIV who are more than 12 months of age and who are unlikely to have active TB should receive six months of IPT as part of a comprehensive package of HIV care. For children less than 12 months of age, only those who have been evaluated for TB (using investigations) should receive six months IPT if the evaluation shows no TB disease. In contrast to adults and adolescents, there is no evidence to support the use of INH for longer than six months in children. Therefore, the Guidelines Group concluded that until more data are available, INH for children could not
13
be recommended for more than six months. Similarly, there is no evidence on whether repeating a course of IPT is beneficial for children. INH should be given at a dose of 10 mg/kg body weight and it is desirable that vitamin B6 be supplied
with INH at a dose of 25 mg daily. All available data to date suggest that INH is not toxic for children, even in those receiving ART. The following table shows a simplified dosing schedule for children (total dose 10 mg of INH/kg/day).
Weight range (kg)

<5 5.19.9 1013.9 1419.9 2024.9 >25
Number of 100 mg tablets of INH to be administered per dose (total dose 10 mg/kg/day)
tablet 1 tablet 1 2 tablet tablets 2 tablets 3 tablets or one adult tablet
Dose given (mg)

50 100 150 200 250 300
3.3 Secondary prophylaxis and IPT with ART in children
he Guidelines Group noted that there is no evidence on the use of IPT in children living with HIV after successful completion of TB treatment. However, like adults, children living with HIV are exposed to reinfection and recurrence of TB. Therefore, the Group conditionally recommends that all children living with HIV who have been successfully treated for TB and are
T T
3.3.1 Secondary prophylaxis

living in settings with a high TB prevalence and transmission should receive IPT for an additional six months. IPT can be started immediately after the last dose of anti-TB therapy or at a later date. TB screening should be carried out for all children living with HIV, regardless of history of TB treatment, during each contact of the child with a health-care worker (Annex 13).
3.3.2 IPT with ART in children

and adolescents to children. Therefore, the Guidelines Group conditionally recommends the combined use of IPT with ART for all children. The Guidelines Group also emphasized that ART should not be delayed while starting or completing a course of IPT.[45]
he Guidelines Group concluded that there are no data regarding the efficacy of IPT for children stratified by degree of immunosuppression. However, it was noted that there is biological plausibility in extrapolating what is known for adults
3.4 The role of TST and IGRA in evaluating children for IPT
he Guidelines Group, as in the recommendation for adults, concluded that TST is not required to initiate IPT in children and should not be routinely used as part of the process to determine eligibility for IPT (Annex 9). However, the Group noted that TST may provide important additional information in assessing a child with suspected TB, especially if there is no positive contact history. Although a positive TST may indicate infection with mycobacteria, usually Mycobacterium tuberculosis, it is not a reliable marker of TB disease activity. The
main limitation of TST in the diagnosis of TB in HIVinfected children is its variable sensitivity. Important clinical causes of false-negative results include severe malnutrition, severe TB disease and HIV infection. Therefore, in settings where it is available, TST may be used for the diagnosis of active TB in children and may also have a role in screening for LTBI. Like TST, IGRA cannot distinguish between M. tuberculosis infection and active TB disease.
14
Intensified tuberculosis case-finding and prevention in children living with HIV
Encouraging data show that IGRA are more sensitive than TST in HIV-infected children, including those with a low CD4 count and/or malnutrition.[5052]. In addition, excellent specificity for M. tuberculosis infection has been reported and, unlike TST, IGRA are unaffected by prior BCG vaccination or exposure to environmental mycobacteria. However, more evidence is needed and implementation issues
affecting most HIV-prevalence settings (cost, specific laboratory equipment and the need for a venous blood sample) have to be addressed. Therefore, the Guidelines Group strongly recommends that there is currently insufficient evidence to support the use of IGRA to identify children eligible for IPT outside research settings with laboratory-validated procedures.[53]
3.5 Figure 2: Algorithm for TB screening in children more than one year of age and living with HIV
Child more than 12 months of age and living with HIV* Screen for TB with any one of the following symptoms: Poor weight gain Fever Current cough Contact history with a TB case No Assess for contraindications to IPT No Give IPT Yes Defer IPT Yes Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Screen regularly for TB Not TB Follow up and consider IPT TB Treat for TB
FOOTNOTES TO ALGORITHM FOR CHILDREN

* All children and infants less than one year of age should be provided with IPT if they have a history of household contact with a TB case. Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weightfor-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening. Contraindications include: active hepatitis (acute or chronic) and symptoms of peripheral neuropathy. Past history of TB should not be a contraindication for starting IPT. Although not a requirement for initiating IPT, TST may be done as a part of eligibility screening in some settings. Investigations for TB must be done in accordance with existing national guidelines.
15
4. Research gaps
research, donors and the scientific community expedite the implementation of research to respond to these gaps in order to inform policy formulation and programme implementation. Along with global TB and HIV stakeholders, WHO has developed a document that summarizes the overall research priorities around TB/HIV and addresses the broader context of research gaps.[54]
he review of the evidence for formulating the recommendations exposed important unmet research needs (Annex 15). The Guidelines Group discussed the priority research gaps that need to be addressed in order to update these guidelines. The following are the key questions identified by the Guidelines Group in all the areas included in these guidelines. It is imperative that

A new point-of-care test is needed to identify those with active TB, LTBI and those not infected with M. tuberculosis; particular emphasis should be placed on new diagnostics for children. The role of IGRA in people living with HIV who are infected with M. tuberculosis with or without active TB; information is needed about the association between performance of IGRA and immune status. The use of TST testing in people living with HIV and receiving ART, with a particular emphasis on the frequency of performing TST to determine immune reconstitution and/or boosting in those who were initially TST negative. The optimal TB screening algorithm to be used across different settings with different TB and HIV disease burdens to safely initiate preventive therapy The optimal frequency of screening people with HIV for active TB with a symptom-based questionnaire Evaluation of the WHO-recommended algorithm to diagnose TB using new technological advances such as LED microscopy, rapid culture and polymerase chain reaction (PCR)-based methods Further validation of the screening algorithm in various programmatic settings Effect of ICF on nosocomial transmission, in particular, among people living with HIV, healthcare workers and/or their families Optimal diagnostic algorithm for diagnosis of TB following TB screening for IPT.
4.2 Preventive treatment for TB

Optimal duration, safety, efficacy and costeffectiveness of IPT alone or in conjunction with ART in reducing the risk of active TB, compared to ART alone among people living with HIV, particularly under programme conditions Co-formulation as a fixed-dose combination of isoniazid and vitamin B6 with co-trimoxazole, and with antiretrovirals, and evaluation of the efficacy and effectiveness of such fixed-dose combinations Further evaluate the role of vitamin B6 in people living with HIV. Evaluate the efficacy and feasibility of long-term IPT in children. Study the efficacy of IPT in people with HIV and hepatitis C virus (HCV) coinfection. Determine the best regimen for and approach to IPT for those with drug-resistant or suspected to have drug-resistant M. tuberculosis. Outcomes of TB treatment for breakthrough TB in people living with HIV Optimal timing for initiation of IPT in relation to initiation of ART For those on lifelong IPT, is there value in discontinuing it after immune reconstitution? Interaction between IPT and other medications, particularly in the context of coinfection with viral hepatitis Modelling studies to estimate the risks and benefits of IPT key considerations include the incidence and prevalence of HIV and TB, risk for TB by immune status, impact of ART on prevention of both HIV and TB, added benefit of IPT, optimum duration of IPT, prevalence of INH and rifampicin resistance, immune status, TST status.
16
Research gaps
4.3 Operational research

Potential limitations of IPT in populations with a high prevalence of INH-resistant TB Risks and benefits of administering INH (in error) to undiagnosed people with active TB Effectiveness of IPT programmes in resourcelimited settings; cost-effectiveness and cost benefit from the health systems and patients perspectives IPT and special populations: benefits of and duration for health-care workers living with HIV; frequency of screening; benefits for TST-negative health-care workers; HIV-exposed children How to operationalize short-term and lifelong IPT with a particular focus on monitoring programmes and individuals (i.e. clinical status and adherence) Population-based drug-resistance surveillance to determine the impact of IPT programmes on drug-resistant TB in the community, including increases or decreases in mono-INH and monorifampicin resistance, and MDR TB Evaluate the best national programmes or services to lead the implementation of IPT (e.g. HIV, maternal and child health [MCH], TB, all programmes). Optimal delivery of IPT and other HIV care for special groups including women and children.
17
5. References
1. WHO. Global tuberculosis control: a short update to the 2010 Report. December 2009. Geneva, Switzerland, World Health Organization, 2010. 2. Getahun H et al. HIV infection associated tuberculosis: the epidemiology and the response. Clinical Infectious Diseases, 2010, 50:S201S207; doi:10.1086/651492. 3. WHO. Interim policy on collaborative TB/HIV activities. Geneva, Switzerland, World Health Organization, 2004 (WHO/HTM/TB/2004.330; WHO/HTM/HIV/2004.1). 4. WHO Three Is for HIV/TB Meeting Report. Intensified case-finding (ICF), isoniazid preventive therapy (IPT) and TB infection control (IC) for people living with HIV. Geneva, Switzerland, World Health Organization, 2008. 5. Wood R et al. Undiagnosed tuberculosis in a community with high HIV prevalence: implications for tuberculosis control. American Journal for Respiratory and Critical Care Medicine, 2007, 175:8793. 6. Kimerling ME et al. Prevalence of pulmonary tuberculosis among HIV-infected persons in a home care program in Phnom Penh, Cambodia. International Journal of Tuberculosis and Lung Disease, 2002, 6:988994. 7. De Cock KM, Chaisson RE. Will DOTS do it? A reappraisal of tuberculosis control in countries with high rates of HIV infection. International Journal of Tuberculosis and Lung Disease, 1999, 3:457465. 8. Lawn SD, Shattock RJ, Griffin GE. Delays in the diagnosis of tuberculosis: a great new cost. International Journal of Tuberculosis and Lung Disease, 1997, 1:485486. 9. Nachega J et al. Tuberculosis active case-finding in a mother-to-child HIV transmission prevention programme in Soweto, South Africa. AIDS, 2003, 17:13981400. 10. Burgess AL et al. Integration of tuberculosis screening at an HIV voluntary counselling and testing centre in Haiti. AIDS, 2001, 15:18751879. 11. WHO/UNAIDS. WHO and UNAIDS policy statement on preventive therapy against tuberculosis in people living with HIV. Weekly Epidemiological Record, 1999, 74:385400. 12. WHO. TB/HIV: a clinical manual. Geneva, Switzerland, World Health Organization, 2004 (WHO/HTM/TB/2004.329). 13. WHO. Handbook for guidelines development. Geneva, Switzerland, World Health Organization, 2009. 14. Atkins D et al. Grading quality of evidence and strength of recommendations. British Medical Journal, 2004, 328:1490. 15. Getahun H et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource constrained settings: individual participant data meta-analysis of observational studies. PLoS Medicine, 2011, 8(1): e1000391. doi:10.1371/journal.pmed.1000391. 16. Akolo C et al. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database of Systematic Reviews, 2010, 1:CD000171. 17. Samandari TM et al.; and IPT Trial Study Group. Randomized, placebo-controlled trial of 6 vs 36 months isoniazid TB preventive therapy for HIV-infected adults in Botswana. 17th Conference on Retroviruses and Opportunistic Infections, San Francisco, 1619 February 2010 [Paper 104LB]. 18. Martinson NB et al. Novel regimens for treating latent TB in HIV-infected adults in South Africa: a randomized clinical trial. 16th Conference on Retroviruses and Opportunistic Infections, Montreal, 811 February 2009 [Paper 36bLB]. 19. Golub JE et al. The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis incidence in HIV-infected patients in Rio de Janeiro, Brazil. AIDS, 2007, 21:14411448. 20. Golub JE et al. Isoniazid preventive therapy, HAART and tuberculosis risk in HIV-infected adults in South Africa: a prospective cohort. AIDS, 2009, 23:631636. 21. WHO. Antiretroviral therapy for HIV infection in adults and adolescents. Recommendations for a public health approach, 2009 revision. Geneva, Switzerland, World Health Organization, 2009. 22. Gupta A et al. Postpartum tuberculosis incidence and mortality among HIV-infected women and their infants in Pune, India, 20022005. Clinical Infectious Diseases, 2007, 45:241249. 23. Pillay T et al. Perinatal tuberculosis and HIV-1: considerations for resource-limited settings. Lancet Infectious Diseases, 2004, 4:155165. 24. Perriens JH et al. Pulmonary tuberculosis in HIV-infected patients in Zaire. A controlled trial of treatment for either 6 or 12 months. New England Journal of Medicine, 1995, 332:779784. 25. Haller L et al. Isoniazid plus sulphadoxinepyrimethamine can reduce morbidity of HIV-positive patients treated for tuberculosis in Africa: a controlled clinical trial. Chemotherapy, 1999, 45:452465. 26. Fitzgerald DW et al. Effect of post-treatment isoniazid on prevention of recurrent tuberculosis in HIV-1-infected individuals: a randomised trial. Lancet, 2000, 356:14701474. 27. Churchyard GJ et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17:20632070. 28. WHO. Policy guidelines for collaborative TB and HIV services for injecting and other drug users: an integrated approach. Geneva, Switzerland, World Health Organization, 2008 (WHO/HTM/TB/2008.404; WHO/HIV/2008.750). 29. Pai M, Kalantri S, Dheda K. New tools and emerging technologies for the diagnosis of tuberculosis: part I. Latent
18
References
tuberculosis. Expert Review of Molecular Diagnostics, 2006, 6:413422. 30. Moline JM, Markowitz SB. Medical surveillance for workers exposed to tuberculosis. Occupational Medicine, 1994, 9:695721. 31. Markowitz N et al. Tuberculin and anergy testing in HIV-seropositive and HIV-seronegative persons. Pulmonary Complications of HIV Infection Study Group. Annals of Internal Medicine, 1993, 119:185193. 32. Aichelburg MC et al. Detection and prediction of active tuberculosis disease by a whole-blood interferon-gamma release assay in HIV-1-infected individuals. Clinical Infectious Diseases, 2009, 48:954962. 33. Clark SA et al. Tuberculosis antigen-specific immune responses can be detected using enzyme-linked immunospot technology in human immunodeficiency virus (HIV)-1 patients with advanced disease. Clinical and Experimental Immunology, 2007, 150:238244. 34. Dheda K et al. T-cell interferon-gamma release assays for the rapid immunodiagnosis of tuberculosis: clinical utility in high-burden vs low-burden settings. Current Opinion in Pulmonary Medicine, 2009, 15:188200. 35. WHO. Three interlinked patient monitoring systems for HIV Care/ART, MCH/PMTCT and TB/HIV: standardized minimum data set and illustrative tools. Geneva, Switzerland, World Health Organization, 2009. 36. WHO, UNAIDS and OGAC. A guide to monitoring and evaluation for collaborative TB/HIV activities. Geneva, Switzerland, World Health Organization, 2009 (WHO/HTM/TB/2009.414; WHO/HTM/HIV 09.01). 37. Granich R et al. Prevention of tuberculosis in people living with HIV. Clinical Infectious Diseases, 2010, 50:S215 S222; doi:10.1086/651494. 38. Balcells ME et al. Isoniazid preventive therapy and risk for resistant tuberculosis. Emerging Infectious Diseases, 2006, 12:744751. 39. Van Halsema CL et al. Tuberculosis outcomes and drug susceptibility in individuals exposed to isoniazid preventive therapy in a high HIV prevalence setting. AIDS, 2010, 24:10511055. 40. Halsey NA et al. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet, 1998, 351:786792. 41. Souza CT et al. Effectiveness and safety of isoniazid chemoprophylaxis for HIV-1 infected patients from Rio de Janeiro. Memrias do Instituto Oswaldo Cruz, 2009, 104:462467. 42. Hawken MP et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS, 1997, 11:875882. 43. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 44. Whalen CC et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. UgandaCase Western Reserve University Research Collaboration. New England Journal of Medicine, 1997, 337:801808. 45. WHO. Antiretroviral therapy for HIV infection in infants and children: towards universal access. Recommendations for a public health approach. 2010 revision. Geneva, Switzerland, World Health Organization, 2010. 46. Edwards DJ, Kitetele F, Van Rie A. Agreement between clinical scoring systems used for the diagnosis of pediatric tuberculosis in the HIV era. International Journal of Tuberculosis and Lung Disease, 2007, 11:263269. 47. Song R et al. Evaluation of TB screening approaches among HIV-infected children Rwanda, 2008. 5th IAS conference on HIV pathogenesis and treatment, 1922 July 2009 [Abstract no TUPEB132]. 48. Marais BJ et al. A refined symptom-based approach to diagnose pulmonary tuberculosis in children. Pediatrics, 2006, 118:e13501359. 49. Zar HJ et al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomised controlled trial. British Medical Journal, 2007, 334:136. 50. Mandalakas AM et al. High level of discordant IGRA results in HIV-infected adults and children. International Journal of Tuberculosis and Lung Disease, 2008, 12:417423. 51. Liebeschuetz S et al. Diagnosis of tuberculosis in South African children with a T-cell-based assay: a prospective cohort study. Lancet, 2004, 364:21962203. 52. Davies MA et al. Detection of tuberculosis in HIV-infected children using an enzyme-linked immunospot assay. AIDS, 2009, 23:961969. 53. WHO/The Union. Guidance for national tuberculosis and HIV programmes on the management of tuberculosis in HIV-infected children: recommendations for a public health approach. 2010. Available at: http://www.tbcta.org// Uploaded_files/Zelf/TBHIVchildfinaldoc1203101268950707.pdf (accessed on 26 October 2010). 54. WHO. Priority research questions for TB/HIV in HIV-prevalent and resource-limited settings. Geneva, Switzerland, World Health Organization, 2010.
19
6. Selected GRADE profiles

GRADE profile table 1: TB screening for adults and adolescents
What is the best combination of symptoms with or without radiology that can be used as a screening tool to identify people living with HIV who are eligible for treatment of LTBI and for diagnostic work-up for active TB? Bibliography: Ayles et al. 2009; Corbett et al. 2010; Cain et al. 2010; Corbett et al. 2007; Lewis et al. 2009; Shah et al. 2009; Kimerling et al. 2002; Lawn et al. 2009; Chheng et al. 2008; Getahun et al. (in press)
Any one of current cough, fever, night sweats, weight loss as the best combination of symptoms for screening Values and uncertainty around these Negative predictive value 0.97 (95% CI: 0.97, 0.98) Sensitivity 0.79 (95% CI: 0.75, 0.82) Specificity 0.49 (95% CI: 0.29, 0.70) 8148 (9 studies) Moderate Important Any one of current cough, fever, night sweats, weight loss or abnormal chest X-ray findings as the best combination of symptoms for screening Negative predictive value 0.98 (95% CI: 0.97, 0.99) Sensitivity 0.90 (95% CI: 0.66, 0.97) Specificity 0.38 (95% CI: 0.12, 0.73) 2805 (4 studies) Moderate Important 2805 (4 studies) Moderate Critical 2805 (4 studies) Moderate Critical 8148 (9 studies) Moderate Critical 8148 (9 studies) Moderate Critical Number of participants (studies) Quality of evidence Importance
20
Selected GRADE profiles
GRADE profile table 2: TB screening for children

What is the best combination of symptoms and diagnostic tools that can be used as a screening tool to identify HIVinfected children eligible for treatment of LTBI? Bibliography: Song et al. 2009
Quality assessment
Quality assessment
(number of participants)
Design
Limitations
Inconsistency
Indirectness
Imprecision
Other considerations
Quality of evidence
Any one of cough 2 weeks, fever or failure to thrive Negative predictive value 0.99 1 Sensitivity 0.90 1 Specificity 0.65 1 Observational study (303) Observational study (303) Serious limitation* Serious limitation* No serious inconsistency No serious inconsistency No serious indirectness No serious indirectness No serious imprecision# No serious imprecision# Low Observational study (303) Serious limitation* No serious inconsistency No serious indirectness No serious imprecision# Low Observational study (303) Serious limitation* No serious inconsistency No serious indirectness No serious imprecision# Low
Positive predictive value 0.15 1 Low
A combination of culture and radiological appearance was used as a gold standard, which is not a perfect gold standard. The study did not qualify for the highest quality of evidence since it was an observational study and did not have a well-defined gold standard. * The reference standard used is unlikely to correctly classify all the children with disease as having the disease. Moreover, sputum was collected only from children having signs and symptoms suggestive of TB or abnormal chest X-ray findings. # Confidence intervals for sensitivity and specificity were not reported.
21
GRADE profile table 3: Efficacy of INH vs placebo in persons with any TST status
Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Rivero et al. 2003; Whalen et al.1997 anergy
Quality assessment
No. of studies 8 Design Limitations Inconsistency Indirectness Imprecision Other considerations None
Active TB incidence (probable, possible, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) Active TB incidence (probable, possible, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) Randomized trials No serious limitations No serious inconsistency1 No serious indirectness No serious imprecision
Confirmed TB (follow up 13 years; culture-proven) 5 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision None
Mortality (any cause) (follow up 13 years; review of hospital records) 7 Randomized trials No serious limitations Serious2 No serious indirectness No serious imprecision None
HIV disease progression (follow up 1-3 years; clinical and immunological criteria) 2 Randomized trials No serious limitations Serious2 No serious indirectness No serious imprecision None
Adverse drug reaction leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) 7 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision None
22
Three out of 8 studies showed an opposite direction of the effect
Different direction of the effect across the studies
INH prophylaxis
No. of patients
Control
Summary of findings Effect

Relative risk (95% CI)
Importance
Absolute
Quality
85/2152 (3.9%)
123/1984 (6.2%)
RR 0.67 (0.510.87)
20 fewer per 1000 (from 8 fewer to 30 fewer) 7 fewer per 1000 (from 3 fewer to 10 fewer) 165 fewer per 1000 (from 65 fewer to 245 fewer)
HIGH
CRITICAL
2% 50%
34/1037 (3.3%)
123/1984 (6.2%) 2% 50%
RR 0.72 (0.471.11)
13 fewer per 1000 (from 24 fewer to 5 more) 6 fewer per 1000 (from 11 fewer to 2 more) 140 fewer per 1000 (from 265 fewer to 55 more)
HIGH
CRITICAL
427/2152 (19.8%)
419/1984 (21.1%) 5% 50%
RR 0.95 (0.851.06)
11 fewer per 1000 (from 32 fewer to 13 more) 3 fewer per 1000 (from 7 fewer to 3 more)
O MODERATE
CRITICAL
25 fewer per 1000 (from 75 fewer to 30 more) RR 0.88 (0.61.28) 30 fewer per 1000 (from 101 fewer to 70 more) 12 fewer per 1000 (from 40 fewer to 28 more) 60 fewer per 1000 (from 200 fewer to 140 more) O MODERATE CRITICAL
41/184 (22.3%)
43/171 (25.1%) 10% 50%
56/2026 (2.8%)
33/1873 (1.8%) 0% 20%
RR 1.66 (1.092.51)
12 more per 1000 (from 2 more to 27 more) 0 more per 1000 (from 0 more to 0 more) 132 more per 1000 (from 18 more to 302 more)
HIGH
CRITICAL
23
GRADE profile table 4: Efficacy of INH vs placebo in persons who are TST positive
Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998
Quality assessment
Active TB incidence (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) Active TB incidence (probable, possible, confirmed) (follow up 1-3 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision
Confirmed TB (follow up 13 years; culture-proven) 1 Randomized trials No serious limitations Serious1 No serious indirectness Serious2 None
Mortality (any cause) (follow up 13 years; review of hospital records) 3 Randomized trials No serious limitations No serious inconsistency3 No serious indirectness No serious imprecision None
HIV disease progression (follow up 13 years; clinical and immunological criteria) 1 Randomized trials No serious limitations Serious1 No serious indirectness No serious imprecision None
Only one study available to address this outcome Small sample size and wide CI 3 Mwinga et al. report an opposite direction of the effect
1 2
24
INH prophylaxis
No. of patients
Control

Importance
Absolute
Quality
CRITICAL
18/693 (2.6%)
46/618 (7.4%) 2% 50%
RR 0.36 (0.220.61)
48 fewer per 1000 (from 29 fewer to 58 fewer) 13 fewer per 1000 (from 8 fewer to 16 fewer) 320 fewer per 1000 (from 195 fewer to 390 fewer) 58 fewer per 1000 (from 66 fewer to 88 more) 17 fewer per 1000 (from 20 fewer to 26 more) 435 fewer per 1000 (from 495 fewer to 660 more)
HIGH
0/52 (0%)
4/60 (6.7%)
RR 0.13 (0.012.32)
OO LOW
CRITICAL
2%
50%
71/693 (10.2%)
84/618 (13.6%) 2% 50%
RR 0.74 (0.551)
HIGH
CRITICAL
6/11 (54.5%)
38/25 (152%) 10% 50%
RR 0.36 (0.150.85) 973 fewer per 1000 (from 228 fewer to 1292 fewer)
O MODERATE
CRITICAL
64 fewer per 1000 (from 15 fewer to 85 fewer) 320 fewer per 1000 (from 75 fewer to 425 fewer)
25
GRADE profile table 5: Efficacy of INH vs placebo in persons who are TST negative
Bibliography: Fitzgerald et al. 2001; Gordin et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Pape et al. 1993; Rivero et al. 2003; Whalen et al.1997 anergy
Quality assessment
Active TB incidence (probable, possible, (follow up (follow up clinical examination, chest X-ray, sputum for AFB) Active TB (possible, probable, confirmed) confirmed) 13 years; 1-3 years; clinical examination, chest X-ray, sputum for AFB) Randomized trials No serious limitations No serious inconsistency1 No serious indirectness No serious imprecision
Confirmed TB (follow up 13 years; culture-proven) 3 Randomized trials No serious limitations Serious2 No serious indirectness No serious imprecision None
Mortality (any cause) (follow up 13 years; review of hospital records) 7 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision None
HIV disease progression (follow up 13 years; clinical and immunological criteria) 2 Randomized trials No serious limitations Serious3 No serious indirectness No serious imprecision None
Fitzgerald et al. and Hawken et al. showed an opposite direction of the effect Different direction of the effect across studies 3 Opposite direction of the effect
1 2
26
INH prophylaxis
No. of patients
Control

Importance
Absolute
Quality
CRITICAL
49/1297 (3.8%)
54/1193 (4.5%) 2% 50%
RR 0.86 (0.591.26)
6 fewer per 1000 (from 19 fewer to 12 more) 3 fewer per 1000 (from 8 fewer to 5 more) 70 fewer per 1000 (from 205 fewer to 130 more) 7 fewer per 1000 (from 19 fewer to 18 more) 5 fewer per 1000 (from 13 fewer to 12 more) 120 fewer per 1000 (from 320 fewer to 305 more)
HIGH
12/521 (2.3%)
15/500 (3%)
RR 0.76 (0.361.61)
O MODERATE
CRITICAL
2% 50%
328/1297 (25.3%)
298/1193 (25%) 2% 50%
RR 1.02 (0.91.16)
5 more per 1000 (from 25 fewer to 40 more) 0 more per 1000 (from 2 fewer to 3 more)
HIGH
CRITICAL
10 more per 1000 (from 50 fewer to 80 more) 22 more per 1000 (from 61 fewer to 151 more) O MODERATE CRITICAL
35/146 (24%)
32/146 (21.9%) 10%
RR 1.10 (0.721.69)
50%
27
GRADE profile table 6: Efficacy of INH vs placebo in persons with unknown TST status
Bibliography: Mwinga et al. 1998; Hawken et al. 1997
Quality assessment
Active TB incidence (probable, possible, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) Randomized trials No serious limitations Serious1 No serious indirectness No serious imprecision
Confirmed TB (follow up 13 years; culture-proven) 2 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision None
Mortality (any cause) (follow up 13 years; review of hospital records) 2 Randomized trials No serious limitations No serious inconsistency No serious indirectness No serious imprecision None
HIV disease progression (follow up 13 years; clinical and immunological criteria) 0 No evidence available None
Opposite direction of the effect
28
INH prophylaxis
No. of patients
Control

Importance
Absolute
Quality
O MODERATE CRITICAL
18/162 (11.1%)
23/173 (13.3%) 2%
RR 0.86 (0.481.52)
19 fewer per 1000 (from 69 fewer to 69 more)
50%
22/464 (4.7%)
28/466 (6%) 2% 50%
RR 0.79 (0.461.36)
HIGH
CRITICAL
28/162 (17.3%)
37/173 (21.4%) 2% 50%
RR 0.81 (0.521.27)
HIGH
CRITICAL
0/0 (0%)
0/0 (0%)
RR 0 (00)
CRITICAL
0%
29
GRADE profile table 7: Duration of IPT in adults INH 6 months vs 36 months

Bibliography: Martinson et al.2009; Samandari et al. 2009
Quality assessment
Active TB incidence (probable, possible, confirmed) (follow up mean 36 months; clinical assessment, chest X-ray, sputum for AFB) Randomized trials No serious limitations1 No serious inconsistency Serious2, 3, 4 No serious imprecision
Confirmed TB (follow up mean 36 months; culture-proven) 1 Randomized trials No serious limitations No serious inconsistency Serious No serious imprecision None
Mortality (any cause) (follow up 36 months; review of hospital records and patients files) 2 Randomized trials No serious limitations1 No serious inconsistency Serious2, 3, 4 No serious imprecision None
HIV disease progression 05 No evidence available None
Adverse drug reactions leading to treatment interruption (follow up 36 months; laboratory monitoring and clinical assessment) 2 Randomized trials No serious limitations Serious6 No serious indirectness No serious imprecision None
The Soweto trial was not a head-to-head comparison but a four-arm study designed to compare the efficacy of different regimens as well. The Soweto trial considered TST-positive patients while the BOTUSA trial enrolled TST-positive and -negative patients. 3 Mean CD4 count at baseline was >500 cells/mm3 for the Soweto trial and around 200 cells/mm3 for the BOTUSA trial. 4 The Soweto trial enrolled patients not eligible for ART, while in the Botusa trial about 40% of the patients had started ART. 5 Subanaylsis on this outcome is expected to be performed soon. 6 The number of the affected is quite different between studies.
1 2
30
Continuous INH prophylaxis 20/997 (2%)
No. of patients
6 months INH prophylaxis 46/1150 (4%)

Importance
Absolute
Quality
O MODERATE CRITICAL
RR 0.50 (0.290.84)
20 fewer per 1000 (from 6 fewer to 28 fewer) 10 fewer per 1000 (from 3 fewer to 14 fewer) 250 fewer per 1000 (from 80 fewer to 355 fewer)
2% 50%
14/997 (1.4%)
33/1150 (2.9%) 2% 50%
RR 0.48 (0.260.9)
10 fewer per 1000 (from 2 fewer to 15 fewer) 260 fewer per 1000 (from 50 fewer to 370 fewer) RR 0.43 (0.240.78) 20 fewer per 1000 (from 8 fewer to 26 fewer) O MODERATE CRITICAL
15 fewer per 1000 (from 3 fewer to 21 fewer)
O MODERATE
CRITICAL
15/997 (1.5%)
40/1150 (3.5%) 10% 50%
57 fewer per 1000 (from 22 fewer to 76 fewer) 285 fewer per 1000 (from 110 fewer to 380 fewer) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 5.02 (2.749.198) 57 more per 1000 (from 25 more to 116 more) 40 more per 1000 (from 17 more to 82 more) O MODERATE CRITICAL CRITICAL
0/0 (0%)
0/0 (0%)
0%
70/983 (7.1%)
12/846 (1.4%) 1%
31
GRADE profile table 8: Duration of IPT in adults INH 6 months vs 12 months in those with any TST status
Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Rivero et al. 2003; Whalen et al.1997 anergy
Quality assessment
Active TB--possible, probable, confirmed (follow up 13 years; clinical assessment, chest X-ray, sputum for AFB) Randomized trials Very serious No serious inconsistency No serious indirectness No serious imprecision
Confirmed TB (follow up 13 years; culture-proven) 0 No evidence available None
Mortality (any cause) (follow up 13 years; review of hospital records and patients files) 2 Randomized trials Very serious No serious inconsistency No serious indirectness No serious imprecision None
HIV disease progression 0 No evidence available None
Adverse drug reactions leading to treatment interruption (follow up 13 years; laboratory monitoring and clinical assessment) 12 Randomized trials Very serious No serious inconsistency No serious indirectness No serious imprecision None
Not estimable due to the lack of events in the 12 months group.
32
6 months of INH prophylaxis 57/1806 (3.2%)
No. of patients
12 months of INH prophylaxis 10/184 (5.4%) 2% 50%

Importance
Absolute
Quality
RR 0.58 (0.31.12)
OO LOW
CRITICAL
0/0 (0%)
0/0 (0%) 0% 0%
RR 0 (00)
CRITICAL
375/1806 (20.8%)
24/184 (13%) 10% 50%
RR 1.59 (1.0852.34) 77 more per 1000 (from 11 more to 175 more) 59 more per 1000 (from 9 more to 134 more) 295 more per 1000 (from 43 more to 670 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0 (00)1 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer)
OO LOW
CRITICAL
0/0 (0%)
0/0 (0%) 0%
CRITICAL
56/1968 (2.8%)
0/58 (0%) 0%
OO LOW
CRITICAL
33
GRADE profile table 9: Efficacy in children INH 6 months vs placebo

Bibliography: Zar et al. 2007; Madhi et al. 2008
Quality assessment
Active TB (follow up 5.79 months; clinical algorithm criteria, chest X-ray, bacteriological isolates from any site) Randomized trials No serious limitations Serious1 No serious indirectness2 No serious imprecision
Confirmed TB (follow up 5.79 months; culture-proven) 1 Randomized trials No serious limitations Serious3 No serious indirectness Serious4 None
Mortality (all causes) (follow up 5.79 months; review of hospital records and patients files) 2 Randomized trials No serious limitations Serious1 No serious indirectness2 No serious imprecision None
Adverse reaction (grade 3 or 4 toxicity) (follow up 5.79 months; clinical and laboratory monitoring) 2 Randomized trials No serious limitations No serious inconsistency No serious indirectness5 No serious imprecision None
HIV disease progression 0 None
Opposite direction of the effect. P1041 represents an optimal HIV care setting, with good facilities to rule out active TB: children were younger, healthier and presented in a less advanced stage of disease; the study by Zar et al. represents the most common condition of rural areas with later diagnosis of TB, fewer resources, and children presenting with more advanced disease and challenging TB diagnosis. 3 One trial available 4 Wide confidence intervals 5 Raw data are missing for P1041 (but no significant difference was reported between the two groups).
1 2
34
INH prophylaxis (6 months) 44/358 (12.3%)
No. of patients
Placebo

Importance
Absolute
Quality
O MODERATE CRITICAL
45/357 (12.6%) 5%
RR 0.97 (0.6609 1.4384)
3/226 (1.3%)
3/226 (1.3%) 0.9%
RR 1.5 (0.258.89)
OO LOW
CRITICAL
26/358 (7.3%)
31/357 (8.7%)
RR 0.84 (0.511.37)
O MODERATE
CRITICAL
10%
5/132 (3.8%)
8/131 (6.1%)
RR 0.62 (0.211.85)
23 fewer per 1000 (from 48 fewer to 52 more) 0 fewer per 1000 (from 0 fewer to 0 fewer)
HIGH
CRITICAL
0%
0/0 (0%)
0/0 (0%)
RR 0 (00)
CRITICAL
0%
35
GRADE profile table 10: Drug resistance and use of preventive therapy
Bibliography: Hawken 1997; Johnson et al. 2001; Pape et al. 1993; Rivero et al. 2003; Saenghirunvatta 1996; Zar et al. 2007; le Roux et al. 2009; Mwinga et al. 1998; Halsey et al. 1998; Gordin et al. 2000
Quality assessment
Mono-resistance to INH vs placebo (IPT intervention vs placebo) Randomized trials Serious1 No serious inconsistency No serious indirectness Serious2
Mono-resistance to INH vs rifampicin (IPT intervention vs rifampicin as control) 3 Randomized trials Serious1 No serious inconsistency No serious indirectness Very serious3 None
Incomplete accounting of patients and outcomes Small number of cases and patients 3 Small number of patients
1 2
GRADE profile table 11: Secondary prophylaxis

Bibliography: Perriens et al. 1995; Haller et al. 1999; Fitzgerald et al. 2000; Churchyard et al. 2003
Quality assessment
No. of studies 1 Design Limitations Inconsistency Indirectness Imprecision Other considerations Strong association
TB recurrence (observational) (follow up 0.91 vs 0.41 patient-years; isoniazid vs co-trimoxazole) Observational study Serious No serious inconsistency Serious No serious imprecision
TB recurrence (randomized) 3 Randomized trials No serious limitations No serious inconsistency No serious indirectness1 Serious2 None
1 2
The study by Perriens et al. provided INH and rifampicin for six months instead of INH alone. Small numbers in study
36
Anti-TB medications
No. of patients
No medications

Relative risk (95% CI) RR 1.87 (0.655.38)
Importance
Absolute
Quality
O MODERATE CRITICAL
11/1255(0.9%)
5/1069 (0.5%)
4 more per 1000 (from 2 fewer to 20 more)
3/1469 (0.2%)
1/1469 (0.1%)
RR 2 (0.1822.03)
1 more per 1000 (from 1 fewer to 14 fewer)
OOO VERY LOW
LESS CRITICAL
Secondary treatment of LTBI 28/338 (8.3%)
No. of patients

Importance
Absolute
Control
Quality
23/221 (10.4%)
RR 0.45 (0.260.78)
OOO VERY LOW
CRITICAL
7/275 (2.5%)
31/286 (10.8%)
RR 0.23 (0.110.52)
O MODERATE
LESS CRITICAL
37
Bibliography
Ayles H et al. Prevalence of tuberculosis, HIV and respiratory symptoms in two Zambian communities: implications for tuberculosis control in the era of HIV. PLOS One, 2009, 4:e5602. Cain KP et al. An algorithm for tuberculosis screening and diagnosis in people with HIV. New England Journal of Medicine, 2010, 362:707716. Chheng P et al. Pulmonary tuberculosis among patients visiting a voluntary confidential counseling and testing center, Cambodia. International Journal of Tuberculosis and Lung Disease, 2008, 12:5462. Churchyard et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17: 20632070. Corbett EL et al. Provider-initiated symptom screening for tuberculosis in Zimbabwe: diagnostic value and the effect of HIV status. Bulletin of the World Health Organization, 2010, 88:1321. Corbett EL et al. Epidemiology of tuberculosis in a high HIV prevalence population provided with enhanced diagnosis of symptomatic disease. PLOS Medicine, 2007, 4:e22. Fitzgerald DW et al. No effect of isoniazid prophylaxis for purified protein derivative-negative HIV infected adults living in a country with endemic tuberculosis: results of a randomized trial. Journal of Acquired Immune Deficiency Syndromes, 2001, 28:305307. Fitzgerald DW et al. Effect of post-treatment isoniazid on prevention of recurrent tuberculosis in HIV-1-infected individuals: a randomised trial. Lancet, 2000, 356:14701474. Getahun H et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource constrained settings: individual participant data meta-analysis of observational studies. PLoS Medicine, 2011, 8(1): e1000391. doi:10.1371/journal.pmed.1000391. Gordin FM et al. A controlled trial of isoniazid in persons with anergy and human immunodeficiency virus infection who are at high risk for tuberculosis. New England Journal of Medicine, 1997, 337:315320. Haller L et al. Isoniazid plus sulphadoxinepyrimethamine can reduce morbidity of HIV-positive patients treated for tuberculosis in Africa: a controlled clinical trial. Chemotherapy, 1999, 45:452465. Halsey NA et al. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet, 1998, 351:786792. Hawken MP et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS, 1997, 11:875882. Johnson JL et al.; for the UgandaCase Western Reserve University Research Collaboration. Duration of efficacy of treatment of latent tuberculosis infection in HIV-infected adults. AIDS, 2001, 15:21372147. Kimerling ME et al. Prevalence of pulmonary tuberculosis among HIV-infected persons in a home care program in Phnom Penh, Cambodia. International Journal of Tuberculosis and Lung Disease, 2002, 6:988994. Lawn S et al. Urine lipoarabinomannan assay for tuberculosis screening prior to antiretroviral therapy: diagnostic yield and association with immune reconstitution disease. AIDS, 2009, 23:18751880. le Roux SM et al. Adherence to isoniazid prophylaxis among HIV-infected children: a randomized controlled trial comparing two dosing schedules. BMC Medicine, 2009, 7:67. Lewis JJ et al. HIV infection does not affect active case finding of tuberculosis in South African gold miners. American Journal of Respiratory and Critical Care Medicine, 2009, 180:12711278.
38
Bibliography
Madhi SA et al.; and the P1041 Team. Lack of efficacy of primary isoniazid (INH) prophylaxis in increasing tuberculosis (TB) free survival in HIV-infected (HIV+) South African children. 48th ICAAC/IDSA 46th annual meeting, 2008 [G21346a]. Martinson NB et al. Novel regimens for treating latent TB in HIV-infected adults in South Africa: a randomized clinical trial. 16th Conference on Retroviruses and Opportunistic Infections, Montreal, 811 February 2009 [Paper 36bLB]. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:24472457. Pape JW et al. Effect ofisoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection. The Lancet, 1993, 342:26872. Perriens JH et al. Pulmonary tuberculosis in HIV-infected patients in Zaire. A controlled trial of treatment for either 6 or 12 months. New England Journal of Medicine, 1995, 332:779784. Rivero A et al. [A randomized trial of three regimens to prevent tuberculosis in HIV-infected patients with anergy]. Enfermedades infecciosas y microbiologia clinica, 2003, 21:287292. Saenghirunvattana S. [Effect of isoniazid prophylaxis on incidence of active tuberculosis among Thai HIV-infected individuals]. Journal of the Medical Association of Thailand, 1996, 79:285287. Samandari T et al.; on behalf of BOTUSA IPT study. Preliminary results of the Botswana IPT Trial: 36 months vs. 6 months isoniazid for TB prevention in HIV-infected adults. 40th Union World Lung Conference, Cancun, 2009. Shah S et al. Intensified tuberculosis case finding among HIV-infected persons from a voluntary counseling and testing center in Addis Ababa, Ethiopia. Journal of Acquired Immune Deficiency Syndromes, 2009, 50:537545. Song R et al. Evaluation of TB screening approaches among HIV-infected children Rwanda, 2008. 5th IAS conference on HVI pathogenesis and treatment, 1922 July 2009 [Abstract no TUPEB132]. Whalen CC et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. UgandaCase Western Reserve University Research Collaboration. New England Journal of Medicine, 1997, 337:801808. Zar HJ et al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomised controlled trial. British Medical Journal, 2007, 334:136.
39
Tuberculosis (TB) is responsible for more than a quarter of deaths in people living with HIV. Isoniazid Preventive Therapy (IPT) and Intensified tuberculosis Case-Finding (ICF) are key public health interventions that significantly reduce the morbidity and mortality from TB in people living with HIV. IPT and ICF should be part of a TB prevention package along with infection control for TB and the provision of ART. The objective of these guidelines is to provide guidance to national AIDS and tuberculosis programmes and those providing HIV services to accelerate the nationwide implementation of IPT and ICF. They include evidence-based recommendations for adults, children and infants living with HIV, address implementation issues and identify key research gaps in order to scale up TB prevention, diagnosis and treatment as a core component of HIV prevention, treatment and care. They are aimed at policy-makers and health programme managers, governments, nongovernmental organizations, donors, patient support groups working in the field of HIV/AIDS and TB, and health-care workers providing care for people living with HIV.
9789241500708
Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resourceconstrained settings
Annexes
(for web posting and CD-Rom distribution in tandem with the guidelines document)
Department of HIV/AIDS Stop TB Department World Health Organization, Geneva, Switzerland
Contents
Annex 1: WHO guidelines meeting on preventive therapy and case-finding for TB in people living with HIV: list of participants Annex 2: WHO guidelines group Annex 3: WHO meeting on preventive therapy and case-finding for TB among people living with HIV: meeting agenda Annex 4: Report on the consultation for the revision of the WHO/UNAIDS policy statement on preventive therapy against TB for people living with HIV Annex 5: Summary of findings and quality of evidence evaluation: intensified TB case-finding for adults and adolescents living with HIV Annex 6: Summary of findings and quality of evidence evaluation: preventive therapy for adults and adolescents living with HIV Annex 7: Summary of findings and quality of evidence evaluation: secondary prophylaxis Annex 8: Summary of findings and quality of evidence: tuberculin skin test (TST) Annex 9: Summary of findings and quality of evidence evaluation: interferon gamma release assays (IGRA) Annex 10: Summary of findings and quality of evidence: INH resistance Annex 11: Summary of finding and quality of evidence: adherence to preventive therapy Annex 12: Summary of findings and quality of evidence: cost effectiveness Annex 13: Summary of findings and quality of evidence evaluation: intensive TB case-finding for children living with HIV Annex 14: Summary of findings and quality of evidence: preventive therapy for children living with HIV Annex 15: Research gaps 1 5 6 9 24 32 80 86 92 99 107 113 120 128 135
ii
Annex 1: WHO guidelines meeting on preventive therapy

2527 January 2010
Helen Ayles Department of Medicine P.O. Box 50697 Ridgeway Campus, Nationalist Road Lusaka Zambia Draurio Barreira Programa Nacional de Controle da Tuberculose DEVEP/SVS/MS SCS Quadra 4 Bloco A Edif. Principal 3 andar 70.304-000 Braslia - DF Franois-Xavier Blanc Agence Nationale de Recherche sur le SIDA et les Hpatites virales 101 rue de Tolbiac, 75013 Paris France Charlene Brown Technical Leadership & Research Division Office of HIV/AIDS US Agency for International Development United States of America Kevin Cain International Research and Programs Branch Division of Tuberculosis Elimination Centers for Disease Control and Prevention 1600 Clifton Rd., MS-E-10 Atlanta, GA 30333 United States of America Rolando A. Cedillos Proyecto Regional VIH SIDA Centroamricamerica Av. El Espino 65 Urbanizacion Madre Selva Antiguo Cuscatlan, La Libertad El Salvador Richard Chaisson Consortium to Respond Effectively to the AIDS and TB Epidemic Johns Hopkins University Center for Tuberculosis Research 1503 E. Jefferson Street 21231-1003, Baltimore United States of America
and case-finding for TB in people living with HIV: list of participants
Mean Chhivun National Center for HIV/AIDS Dermatology and STD Ministry of Health No 266, St 1019, SK Phnom Penh Thmey Khan Roesseykeo Phnom Penh Cambodia Anupong Chitwarakorn HIV/STI/TB) Bureau of AIDS/TB/STI Department of Disease Control Ministry of Public Health 1100 Nonthaburi Thailand Gavin Churchyard Aurum Institute for Health Research P.O. Box 61587, 2107, Marshalltown Republic of South Africa Mark Cotton Stellenbosch University Department of Paediatrics and Child Health Faculty of Health Sciences P.O. Box 19063, 7505 Tygerberg Republic of South Africa Anand Date Global AIDS Program, TB/HIV Team, HIV Care and Treatment Branch Centers for Disease Control and Prevention 1600 Clifton Road, N.E. Mailstop E-04 Atlanta, GA 30333 United States of America Kevin De Cock (Co-chair) Centers for Disease Control and Prevention P.O. Box 54840-00200 Kenya Dmytro Donchuk State Medical University Ukraine
Annex 1
Wafaa El-Sadr International Center for AIDS Care and Treatment Programs Mailman School of Public Health Columbia University 722 West 168th Street, Room 715 New York, NY 10032 United States of America Peter Godfrey-Faussett Department of Infection & Tropical Diseases London School of Hygiene & Tropical Medicine Keppel Street, WC1E 7HT, London United Kingdom of Great Britain and Northern Ireland Olga Petrovna Frolova TB/HIV Health Care Centre of Ministry of Health and Social Development 107014, Moscow, 3, Barbolina Street Russian Federation Paula Fujiwara International Union Against Tuberculosis & Lung Disease 68 boulevard Saint-Michel, 75006 Paris France Alison Grant Clinical Research Unit London School of Hygiene & Tropical Medicine Keppel Street London WC1E 7HT United Kingdom of Great Britain and Northern Ireland Mark Harrington Treatment Action Group 611 Broadway, Suite 612 New York, NY 10012 United States of America Catherine Hewison Mdecins Sans Frontires 8 Rue Saint-Sabin 75011, Paris France Maureen Kamene Kimenye Ministry of Public Health P.O. Box 10016 Nairobi Kenya
Michael Kimerling Bill and Melinda Gates Foundation PO Box 23350 Seattle, WA 98102 USA Steve Lawn Institute of Infectious Diseases and Molecular Medicine, Faculty of Health Desmond Tutu HIV Centre Anzio Road Observatory 7925 Cape Town Republic of South Africa Gary Maartens University of Cape Town, Faculty of Health Sciences Division of Clinical Pharmacology Department of Medicine Observatory 7925 Republic of South Africa Barbara Jean Marston Global AIDS Program, Care and Treatment Branch Centers for Disease Control and Prevention 1600 Clifton Road, N.E. Mailstop E04 Atlanta, GA 30333 United States of America Thombile Mbengashe National AIDS Programme Cluster National Department of Health Private Bag X828 Pretoria 0001 Republic of South Africa Zenebe Melaku International Center for AIDS Care and Treatment Programs Addis Ababa Ethiopia Peter Mgosha Ministry of Health and Social Welfare National AIDS Control Programme P.O. Box 11578 Tanzania Muhamed Mulongo Tropical Medical and Maternity Center District P.O. Box 17, Sironko Uganda
Annex 1: WHO guidelines meeting on preventive therapy and case-finding for TB in people living with HIV: list of participants
Sharon Nachman Health Science Center State University New York Stony Brook NY 11794-8111 United States of America Alasdair Reid HIV/TB Adviser UNAIDS 20 Avenue Appia CH -1211 Geneva 27 Switzerland Stewart Reid Centre for Infectious Disease Research in Zambia P.O. Box 34681 Plot 2374 , Counting House Square Thabo Mbeki Road Lusaka Zambia Taraz Samandari Centers for Disease Control and Prevention 1600 Clifton Road MS-E-10 Atlanta, GA 30333 United States of America Paula Isabel Samo Gudo National Tuberculosis Program Ministry of Public Health Mozambique Mauro Schechter Hospital Universitario Clementino Fraga Filho Universidade Federal do Rio de Janeiro Rua Professor Rodolpho Paulo Rocco, n 255 Ilha do Fundo Rio de Janeiro RJ Brazil 21941.590
Holger Schnemann Department of Clinical Epidemiology & Biostatistics McMaster University of Health Sciences Centre Room 2C10B, 1200 Main Street West Hamilton, ON, L8N 3Z5 Canada Wim Vandevelde European Community Advisory Board European AIDS Treatment Group Place Raymond Blyckaerts 13 B-1050 Brussels Belgium Eric van Praag Family Health International Off Haile Selassie Road Plot No. 8/10, Oysterbay PO Box 78735 Dar es Salaam Tanzania Jay K. Varma U.S. CDC International Emerging Infections Program CDC-GAP China Office 23 Dongzhimenwai Dajie Beijing 100600 Peoples Republic of China Fujie Zhang Fujie Zhang National Center for AIDS/STD Control and Prevention Chinese CDC No. 27 Nanwei Road Beijing, P.R. China 100050
Annex 1
WHO REGIONAL OFFICES

AFRO Frank Lule AMRO Rafael Lopez Olarte WPRO Massimo Ghidinelli SEARO Puneet Dewan
WHO SECRETARIAT, HEADQuARTERS

Lopold Blanc STOP TB Department Siobhan Crowley HIV/AIDS Department Colleen Daniels STOP TB Department Andrew Doupe HIV/AIDS Department Haileyesus Getahun (Guidelines Coordinator) STOP TB Department Sandy Gove HIV/AIDS Department Reuben Granich (Guidelines Coordinator) HIV/AIDS Department Teguest Guerma HIV/AIDS Department Malgorzata Grzemska STOP TB Department Christian Gunneberg STOP TB Department Suzanne Hill (Co-chair) Policy, Access and Rational Use Ying-Ru Lo HIV/AIDS Department Lulu Muhe Child and Adolescent Health Eyerusalem Negussie HIV/AIDS Department Rose Pray STOP TB Department Mario Raviglione STOP TB Department Delphine Sculier STOP TB Department Yves Souteyrand HIV/AIDS Department Diana Weil STOP TB Department
WHO CONSuLTANTS
Christopher Akolo 9040 I Steinberg Way Laurel MD 29723 United States of America Martina Penazzato Via E. Forcellini 43 35128 Padova Italy Georgina Russell 16 Oliphant Street, London W10 4EG United Kingdom of Great Britain and Northern Ireland Caoimhe Smyth WHO HIV/AIDS Department
Annex 2: WHO guidelines group

The following group represents experts from the fields of HIV, TB, HIV/TB, sexually transmitted infections, child health, infectious and tropical diseases, clinical research, maternal health, infectious disease research, clinical epidemiology and biostatistics.
Helen Ayles (ZAMBART Project, Zambia), Draurio Barreira (National TB Program, Brazil), Franois-Xavier Blanc (Agence Nationale de Recherche sur le SIDA et les Hpatites virales, France), Charlene Brown (US Agency for International Development [USAID], United States of America [USA]), Kevin Cain (Centers for Disease Control and Prevention [CDC], USA), Rolando Cedillos (Proyecto Regional VIH SIDA para Centroamrica, El Salvador), Richard Chaisson (Johns Hopkins University, USA), Mean Chhivun (National AIDS Programme, Cambodia), Anupong Chitwarakorn (Ministry of Public Health, Thailand), Gavin Churchyard (Aurum Institute for Health Research, Republic of South Africa), Mark Cotton (Stellenbosch University, Republic of South Africa), Anand Date (CDC, USA), Dmytro Donchuk (State Medical University, Ukraine), Wafaa El-Sadr (International Center for AIDS Programs, Columbia University, USA), Peter Godfrey-Faussett (London School of Hygiene and Tropical Medicine, UK), Olga Petrovna Frolova (Ministry of Health and Social Development, Russian Federation), Paula Fujiwara (International Union Against Tuberculosis and Lung Disease [The Union], France), Alison Grant (London School of Hygiene and Tropical Medicine, UK), Mark Harrington (Treatment Action Group, USA), Catherine Hewison (Medecins sans Frontieres [MSF], France), Maureen Kamene Kimenye (Ministry of Public Health, Kenya), Michael Kimerling (Bill and Melinda Gates Foundation, USA), Stephen D. Lawn, (University of Cape Town, South Africa), Gary Maartens (University of Cape Town, South Africa), Barbara Jean Marston (CDC, USA), Thombile Mbengashe (National Department of Health, Republic of South Africa), Zenebe Melaku (International Center for AIDS Care and Treatment Programs [ICAP], Ethiopia), Peter Mgosha (Ministry of Health and Social Welfare, Tanzania), Muhamed Mulongo (Tropical Medical and Maternity Centre, Uganda), Sharon Nachman (Stony Brook University Medical Center, USA), Alasdair Reid (Joint United Nations Programme on HIV/AIDS [UNAIDS], Geneva), Stewart Reid (Centre for Infectious Disease Research in Zambia [CIDRZ], Zambia), Taraz Samandari (CDC, USA), Paula Isabel Samo Gudo (Ministry of Public Health, Mozambique), Mauro Schechter (AIDS Research Laboratory, Brazil), Wim Vandevelde (European Community Advisory Board, European AIDS Treatment Group, Belgium), Eric van Praag (Family Health International, Tanzania), Jay K. Varma (CDC, USA), Fujie Zhang (National Center for AIDS/STD Control and Prevention, Peoples Republic of China) Peer reviewers Jesus Maria Garcia Calleja (HIV/AIDS Department, WHO), Jacob Creswell (Stop TB Department, WHO), Irina Eramova (WHO EURO), Robert Gie (University of Stellenbosch, South Africa), Steve Graham (The Union, Australia), Prakash Kudur Hanumaiah (Karnataka Health Promotion Trust, India), Cathy Hewisson (MSF, France), Charles Mwansambo (Kamuzu Central Hospital, Malawi), Nguyen Viet Nhung (National TB programme, Viet Nam), Carla Obermeyer (HIV/AIDS Department, WHO), Ikushi Onozaki (Stop TB Department, WHO), Cyril Pervilhac (HIV/AIDS Department, WHO), Renee Ridzon (Bill and Melinda Gates Foundation, USA), Matji Rifiloe (Ministry of Health, South Africa), Quaid Saeed (WHO EMRO), Fabio Scano (WHO WPRO), Sahu Suvanand (Stop TB Partnership), Risard Zaleskis (WHO EURO) WHO Headquarters and Regional offices Lopold Blanc (Stop TB Department, WHO), Colleen Daniels (Stop TB Department, WHO), Puneet Dewan (WHO SEARO), Massimo Ghidinelli (WHO WPRO), Sandra Gove (HIV/AIDS Department, WHO), Malgorzata Grzemska (Stop TB Department, WHO), Teguest Guerma (HIV/AIDS Department, WHO), Christian Gunneberg (Stop TB Department, WHO), Rafael Lopez Olarte (WHO AMRO), Frank Lule (WHO AFRO), Eyerusalem Negussie (HIV/AIDS Department, WHO), Rose Pray (Stop TB Department, WHO), Mario Raviglione (Stop TB Department, WHO) Coordinated by Haileyesus Getahun and Reuben Granich
Annex 3: WHO guidelines meeting on preventive therapy

2527 January 2010, Geneva, Switzerland
08:3009:00 09:0009:15 09:1509:30 Registration
and case-finding for TB in people living with HIV: meeting agenda
Day 1: 25 January 2010

T. Guerma M. Raviglione
Welcome remarks Meeting overview and introductions
09:3009:50 09:5010:00 10:0010:30 10:3011:30
Rationale and review process of guidelines Overview of GRADE methodology and review of PICOT questions for systematic reviews of scientific evidence Coffee break
K. De Cock (Co-chair) H. Shnemann (Co-chair) R. Granich H. Getahun
M. Penazzato
GRADE profiles, systematic reviews and draft recommendations for M. Penazzato PICOT Question 1: Define the optimal duration and drug regimen (e.g. INH, rifampicin, etc.) for treatment of LTBI to reduce the risk of developing TB Community commentary: W. Vandevelde, Belgium Country/national programme commentary: A. Chitwarakorn, Thailand
11:3012:30
Discussion: Open to the floor
GRADE profiles, systematic reviews and draft recommendations for M. Penazzato PICOT Question 2: Define the optimal time to start considering IPT (i.e. should immune status be considered and should IPT be started with ART). Community commentary: W. Vandevelde, Belgium Country/national programme commentary: N. Muraguri, Kenya
12:3014:00 14:0015:00
Discussion: Open to the floor Lunch A. Sharma GRADE profiles, systematic reviews and draft recommendations for PICOT Question 3: Determine whether people living with HIV who had received TB treatment in the past should be provided secondary treatment for LTBI to prevent reinfection or recurrence of tuberculosis. Community commentary: M. Harrington, United States of America Country/national programme commentary: T. Mbengashe, Republic of South Africa Discussion: Open to the floor
Annex 3: WHO meeting on preventive therapy and case-finding for TB in people living with HIV: meeting agenda
15:0016:30
16:3017:30
Working group session Group A/PICOT 1 Optimal duration/drug regimen Group B/PICOT 2 Optimal time to start Group C/PICOT 3 Secondary treatment of LTBI
18:00
Report back from Groups A, B, C: Group A/PICOT 1 (15) Optimal duration/drug regimen Group B/PICOT 2 (15) Optimal time to start Group C/PICOT 3 (15) Secondary treatment of LTBI Reception: Cafeteria, WHO/UNAIDS D Building
K. De Cock (Co-chair) H. Shnemann (Co-chair)
09:0009:10
Overview of Day 2
Day 2: 26 January
09:1010:00
GRADE profiles, systematic reviews and draft recommendations for PICOT Question 8: Determine the best combination of signs, symptoms and diagnostic procedures (e.g. radiography, serum-based tests such as IGRA, etc.) that can be used as screening tools to determine eligibility for LTBI treatment and to diagnose TB among people living with HIV (feasibility considerations included). Community commentary: M. Harrington, United States of America Country/national programme commentary: Z. Melaku, Ethiopia
K. De Cock (Co-chair) H. Shnemann (Co-chair) A. Date
10:0010:30 10:3012:30
Discussion: Open to the floor Coffee break GRADE profiles, systematic reviews and draft recommendations A. Sharma for PICOT Question 4: Determine whether treatment for LTBI among people living with HIV leads to significant development of mono-resistance against the drugs used for LTBI treatment. Community commentary: W. Vandevelde, Belgium Country/national programme commentary: R. Cedillos, El Salvador
12:0012:30
Discussion: Open to the floor
GRADE profiles, systematic reviews and draft recommendations for PICOT Question 7: TST in resource-limited settings Community commentary: M. Harrington, United States of America Country/national programme commentary: D. Barreria, Brazil
C. Akolo
12:3013:30 15:3015:45
Discussion: Open to the floor Lunch Working coffee
Annex 3
15:4517:00
17:0018:30
Working group session (contd) Group A/PICOT 4+5 Group B/PICOT 6+7 Group C/PICOT 8
Report back from Groups A, B, C: Group A/PICOT 4 +5 (20) Resistance Group B/PICOT 6+7 (20) TST Group C/PICOT 8 (20) Signs, symptoms/diagnostics
K. De Cock (Co-chair) S. Hill (Co-chair)
09:0009:15
Overview of Day 3
Day 3: 27 January
09:1510:00 10:0010:15 10:1511:15 11:1512:00
Paediatrics and IPT Coffee break
S. Nachman M. Cotton K. De Cock (Co-chair)
Review of recommendations Research priorities: Report back from each group of research issues that emerged from working group discussions Group A (5) Group B (5) Group C (5)
12:0012:15 12:1512:30
Wrap-up of major conclusions, recommendations Closing and Next steps
R. Granich H. Getahun
T. Guerma M. Raviglione
Consultation for the Revision of the WHO/uNAIDS Policy Statement on Preventive Therapy against TB for People Living with HIV
Report on the consultation for the revision of the WHO/uNAIDS policy statement on preventive therapy against TB for people living with HIV
Annex 4:
16 November11 December 2009
Consultation for the Revision of the WHO/uNAIDS Policy Statement on Preventive Therapy against TB for People Living with HIV
Global Network of People Living with HIV December 2009
Contents
Contents
Acknowledgments Executive Summary Introduction Methodology Results and findings Survey Introduction Design Implementation Results and findings Conclusions and Recommendations Annexes 13 13 14 14 14 15 15 15 16 16 17 18
12
Annex 4. Report on the consultation for the revision of the WHO/UNAIDS policy statement on preventive therapy against TB for people living with HIV
Acknowledgements
he Global Network of People living with HIV (GNP+) would like to thank and acknowledge the following people for their assistance in and contributions to making this e-consultation a success: Rose Bradbury; Georgina Caswell; Chris Mallouris; Thomas Paterson; all of the people
living with HIV who donated their time and made valuable contributions to this consultation; and the World Health Organization (WHO) for its support. The e-consultation and report was coordinated and written by Jason Farrell, CEO, Harm Reduction Consulting Services, Inc.
Executive summary
his report represents the key findings and recommendations that emerged during an online e-consultation to revise the WHO/UNAIDS policy statement on preventive therapy against TB for people living with HIV. GNP+ organized an e-consultation (three weeks) as well as an online survey to solicit comments and recommendations from people living with HIV on the WHO/UNAIDS policy statement. The online e-consultation organized by GNP+ and hosted by NAM was held between 16 November and 6 December 2009. The GNP+ e-consultations were designed to gather the perspectives and recommendations from people living with HIV. Each week, the consultation focused on a different topic with several questions that covered various aspects of the guidelines: Week 1: (1622 November) Treatment initiation and adherence Week 2: (2329 November) Access to treatment for people living with HIV Week 3: (30 November6 December) Resolving barriers and building support for expanded treatment GNP+ invited 306 advocates among people living with HIV to participate in the consultation via e-mail. Approximately 53 registered, representing 52 countries. Of those who registered, four were male, 30 female and 19 others chose not to identify their gender. Based on the numerous responses we received, a few critical key points stood out. What was strongly agreed to by a significant number of PLHIV advocates was that WHOs overall goal should be to eliminate TB infection worldwide; deal more efficiently with poor outcomes of treatment; and make all efforts to provide TB diagnosis and treatment as early as possible. More importantly, we found that responses to our consultation reflected many of the prerequisites for
preventive therapy as indicated in the 1998 Policy statement on preventive therapy against TB for people living with HIV. As suggested in 1998, today we find that the provision of co-located and integrated TB/HIV care should be made available where and when possible, and there should be an appropriate number of skilled and efficient medical providers knowledgeable in current diagnostics, treatment and care. Many of the GNP+ contributors felt that the quality of care tends to vary between doctors, nurses and counsellors. It was suggested that all medical providers working with people living with HIV should have a standardized level of current knowledge on TB prevention and treatment. In addition to the e-consultation, GNP+ also developed a survey to capture information on how being diagnosed with TB, participating in medical treatment and maintaining TB treatment impacts the daily living situation of someone with TB. The survey was designed to provide us with a better understanding of the quality of care being provided and how the relationship between the medical provider and patient can impact treatment outcomes. The GNP+ survey was posted during the last week of the e-consultation (30 November7 December 2009) using the survey monkey program. Forty HIV-positive advocates filled out the survey. The findings showed that having a supportive relationship with the medical provider as well as with a family member, friend or spouse is critical to successful TB treatment including treatment adherence. Many of those who participated in the survey reported being HIV-positive for an average of 1015 years and were in a good medical condition, indicating that they had undetectable viral loads and did not have significant side-effects from medications for TB prevention and/or treatment.
13
Annex 4
Introduction
he online e-consultation organized by GNP+ and hosted by NAM was conducted from 16 November to 6 December 2009. GNP+ e-consultations were designed to gather perspectives and recommendations from people living with HIV. Each week, the e-consultation focused on a different topic, with several questions that covered various aspects of the guidelines: Week 1: (1622 November) Treatment initiation and adherence
Week 2: (2329 November) Access to treatment for people living with HIV Week 3: (30 November6 December) Resolving barriers and building support for expanded treatment
GNP+ invited 306 HIV-positive advocates to participate in the consultation via e-mail. Approximately 53 registered, representing 52 countries. Of those who registered, four were male, 30 female and 19 others chose not to identify their gender.
Methodology
uilding on the experience of undertaking e-consultations, GNP+ employed a userfriendly, accessible and interactive webbased tool to gain the views and experiences of people living with HIV globally on TB. The GNP+ e-consultation included a range of questions and issues requested by WHO. All questions were open-ended and aimed at gathering qualitative information on the experiences and perspectives of people living with HIV.
During a three-week period, a series of questions were made available to HIV-positive advocates through the use of a password and user name for an online access system. Each week, a different set of questions was posted to obtain responses from those who had registered. To ensure timely responses to questions and concerns regarding the consultation, GNP+ engaged a consultant to moderate the site and discussions, and respond to any problem or question pertaining to TB policy and the consultation process.
Results and findings

hile responding to each weeks questions, participants shared similar opinions and/or beliefs as to what constitutes acceptable and appropriate treatment. Keeping in mind that all those who responded came from different geographical locations/regions and varied types of settings and/or delivery systems for medical care, it is the quality and provision of care that is a cause for concern among many of the participants. In addition to the quality and provision of care, stigma due to their having TB and lack of acceptance by others remain barriers for many who are interested in seeking screening diagnosis, treatment and maintaining ongoing medical care. Not only were concerns regarding the quality of care and stigma consistently noted among our participants, but it was also strongly suggested that TB screening be co-located within HIV testing sites. Combination treatment should be client-centred, and tailored to the patients specific needs such as living conditions as well as the potential adverse effects of combining
TB medications with other prescribed medications and/or any illicit drugs taken. One of our participants reported that I would prefer to receive treatment at a hospital as this creates the way to incorporate other services while being treated and this would also prevent some amount of stigma and discrimination... in my country, this exists a lot so the hospital setting would be perfect for me. As stigma can deter many from accessing care even in facilities with integrated care systems, anonymity of service provision needs to be taken into consideration. Therefore, medical centres or facilities need to be inconspicuous and not identifiable as HIV or TB clinics. As stigma was mentioned throughout the consultation, we want to draw attention to the provision of comprehensive care. When we asked what are the most essential outcomes sought from TB preventive therapy, not only did we receive many responses
14
that TB should be eradicated worldwide, but we also received a number of responses identifying the need to assess and treat mental health conditions. I know many people living with HIV in my country and other countries in the xxx region who have suffered from TB, and have been treated very badly by the healthcare providers and some of them died after few days. I think, as people living with HIV, when they got TB, they faced a double stigma from their surroundings which forced them into a depressive period and really affected their status. Therefore, as we focus on the provision of comprehensive care and treatment adherence, there
is clearly an identified need for attention to be devoted to mental health support and establishing social support systems for patients. Without appropriate support systems, recently diagnosed individuals and patients currently engaged in preventive therapy are at high risk for developing depression and terminating treatment and/or making any contact with medical providers as well as family and friends. This is especially the case when treating adolescents. As reported in our consultation, adolescents are subjected to peer pressure and stigma. These as well as economic status can adversely affect treatment or outcomes of preventive therapy.
Survey
Introduction
n addition to the e-consultation, GNP+ developed a survey to capture information on how being diagnosed with TB participating in preventive therapy, continuing with ongoing medical treatment and treatment adherence impacts the daily living situation of people living with HIV. Additionally, the goal of our survey was to gain a better understanding of the provision of care, specifically the quality of care, and how the relationship between the patient and medical provider impacts treatment outcomes. Our survey was posted during the last week of the e-consultation (30 November7 December 2009)
using the survey monkey program. Forty HIVpositive advocates joined us in filling out the survey questionnaires. Based on the survey findings, having a supportive relationship with the medical provider as well as with a family member, friend or spouse is critical to successful TB treatment, including treatment adherence. Many of those who participated in the survey reported being HIV-positive for an average of 1015 years, and were in a good medical condition, indicating that they had undetectable viral loads and did not have significant side-effects from TB prevention and/or treatment medications.
Design
he concept of conducting a survey was based on the desire to gain a better understanding of how TB affects or impacts the lifestyle of people living with HIV. The e-consultation questions were primarily based on WHO interests, focused more on clinical issues such as preventive therapy and treatment adherence. Therefore, we felt that offering an alternative, less clinical approach would be successful in garnering a broad range of pertinent viewpoints and opinions from the community of people living with HIV.
Many of the questions included in the survey were designed to elicit information on how the doctor patient relationship can effect treatment; what support systems are needed or are beneficial for maintaining treatment adherence, and who becomes the primary source of support for and information on TB. The survey consisted of five sections: demographic information; risk information; medical provider information; TB treatment and care information; and social support information.
15
Annex 4
Implementation
he survey prepared by GNP+ was made available during the last week of our e-consultation (30 November7 December). HIV-positive advocates were invited to participate in the survey by using the same password and user
name provided for the e-consultation questions. The survey was conducted by using the survey monkey program to rank the critical outcomes that people living with HIV seek from TB preventive therapy.
Results and findings
ur survey provided interesting results. Basic demographic information of the survey respondents showed that: Those taking part in the survey were mostly from Africa, Asia, Europe and Eastern Europe. Of them, 54% were male and 46% female. Many reported being HIV-positive for 1015 years and were predominately between 40 and 50 years of age. Of those who completed the survey, 54% indentified as being men who have sex with men (MSM), 29% as drug users and 17% had a history of incarceration. Individuals who reported drug use did not mention if they had notified their medical provider about this due to fear of the treatment being terminated or a change in the relationship between them. Regarding the quality of medical care, 54% of survey participants reported receiving medical care regularly on a monthly basis and were happy with their medical provider. Surprisingly, 55% reported having a mutually respectful relationship with their medical provider and, of those individuals, eightyfive per cent took HIV medications and had undetectable viral loads. Fifty per cent of survey respondents said that the relationship with the medical provider was helpful for treatment adherence and, because of this relationship, they felt comfortable disclosing when medication doses were missed. Of those who reported taking TB medications, none reported any side-effects that required discontinuation or switching of medications.
obtained from preventive therapy. Additionally, all survey participants reported having a family member, spouse or friend to provide support, and all were stably housed. These four elements of having (1) a good relationship with the medical provider, (2) responding well to HIV treatment, (3) having support systems, and (4) stable housing are key factors for successful treatment outcomes. However, as substantiated by the findings of the survey, it is apparent that the individuals who participated were well connected to medical care and all of them indicated that they had support either from the family, spouse or friends. Because of these key factors, many responded well to HIV treatment and preventive therapy for TB. Further consultations and additional funding are required to obtain information from individuals who are not connected to care and support systems. The population of people living with HIV which does not have internet access or computers need to be reached. This target group consists of individuals who are at high risk for and vulnerable to resistant strains of TB, poor care, unstable housing and substance use. Therefore, we need to obtain information from the population that we could not connect with. The information obtained clearly shows that if there is a good relationship with the doctor, and a supportive family environment, people will respond well to any medical treatment. It would thus be worthwhile for WHO consider providing additional funding to conduct specific outreach efforts targeting the unreached community of people living with HIV at risk for TB, and to conduct focus groups designed to obtain the information required to develop appropriate interventions to successfully connect this group with comprehensive preventive therapy.
The survey findings showed that having a respectful and/or good relationship with the medical provider is critical for successful treatment outcomes. Individuals who disclosed taking TB medications also reported having undetectable viral loads and were clinically in a good medical condition. They self-reported that this increased the benefits they
16
GNP+ recommends that future consultations with people living with HIV be designed and implemented such that they draw on the experiences of HIVpositive people at the community level, people who may not have been reached through the GNP+ e-consultation referred to in this document.
Networks of people living with HIV and/or TB-related organizations at country and/or regional level may have to be supported to develop appropriate methodologies to collect, analyse and present the recommendations from people living with HIV at the community level.
Conclusions and Recommendations
The provision of preventive therapy for TB needs to be offered in a safe, stigma-free environment that is not identified as a TB/HIV clinic. The anonymity of the patient needs to be taken into consideration at all times. HIV testing locations must have on-site screening and preventive therapy for TB available. Opportunities for early TB screening must be sought whenever possible. The quantity of medication provided to each patient should be based solely upon the doctors prescription, and the distance from medical provider/clinic. Those who live near a clinic should receive a one-month supply and those who live at a greater distance should receive a three-month supply. Regional and/or local HIV treatment adherence rates should not be used to determine if TB preventive services should be made available. They should be used to examine why adherence rates are low and this information should be used to improve adherence to TB treatment. An individuals state of health should be taken into consideration to determine the best time and health condition to start treatment for better results and fewer problems from side-effects.
e provide WHO with the following recommendations and highlight key issues for further discussion.
Personal perceptions regarding treatment efficacy and its benefits can impact adherence and overall treatment outcomes. Therefore, mental health screening should be conducted to determine if depression and/or personal perception could affect treatment outcome and overall well-being. In addition to standard testing to determine if the prescribed TB prevention medication is working efficiently, additional testing such as polymerase chain reaction (PCR) HIV viral load tests should be provided to determine if TB has been eradicated entirely from the patient. Treatment outcomes for adolescents can be affected by peer support/pressure as well as stigma and economic status. When treating injecting drug users and those who use other drugs, drug interactions need to be discussed and monitored. Bioavailability and safety issues pertaining to interactions between TB medications and illicit drugs were noted as a concern. The number of pills required for TB prevention and/or treatment can present significant barriers to successful treatment. Substance use among individuals who have been prescribed TB preventive therapy has not been reported as a barrier to successful treatment. However, each persons substance use situation can vary and should be assessed based on the individuals stability.
17
Annexes
Annex 1: e-Consultation questions Week 1: Treatment initiation and adherence 1. What do you think are the most essential outcome or outcomes from TB preventive therapy? 2. In your opinion what do you think the best TB preventive treatments would be: isoniazid (INH), rifampicin (RIF), or pyrazinamide (PZA)? 3. How long do you think would be the best number of days/months needed to take it? 4. Are there benefits of shorter treatment time periods versus potential risks of developing drug-resistant TB? 5. Would providing a three-month supply of medication as opposed to a one-month supply each month improve treatment adherence? 6. What are the acceptable and unacceptable uncertainties relating to TB preventive therapy? (e.g. starting too early; not having long-term safety data on drug contraindications with antiretroviral drugs; hepatitis coinfection and liver toxicity) 7. What have been some of your fears and uncertainties about TB drug resistance, both of being (re)infected with resistant TB and developing drug resistance while taking TB preventive therapy? Week 2: Access to treatment for people living with HIV 1. What issues or problems, if any, do you think affect people living with HIV who may be diagnosed with TB or are taking TB preventive therapy? 2. Should people living with HIV who received TB treatment in the past be given secondary treatment for latent TB infection to prevent reinfection or recurrence of TB? 3. What would be the best time to start considering preventive therapy (isoniazid preventive treatment; IPT)? 4. Should the condition of a persons immune status be considered as a factor for when to start? (e.g. T4/T8 cell count, HIV viral load) 5. Should preventive therapy (IPT) be started at the same time as ART? What would you see as personal and medical treatment benefits and trade-offs between earlier and later start of IPT? 6. Do you think low adherence rates to TB preventive therapy should be a barrier to implementation of TB prevention programmes among people living with HIV? 7. What is the level and quality of TB-related health care you receive from your HIV-related healthcare provider? (e.g. doctors, nurses, counsellors, home-based care providers, community and peer outreach workers, etc.) 8. Where would you prefer to receive TB preventive treatment services? (e.g. home, TB clinic, HIV clinic, general services clinic) Week 3: Resolving barriers and building support for expanded treatment 1. Do you think peoples perceptions regarding screening for and diagnosis of TB will have any impact regarding interest of people living with HIV in receiving TB screening? (e.g. healthcare workers, family, work colleagues, friends, community perceptions, etc.) What are those perceptions? And what is their impact? 2. Do you think peoples perceptions regarding adherence to TB preventive therapy will have any impact on adherence rates among people living with HIV? What are those perceptions? And what impact? 3. What issues or problems, if any, do you think affect children who may be taking TB preventive therapy? 4. What issues or problems, if any, do you think affect adolescents in taking TB preventive therapy? 5. What issues or problems, if any, do you think affect injection drug users and other drug users including those who may use stimulants (yaba, crystal, etc.) in taking TB preventive therapy?
18
6. What can be or have been barriers to you or someone you know to successfully taking TB preventive therapy? (e.g. number of pills, pill burden, important drug interactions, side-effects, need for toxicity monitoring, etc.)
Annex 2: TB patient survey 2009 Recommendations for preventive therapy against tuberculosis for people living with HIV 1. Demographic information: Gender: Race: O Male O White O Female O Asian O Lesbian O African O Latin
Country of residence: O Africa O North America
O South America
In the past six months where did you live most of the time? O Your house or apartment O Parents house or apartment O Relatives house or apartment O Someone elses house or apartment O Hotel, rooming house or shelter O In a squat O On the streets O In jail How would you consider your housing situation? O Permanent O Temporary O Transitional O Homeless 2. Risk information: What TB risk group do you identify with? Check all that apply. O Drug user O MSM O Prison/jail history Past and current drug use: O Never O Less than six times O 3x or more daily O Monthly In past 30 days what drugs were used? In past six months what drugs were used? Do you live in a country that has a high rate of TB and HIV? Please indicate country.
O All O 23 times per week
O Weekly O 23 times per month
Do you know how you contracted TB?
19
Annexes
3. Medical provider information When was the last time you received any medical services? O Past 30 days O 6 months to a year ago O 12 years ago O Emergency room O 2+ years ago O Other O Other
Where did your medical visit take place? O Clinic/doctors office O Walk-in clinic How would you rate your last medical visit? O Excellent O Good O Fair O Poor
Do you see a health professional (doctor, nurse, etc.) for regular check-ups? O Yes O No O Dont know If no, why dont you see a health professional for regular check-ups? O They are disrespectful O They assume any/all health problems are due to drug use O They always tell me to get drug treatment O No health insurance O Dont know of a doctor O Other If you use illicit drugs does your doctor know? O Yes O No O Dont know If your health professional knows about your drug use has it affected you treatment? O Yes O No In what way has it affected your treatment? O Provider became hostile O Presumed all symptoms caused by drug use O No difference/no impact on treatment When did you find out you were HIV-positive? Where were you tested? O HIV testing site O Doctors office/clinic/jail O Other O Stopped treatment/referred to other provider O Always talked about stop using O Other
How many years have you been living with HIV infection? O Less than one year O 15 years O 510 years O 1015 years O 15+ years What is your current viral load? What is your current T-cell count? Are you taking any HIV medications? O Yes O No If taking HIV medication, what medications are you taking? Has your doctor informed you about new treatments? O Yes O No Would you recommend your doctor to a friend? O Yes O No
20
4. TB treatment and care information Upon noticing TB-related symptoms how long did you wait until you went to a doctor? O 1 week O 30 days O 2 months O 36 months O 69 months O 912 months O 1 year + During the doctors visit did he or she offer any information or explanation about TB diagnosis and treatments? O Yes O No What type of testing was provided to determine TB infection? O Blood tests O Chest X-ray O Both O Other How much information did the doctor provide? O As much as I asked for O Just enough O Very little O None O Internet
If the doctor did not offer information, where did you find TB treatment information? O Friends O Family O Partner/boyfriend/girlfriend O Husband/wife O Educational pamphlets O Outreach workers O Other How was the relationship with your medical provider? O Very good O Good O Acceptable Was your medical provider respectful? O Very respectful O Somewhat respectful O Tolerable O Disrespectful O Not acceptable
O Poor
O Reasonable
Did your relationship with medical provider affect treatment adherence? O Yes O No Were you comfortable telling the medical provider when you missed taking medication? O Yes O No Did the number of pills required to be taken make it difficult to manage treatment? O Yes O No What TB medications where you taking? For treatment or prevention? Please list.
Did you develop any side-effects from TB medications? Describe. Did the side-effects at any time cause you to discontinue or miss treatment? O Yes O No If side-effects caused missing medication how long was it before you started again? O 13 days O 37 days O 12 weeks O 1 month O 12 months Did you take any drugs aside from those prescribed to help with side-effects? O Yes O No
21
Annexes
If yes, what illicit drugs did you take for relief from side-effects? Did you ever switch TB medications to reduce side-effects? O Yes O No If yes, what medication did you stop taking and switch to with less side-effects? Stopped taking Switched to 5. Social support information Did you inform others about your TB diagnosis? O Yes O No How long after being diagnosed with TB did you feel comfortable telling others? O 13 days O 37 days O 12 weeks O 1 month O 12 months Who were the first people or first person you disclosed your TB infection? O Boyfriend O Girlfriend O Mother O Husband O Wife O All listed Why were these people the first you disclosed? O Trusted them O Would receive support How supportive was your family? O Very supportive O Supportive O Barely supportive O Not supportive O Offered information O All
O Somewhat supportive
Was your husband, wife, boyfriend, girlfriend, best friend or employer supportive? O Yes O No If not how did they react to your TB diagnosis and treatment period? O Avoided you due to fear of exposure O Stopped talking to you O Embarrassed/ashamed to know you O Disclosed diagnosis without consent O Used different plates and utensils O Cleaned/wiped everything you touched O Discontinued/ended relationship O Other Did friends, family, spouse, and/or employer support help with treatment adherence? O Yes O No How would or does support from friends family, spouse, and/or employer help with treatment adherence? Please describe. When you meet people, do you feel comfortable disclosing that you have/had TB? O Yes O No If not, why, did people discriminate or make you uncomfortable? O Yes O No Did the people who knew you had/have TB treat you differently? If yes, please describe. O Yes O No
22
Please describe. It is important for us to document all discrimination and mistreatment.
23
Annex 5: Summary of findings and quality of evidence evaluation:

intensified TB case-finding for adults and adolescents living with HIV
PICOT Question: What is the best combination of symptoms (with or without abnormal chest radiograph findings) that can be used as a screening tool to identify Population: Adults and adolescents living with HIV Intervention: Combination of signs and symptoms and diagnostic tools that can be used as a screening tool to identify adults and adolescents living with HIV who are eligible for treatment of LTBI and diagnosis of active TB adults living with HIV who are eligible for treatment of latent TB infection (LTBI) and for diagnostic work-up of active TB? Comparison: TB prevalence without intervention Outcomes: Negative predictive value, sensitivity, specificity, positive predictive value Timeline: 12 months during work-up for TB
1. Outcomes of interest
Outcomes
Negative predictive value (to identify those eligible for treatment of LTBI)
Relative importance (rank 1 9 most critical)

9
Comment
Critical Critical Important
Sensitivity (to identify patients for further 9 diagnostic work-up) Specificity Positive predictive value
Important
2. Literature search strategy and information retrieval

Search criteria PubMed Search (Tuberculosis[Mesh]) AND HIV[Mesh]) AND (Diagnosis[Mesh]) OR (TB screening, intensified case-finding, active case-finding, HIV, tuberculosis) Leading researchers in the area were also contacted to provide any unpublished data fulfilling the search criteria.
2119
articles
Limits Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial,
multicentre study, adolescent: 1318 years, adult: 1944 years, middle-aged: 4564 years, middleaged + aged: 45+ years, aged: 65+ years, 80 and over: 80+ years
24
Annex 5. Summary of findings and quality of evidence: Intensified TB case finding for adults and adolescents living with HIV
258
By title
53
By abstract
26
Of interest
21
primary data meta-analysis entitled Development of a standardized screening rule for tuberculosis in people living with HIV in resource constrained settings: individual participant data meta-analysis of observational studies was used to identify the best symptoms to screen a person living with HIV for treatment of LTBI. The meta-analysis was conducted in collaboration with WHO, CDC and principal investigators of 12 studies that met the inclusion criteria [112]. The inclusion criteria for the meta-analysis included: systematic collection of sputum specimens regardless of signs or symptoms, at least one mycobacterial culture, clinical symptoms, and HIV and TB disease status. Studies which looked at the performance of individual or a combination of symptoms with or without chest X-ray as a screening tool were also examined but they were not considered for evidence retrieval as their participants were limited to TB suspects (cough >2 weeks) or only smear-negative TB patients or they did not use culture as the gold standard [1320]. Twelve studies [112] were included in the primary data meta-analysis [20]. In addition to these studies, other studies were found, which examined the role of symptoms, individually or in combination with or
without chest radiology, as a screening tool for TB [14, 1519]. Of these, two studies [14,15] collected information on signs and symptoms only among TB suspects, defined as patients with cough >2/3 weeks duration. Many studies have demonstrated that cough alone is not a sensitive screening tool and would miss a large number of potential TB cases. Hence, these studies would have missed a large number of TB cases even before assessing the utility of other symptoms and would not have allowed proper assessment of the role of symptoms or their combination. Three studies [14, 16, 18] did not use culture as a gold standard for the entire or a part of their study population. One study [18] focused only on smear-negative TB, thus limiting their findings to patients with smear-negative disease only. Findings from one study [13] were presented at the 16th Conference on Retroviruses and Opportunistic Infections (CROI) as a poster and had not yet been published in a peer-reviewed journal at the time the guidelines were developed. The poster did not provide details of the symptoms used other than cough. The findings of the study for combination of symptoms with abnormal findings on chest X-ray were similar to the results of the primary patient meta-analysis.
25
Annex 5
3. Findings and GRADE profile

half of the patients would screen negative and could be considered for treatment of LTBI. Of these 4375 individuals, 107 (2.4%) TB patients might be wrongly put on treatment for LTBI as the negative predictive value is not 1. This screening rule had a sensitivity of 0.79 (95% CI: 0.58, 0.91) with a specificity of 0.49 (95% CI: 0.29, 0.70). Using the combination of symptoms proposed in a population of people living with HIV with a 5% TB prevalence requires the investigation of 11 extra cases for every TB case detected; a ratio of TB suspects to a TB case identified that is not much different from what national TB control programmes target in the general population. The meta-analysis also showed that the addition of chest radiography with abnormal findings to the combination increased the sensitivity substantially from 0.79 to 0.90; however, the specificity reduced by 11% (from 50% to 39%). There was no significant gain in the negative predictive value.
he primary patient meta-analysis and other studies were examined using GRADE criteria. Although all the studies included in the metaanalysis were observational; all the studies except two [6,9] met the criteria (population of interest, uncertain diagnosis, consecutive enrolment and comparison with a gold standard), and therefore qualified as being of the highest quality before other factors for determining the final strength of the evidence were considered [21]. The primary patient meta-analysis showed that at a TB prevalence of 5%, which is commonly found in many heavily impacted settings, absence of all of current cough, night sweats, fever and weight loss could identify a subset of people living with HIV who have a very low probability of having TB disease (negative predictive value 0.97 [95% CI: 0.97, 0.98]) and could be considered for treatment of LTBI based on their eligibility. Using this screening tool, more than
GRADE profile table 1: TB screening for adults and adolescents: What is the best combination of symptoms with or without radiology that can be used as a screening tool to identify people living with HIV who are eligible for treatment of LTBI and diagnostic work-up for active TB?
Any one of current cough, fever, night sweats, weight loss as the best combination of symptoms for screening
Values and uncertainty around these Negative predictive value 0.97 (95% CI: 0.97, 0.98) Sensitivity Number of participants (studies) 8148 (9 studies) 8148 (9 studies) 8148 (9 studies) Quality of evidence Importance Moderate Moderate Moderate Critical Critical Important
0.79 (95% CI: 0.75, 0.82) Specificity 0.49 (95% CI: 0.29, 0.70)
Any one of current cough, fever, night sweats, weight loss and abnormal chest X-ray findings as the best combination of symptoms for screening
Values and uncertainty around these Negative predictive value 0.98 (95% CI: 0.97, 0.99) Sensitivity Number of participants (studies) 2805 (4 studies) 2805 (4 studies) 2805 (4 studies) Quality of evidence Importance Moderate Moderate Moderate Critical Critical Important
0.90 (95% CI: 0.66, 0.97) Specificity 0.38 (95% CI: 0.12, 0.73)
26
4. Risk and benefit assessment

Adults and adolescents living with HIV who do not have current cough, fever, weight loss or night sweats are unlikely to have active TB and should be offered isoniazid preventive therapy (IPT)
Population: Adults and adolescents living with HIV Intervention: Symptom screening using a combination of symptoms comprising current cough, fever, night sweats or weight loss as a screening tool Factor Quality of evidence Decision Moderate Explanation The GRADE assessment of the evidence relies on a primary patient meta-analysis of 12 studies including over 8000 patients. The meta-analysis assessed a combination of different symptoms (with or without abnormal X-ray findings). The four-symptom rule has a negative predictive value of 98% and the probability of disease among those who screen negative is very low. Risks or undesired effects Strong (Benefits outweigh risks) Values and preferences Strong Identifies people living with HIV eligible for treatment of LTBI (and potential for reduced TB-related morbidity/ transmission). Identifies people living with HIV not in need of treatment for active TB. Identifies people living with HIV not in need of further evaluation with diagnostic tests for active TB disease. Increases health-care worker focus on screening. Some patients with active TB may not be given treatment for TB. Some patients with active TB may be given treatment for LTBI (see discussion on scientific evidence of low risk for developing drug resistance).
Benefits or desired effects
Costs
Feasibility
Moderate
Increased by: Additional staff time required for screening Cost of supplying isoniazid and pyridoxine Reduced by: Limited resources needed for symptom screening among persons who regularly attend clinical services for HIV Avoiding costs of additional diagnostic tests including chest X-ray Avoiding costs that would have been associated with treatment of TB Implementation of symptom screening would be feasible as Screening tool includes common TB-associated symptoms and health-care workers are familiar with these Avoids the need for repeat follow up and/or additional diagnostic tests But Would need improved coordination between TB and HIV programmes Would need operational aspects of implementing IPT to be worked out.
Improved quality of HIV care Access to treatment for LTBI Prevention of life-threatening disease and transmission to family, friends and the larger community
Overall ranking of recommendation
Strength of recommendation Strong
27
Annex 5
Adults and adolescents living with HIV who have any one symptom of current cough, fever, weight loss or night sweats may have active TB and should be evaluated for TB and other diseases.
Population: Adults and adolescents living with HIV Intervention: Symptom screening using a combination of symptoms comprising current cough, fever, night sweats or weight loss as a screening tool Factor Quality of evidence Decision Moderate Explanation The GRADE assessment of the evidence relies on a primary patient meta-analysis of 12 studies including over 8000 patients. The meta-analysis assessed a combination of different symptoms (with or without abnormal X-ray findings). The four-symptom rule has a negative predictive value of 98% and the probability of disease among those who screen negative is very low. Risks or undesired effects Conditional Values and preferences Strong Early identification of TB suspects followed by treatment of identified TB cases reduces TB-associated morbidity and mortality among people living with HIV. Identification and treatment of non-tuberculous pulmonary infections Increase in demand for availability of clinical and diagnostic services Increased volume and attention to importance of TB may result in improved quality of TB diagnostic services. Because about half of all HIV-infected patients in high HIV/ TB prevalence settings will have symptoms, there could be a substantial burden on clinical and diagnostic services. If total diagnostic capacity is compromised, then increased demand could result in the displacement of patients in greatest need of diagnostics. Volume may negatively influence the quality of laboratory diagnostic services, which might lead to missed diagnoses. Not all patients with a positive screen will have TB so, for some patients, there will be inconvenience/cost of tests without benefit. Patients generally desire an accurate diagnosis of disease and may be willing to undergo diagnostic evaluation or treatment in an effort to prevent morbidity/mortality and transmission.
Costs
Conditional
Feasibility
Increased by: Routine screening will require increased staff time. Additional diagnostic evaluations (staff, reagents, transport, laboratory capacity) Increased quality assurance of diagnostic services Reduced by: Using a relatively low-cost screen to identify persons in need for further diagnostic evaluation avoids the costs of using more expensive tests more broadly
An additional burden of diagnostic tests will be placed on already overburdened laboratories. It requires an increased need for referral and a better system for tracking of referrals, which is already compromised. Diagnostic services for those identified as TB suspects might not be available. Quality of diagnostic services needs to be improved. Although conducting a symptom screen may be more feasible than some other diagnostic interventions, it still requires time from already overburdened health-care workers.
28
In resource-limited settings where chest radiology is easily available and affordable, chest X-ray should be added to the screening tool to improve the sensitivity of screening for TB among adults and adolescents living with HIV.
Population: Adults and adolescents living with HIV Intervention: Screening using a combination of symptoms comprising current cough, fever, night sweats, weight loss or abnormal chest X-ray as a screening tool Factor Quality of evidence Decision Moderate Explanation The GRADE assessment of the evidence relies on a primary patient meta-analysis of 12 studies including over 8000 patients. The meta-analysis assessed a combination of different symptoms (with or without abnormal X-ray findings). The GRADE assessment of the evidence showed that the addition of abnormal chest radiograph findings to the four symptom-based rule increases the sensitivity from 79% to 91% with a drop in specificity from 50% to 39%. Increase in the number of patients identified and evaluated as being likely to have TB disease who might have been missed by symptom screening alone Early identification of TB suspects among HIV-infected people followed by appropriate treatment of identified TB cases which, in turn, can reduce TB-associated morbidity and mortality among people living with HIV. Identification and treatment of non-tuberculous pulmonary infections Substantial burden on the radiology technicians and radiologists reading the X-rays Issues with reading of X-rays and inter-reader variability In an effort to identify all patients with TB, screening could result in the displacement of patients in greatest need of diagnostics if the total diagnostic capacity is limited. Not all patients who screen positive will have TB so, for some patients, there will be inconvenience, unnecessary exposure to radiation, and cost of tests without benefit.
Risks or undesired effects Conditional Values and preferences
Costs
Weak (very costly)
Feasibility
Conditional
Increased by: Routine screening will require staff time. Chest X-ray for every patient will require additional resources. Identification of additional persons requiring diagnostic evaluation will require resources (staff, reagents, transport, laboratory capacity). Reduced by: Treatment benefits patients and the society. Feasible only if enough financial and human resources
Patients generally desire an accurate diagnosis of disease and may be willing to undergo diagnostic evaluation or treatment in an effort to prevent morbidity/mortality and TB transmission to family, friends and the community.
29
Annex 5
Figure 1. Algorithm for TB screening in adults and adolescents living with HIV in HIV-prevalent and resource-constrained settings
Adults and adolescents living with HIV* Positive screen for TB with any one of the following: Current cough Fever Weight loss Night sweats No Assess for contraindications to IPT No Give IPT Yes Defer IPT Yes Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Not TB Follow up and consider IPT TB Treat for TB
FOOTNOTES TO ADuLT ALGORITHM

* Every adult and adolescent needs to be evaluated for eligibility to receive antiretroviral therapy (ART) and infection control measures should be prioritized to reduce M. tuberculosis transmission in all settings that provide care. Chest radiography can be done if available, but it is not required to classify patients into TB and non-TB groups. In high HIVprevalence settings with a high TB prevalence among people living with HIV, strong consideration must be given to performing additional sensitive investigations. Contraindications include: active hepatitis (acute or chronic) or regular and heavy alcohol consumption or symptoms of peripheral neuropathy. Past history of TB and current pregnancy should not be contraindications for starting IPT. Although not a requirement for initiating IPT, tuberculin skin testing (TST) may be done as part of eligibility screening in some settings. Investigations for TB should be done in accordance with existing national guidelines.
30
References
1. Ayles H et al. Prevalence of tuberculosis, HIV and respiratory symptoms in two Zambian communities: implications for tuberculosis control in the era of HIV. PLoS One, 2009, 4(5):e5602. 2. Cain KP et al. An algorithm for tuberculosis screening and diagnosis in people with HIV. New England Journal of Medicine, 2010, 362(8):707716. 3. Chheng P et al. Pulmonary tuberculosis among patients visiting a voluntary confidential counseling and testing center, Cambodia. International Journal of Tuberculosis and Lung Disease, 2008, 12(3 Suppl 1):5462. 4. Corbett EL et al. Epidemiology of tuberculosis in a high HIV prevalence population provided with enhanced diagnosis of symptomatic disease. PLoS Medicine, 2007, 4(1):e22. 5. Corbett EL et al. Provider-initiated symptom screening for tuberculosis: diagnostic value, and the impact of HIV. Bulletin of the World Health Organization, 2010, 88:1321. 6. Day JH et al. Screening for tuberculosis prior to isoniazid preventive therapy among HIV-infected gold miners in South Africa. International Journal of Tuberculosis and Lung Disease, 2006, 10(5):523529. 7. Kimerling ME et al. Prevalence of pulmonary tuberculosis among HIV-infected persons in a home care program in Phnom Penh, Cambodia. International Journal of Tuberculosis and Lung Disease, 2002, 6(11):988994. 8. Lawn SD et al. Urine lipoarabinomannan assay for tuberculosis screening before antiretroviral therapy diagnostic yield and association with immune reconstitution disease. AIDS, 2009, 23(14):18751880. 9. Lewis JJ et al. HIV infection does not affect active case finding of tuberculosis in South African gold miners. American Journal of Respiratory and Critical Care Medicine, 2009, 180(12):12711278. 10. Mohammed A et al. Screening for tuberculosis in adults with advanced HIV infection prior to preventive therapy. International Journal of Tuberculosis and Lung Disease, 2004, 8(6):792795. 11. Shah S et al. Intensified tuberculosis case finding among HIV-Infected persons from a voluntary counseling and testing center in Addis Ababa, Ethiopia. Journal of Acquired Immune Deficiency Syndromes, 2009, 50(5):537545. 12. Wood R et al. Undiagnosed tuberculosis in a community with high HIV prevalence: implications for tuberculosis control. American Journal of Respiratory and Critical Care Medicine, 2007, 175(1):8793. 13. Bassett IV et al. Intensive tuberculosis screening for HIV-infected patients starting antiretroviral therapy in Durban, South Africa: limitations of WHO guidelines. In: Proceedings of the 16th Conference on Retroviruses and Opportunistic infections; Montreal, Canada, February 2009 [Abstract #779]. 14. Burgess AL et al. Integration of tuberculosis screening at an HIV voluntary counselling and testing centre in Haiti. AIDS, 2001, 15(14):18751879. 15. Espinal MA et al. Screening for active tuberculosis in HIV testing centre. Lancet, 1995, 345(8954):890893. 16. Samb B et al. Methods for diagnosing tuberculosis among in-patients in eastern Africa whose sputum smears are negative. International Journal of Tuberculosis and Lung Disease, 1997, 1(1):2530. 17. Tamhane A et al. Predictors of smear-negative pulmonary tuberculosis in HIV-infected patients, Battambang, Cambodia. International Journal of Tuberculosis and Lung Disease, 2009, 13(3):347354. 18. Were W et al. A simple screening tool for active tuberculosis in HIV-infected adults receiving antiretroviral treatment in Uganda. International Journal of Tuberculosis and Lung Disease, 2009, 13(1):4753. 19. Wilson D et al. Diagnosing smear-negative tuberculosis using case definitions and treatment response in HIVinfected adults. International Journal of Tuberculosis and Lung Disease, 2006, 10(1):3138. 20. Getahun H et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource constrained settings: individual participant data meta-analysis of observational studies. PLoS Medicine, 2010, 8(1): e1000391. doi:10.1371/journal.pmed.1000391. 21. Schunemann HJ et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. British Medical Journal, 2008, 336(7653):11061110.
31

preventive therapy for adults and adolescents living with HIV 6.1 Optimal duration of and drug regimen for the treatment of latent TB infection (LTBI)
PICOT Question: What is the optimal duration of and drug regimen (e.g. INH, rifampicin [RIF], pyrazinamide Efficacy Population: Adults and children living with HIV Intervention: IPT (any regimen) Comparison: No IPT Outcomes: Active TB incidence, mortality, progression of HIV disease, adverse effects (AEs), adherence, TB drug resistance, interval to active TB (Atb), interval to death Timeline: Lifetime Q6.1.1 Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: No IPT Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.1.2 Population: Adults living with HIV Intervention: IPT (INH+RIF) Comparison: No IPT Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q.6.1.3 Population: Adults living with HIV Intervention: IPT (RIF+pyrazinamide [PZA]) Comparison: No IPT Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime [PZA], etc.) for the treatment of LTBI to reduce the risk of developing TB among people living with HIV/AIDS? Q6.1.4 Population: Adults living with HIV Intervention: IPT (INH+RIF+PZA) Comparison: No IPT Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.2 Optimal duration (a) Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: IPT (12 months INH) Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.2 Optimal duration (b) Population: Adults living with HIV Intervention: IPT (continuous INH) Comparison: IPT (6 months INH) Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.3 Optimal regimen Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: Any other regimen Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime
32
Annex 6: Summary of findings and quality of evidence evaluation: preventive therapy for adults and adolescents living with HIV
Q6.3.1 Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: RIF+PZA Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.3.2 Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: INH+RIF Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.3.3 Population: Adults living with HIV Intervention: IPT (6 months INH) Comparison: INH+RIF+PZA Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime
Q6.3.4 Population: Adults living with HIV Intervention: INH+RIF Comparison: INH+RIF+PZA Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.3.5 Population: Adults living with HIV Intervention: INH+RIF Comparison: RIF+PZA Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime Q6.3.6 Population: Adults living with HIV Intervention: INH+RPT Comparison: INH Outcomes: Active TB incidence, mortality, progression of HIV disease, AEs, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime
Outcomes Active TB incidence (suspected, probable, confirmed) Confirmed TB Mortality 9 9 9 8 8 7 7 7 6 6 Relative importance (rank 1 9 most critical)
Comment Critical Critical Critical Critical Critical Critical (addressed by Annex 11) Critical (addressed by Annex 10) Critical (addressed by Annex 12) Less critical Less critical
Progression of HIV disease Adverse events Adherence
TB drug resistance Cost-effectiveness Interval to death Interval to active TB
33
Annex 6

2.1 Pubmed Search (Tuberculosis[Mesh] AND HIV Infections[Mesh]) AND Therapeutics[Mesh]
1375
articles
Limits Clinical trial, meta-analysis, randomized controlled trial, review, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial, multicentre study,
adolescent: 1318 years, adult: 1944 years, middle-aged: 4564 years, middle-aged + aged: 45+ years, aged: 65+ years, 80 and over: 80+ years
219
By title
30
By abstract
19
Of interest
16
International conferences (CROI [Conference on Retroviruses and Opportunistic Infections], International AIDS Society [IAS], Interscience Conference on Antimicrobial Agents and Chemotherapy [ICAAC], World AIDS conference), excluding published papers: ICAAC 2009 and previous 0 IAS conferences 9 CROI 2009 and previous 4
34
he WHO Guidelines Group reviewed the evidence on a wide range of regimens and their duration for TB prevention among people living with HIV, including results from three unpublished trials. The evidence reviewed studies of drug combinations used for prevention including INH, rifampicin, pyrazinamide and rifapentine. A total of seven studies [17] compared INH alone with other regimens, and found that regimens which included pyrazinamide, rifampicin and rifapentine were as
efficacious as INH alone, but were associated with higher rates of toxicity (Table 1). The WHO Guidelines Group concluded that INH at a dose of 300 mg/day remains the drug of choice for chemotherapy to prevent TB in adults living with HIV. The WHO Guidelines Group reviewed and graded the evidence on the duration and durability of effect of IPT in people living with HIV. The critical outcome of interest considered was the efficacy of IPT in preventing active TB, relapse, reinfection and toxicity.
Table 1: Comparison of the efficacy of different drug regimens

Intervention INH INH INH INH and rifampicin Comparator Rifampicin and pyrazinamide INH and rifampicin RR (95% Cl) 1.03 (0.751.4) Quality of evidence Moderate Moderate Low Moderate Moderate
INH, rifampicin and pyrazinamide 0.69 (0.231.57) INH, rifampicin and pyrazinamide 0.75 (0.211.82) INH 1.05 (0.561.97)
0.97 (0.521.83)
INH and rifapentine
he WHO Guidelines Group considered the existing evidence on the optimal duration of IPT (6, 9, 12 and 36 months). Although nine months of IPT is supported by evidence and recommended in some guidelines, there are no studies that have directly compared six and nine months of IPT. Therefore, the discussion focused on the choice between a six- versus 12-month duration of IPT, which showed no significant difference in efficacy. This led the Guidelines Group to strongly recommend the sixmonth duration. The durability of effect of IPT ranges from six months to five years. The Guidelines Group also reviewed unpublished evidence from two studies that suggest increased benefit with a 36-month or longer duration of IPT, particularly in people who are TST positive.[8,9] The Guidelines Group considered at least 36 months duration as a surrogate for lifelong treatment and emphasized the potential benefit of this regimen for people living with HIV in settings with a background of high TB prevalence and transmission. Given the preliminary nature of the evidence, feasibility concerns and potential adverse events, the Guidelines Group conditionally recommends 36 months duration of IPT for people living with HIV in settings with high TB prevalence and transmission. The Guidelines Group reviewed available data on the initiation of IPT and immune status, including concomitant use with ART. Six studies were examined [4,913] and showed contrasting results regarding the reduction of TB risk by immune status.
Additional protective benefits of concomitant use of IPT with ART were demonstrated in two observational studies from Brazil and South Africa [12,13], other studies from South Africa [4,12,14,15], and a subanalysis of data from an unpublished randomized clinical trial from Botswana. Based on this evidence and the potential benefit of concomitant use of IPT with ART, the Guidelines Group strongly recommend that IPT be given irrespective of immune status and whether or not the person is on ART. IPT initiation or completion should not delay commencement of ART in eligible people living with HIV. However, the Guidelines Group recognizes the absence of evidence on whether concomitant initiation of IPT with ART or delayed initiation of IPT is better in terms of efficacy and reduction in toxicity. Pregnant women living with HIV are at risk for TB, which can impact on maternal and perinatal outcomes, ranging from death of the mother and the newborn, to prematurity and low birth weight. The Guidelines Group stressed the importance of screening pregnant women living with HIV for active TB using the standard clinical algorithm for adults and adolescents as listed in Annex 5. The Group concluded that existing evidence and experience in the pre-HIV and HIV era suggest that IPT is safe in pregnant women. Therefore, the Guidelines Group strongly recommends that pregnancy should not exclude women living with HIV from symptom-based screening for TB and receiving IPT.
35
Annex 6
The Guidelines Group reviewed the evidence and discussed the use of IPT as secondary prophylaxis for people who have previously been successfully treated for TB. The evidence was analysed using the GRADE process. Evidence from four studies including three randomized controlled trials and one observational study showed the value of providing IPT immediately after successful completion of TB treatment. The WHO Guidelines Group strongly recommends that adults and adolescents living with
HIV who have successfully completed TB treatment should continue to receive INH for six months and should conditionally receive it for 36 months based on the local situation and existing national guidelines. As there was no evidence on the potential role of IPT for a patient who had successfully completed treatment for multidrug-resistant (MDR) or extensively drugresistant (XDR) TB, the Guidelines Group has not made any recommendation on the use of IPT after successful treatment for MDR or XDR TB.
2. Pubmed Search ((Child[Mesh] OR Child, Preschool[Mesh]) OR Infant[Mesh]) AND Tuberculosis[Mesh] AND HIV Infections[Mesh] AND Therapeutics[Mesh])
187
articles
Limits Clinical trial, meta-analysis, randomized controlled trial, review, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial, evaluation studies,
multicentre study, Medline, PubMed Central, all infants: birth23 months, newborn: birth1 month, infant: 123 months, preschool child: 25 years, child: 612 years
54
By title
By abstract
Of interest
International conferences (CROI, IAS, ICAAC, World AIDS conference) excluding published papers: ICAAC 2009 and previous 1 IAS conferences 2 CROI 2009 and previous 1
36
wo studies [14,16] were considered for the GRADE analysis. One study suggested significant benefits for children receiving INH for six months, in particular, with regard to significant reductions in mortality. However, findings from a randomized controlled trial [14] conducted in South Africa showed that when HIV-infected infants with no known exposure to a TB source case are identified in the first three to four months of life, given rapid access to ART and carefully monitored for new TB exposure or disease on a monthly basis, there is no benefit from IPT. Therefore, based on this, the Guidelines Group recommends that all children more than 12 months of age who are living with HIV and who are unlikely to have active TB should receive six months of IPT as part of a comprehensive package of HIV care.
For children less than 12 months of age, only those with a history of contact with a TB case should receive six months of IPT. In contrast to adults and adolescents, there is no evidence to support the use of INH for longer than six months in children. Therefore, the Guidelines Group concludes that until more data are available, INH for children could not be recommended for more than six months. Similarly, there is no evidence on whether repeating a course of IPT is beneficial for children. INH should be given at a dose of 10 mg/kg of body weight/per day and it is desirable that 25 mg of vitamin B6 be supplied daily with INH. All available data to date suggest that INH is not toxic for children, even in those receiving ART. The following table shows a simplified dosing schedule for INH at a dose of 10 mg/kg/day.
Table 2: Dosing schedule for INH in children

Weight range (kg) <5 Number of 100 mg tablets of INH to be administered per dose tablet 1 tablet Dose given (mg) 50
5.19.9
1013.9 1419.9 2024.9
1 tablet 2 tablets 2 tablets
100 150 200 250 300
>25
3 tablets or one adult tablet
he Guidelines Group noted that there is no evidence on the use of IPT in children after successful completion of TB treatment. However, like adults, children living with HIV are exposed to reinfection and recurrence of TB. Therefore, the Group recommends that children who have been successfully treated for TB and are living in settings with high TB transmission should receive IPT for an additional six months. IPT can be started immediately following the last doses of antituberculosis therapy. TB screening should be carried out regardless of a history of TB treatment during each contact of the child with a health-care worker.
The Guidelines Group also concluded that there are no data regarding the efficacy of IPT for children stratified by the degree of immunosuppression. However, the Group noted that there is biological plausibility in extrapolating what is known for adults and adolescents to children. Therefore, it conditionally recommends the combined use of IPT with ART for all children. The Group emphasizes that ART should not be delayed while starting or completing a course of IPT.
37
Annex 6
Q 1.1 Efficacy Q 1.1.1 ANY TB DRuG A. Question: Should preventive therapy (any TB drug) be used in people living with HIV with any PPD status? Settings: Any TB/HIV prevalence Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Mohammed et al. 2007; Rivero et al. 2003; Gordin et al. 2000; Hasley et al. 1998; Martinez et al. 2000; Rivero 2007
No. of patients
Control Relative risk (95% CI) Absolute
Quality assessment
Indirectness Preventive therapy (any TB drug)
142/3728 (3.8%) MODERATE 2% 50% 160 fewer per 1000 (from 75 fewer to 230 fewer) RR 0.73 (0.491.08) CRITICAL HIGH 12 fewer per 1000 (from 23 fewer to 4 more) 5 fewer per 1000 (from 10 fewer to 2 more) 108 fewer per 1000 (from 204 fewer to 32 more) RR 0.94 (0.851.05) CRITICAL LOW 1 fewer per 1000 (from 3 fewer to 1 more) 30 fewer per 1000 (from 75 fewer to 25 more) RR 0.88 (0.61.28) 1% 50% CRITICAL HIGH 1 fewer per 1000 (from 4 fewer to 3 more) 60 fewer per 1000 (from 200 fewer to 140 more) 140/3602 (3.9%) 33/1923 (1.7%) 0% RR 2.55 (1.73.85) CRITICAL HIGH 0 more per 1000 (from 0 more to 0 more) 27 more per 1000 (from 12 more to 49 more) 19 fewer per 1000 (from 63 fewer to 44 more) 13 fewer per 1000 (from 31 fewer to 10 more) 6 fewer per 1000 (from 3 fewer to 9 fewer) 129/2034 (6.3%) RR 0.68 (0.540.85) 20 fewer per 1000 (from 10 fewer to 29 fewer)
No. of studies
No serious indirectness2 No serious imprecision None
Design
Limitations
Inconsistency
Imprecision
Summary of findings Effect Quality

CRITICAL
Importance
Active TB incidence (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for acid-fast bacilli [AFB])
12
Randomized trials
Serious1
No serious inconsistency
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness2 50/1547 (3.2%) 2% 40% 47/1026 (4.6%) No serious imprecision None
3. Findings and GRADE profiles
Randomized trials
No serious limitations1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness2 633/3728 (17%) 2% 50% 427/2034 (21%) No serious imprecision None
Randomized trials
Serious1
Serious3
HIV disease progression (follow up 13 years; clinical and immunological criteria) No serious indirectness No serious imprecision None 41/184 (22.3%) 43/272 (15.8%)
Randomized trials
No serious limitations
Adverse drug reaction leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None
Randomized trials
38
Concealment of allocation was adequate in five studies and unclear in the remaining seven; in seven trials both providers and participants were blinded; the inclusion of randomized participants was adequate in all the included trials; intention-to-treat (ITT) analysis has been performed in 10 trials; participants loss to follow up ranged from 0 to 31%. Mohammed et al. 2007 was not powered to assess the effectiveness of IPT. 2 A different TST status was considered across the studies; however, TST -negative patients were considered only in settings with >30% of LTBI. 3 Four out of nine studies provided an opposite direction of the effect.
39
39/1760 (2.2%) 2% 50% 310 fewer per 1000 (from 215 fewer to 375 fewer) RR 0.30 (0.061.57) CRITICAL LOW 14 fewer per 1000 (from 19 fewer to 11 more) 350 fewer per 1000 (from 470 fewer to 285 more) RR 0.80 (0.631.02) CRITICAL MODERATE 10 fewer per 1000 (from 19 fewer to 1 more) 50% 100 fewer per 1000 (from 185 fewer to 10 more) 11/25 (44%) 5% 50% RR 0.36 (0.150.85) CRITICAL HIGH 32 fewer per 1000 (from 7 fewer to 43 fewer) 320 fewer per 1000 (from 75 fewer to 425 fewer) 282 fewer per 1000 (from 66 fewer to 374 fewer) 27 fewer per 1000 (from 50 fewer to 3 more) 47 fewer per 1000 (from 63 fewer to 38 more) 12 fewer per 1000 (from 9 fewer to 15 fewer) 46/618 (7.4%) RR 0.38 (0.250.57)
B. Question: Should preventive therapy (any TB drug) be used in TST-positive people living with HIV? Settings: Different TB/HIV prevalences Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998
No. of patients
Quality assessment
Imprecision Other considerations
None 46 fewer per 1000 (from 32 fewer to 56 fewer)
No. of studies
No serious indirectness No serious imprecision
Design
Limitations
Inconsistency
Importance
Active TB incidence (possible, probable or confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% Serious2 None 2/101 (2%) 4/60 (6.7%)
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 195/1760 (11.1%) 5% 84/618 (13.6%) No serious imprecision None
Randomized trials
Serious3
HIV disease progression (follow up 13 years; clinical and immunological criteria) No serious indirectness No serious imprecision None 6/38 (15.8%)
Randomized trial
Only one study is available to address this outcome. Small sample size and wide confidence interval 3 One of the four studies showed an opposite direction of the effect.
Annex 6
C. Question: Should preventive therapy (any TB drug) be used in TST-negative people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Fitzgerald et al. 2001; Gordin et al. 1997; Hawken et al. 1997; Mohammed et al. 2007; Mwinga et al. 1998; Pape et al. 1993; Rivero et al. 2003; Whalen et al.1997 anergy
No. of patients
Quality assessment
75/1677 (4.5%) 2% 50% 55 fewer per 1000 (from 180 fewer to 120 more) RR 0.74 (0.381.45) 2 fewer per 1000 (from 7 fewer to 5 more) 60/1243 (4.8%) RR 0.89 (0.641.24)
No. of studies
No serious indirectness 5 fewer per 1000 (from 17 fewer to 12 more) No serious imprecision None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB incidence (follow up 13 years; clinical presentation, chest X-ray, sputum for AFB, response to TB treatment) MODERATE CRITICAL
Randomized trials
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 40% No serious imprecision None 22/853 (2.6%) 15/500 (3%) 8 fewer per 1000 (from 19 fewer to 14 more) 5 fewer per 1000 (from 12 fewer to 9 more) 104 fewer per 1000 (from 248 fewer to 180 more) RR 1.01 (0.891.14) CRITICAL MODERATE 0 more per 1000 (from 2 fewer to 3 more) 50% 5 more per 1000 (from 55 fewer to 70 more) 32/146 (21.9%) 2% 50% RR 1.10 (0.721.1) CRITICAL MODERATE 2 more per 1000 (from 6 fewer to 2 more) 50 more per 1000 (from 140 fewer to 50 more) 22 more per 1000 (from 61 fewer to 22 more) 3 more per 1000 (from 28 fewer to 36 more) MODERATE CRITICAL
Randomized trials
Serious2
Mortality (any cause) (follow up 13 years) No serious indirectness 2% No serious imprecision None 387/1677 (23.1%) 316/1243 (25.4%)
Randomized trials
Serious3
HIV disease progression (follow up 13 years; clinical and immunological criteria) No serious indirectness No serious imprecision None 35/146 (24%)
Randomized trials
Serious4
Three studies out of eight showed a different direction of the effect. One of the three studies available presented an opposite direction of the effect. 3 The direction of effect was different across the studies. 4 Opposite direction of the effect
40
41
28/291 (9.6%) LOW 2% 50% 95 fewer per 1000 (from 255 fewer to 170 more) RR 0.79 (0.471.32) CRITICAL HIGH 4 fewer per 1000 (from 11 fewer to 6 more) 84 fewer per 1000 (from 212 fewer to 128 more) RR 0.84 (0.581.24) CRITICAL HIGH 3 fewer per 1000 (from 8 fewer to 5 more) 50% 80 fewer per 1000 (from 210 fewer to 120 more) 34 fewer per 1000 (from 90 fewer to 51 more) 13 fewer per 1000 (from 32 fewer to 19 more) 4 fewer per 1000 (from 10 fewer to 7 more) 23/173 (13.3%) RR 0.81 (0.491.34) 25 fewer per 1000 (from 68 fewer to 45 more)
D. Question: Should preventive therapy (any TB drug) be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998; Hawken et al. 1997
No. of patients
Quality assessment
None
No. of studies
No serious indirectness Serious1
Design
Limitations
Inconsistency
Importance
Active TB incidence (follow up 13 years; clinical presentation, chest X-ray, sputum for AFB, response to TB treatment) CRITICAL
Randomized trials
Serious
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 40% No serious imprecision None 26/593 (4.4%) 28/466 (6%)
Randomized trials
Randomized trials
Annex 6
Q 1.1.2 INH vs PLACEBO A. Question: Should INH prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Rivero et al. 2003; Whalen et al.1997 anergy
No. of patients
Quality assessment
Indirectness INH prophylaxis
85/2152 (3.9%) 2% 50% 7 fewer per 1000 (from 3 fewer to 10 fewer) 165 fewer per 1000 (from 65 fewer to 245 fewer) RR 0.72 (0.471.11) 123/1984 (6.2%) RR 0.67 (0.510.87)
No. of studies
No serious indirectness 20 fewer per 1000 (from 8 fewer to 30 fewer) No serious imprecision None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB incidence ( probable, possible, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) HIGH CRITICAL
Randomized trials
No serious inconsistency1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 34/1037 (3.3%) 2% 50% 47/1026 (4.6%) No serious imprecision None 13 fewer per 1000 (from 24 fewer to 5 more) 6 fewer per 1000 (from 11 fewer to 2 more) 140 fewer per 1000 (from 265 fewer to 55 more) RR 0.95 (0.851.06) CRITICAL MODERATE 3 fewer per 1000 (from 7 fewer to 3 more) 50% 25 fewer per 1000 (from 75 fewer to 30 more) 43/272 (15.8%) 10% 50% RR 0.88 (0.61.28) CRITICAL MODERATE 12 fewer per 1000 (from 40 fewer to 28 more) 60 fewer per 1000 (from 200 fewer to 140 more) None 56/2026 (2.8%) 33/1873 (1.8%) 0% 20% RR 1.66 (1.092.51) CRITICAL HIGH 0 more per 1000 (from 0 more to 0 more) 132 more per 1000 (from 18 more to 302 more) 12 more per 1000 (from 2 more to 27 more) 30 fewer per 1000 (from 101 fewer to 70 more) 11 fewer per 1000 (from 32 fewer to 13 more) 5% HIGH CRITICAL
Randomized trials
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 427/2152 (19.8%) No serious imprecision None 419/1984 (21.1%)
Randomized trials
Serious2
Randomized trials
Serious2
Adverse drug reaction leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision
Randomized trials
42
Three out of eight showed an opposite direction of the effect. Different direction of the effect across the studies
43
18/693 (2.6%) 2% 50% 320 fewer per 1000 (from 195 fewer to 390 fewer) RR 0.13 (0.012.32) CRITICAL LOW 17 fewer per 1000 (from 20 fewer to 26 more) 435 fewer per 1000 (from 495 fewer to 660 more) RR 0.74 (0.551) CRITICAL HIGH 5 fewer per 1000 (from 9 fewer to 0 more) 50% 130 fewer per 1000 (from 225 fewer to 0 more) 38/25 (152%) 10% 50% RR 0.36 (0.150.85) CRITICAL MODERATE 64 fewer per 1000 (from 15 fewer to 85 fewer) 320 fewer per 1000 (from 75 fewer to 425 fewer) 973 fewer per 1000 (from 228 fewer to 1292 fewer) 35 fewer per 1000 (from 61 fewer to 0 more) 58 fewer per 1000 (from 66 fewer to 88 more) 13 fewer per 1000 (from 8 fewer to 16 fewer) 46/618 (7.4%) RR 0.36 (0.220.61)
B. Question: Should INH prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998
No. of patients
Quality assessment
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB incidence (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 0/52 (0%) 2% 50% 4/60 (6.7%) Serious2 None
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 2% No serious imprecision None 71/693 (10.2%) 84/618 (13.6%)
Randomized trials
Randomized trial
Serious1
Only one study was available to address this outcome. Small sample size and wide CI 3 Mwinga et al. reported an opposite direction of the effect.
Annex 6
C. Question: Should INH prophylaxis be used in people living with HIV who are TST -negative? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Fitzgerald et al. 2001; Gordin et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Pape et al. 1993; Rivero et al. 2003; Whalen et al.1997 anergy
No. of patients
Quality assessment
49/1297 (3.8%) HIGH 2% 50% 70 fewer per 1000 (from 205 fewer to 130 more) RR 0.76 (0.361.61) 3 fewer per 1000 (from 8 fewer to 5 more) 54/1193 (4.5%) RR 0.86 (0.591.26) 6 fewer per 1000 (from 19 fewer to 12 more)
No. of studies
No serious indirectness No serious imprecision None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% No serious imprecision None 12/521 (2.3%) 15/500 (3%) 7 fewer per 1000 (from 19 fewer to 18 more) 5 fewer per 1000 (from 13 fewer to 12 more) 120 fewer per 1000 (from 320 fewer to 305 more) RR 1.02 (0.91.16) CRITICAL HIGH 0 more per 1000 (from 2 fewer to 3 more) 50% 10 more per 1000 (from 50 fewer to 80 more) 32/146 (21.9%) 10% 50% RR 1.10 (0.721.69) CRITICAL MODERATE 10 more per 1000 (from 28 fewer to 69 more) 50 more per 1000 (from 140 fewer to 345 more) 22 more per 1000 (from 61 fewer to 151 more) 5 more per 1000 (from 25 fewer to 40 more) MODERATE CRITICAL
Randomized trials
Serious2
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 2% No serious imprecision None 328/1297 (25.3%) 298/1193 (25%)
Randomized trials
HIV disease progression (follow up 13 years; clinical and immunological criteria) No serious indirectness No serious imprecision None 35/146 (24%)
Randomized trials
Serious3
Fitzgerald et al. and Hawken et al. showed an opposite direction of the effect. Different direction of the effect across the studies 3 Opposite direction of the effect
44
45
18/162 (11.1%) 2% 50% 70 fewer per 1000 (from 260 fewer to 260 more) RR 0.79 (0.461.36) CRITICAL HIGH 4 fewer per 1000 (from 11 fewer to 7 more) 105 fewer per 1000 (from 270 fewer to 180 more) RR 0.81 (0.521.27) CRITICAL HIGH 4 fewer per 1000 (from 10 fewer to 5 more) 50% 95 fewer per 1000 (from 240 fewer to 135 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 41 fewer per 1000 (from 103 fewer to 58 more) 13 fewer per 1000 (from 32 fewer to 22 more) 3 fewer per 1000 (from 10 fewer to 10 more) 23/173 (13.3%) RR 0.86 (0.481.52)
D. Question: Should INH prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998; Hawken et al. 1997
No. of patients
Quality assessment
None 19 fewer per 1000 (from 69 fewer to 69 more)
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) MODERATE CRITICAL
Randomized trials
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 22/464 (4.7%) 2% 50% 28/466 (6%) No serious imprecision None
Randomized trials
Randomized trials
HIV disease progression (follow up 13 years; clinical and immunological criteria) None 0/0 (0%)
No evidence available
Annex 6
Q 1.1.3 INH+RIF vs PLACEBO A. Question: Should INH+RIF prophylaxis be used in people living with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997; Rivero et al. 2003
No. of patients
Quality assessment
Indirectness INH+RIF prophylaxis
12/638 (1.9%) 2% 50% 295 fewer per 1000 (from 95 fewer to 395 fewer) RR 0.70 (0.163.05) 12 fewer per 1000 (from 4 fewer to 16 fewer) 25/541 (4.6%) RR 0.41 (0.210.81)
No. of studies
None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 27 fewer per 1000 (from 9 fewer to 37 fewer) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 0% 50% Serious2 None 3/82 (3.7%) 4/77 (5.2%) 16 fewer per 1000 (from 44 fewer to 106 more) 0 fewer per 1000 (from 0 fewer to 0 more) 150 fewer per 1000 (from 420 fewer to 1025 more) RR 0.69 (0.50.95) CRITICAL HIGH 31 fewer per 1000 (from 5 fewer to 50 fewer) 50% 155 fewer per 1000 (from 25 fewer to 250 fewer) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 28/638 (4.4%) 1/541 (0.2%) 0% RR 16.72 (3.2984.89) CRITICAL HIGH 0 more per 1000 (from 0 more to 0 more) 29 more per 1000 (from 4 more to 155 more) CRITICAL 43 fewer per 1000 (from 7 fewer to 69 fewer) LOW CRITICAL
Randomized trial
Serious1
Randomized trials
Randomized trials
Only one study is available for this outcome. Small sample size and wide CI
46
47
9/556 (1.6%) 2% 50% 320 fewer per 1000 (from 115 fewer to 415 fewer) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0.74 (0.531.04) CRITICAL MODERATE 26 fewer per 1000 (from 47 fewer to 4 more) 130 fewer per 1000 (from 235 fewer to 20 more) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 36 fewer per 1000 (from 65 fewer to 6 more) CRITICAL 13 fewer per 1000 (from 5 fewer to 17 fewer) 21/464 (4.5%) RR 0.36 (0.170.77)
B. Question: Should INH+RIF prophylaxis be used in people living with HIV who are TST-positive? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
No. of patients
Quality assessment
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB disease (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) MODERATE CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) None 0/0 (0%) 0% 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 57/556 (10.3%) 10% 50% 64/464 (13.8%) No serious imprecision None
Randomized trial
Serious2
HIV disease progression (follow up 13 years; clinical and immunological criteria) None 0/0 (0%) 0/0 (0%)
Only one study addressed this outcome. No explanation was provided.
Annex 6
C. Question: Should INH+RIF prophylaxis be used in people living with HIV who are TST -negative? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Rivero et al. 2003
No. of patients
Quality assessment
No. of studies
Serious2 None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness 16 fewer per 1000 (from 44 fewer to 106 more) LOW CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 3/82 (3.7%) 2% 50% 4/77 (5.2%) Serious2 None 16 fewer per 1000 (from 44 fewer to 106 more) 6 fewer per 1000 (from 17 fewer to 41 more) 150 fewer per 1000 (from 420 fewer to 1025 more) RR 0.34 (0.111.03) 94 fewer per 1000 (from 127 fewer to 4 more) 66 fewer per 1000 (from 89 fewer to 3 more) 50% 330 fewer per 1000 (from 445 fewer to 15 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL CRITICAL LOW CRITICAL
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 10% Serious2 None 4/82 (4.9%) 11/77 (14.3%)
Randomized trial
Serious3
Only one study addressed this outcome. Small sample size and wide CI 3 No explanation was provided.
48
D. Question: Should INH+RIF prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
49
Indirectness INH+RIF+PZA prophylaxis
10/462 (2.2%) 2% 50% 260 fewer per 1000 (from 385 fewer to 0 more) RR 0.91 (0.651.27) CRITICAL LOW 2 fewer per 1000 (from 7 fewer to 5 more) 45 fewer per 1000 (from 175 fewer to 135 more) RR 0.69 (0.50.95) 12 fewer per 1000 (from 48 fewer to 37 more) 12 fewer per 1000 (from 48 fewer to 37 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 1/464 (0.2%) 0% RR 26.11 (3.56191.63) CRITICAL MODERATE 0 more per 1000 (from 0 more to 0 more) 54 more per 1000 (from 6 more to 411 more) CRITICAL OO LOW CRITICAL 12 fewer per 1000 (from 48 fewer to 37 more) 10 fewer per 1000 (from 15 fewer to 0 more) 21/464 (4.5%) RR 0.48 (0.231)
Q 1.1.4 INH+RIF+PZA vs placebo A. Question: Should INH+RIF+PZA prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
No. of patients
Quality assessment
No. of studies
Serious3 No serious imprecision
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years1; clinical examination, chest X-ray, sputum for AFB) LOW CRITICAL
Randomized trial
Serious2
Confirmed TB (follow up 13 years; culture-proven) Serious3 58/462 (12.6%) 2% 50% 64/464 (13.8%) No serious imprecision None
Randomized trial
Serious2
Mortality (any cause) (follow up 13 years1; review of hospital records) Serious4 13.8% No serious imprecision None 58/462 (12.6%) 64/464 (13.8%)
Randomized trial
Serious2
Adverse drug reaction leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None 26/462 (5.6%)
Randomized trial
Serious2
Johnson et al. 2001 provided data on long-term follow up. Only one study is available to assess this outcome. 3 Only TST-positive subjects were considered. 4 No explanation was provided.
Annex 6
B. Question: Should INH+RIF+PZA prophylaxis be used in people living with HIV who are TST-positive? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et el. 1997
No. of patients
Quality assessment
Indirectness INH+RIF+PZA prophylaxis
10/462 (2.2%) 2% 50% 260 fewer per 1000 (from 385 fewer to 0 more) RR 0.91 (0.651.27) 10 fewer per 1000 (from 15 fewer to 0 more) 21/464 (4.5%) RR 0.48 (0.231)
No. of studies
Design
Limitations
Inconsistency
Imprecision
Importance
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 58/462 (12.6%) 2% 50% 64/464 (13.8%) No serious imprecision None 12 fewer per 1000 (from 48 fewer to 37 more) 2 fewer per 1000 (from 7 fewer to 5 more) 45 fewer per 1000 (from 175 fewer to 135 more) RR 0 (00) CRITICAL MODERATE 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 0 fewer per 1000 (from 0 fewer to 0 fewer) MODERATE CRITICAL
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 0% No serious imprecision None 0/0 (0%) 0/0 (0%)
Randomized trial
Serious1
Only one study is available to assess this outcome.
C. Question: Should INH+RIF+PZA prophylaxis be used in people living with HIV who are TST-negative? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
50
D. Question: Should INH+RIF+PZA prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
51
Indirectness RIF+PZA prophylaxis
26/428 (6.1%) 2% 50% 230 fewer per 1000 (from 70 fewer to 330 fewer) RR 0.69 (0.341.38) CRITICAL HIGH 0 fewer per 1000 (from 0 fewer to 0 more) 155 fewer per 1000 (from 330 fewer to 190 more) RR 1.04 (0.771.41) CRITICAL MODERATE 4 more per 1000 (from 23 fewer to 41 more) 50% 20 more per 1000 (from 115 fewer to 205 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 27/428 (6.3%) 3/427 (0.7%) 0% RR 7.84 (2.623.67) CRITICAL HIGH 0 more per 1000 (from 0 more to 0 more) 48 more per 1000 (from 11 more to 159 more) CRITICAL 6 more per 1000 (from 37 fewer to 66 more) 14 fewer per 1000 (from 29 fewer to 17 more) 9 fewer per 1000 (from 3 fewer to 13 fewer) 48/427 (11.2%) RR 0.54 (0.340.86)
Q 1.1.5 RIF+PZA vs placebo A. Question: Should RIF+PZA prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998; Rivero et al. 2003
No. of patients
Quality assessment
None
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 52 fewer per 1000 (from 16 fewer to 74 fewer) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 0% 50% No serious imprecision None 13/428 (3%) 19/427 (4.4%)
Randomized trials
Randomized trials
Serious1
Randomized trials
Annex 6
B. Question: Should RIF+PZA prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998
No. of patients
Quality assessment
2/49 (4.1%) 2% 50% 390 fewer per 1000 (from 20 fewer to 475 fewer) RR 0.61 (0.123.2) 16 fewer per 1000 (from 1 fewer to 19 fewer) 11/60 (18.3%) RR 0.22 (0.050.96)
No. of studies
None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 143 fewer per 1000 (from 7 fewer to 174 fewer) MODERATE CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2/49 (4.1%) 2% 50% 4/60 (6.7%) Serious2 None 26 fewer per 1000 (from 59 fewer to 147 more) 8 fewer per 1000 (from 18 fewer to 44 more) 195 fewer per 1000 (from 440 fewer to 1100 more) RR 2.76 (0.98.41) CRITICAL LOW 176 more per 1000 (from 10 fewer to 741 more) 50% 880 more per 1000 (from 50 fewer to 3705 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 117 more per 1000 (from 7 fewer to 494 more) LOW CRITICAL
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 10% Serious2 None 9/49 (18.4%) 4/60 (6.7%)
Randomized trial
Serious1
Only one study available to address this outcome Small sample size and wide CI
52
53
14/250 (5.6%) 2% 50% 180 fewer per 1000 (from 330 fewer to 115 more) RR 0.76 (0.282.01) CRITICAL MODERATE 5 fewer per 1000 (from 14 fewer to 20 more) 120 fewer per 1000 (from 360 fewer to 505 more) RR 1.02 (0.681.52) CRITICAL MODERATE 2 more per 1000 (from 32 fewer to 52 more) 50% 10 more per 1000 (from 160 fewer to 260 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 3 more per 1000 (from 51 fewer to 83 more) 9 fewer per 1000 (from 27 fewer to 37 more) 7 fewer per 1000 (from 13 fewer to 5 more) 21/243 (8.6%) RR 0.64 (0.341.23)
C. Question: Should RIF+PZA prophylaxis be used in people living with HIV who are TST-negative? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998; Rivero et al. 2003
No. of patients
Quality assessment
None
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 31 fewer per 1000 (from 57 fewer to 20 more) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% No serious imprecision None 7/250 (2.8%) 9/243 (3.7%)
Randomized trials
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 41/250 (16.4%) 10% 39/243 (16%) No serious imprecision None
Randomized trials
Serious1
Annex 6
D. Question: Should RIF+PZA prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998
No. of patients
Quality assessment
No. of studies
None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 52 fewer per 1000 (from 93 fewer to 35 more) MODERATE CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% No serious imprecision None 4/129 (3.1%) 6/124 (4.8%) 17 fewer per 1000 (from 39 fewer to 59 more) 7 fewer per 1000 (from 16 fewer to 24 more) 180 fewer per 1000 (from 405 fewer to 610 more) RR 0.85 (0.511.41) CRITICAL MODERATE 15 fewer per 1000 (from 49 fewer to 41 more) 10% 50% 75 fewer per 1000 (from 245 fewer to 205 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 31 fewer per 1000 (from 103 fewer to 86 more) MODERATE CRITICAL
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness No serious imprecision None 23/129 (17.8%) 26/124 (21%)
Randomized trial
Serious1
Only one study is available to address this outcome.
54
55
77/1705 (4.5%) 2% 50% 1 more per 1000 (from 5 fewer to 8 more) 15 more per 1000 (from 125 fewer to 200 more) RR 1.02 (0.671.55) CRITICAL MODERATE 0 more per 1000 (from 7 fewer to 11 more) 10 more per 1000 (from 165 fewer to 275 more) RR 1.03 (0.891.19) CRITICAL HIGH 10% 50% 3 more per 1000 (from 11 fewer to 19 more) 15 more per 1000 (from 55 fewer to 95 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 73/1705 (4.3%) 114/1704 (6.7%) 0% RR 0.63 (0.480.84) CRITICAL HIGH 0 fewer per 1000 (from 0 fewer to 0 fewer) 25 fewer per 1000 (from 11 fewer to 35 fewer) CRITICAL 6 more per 1000 (from 20 fewer to 35 more) 1 more per 1000 (from 9 fewer to 14 more) 75/1704 (4.4%) RR 1.03 (0.751.4)
Q 1.2 OPTIMAL REGIMEN Q.1.2.1 INH vs RIF+PZA A. Question: Should INH vs RIF+PZA prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Gordin et al. 2000; Halsey et al. 1998; Mwinga et al. 1998; Rivero et al. 2007; Rivero et al. 2003
Importance
Absolute
Quality assessment
Imprecision RIF+PZA prophylaxis Relative risk (95% CI) Other considerations
None
No. of studies
Design
Limitations
Inconsistency
No. of patients Quality
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 1 more per 1000 (from 11 fewer to 18 more) MODERATE CRITICAL
Randomized trials
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 43/1597 (2.7%) 2% 50% 42/1599 (2.6%) No serious imprecision None
Randomized trials
Serious2
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 299/1597 (18.7%) No serious imprecision None 283/1540 (18.4%)
Randomized trials
Adverse drug reactions leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None
Randomized trials
Different direction of the effect across the studies Gordin et al. 2000 showed an opposite direction of the effect.
Annex 6
B. Question: Should INH vs RIF+PZA prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Gordin et al. 2000; Halsey et al. 1998; Mwinga et al. 1998; Rivero et al. 2007
No. of patients
RIF+PZA prophylaxis
RR 1 (0.681.47) 0 fewer per 1000 (from 6 fewer to 9 more) 0 fewer per 1000 (from 160 fewer to 235 more) RR 1 (0.621.63) CRITICAL MODERATE 0 fewer per 1000 (from 8 fewer to 13 more) 0 fewer per 1000 (from 190 fewer to 315 more) RR 1.09 (0.931.29) CRITICAL MODERATE 2 more per 1000 (from 1 fewer to 6 more) 50% 45 more per 1000 (from 35 fewer to 145 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 16 more per 1000 (from 13 fewer to 52 more) 0 fewer per 1000 (from 10 fewer to 16 more)
Quality assessment
51/1322 (3.9%) 2% 50% 51/1325 (3.8%)
No. of studies Relative risk (95% CI) Absolute

None
Design
Limitations
Inconsistency
Imprecision
Importance
Randomized trials
Serious1
Randomized trials
Serious2
Mortality (any cause) (follow up 13 years) No serious indirectness 238/1214 (19.6%) 10% 219/1220 (18%) No serious imprecision None
Randomized trials
Serious2
Different direction of the effect across the studies One out of the three studies showed an opposite effect.
56
57
17/261 (6.5%) 2% 50% 85 more per 1000 (from 205 fewer to 660 more) RR 1.24 (0.473.28) CRITICAL HIGH 5 more per 1000 (from 11 fewer to 46 more) 120 more per 1000 (from 265 fewer to 1140 more) RR 0.80 (0.541.17) CRITICAL HIGH 4 fewer per 1000 (from 9 fewer to 3 more) 50% 100 fewer per 1000 (from 230 fewer to 85 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 43 fewer per 1000 (from 99 fewer to 36 more) 7 more per 1000 (from 15 fewer to 64 more) 3 more per 1000 (from 8 fewer to 26 more) 14/250 (5.6%) RR 1.17 (0.592.32)
C. Question: Should INH vs RIF+PZA prophylaxis be used in TST-negative people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998; Rivero et al. 2003
No. of patients
RIF+PZA prophylaxis Relative risk (95% CI) Absolute
Quality assessment
None
No. of studies
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 10 more per 1000 (from 23 fewer to 74 more) HIGH CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 9/261 (3.4%) 2% 50% 7/250 (2.8%) No serious imprecision None
Randomized trials
Randomized trials
Annex 6
D. Question: Should INH vs RIF+PZA prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Mwinga et al. 1998
No. of patients
RIF+PZA prophylaxis
RR 0.95 (0.42.26) 1 fewer per 1000 (from 12 fewer to 25 more) 25 fewer per 1000 (from 300 fewer to 630 more) RR 0.79 (0.183.47) CRITICAL MODERATE 0 fewer per 1000 (from 0 fewer to 0 more) RR 0.87 (0.51.52) CRITICAL MODERATE 13 fewer per 1000 (from 50 fewer to 52 more) 65 fewer per 1000 (from 250 fewer to 260 more) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 23 fewer per 1000 (from 89 fewer to 93 more) 7 fewer per 1000 (from 25 fewer to 77 more)
Quality assessment
9/122 (7.4%) 2% 50% 10/129 (7.8%)

Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 1-3 years; clinical examination, chest X-ray, sputum for AFB) MODERATE CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 0% No serious imprecision None 3/122 (2.5%) 4/129 (3.1%)
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 19/122 (15.6%) 10% 50% 23/129 (17.8%) No serious imprecision None
Randomized trial
Serious1
58
59
18/791 (2.3%) 2% 50% 15 fewer per 1000 (from 240 fewer to 415 more) RR 1.49 (0.494.5) CRITICAL LOW 10 more per 1000 (from 10 fewer to 70 more) 245 more per 1000 (from 255 fewer to 1750 more) RR 1.09 (0.81.5) CRITICAL LOW 9 more per 1000 (from 20 fewer to 50 more) 50% 45 more per 1000 (from 100 fewer to 250 more) 40/0 (0%) 0% RR 0.79 (0.51.23) CRITICAL HIGH 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 9 more per 1000 (from 19 fewer to 48 more) 16 more per 1000 (from 17 fewer to 116 more) 1 fewer per 1000 (from 10 fewer to 17 more) 19/810 (2.3%) RR 0.97 (0.521.83)
Q 1.2.2 INH vs INH+RIF A. Question: Should INH vs RIF+INH prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Martinez et al. 2000; Rivero et al. 2007; Rivero et al. 2003; Whalen et al. 1997
No. of patients
RIF+INH prophylaxis Relative risk (95% CI) Absolute
Quality assessment
None
No. of studies
Design
Limitations
Inconsistency
Importance
Randomized trials
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% Serious3 None 7/147 (4.8%) 5/151 (3.3%)
Randomized trials
Serious2
Randomized trials
Serious4
Serious5
Adverse drug reactions leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None 31/0 (0%)
Randomized trials
Martinez et al. 2000 showed an opposite direction of the effect. One out of the three studies showed an opposite direction of the effect. 3 Small sample size and wide CI 4 In Martinez et al. 2000 RR was not estimable for lack of events. 5 Different direction and size of the effect across studies 6 Martinez et al. 2000 found an opposite direction of the effect.
Annex 6
B. Question: Should INH vs RIF+INH prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Martinez et al. 2000; Rivero et al. 2007; Whalen et al. 1997
No. of patients
RIF+INH prophylaxis
RR 0.97 (0.521.83) 1 fewer per 1000 (from 11 fewer to 19 more) 15 fewer per 1000 (from 265 fewer to 485 more) RR 3.71 (0.4233.15) CRITICAL LOW 104 more per 1000 (from 22 fewer to 1237 more) 54 more per 1000 (from 12 fewer to 643 more) 1355 more per 1000 (from 290 fewer to 16075 more) RR 1.06 (0.751.49) CRITICAL MODERATE 6 more per 1000 (from 25 fewer to 49 more) 50% 30 more per 1000 (from 125 fewer to 245 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 6 more per 1000 (from 25 fewer to 48 more)
Quality assessment
14/665 (2.1%) 2% 50% 15/685 (2.2%)

None
Design
Limitations
Inconsistency
Imprecision
Importance
Randomized trials
Serious1
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 3/21 (14.3%) 2% 50% 1/26 (3.8%) Serious3 None
Randomized trials
Serious2
Randomized trials
Serious4
Martinez et al. 2000 showed an opposite direction of the effect. Only one study is available to address this outcome. 3 Small sample size and wide CI 4 In Martinez et al. 2000 the RR was not estimable.
60
61
4/126 (3.2%) 2% 50% 5 fewer per 1000 (from 375 fewer to 1435 more) RR 0.99 (0.253.87) CRITICAL MODERATE 0 fewer per 1000 (from 15 fewer to 57 more) 5 fewer per 1000 (from 375 fewer to 1435 more) RR 1.29 (0.592.84) CRITICAL LOW 6 more per 1000 (from 8 fewer to 37 more) 50% 145 more per 1000 (from 205 fewer to 920 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 23 more per 1000 (from 33 fewer to 147 more) 0 fewer per 1000 (from 24 fewer to 92 more) 0 fewer per 1000 (from 15 fewer to 57 more) 4/125 (3.2%) RR 0.99 (0.253.87)
C. Question: Should INH vs RIF+INH prophylaxis be used in TST-negative people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Martinez et al. 2000; Rivero et al. 2007
No. of patients
RIF+INH prophylaxis Relative risk (95% CI) Absolute
Quality assessment
None
No. of studies
Serious1
Design
Limitations
Inconsistency
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness 0 fewer per 1000 (from 24 fewer to 92 more) MODERATE CRITICAL
Randomized trials
Confirmed TB (follow up 13 years; culture-proven) No serious indirectness 2% 50% Serious1 None 4/126 (3.2%) 4/125 (3.2%)
Randomized trials
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 13/126 (10.3%) 2% 10/125 (8%) Serious1 None
Randomized trials
Serious2
Small sample size and wide CI Opposite direction of the effect
D. Question: Should INH vs RIF+INH prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
Annex 6
Q 1.2.3 INH vs INH+RIF+PZA A. Question: Should INH vs INH+RIF+PZA prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
No. of patients
INH+RIF+PZA prophylaxis
RR 0.60 (0.231.57) 8 fewer per 1000 (from 15 fewer to 11 more) 200 fewer per 1000 (from 385 fewer to 285 more) RR 0.86 (0.611.21) CRITICAL MODERATE 3 fewer per 1000 (from 8 fewer to 4 more) 70 fewer per 1000 (from 195 fewer to 105 more) RR 0.86 (0.611.21) CRITICAL LOW 14 fewer per 1000 (from 39 fewer to 21 more) 50% 70 fewer per 1000 (from 195 fewer to 105 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 3/536 (0.6%) 26/462 (5.6%) 2% 50% RR 0.10 (0.030.33) CRITICAL MODERATE 18 fewer per 1000 (from 13 fewer to 19 fewer) 450 fewer per 1000 (from 335 fewer to 485 fewer) 51 fewer per 1000 (from 38 fewer to 55 fewer) CRITICAL 18 fewer per 1000 (from 49 fewer to 26 more) 18 fewer per 1000 (from 49 fewer to 26 more)
Quality assessment
7/536 (1.3%) 2% 50% 10/462 (2.2%)

Serious2 No serious imprecision None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) 9 fewer per 1000 (from 17 fewer to 12 more) LOW CRITICAL
Randomized trial
Serious1
Randomized trial
Serious1
Mortality (any cause) (follow up 13 years; review of hospital records) Serious2 10% No serious imprecision None 58/536 (10.8%) 58/462 (12.6%)
Randomized trial
Serious1
Adverse drug reactions leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None
Randomized trial
Serious1
Only one study assessed this outcome. Only TST-positive patients
62
63
7/536 (1.3%) 2% 50% 200 fewer per 1000 (from 385 fewer to 285 more) RR 0.86 (0.611.21) CRITICAL MODERATE 3 fewer per 1000 (from 8 fewer to 4 more) 70 fewer per 1000 (from 195 fewer to 105 more) RR 0 (00) CRITICAL MODERATE 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 0 fewer per 1000 (from 0 fewer to 0 fewer) 18 fewer per 1000 (from 49 fewer to 26 more) 8 fewer per 1000 (from 15 fewer to 11 more) 10/462 (2.2%) RR 0.60 (0.231.57)
B. Question: Should INH vs INH+RIF+PZA prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
No. of patients
INH+RIF+PZA prophylaxis Relative risk (95% CI) Absolute
Quality assessment
No. of studies
Design
Limitations
Inconsistency
Importance
Randomized trial
Serious1
Randomized trial
Serious1
Mortality (any cause) (review of hospital records) No serious indirectness 0/0 (0%) 0% 0/0 (0%) No serious imprecision None
Randomized trial
Serious1
HIV disease progression (clinical and immunological criteria) None 0/0 (0%) 0/0 (0%)
Only one study assessed this outcome.
C. Question: Should INH vs INH+RIF+PZA prophylaxis be used in TST-negative people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
D. Question: Should INH vs INH+RIF+PZA prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
Annex 6
Q 1.2.4 INH+RIF vs INH+RIF+PZA A. Question: Should INH+RIF vs INH+RIF+PZA prophylaxis be used in people living with HIV with any TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
No. of patients
INH+RIF+PZA prophylaxis
RR 0.75 (0.311.82) 5 fewer per 1000 (from 14 fewer to 16 more) 125 fewer per 1000 (from 345 fewer to 410 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0.82 (0.581.15) CRITICAL LOW 18 fewer per 1000 (from 42 fewer to 15 more) 90 fewer per 1000 (from 210 fewer to 75 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 26/462 (5.6%) 0% RR 0.42 (0.220.8) CRITICAL MODERATE 0 fewer per 1000 (from 0 fewer to 0 fewer) 33 fewer per 1000 (from 11 fewer to 44 fewer) CRITICAL 23 fewer per 1000 (from 53 fewer to 19 more) CRITICAL
Quality assessment
9/556 (1.6%) 2% 50% 10/462 (2.2%)

Serious2 None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (possible, probable, confirmed) (follow up 13 years; clinical examination, chest X-ray, sputum for AFB) No serious imprecision 5 fewer per 1000 (from 15 fewer to 18 more) LOW CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) None 0% 0/0 (0%) 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records) Serious2 57/556 (10.3%) 10% 50% 58/462 (12.6%) No serious imprecision None
Randomized trial
Serious1
Adverse drug reactions leading to treatment interruption (follow up 13 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None 13/556 (2.3%)
Randomized trial
Serious
Only one study is available to assess this outcome. Only TST-positive patients
64
65
9/556 (1.6%) 2% 50% 125 fewer per 1000 (from 345 fewer to 410 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 0.82 (0.581.15) CRITICAL MODERATE 18 fewer per 1000 (from 42 fewer to 15 more) 90 fewer per 1000 (from 210 fewer to 75 more) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 23 fewer per 1000 (from 53 fewer to 19 more) CRITICAL 5 fewer per 1000 (from 14 fewer to 16 more) 10/462 (2.2%) RR 0.75 (0.311.82)
B. Question: Should INH+RIF vs INH+RIF+PZA prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Rivero et al. 2007
No. of patients
INH+RIF+PZA prophylaxis Relative risk (95% CI) Absolute
Quality assessment
No. of studies
Design
Limitations
Inconsistency
Importance
Randomized trial
Serious1
Confirmed TB (follow up 13 years; culture-proven) None 0% 0/0 (0%) 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records) No serious indirectness 10% 50% No serious imprecision None 57/556 (10.3%) 58/462 (12.6%)
Randomized trial
Serious1
C. Question: Should INH+RIF vs INH+RIF+PZA prophylaxis be used in TST-negative people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Whalen et al. 1997
D. Question: Should INH+RIF vs RIF+PZA+INH prophylaxis be used in people living with HIV with unknown TST status? Settings: High TB prevalence settings (>30% LTBI) Bibliography: No studies available
Annex 6
Q 1.2.5 INH vs rifapentine (RPT)+INH A. Question: Should INH+RPT vs INH prophylaxis be used in TST-positive people living with HIV? Settings: High TB prevalence settings (>30% LTBI) Bibliography: Martinson et al. 2009
No. of patients
INH prophylaxis
RR 1.05 (0.561.97)2 1 more per 1000 (from 9 fewer to 19 more) 25 more per 1000 (from 220 fewer to 485 more) RR 0.85 (0.451.59)3 CRITICAL MODERATE 3 fewer per 1000 (from 11 fewer to 12 more) 75 fewer per 1000 (from 275 fewer to 295 more) RR 0.66 (0.331.26)4 CRITICAL MODERATE 34 fewer per 1000 (from 67 fewer to 26 more) 50% 170 fewer per 1000 (from 335 fewer to 130 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 4/240 (1.7%) 4/208 (1.9%) 0% RR 0.87 (0.223.42)3 CRITICAL MODERATE 0 fewer per 1000 (from 0 fewer to 0 more) 2 fewer per 1000 (from 15 fewer to 47 more) CRITICAL 35 fewer per 1000 (from 70 fewer to 27 more) 8 fewer per 1000 (from 29 fewer to 31 more)
Quality assessment
Indirectness INH+RPT prophylaxis
24/329 (7.3%) 2% 50% 22/328 (6.7%)

None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB (follow up median 3.91 years; clinical examination, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 3 more per 1000 (from 30 fewer to 65 more) MODERATE CRITICAL
Randomized trial
Serious1
Confirmed TB (follow up median 3.91 years; culture-proven) No serious indirectness 2% 50% No serious imprecision None 20/329 (6.1%) 17/328 (5.2%)
Randomized trial
Serious1
Mortality (any cause) (follow up median 3.91 years; review of hospital records and patients files) No serious indirectness 10% No serious imprecision None 17/250 (6.8%) 25/241 (10.4%)
Randomized trial
Serious1
HIV disease progression (follow up 3.91 years) None 0/0 (0%)
Adverse drug reaction leading to treatment interruption (follow up median 3.91 years; clinical and laboratory monitoring) No serious indirectness No serious imprecision None
Randomized trial
Serious1
Only one trial is available to address this outcome and comparison. Calculated on the basis of TB incidence (per 100 person-years) 3 Calculated on crude numbers 4 Calculated on the basis of death incidence (per 100 person-years)
66
67
Indirectness Continuous INH prophylaxis
20/997 (2%) 2% 50% 10 fewer per 1000 (from 3 fewer to 14 fewer) 250 fewer per 1000 (from 80 fewer to 355 fewer) RR 0.48 (0.260.9) CRITICAL MODERATE 10 fewer per 1000 (from 2 fewer to 15 fewer) 260 fewer per 1000 (from 50 fewer to 370 fewer) RR 0.43 (0.240.78) CRITICAL MODERATE 57 fewer per 1000 (from 22 fewer to 76 fewer) 50% 285 fewer per 1000 (from 110 fewer to 380 fewer) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 70/983 (7.1%) 12/846 (1.4%) 1% RR 5.02 (2.749.198) CRITICAL MODERATE 40 more per 1000 (from 17 more to 82 more) 57 more per 1000 (from 25 more to 116 more) CRITICAL 20 fewer per 1000 (from 8 fewer to 26 fewer) 15 fewer per 1000 (from 3 fewer to 21 fewer) 46/1150 (4%) RR 0.50 (0.290.84)
Q 1.3 Duration Q1.3.1 6 months INH vs 36 months INH A. Question: Should continuous INH vs 6 months INH prophylaxis be used in people living with HIV with any TST status? Settings: High TB/HIV prevalence settings Bibliography: Martinson et al. 2009; Samandari et al. 2009
No. of patients
6 months INH prophylaxis Relative risk (95% CI) Absolute
Quality assessment
None
No. of studies
Serious2,3,4
Design
Limitations
Inconsistency
Importance
Active TB ( possible, probable, confirmed) (follow up mean 36 months; clinical assessment, chest X-ray, sputum for AFB) No serious imprecision 20 fewer per 1000 (from 6 fewer to 28 fewer) MODERATE CRITICAL
Randomized trials
Confirmed TB (follow up mean 36 months; culture-proven) Serious 14/997 (1.4%) 2% 50% 33/1150 (2.9%) No serious imprecision None
Randomized trial
Mortality (any cause) (follow up 36 months; review of hospital records and patients files) Serious2,3,4 15/997 (1.5%) 10% No serious imprecision None 40/1150 (3.5%)
Randomized trials
HIV disease progression None 0/0 (0%)
05
Adverse drug reactions leading to treatment interruption (follow up 36 months; laboratory monitoring and clinical assessment) No serious indirectness No serious imprecision None
Randomized trials
Serious6
The Soweto trial was not a head-to-head comparison but a four-arm study designed to compare the efficacy of different regimens as well. The Soweto trial considered TST-positive patients while the BOTUSA trial enrolled those with TST+/3 Mean CD4 count at baseline was >500 cells/mm3 in the Soweto trial and around 200 cells/mm3 in the BOTUSA trial. 4 The Soweto trial enrolled patients who were not eligible for ART, while in the BOTUSA trial about 40% of the patients had started ART. 5 Sub-anaylsis on this outcome is expected to be performed soon. 6 The difference in the size of the effect is large.
Annex 6
Q 1.3.2 INH 6 months vs 12 months A. Question: Should 6 months of INH vs 12 months of INH prophylaxis be used in people living with HIV with any TST status? Settings: High TB/HIV prevalence settings Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Rivero et al. 2003; Gordin et al. 2000; Martinez et al. 2000; Halsey et al. 1998; Rivero et al. 2007
No. of patients
12 months of INH prophylaxis
RR 0.58 (0.31.12) 8 fewer per 1000 (from 14 fewer to 2 more) 210 fewer per 1000 (from 350 fewer to 60 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 1.59 (1.0852.34) CRITICAL LOW 59 more per 1000 (from 9 more to 134 more) 50% 295 more per 1000 (from 43 more to 670 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 56/1968 (2.8%) 0/58 (0%) 0% RR 0 (0 to 0)1 CRITICAL None LOW 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 77 more per 1000 (from 11 more to 175 more) 10% CRITICAL
Quality assessment
Indirectness 6 months of INH prophylaxis
57/1806 (3.2%) 2% 50% 10/184 (5.4%)

None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB ( possible, probable, confirmed) (follow up 13 years; clinical assessment, chest X-ray, sputum for AFB) No serious indirectness No serious imprecision 23 fewer per 1000 (from 38 fewer to 7 more) LOW CRITICAL
Randomized trials
Very serious
Confirmed TB (follow up 13 years; culture-proven) None 0% 0% 0/0 (0%) 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records and patients files) No serious indirectness 375/1806 (20.8%) No serious imprecision None 24/184 (13%)
Randomized trials
Very serious
Adverse drug reactions leading to treatment interruption (follow up 13 years; laboratory monitoring and clinical assessment) No serious indirectness No serious imprecision
Randomized trials
Very serious
Not estimable due to the lack of events in the 12 months group
68
69
16/655 (2.4%) 2% 50% 270 fewer per 1000 (from 445 fewer to 475 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) RR 1.31 (0.433.98) CRITICAL VERY LOW 31 more per 1000 (from 57 fewer to 298 more) 50% 155 more per 1000 (from 285 fewer to 1490 more) 0/0 (0%) 0% RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 24 more per 1000 (from 45 fewer to 235 more) CRITICAL 11 fewer per 1000 (from 18 fewer to 19 more) 2/38 (5.3%) RR 0.46 (0.111.95)
B. Question: Should 6 months of INH vs 12 months of INH prophylaxis be used in people living with HIV with a positive TST status? Settings: High TB/HIV prevalence settings Bibliography: Pape et al. 1993; Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998
Importance
Absolute
Quality assessment
Imprecision 12 months of INH prophylaxis Relative risk (95% CI) Other considerations
None
No. of studies
Serious1
Design
Limitations
Inconsistency
Active TB ( possible, probable, confirmed) (follow up 13 years; clinical assessment, chest X-ray, sputum for AFB) No serious indirectness 28 fewer per 1000 (from 47 fewer to 50 more) VERY LOW CRITICAL
Randomized trials
Very serious
Confirmed TB (follow up 13 years; culture-proven) None 0% 0% 0/0 (0%) 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records and patients files) No serious indirectness 68/655 (10.4%) 10% 3/38 (7.9%) Serious1 None
Randomized trials
Very serious
Very small sample size in the 12 months group
Annex 6
C. Question: Should 6 months of INH vs 12 months of INH prophylaxis be used in TST -negative people living with HIV? Settings: High TB/HIV prevalence settings Bibliography: Whalen et al. 1997; Hawken et al. 1997; Mwinga et al. 1998; Fitzgerald et al. 2001; Gordin et al. 1997; Rivero et al. 2003; Whallen et al. 1997 anergy
No. of patients
12 months of INH prophylaxis
RR 0.65 (0.311.36) 7 fewer per 1000 (from 14 fewer to 7 more) 175 fewer per 1000 (from 345 fewer to 180 more) RR 0 (00) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL
Quality assessment
41/1151 (3.6%) 2% 50% 8/146 (5.5%)

Serious1 None
Design
Limitations
Inconsistency
Imprecision
Importance
Active TB ( possible, probable, confirmed) (follow up 13 years; clinical assessment, chest X-ray, sputum for AFB) No serious indirectness 19 fewer per 1000 (from 38 fewer to 20 more) VERY LOW CRITICAL
Randomized trials
Very serious
Confirmed TB (culture-proven) None 0/0 (0%) 0/0 (0%)
Mortality (any cause) (follow up 13 years; review of hospital records and patients files) No serious indirectness None 307/1151 (26.7%) 10% 50% 24/146 (16.4%) Serious1 RR 1.62 (1.112.37) 102 more per 1000 (from 18 more to 225 more) 62 more per 1000 (from 11 more to 137 more) 310 more per 1000 (from 55 more to 685 more) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL VERY LOW CRITICAL
Randomized trials
Very serious
HIV disease progression None 0/0 (0%) 0/0 (0%)
Sample size in the 12 months-group was much smaller than that in the 6 months group
70

Adults and adolescents living with HIV, who have an unknown or positive TST status and are unlikely to have active TB, should receive at least six months of IPT as part of a comprehensive package of HIV care. This includes individuals irrespective of degree of immunosuppression, those on ART, those who have previously been treated for TB and pregnant women. TST is not a requirement for initiating IPT among people living with HIV. Where feasible, TST can be used as people with a positive test benefit more from IPT than those with a negative test.
Population: Adults and adolescents living with HIV Factor Decision High Intervention: At least six months of IPT (300 mg plus vitamin B6 25 mg daily) Quality of evidence Explanation Provision of IPT reduces active TB incidence by: 33% regardless of TST status (GRADE high) 64% in TST-positive persons (GRADE high) 14% in TST-negative persons (GRADE high) 14% in those with unknown TST status (GRADE moderate) IPT has been shown to be equally efficacious but less toxic than multidrug regimens in preventing active TB disease (GRADE low/moderate) Risks or undesired effects Strong
Reduction in TB morbidity Reduction in TB transmission Reduction in mortality (even if there is no evidence of a direct effect in terms of death GRADE moderate) Potential reduction in generation of MDR and XDR TB by reducing TB incidence Reduction in prevalence of non-INH mono/drug-resistant strains in the community (e.g. mono-RIF) Avoidance of side-effects of TB treatment Reduction in drugdrug interaction in patients on ART and TB treatment Improved TB infection control in health-care and community settings (particularly in HIV clinics) Fewer side-effects than with a preventive combination regimen (GRADE high) Less drugdrug interaction caused by alternative preventive regimens containing RIF Potential increase in adherence to other treatment (ART, co-trimoxazole, etc.)
Longer regimen compared to a combination preventive regimen Potential for development of greater resistance compared to a combination prevention regimen Increase in number of persons receiving unnecessary treatment (especially when TST is not performed or in case of false-positive TST where TST is performed) Potential INH toxicity Potential development and transmission of INH-resistant strains by treating undiagnosed active TB Potential reduction in adherence to other treatment (ART, co-trimoxazole, etc.)
71
Annex 6
Values and preferences
Strong
Costs
Weak
Increased by: Longer provision as compared to other shorter preventive regimens Costs of INH (including storage, supply and transportation) Potential laboratory monitoring (in case of toxicity) Additional staff time in overburdened HIV care settings Second-line TB treatment needed in case INH resistance occurs and is transmitted Treatment of rare INH-related toxicity (blood tests, hospitalization, patients loss of earning) Additional costs of providing vitamin B6 Decreased by: Less expensive regimen as compared to other preventive regimens Avoids treatment costs of active TB Less hospitalization costs Reduction in social costs and loss of earnings Reduction in costs due to treatment of secondary cases including in health-care workers, which may improve staff retention Reduction in costs related to TBART co-treatment which often includes more expensive regimens Reduction in drugdrug interaction from TB treatment in patients on ART Potential reduction in generation of MDR and XDR TB by reducing the incidence of TB
Less pill burden than in other preventive combination regimens Fewer side-effects than in other preventive combination regimens Easier to take treatment Avoids active TB disease, deaths and transmission to other family members Health-care workers would be less exposed to active TB cases and would feel more protected Infection control benefits for patients, family and the community both in terms of preventing TB and also in preventing infection and/or reinfection Avoids potential need for long TB treatment with a high pill burden and significant side-effects Patients would feel protected against TB, adding value to HIV care services BuT Longer regimen RIF would make it easier to monitor adherence May be concerns regarding the side-effects of INH Pill burden Stigma Increase in number of people attending clinics Adds burden to supply chain
72
Feasibility
Strong
Strong
Clear clinical and public health benefit with strong evidence base Less monitoring needed as compared to other preventive regimens Drug easily available and inexpensive Administration of IPT fits well with the schedule of HIV programmes Does not require laboratory monitoring Health-care workers are already familiar with the drug BuT Longer-term intervention when compared to other preventive regimens Would mean adding an intervention to an overburdened system May increase clinic visits and require more staff time Space required for storage and difficulties in procurement INH as a single formulation less commonly available compared to fixed-dose combination (FDC) TB medicines Opinion leaders are reluctant to use IPT in the face of scientific evidence Requires additional training of health-care workers on benefits of IPT and TB screening
Adults and adolescents living with HIV who have successfully completed their TB treatment should be provided secondary INH prophylaxis for at least six months (strong recommendation) or at least 36 months (conditional recommendation)
Population: People living with HIV who have successfully completed TB treatment Intervention: Secondary INH prophylaxis Factor Quality of evidence Decision Moderate Explanation Benefits or desired effects Strong Risks or undesired effects Values and preferences Costs Weak Strong Feasibility Strong
Evidence is available from three randomized trials and one observational study. One randomized trial compared INH with rifampicin after successful TB treatment. Adverse events and mortality were not addressed (critical outcome) (moderate grade of evidence for recurrent TB outcome) Reduction in incidence of recurrent TB Reduction in TB transmission Potential reduction in generation of MDR strains by reducing TB incidence Improved TB infection control in health-care settings by reducing active TB and also preventing new infections or reinfection (particularly in HIV clinics) Pill burden Reduction in compliance with other treatments (ART, cotrimoxazole, etc.) Lack of motivation in patients as they have already been treated for TB Lack of confidence in efficacy of TB drugs
Costs are unlikely to increase as it can easily be incorporated into routine care for people living with HIV Cost-saving implications through preventing development of recurrent TB It will be part of routine care for people living with HIV
Strength of recommendation Strong (at least 6 months) Conditional (at least 36 months)
73
Annex 6
References
1. Fitzgerald DW et al. No effect of isoniazid prophylaxis for purified protein derivative-negative HIV-infected adults living in a country with endemic tuberculosis: results of a randomized trial. Journal of Acquired Immune Deficiency Syndromes, 2001, 28:305307. 2. Gordin FM et al. A controlled trial of isoniazid in persons with anergy and human immunodeficiency virus infection who are at high risk for tuberculosis. Terry Beirn Community Programs for Clinical Research on AIDS. New England Journal of Medicine, 1997, 337:315320. 3. Hawken MP et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS, 1997, 11:875882. 4. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 5. Pape JW et al. Effect of isoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection. Lancet, 1993, 342:268272. 6. Rivero A et al. [Randomized trial of three regimens to prevent tuberculosis in HIV-infected patients with anergy]. Enfermedades infecciosas y microbiologia clinica, 2003, 21:287292. 7. Whalen CC et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. UgandaCase Western Reserve University Research Collaboration. New England Journal of Medicine, 1997, 337:801808. 8. Martinson N et al. Novel regimens for treating latent TB in HIV-infected patient adults in South Africa: a randomized clinical trial. In: Proceedings of the 16th Annual Conference on Retroviruses and Opportunistic Infections, Montreal, Canada, 811 February 2009. 9. Samandari T et al. Preliminary results of the Botswana Isoniazid Preventive Therapy (IPT) Clinical Trial (6 months vs 36 months). Union World Lung Conference, Cancun, on behalf of BOTUSA IPT Study; 2009. 10. Akolo C et al. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database of Systematic Reviews, 2010, (1):CD000171. 11. Churchyard GJ et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17:20632070. 12. Golub JE et al. Isoniazid preventive therapy, HAART and tuberculosis risk in HIV-infected adults in South Africa: a prospective cohort. AIDS, 2009, 23:631636. 13. Golub JE S et al. The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis incidence in HIV-infected patients in Rio de Janeiro, Brazil. AIDS, 2007, 21:14411448. 14. Madhi SA et al. Lack of efficacy of primary isoniazid (INH) prophylaxis in increasing tuberculosis (TB) free survival in HIV-infected (HIV+) South African children. 48th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, 2528 October 2008. 15. Martinez Alfaro EM et al. [Evaluation of 2 tuberculosis chemoprophylaxis regimens in patients infected with human immunodeficiency virus. The GECMEI Group]. Medicina clinica (Barc), 2000, 115:161165. 16. Zar HJ t al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomized controlled trial. British Medical Journal, 2007, 334:136. 17. Mohammed A et al. Randomised controlled trial of isoniazid preventive therapy in South African adults with advanced HIV disease. International Journal of Tuberculosis and Lung Disease, 2007, 11:11141120. 18. Johnson JL et al.; for the UgandaCase Western Reserve University Research Collaboration. Duration of efficacy of treatment of latent tuberculosis infection in HIV-infected adults. AIDS, 2001, 15:21372147. 19. Gordin F et al. Rifampin and pyrazinamide vs isoniazid for prevention of tuberculosis in HIV-Infected persons. An international randomized trial. Journal of the American Medical Association, 2000, 283:14451450. 20. Halsey NA et al. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet, 1998, 351:786792. 21. Rivero A et al. A randomized clinical trial investigating three chemoprophylaxis regimens for latent tuberculosis infection in HIV-infected patients. Enfermedades infecciosas y microbiologia clinica, 2007, 25:305310.
74
6.2. Timing of initiation: Define the optimal time to start considering IPT (i.e. should immune status be considered and should IPT be started with ART?).
PICOT Question: What is the optimal time to start considering IPT (i.e. should immune status 6.2.1 Immune status Population: People living with HIV Intervention: Starting IPT if immune suppression not severe (CD4 count >200 cells/mm3) Comparison: Starting IPT if immune suppression severe (CD4 count <200 cells/mm3) Outcomes: Active TB incidence, mortality, AEs, progression of HIV disease, adherence, TB drug resistance, interval to active TB (Atb), interval to death Timeline: Lifetime Q6.2.2 (a) ART Population: People living with HIV Intervention: Starting IPT and ART at the same time Comparison: Starting IPT first followed by ART Outcomes: Active TB incidence, mortality, AEs, progression of HIV disease, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime
Outcomes Relative importance (rank 1 9 most critical)
be considered and should IPT be started with ART)? Q6.2.2 (b) ART Population: People living with HIV Intervention: Starting IPT and ART at the same time Comparison: Starting ART first followed by IPT Outcomes: Active TB incidence, mortality, AEs, progression of HIV disease, adherence, TB drug resistance, interval to Atb, interval to death Timeline: Lifetime
Interval to death
Interval to active TB
Cost-effectiveness
TB drug resistance
Adherence
Adverse events
Progression of HIV disease
Mortality
Confirmed TB
Active TB incidence (suspected, probable, confirmed) 9 9 9
Critical Critical Critical
Critical
8 7 8
Critical
Critical (addressed by Annex 12) Less critical Less critical
Critical (addressed by Annex 10)
Critical (addressed by Annex 11)
75
Annex 6
2. Literature search and information retrieval

1. Pubmed Search (Tuberculosis[Mesh] AND HIV Infections[Mesh]) AND Therapeutics[Mesh]
1375
articles
Limits Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical
trial, multicentre study, adolescent: 1318 years, adult: 1944 years, middle-aged: 4564 years, middle-aged + aged: 45+ years, aged: 65+ years, 80 and over: 80+ years
219
By title
30
By abstract
19
Of interest
Evidence retrieval findings The Guidelines Group reviewed the available data regarding the initiation of IPT and immune status, including concomitant use with ART. Six studies [16] were examined and showed contrasting results on the reduction of TB risk by immune status. Additional protective benefits of concomitant use of IPT with ART were demonstrated in two [3,4] observational studies from Brazil and South Africa, and a sub-analysis of data from an unpublished randomized clinical trial from Botswana. Based on this evidence and the potential benefit of concomitant use of IPT with ART, the Guidelines Group strongly recommends that IPT be given irrespective of immune status and whether or not the person is on ART. IPT initiation or completion
should not delay commencement of ART in eligible people living with HIV. However, the Guidelines Group recognizes the absence of evidence as to whether concomitant initiation of IPT with ART or delayed initiation of IPT is better in terms of efficacy and toxicity. A large, randomized controlled study (TEMPRANO study) is currently ongoing in Ivory Coast and will provide data on patients randomized to IPT before ART versus simultaneous initiation of IPT and ART. However, relevant studies providing stratifications/ sub-analysis or some kind of guidance in terms of optimal time of initiation have been selected as follows:
76
STuDIES
Mwinga et al. 1998 [5]
Churchyard et al. 2003 [2] Golub et al. 2007 [4] Brazil
Protection against TB was limited to those with total lymphocyte count of 2x109/l or higher RRs for TB stratified by TST and lymphocyte count are most marked in those with a lymphocyte count of 2x109/l or more, and TST reading of 5 mm or more. The rate of TB was 6/56 = 10.71 per 100 person-years in the placebo group and 2/95 = 2.10 per 100 person-years in the INH and RIF+PZA groups combined (RR = 0.19, 95%CI: 0.04 0.94, P = 0.026) No significant difference in the efficacy of IPT by CD4 stratum (above or below 200 cells/mm3) though absolute TB recurrence rates were substantially higher in the group with CD4 count <200 cells/mm3 as one would expect.
BRIEF DESCRIPTION OF RELEVANT FINDINGS
Samandari et al. 2009 [6] Botswana
The subjects included 11 026 HIV-infected patients receiving medical care at 29 public clinics in Rio de Janeiro, Brazil, between 1 September 2003 and 1 September 2005 Data were collected through a retrospective medical records review. The combination of IPT and ART was associated with significant reduction in incidence of TB for patients with both advanced and early HIV disease. In the population of patients with advanced HIV disease (CD4 cell count <350 cells/ mm3), over a period of two years, ART alone was associated with a significantly reduced incidence of TB, whereas IPT alone was not. In patients with a CD4 cell count >350 cells/mm3 at baseline, ART significantly reduced the risk of TB, IPT reduced the risk but not at a statistically significant level, whereas the combination reduced the risk substantially. The overall incidence of TB was 2.28 cases/100 person-years (95%CI: 2.062.52). Among patients who received neither ART nor IPT, the incidence was 4.01/100 personyears. Patients who received ART had a TB incidence of 1.90/100 person-years (95% CI: 1.662.17) and those treated with IPT had a rate of 1.27/100 person-years (95%CI: 0.412.95). The incidence among patients who received ART and IPT was 0.80/100 person-years (95% CI: 0.381.47). Multivariate Cox proportional hazards modelling revealed a 76% reduction in TB risk among patients receiving both ART and IPT (adjusted relative hazard 0.24; P<0.001) after adjusting for age, previous diagnosis of TB, and CD4 cell counts at baseline.
Akolo et al. 2010 [1] Cochrane Review Churchyard et al. 2010 [2]
Cox regression model that included IPT, baseline CD4 count, baseline TST and ART as a continuous time-dependent variable was performed. For each extra day of ART, the risk of TB decreased by 0.3% (P = 0.018). If provided for 300 days, the risk of TB was reduced by 56%. Overall mortality 1.4% per annum with 2% in the first year Mortality if starting ART: First year: 2.3% Second year: 0.6% Third year: 1.1% From sub-analysis of TST result: Only ART benefits TST-negative persons with a 56% reduction in TB incidence. While ART is important for many reasons, it added only slightly to TB reduction in TSTpositive persons receiving continuous IPT. Among the papers reviewed, the single placebo-controlled trial that assessed the effect of INH by stage of HIV/AIDS at baseline found no difference (with AIDS [RR 0.96, 95% CI: 0.79 to 1.17] and without AIDS [RR 1.07, 95% CI: 0.84 to 1.35]). (Gordin et al. 1997) (Personal communication) Observational study: IPT started within three months of starting ART substantially reduces the risk of death within the first year compared to ART alone, adding further support to the benefits of combining IPT with ART.
77
Annex 6

The GRADE profile was not done since none of the available studies directly assessed the question of interest.

HIV-infected adults and children, who are unlikely to have active TB, should receive IPT as soon as possible irrespective of their degree of immunosuppression.
Population: Adults, adolescents and children living with HIV who are on or not on ART Factor Decision Explanation Intervention: IPT (1020 mg/kg for children, 300 mg for adults, plus vitamin B6 25 mg daily) Quality of evidence Moderate Benefits or desired effects Risks or undesired effects Weak Values and preferences Costs Strong Strong No direct evidence available (likely to be available soon from ongoing randomized controlled trials) Indirect evidence supports the concomitant use of IPT and ART as a more effective intervention than the two alone (Golub et al. 2007; Samandari et al. 2010) Reduction in early mortality (first three months)
Patients with advanced disease present additional challenges for diagnosis and potential onset of immune reconstitution inflammatory syndrome (IRIS). The Botswana trial reported a statistically non-significant 1.6-fold increase in hepatitis due to INH in people who concomitantly received ART. Increased pill burden at ART initiation Not likely to increase cost unless poorer adherence compromises the efficacy of first-line regimen inducing failure Additional costs of toxicity Administration fits well with the schedule of HIV programmes.
Feasibility Overall ranking of recommendation
Strong
78
Figure 1. Algorithm for TB screening in adults and adolescents living with HIV in HIV-prevalent and resource-constrained settings
Adults and adolescents living with HIV* Screen for TB with any one of the following: Current cough Fever Weight loss Night sweats No Assess for contraindications to IPT No Give IPT Yes Defer IPT Yes Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Not TB Follow up and consider IPT TB Treat for TB
FOOTNOTES TO ADuLT ALGORITHM

* Every adult and adolescent needs to be evaluated for eligibility for ART. Infection control measures should be prioritized to reduce M. tuberculosis transmission in all settings providing care. Chest radiography can be done if available, but is not required to classify patients into TB and non-TB groups. In high HIVprevalence settings with a high TB prevalence among people living with HIV (e.g. greater than 10%), strong consideration must be given to conducting additional sensitive investigations. Contraindications include: active hepatitis (acute or chronic) or regular and heavy alcohol consumption or symptoms of peripheral neuropathy. Past history of TB and current pregnancy should not be contraindications for starting IPT. Although not a requirement for initiating IPT, TST may be done as a part of eligibility screening in some settings. Investigations for TB should be done in accordance with existing national guidelines.
References
1. Akolo C et al. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database of Systematic Reviews, 2010, (1):CD000171. 2. Churchyard GJ et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17:20632070. 3. Golub JE et al. Isoniazid preventive therapy, HAART and tuberculosis risk in HIV-infected adults in South Africa: a prospective cohort. AIDS, 2009, 23:631636. 4. Golub JE et al. The impact of antiretroviral therapy and isoniazid preventive therapy on tuberculosis incidence in HIV-infected patients in Rio de Janeiro, Brazil. AIDS, 2007, 21:14411448. 5. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 6. Samandari T et al.; on behalf of BOTUSA IPT study. Preliminary results of the Botswana IPT Trial: 36 months vs. 6 months isoniazid for TB prevention in HIV-infected adults. 40th Union World Lung Conference, Cancun, 2009.
79

secondary prophylaxis
PICOT Question: Should secondary treatment of LTBI be provided to prevent recurrence of TB among Population: Adults living with HIV who have received TB treatment Intervention: IPT (612 months INH) or any regimen Comparison: No IPT or any other regimen immediately after TB treatment people living with HIV who had received TB treatment in the past? Outcomes: Recurrent TB incidence, mortality, progression of HIV disease, AEs, adherence, interval to recurrent TB, interval to death Timeline: Lifetime
Following the GRADE approach, the key outcomes that needed to be considered in making recommendations were chosen and ranked in order to identify the data that had to be sought from the process of evidence retrieval.
Outcomes
Recurrent TB (suspected, probable, confirmed) Recurrent confirmed TB Mortality

9
Comment
Critical Critical
Progression of HIV disease Adverse events Interval to active TB Interval to death
Critical
Critical
Critical
Less critical
Less critical
80
Annex 7: Summary of findings and quality of evidence evaluation: secondary prophylaxis

2.1 Google Scholar and PubMed were used with the following search strategy (Tuberculosis[Mesh] AND HIV Infections[Mesh]) AND ([TB treatment in the past{Mesh}] or ([Recurrence{Mesh}] ([Relapse{Mesh}] or ([Reinfection{Mesh}] ([secondary treatment{Mesh}]) or or
1456
articles
Limits Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial,
multicentre study, adolescent: 1318 years, adult: 1944 years, middle-aged: 4564 years, middle aged + aged: 45+ years, aged: 65+ years, 80 and over: 80+ years
72
By title
57
By abstract
Of interest
he studies of interest include those that compared people living with HIV who had undergone previous treatment for TB and were put on IPT for a duration of six to 12 months with those who did not receive IPT. The Guidelines Group reviewed the evidence and discussed IPT as secondary prophylaxis for people who had previously been successfully treated for TB. GRADE evidence from four studies [14] including three randomized controlled trials and one observational study showed the value of providing IPT immediately after successful completion
of TB treatment. The WHO Guidelines Group strongly recommends that adults and adolescents living with HIV who have successfully completed their TB treatment should continue receiving INH for six months and should conditionally receive it for 36 months based on the local situation and existing national guidelines. No evidence was available on the potential role of IPT for a patient who had successfully completed treatment for MDR or XDR TB. Therefore, the Guidelines Group did not make any recommendation on the use of IPT after successful treatment for MDR or XDR TB.
81
Annex 7

Perriens et al. 1995 Methods/design Population Randomized trial Kinshasa, Zaire. 335 HIV-positive persons with TB => 260 completed TB treatment => 240 enrolled in further prophylaxis study Haller et al. 1999 Randomized trial 280 HIV-positive persons, >15 years old, completed TB treatment, two years follow up (192 person years treatment vs 142 person-years Fitzgerald et al. 2000 Churchyard et al. 2003
Randomized trial Port au Prince, Haiti, >18 years old, HIVpositive, diagnosed and treated for first TB episode, given prophylaxis or placebo after that. Mean follow up 25 months Of 354 with TB => 274 completed treatment => 233 randomized. Of 142 HIV-positive persons => 68 assigned IPT Of 142 => 74 were given placebo
Observational study South Africa, HIVpositive gold miners with documented successful completion of TB treatment
Intervention
Of 260 who 263 eligible HIVcompleted TB positive persons treatment => 121 RIF => 134 INH + co+ INH prophylaxis trimoxazole
338 had received secondary preventive therapy
Comparison
Treatment regimen
Of 260 who completed TB treatment => 119 were given placebo
Of 263=> 129 patients INH, 300 mg once daily plus sulphadoxine pyrimethamine (S, 500 mg/P, 25 mg once weekly), up to 24 months No treatment
221 had no secondary preventive therapy INH 300 mg/daily Co-trimoxazole prophylaxis if their CD4 T cell count was <250x106 cells/l and if symptomatic or <200x106 cells/l regardless of symptoms Recurrent TB Incidence rates 8.6 (INH group) and 19.1 (no treatment group) per 100 personyears; incidence rate ratio, 0.45; 95% confidence interval 0.260.78 55% reduction in prophylaxis group
Six months, twice weekly RIF 600 mg/450 mg, INH 15 mg/kg Placebo
12-month INH 300 mg + vitamin B6 40 mg Vitamin B6 40 mg
Outcome Information on recurrence
Recurrent TB 1/73 had relapse in prophylaxis group (48 lost to follow up) vs 9/83 who relapsed in placebo group (36 lost to follow up)
Recurrent TB 4/134 in prophylaxis group vs 10/129 in control group. The incidence of TB was 2.1 (0.63) observations/100 person-years in the prophylaxis group and 7.0 (3.412.6) observations/100 person-years in the control group (relative risk 0.30 [0.090.94]).
Recurrent TB IPT patients 1.4/100 person-years vs placebo 7.8/100 person-years (relative risk 0.18 [0.040.83], P=0.010)
82
Perriens et al. 1995 Comment HIV-seropositive patients followed for relapse; there was no difference in survival between those assigned to RIF plus INH and those assigned to placebo (P = 0.95). The life-table estimate of relapse rate 18 months after completion of sixmonth treatment was 1.9% for HIV-positive prophylaxis patients lower than the rate of 9% in HIV-positive placebo patients (P<0.01). At 24 months, the HIV-seropositive patients who received extended treatment had a relapse rate of 1.9%, as compared with 9% among the HIV-seropositive patients who received placebo for the second six months (P<0.01). Extended treatment did not improve survival, however.
Haller et al. 1999
Fitzgerald et al. 2000
Churchyard et al. 2003
The TB recurrence Limited population rate was significantly miners lower among the HIV1-positive patients receiving INH than among the HIV-1positive patients receiving placebo.
83
Annex 7
Question 5: Should secondary treatment for LTBI be used for people living with HIV who had received TB treatment in the past to prevent recurrence of TB? Settings: High HIV/TB burden Bibliography: Churchyard et al. (observational) 2003; Perriens et al. 1995; Haller et al. 1999; Fitzgerald et al. 2000
No. of patients
Quality assessment
Indirectness Imprecision Other considerations
Strong association 28/338 (8.3%) 23/221 (10.4%) RR 0.45 (0.260.78) 57 fewer per 1000 (from 23 fewer to 77 fewer)
No. of studies
Serious No serious imprecision
Design
Limitations
Inconsistency
Secondary treatment of LTBI
Importance
TB recurrence (observational) (follow up 0.91 versus 0.41 patient-years; isoniazid vs co-trimoxazole) VERY LOW
Observational study
Serious
TB recurrence (randomized) No serious indirectness1 Serious2 7/275 (2.5%) 31/286 (10.8%) RR 0.23 (0.110.52) 83 fewer per 1000 (from 52 fewer to 96 fewer) None MODERATE
Randomized trials
The study by Perriens et al. 1995 provided INH+ RIF for six months instead of INH Small numbers
84

People living with HIV who have successfully completed their TB treatment should be provided secondary INH prophylaxis.
Population: People living with HIV who have successfully completed TB treatment Intervention: Secondary INH prophylaxis Factor Quality of evidence Decision Moderate Explanation Benefits or desired effects Strong Risks or undesired effects Values and preferences Costs Weak Strong Feasibility Strong
Only three randomized trials and one observational study. One randomized trial compared provision of INH with rifampicin after successful TB treatment. Adverse events and mortality were not addressed (critical outcome). Moderate grade of evidence for recurrent TB outcome
Reduced recurrent TB incidence Reduced TB transmission and improved TB infection control in health-care settings (particularly in HIV clinics) Reduced chances of reinfection Potential reduction in generation of MDR strains by reducing TB incidence Pill burden Reduced adherence to other treatment (ART, cotrimoxazole, etc.)
Lack of motivation among patients as they had just been treated for TB Lack of familiarity with and confidence in efficacy of IPT Costs are unlikely to increase as INH can easily be incorporated into routine care for people living with HIV. Cost-saving implications through prevention of new infections and development of recurrent TB It will be part of routine care for people living with HIV.
References
1. Churchyard GJ et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17:20632070. 2. Fitzgerald DW et al. Effect of post-treatment isoniazid on prevention of recurrent tuberculosis in HIV-1-infected individuals: a randomised trial. Lancet, 2000, 356:14701474. 3. Haller L et al. Isoniazid plus sulphadoxinepyrimethamine can reduce morbidity of HIV-positive patients treated for tuberculosis in Africa: a controlled clinical trial. Chemotherapy, 1999, 45:452465. 4. Perriens JH et al. Pulmonary tuberculosis in HIV-infected patients in Zaire. A controlled trial of treatment for either 6 or 12 months. New England Journal of Medicine, 1995, 332:779784.
85
Annex 8: Summary of findings and quality of evidence: tuberculin

skin test (TST)
PICOT Question: Is it feasible to perform TST for all people living with HIV before treatment of LTBI in resource-poor settings? Population: Adults and adolescents living with HIV Intervention: Feasibility of using TST in resourceconstrained settings Comparison: Using IPT without TST Outcomes: Access to IPT vs minimal or no access Timeline: 540 years
challenges for programmes in resource-poor settings. We evaluated the evidence regarding the feasibility of using TST testing before providing IPT for LTBI in resource-constrained settings.
tudies to date indicate that people living with HIV who are TST-positive are most likely to benefit from preventive therapy. [1] However, TST implementation poses some
Outcomes
Cost associated with TST Failure to return for results

9
Comment
Critical Critical
Stability of purified protein derivative (PPD) and need for cold chain
9 9 9
Measurement issues and interpersonal variability Access to TST

he key feasibility issues assessed were the following: PubMed and Embase were considered as the principal databases. Additional studies were identified by searching reference lists of primary studies and review articles. Web sites and package inserts were also checked for relevant information. Additionally, experts in the field of HIV/TB were consulted for relevant articles. Although each of above issues was not separately searched for, each one was included in the overall review of each possible study/article included.
Cold chain/stability of tuberculin Production quality of tuberculin Price/costs associated with TST Measurement issues/inter-reader variability Loss to follow up/failure to return for TST result Anergy
86
Annex 8: Summary of findings and quality of evidence: tuberculin skin test (TST)
3. Findings
3.1 Search terms
(((Tuberculin[MeSH terms] OR Tuberculin[all fields]) OR TST (all fields) OR PPD[all fields]) AND (hiv[MeSH terms] OR HIV[all fields]) AND (TB[all fields] AND (prevention and control[subheading] OR (prevention[all fields] AND control[all fields]) OR prevention and control[all fields] OR (preventive[all fields] AND therapy[all fields]) OR preventive therapy[all fields])) PPD and cold chain, PPD and cost OR price, tuberculin and stability.
Reports were identified:
references
184
By reading titles:
references
85
By reading abstract:
references
45
By reading paper:
references
29
Articles/ reports of interest:
references
23
87
tudies/articles/reports included were those that evaluated the technical aspects of TST in the general population and HIV-infected
individuals, and were conducted either in highincome or low-income countries.
Annex 8
3.2 Stability of PPD and need for cold chain
ne of the important conditions for maintaining the potency of tuberculin is the need for it to be transported and stored at the appropriate temperature. Preparations of PPD are stable until the indicated date of expiry, if they are kept continuously cold between +2C and +8C and protected from light.[2,3] PPD solution can be adversely affected by exposure to light; therefore, the product should be stored in the dark except when doses are actually being withdrawn from the vial.[2,3] Failure to store and handle PPD as recommended will result in a loss of potency and inaccurate test results.[4] Some studies suggest the possibility of PPD remaining stable even at temperatures outside those recommended.[57]
Another report from India showed that on comparing the mean values of PPD test results obtained by using vials at room temperature with those of refrigerated vials, the differences were not significant although the data from this study pertain to non-refrigeration up to 96 hours only and require further validation under rigorous control conditions.[5] Magnus et al. reported that PPD dilutions can be stored at 20C for some months without significant loss of potency. [7] Although it is possible that PPD solutions can remain stable for at least one year even at 37C, [5,6] performance may still be jeopardized and the recommended temperatures for storage should still be used.
3.3 Costs associated with TST

Company A B C D E F Dose 1 1 1 1 1 1
TST/PPD cost
4.7 8.9 6.1 49 8.7 5.8
Cost/dose (uS$)
Cost/100 000 doses (uS$)* 470 000 890 000 610 000 490 000 870 000 580 000
Price range/100 000 doses: US$ 470 000890 000 *Excluding shipping Type of shipment (shipping takes 510 days) Courier Air freight Air freight
Company G G F
Example: Cost of shipment to Kenya

85 177 125
Cost/1000 doses (uS$)
Cost/100 000 doses (uS$) 8 500 17 700 12 500
uberculin testing costs less than US$ 10 per person,[8,9] but the prevalence of true positive results is relatively low, and screening may cost up to US$ 4500 per person eligible for treatment with isoniazid, and US$ 350 000 per case of TB prevented,[8,9] given the added cost associated with the investigation of false-positive reactions,
NB: Refrigerated item incurs additional shipping charge based on shipping zone.
as well as the costs of treatment and follow up.[10] Though most studies did not state clearly how much cost is associated with TST, the test may also add more costs in terms of the increased workload on personnel, need for patients to return for results to be read and cost of consumables.
88
3.4 Failure to return

field experience in the care of HIV-infected patients may be the best predictor of physician adherence with a recommended TST procedure.[15] However, the feasibility of having to retrain a large population of health-care providers in resource-poor settings on the use of TST also needs to be considered. It has been reported that the need for subjects to be seen a second time so that test results can be read limit the usefulness of TST.[16] Therefore, while useful, there are considerable feasibility challenges around the need for follow up. In addition, accurate interpretation of the results is occasionally not possible, particularly among people with HIV and low CD4 counts.[17]
he use of TST in identifying patients who might benefit from ITP adds further time and expense, and is an additional step where patients are lost to follow up.[11] Most of the included studies showed that about 1121% of patients who had TST did not return for their results to be read. [1214] The burden of having to return to the healthcare facility after 4872 hours of undergoing TST is particularly a great challenge in many resource-poor settings. However, the study by Aisu et al. suggested that the rate of return for TST to be read increased after HIV counsellors were retrained in HIV and TB.[12] This finding may support the suggestion that
3.5 Measurement issues and interpersonal variability

variability.[1821] Evidence that supports other modes of measuring the response is also available, such as the use of flexible calipers.[20] However, measurement methods are imprecise and a number of factors including the time that the results are read after placement of the test may affect the results. The feasibility of training large numbers of health-care workers to place and accurately read TSTs on a large scale is unknown.
actors such as variability between and within readers, need for trained personnel to administer and read the test, and the need for the subjects to be seen a second time so that test results can be read all impact the usefulness of TST.[16] With regard to interpersonal variability in reading the TST, the available evidence shows that groups of health-care workers can be trained to accurately interpret TSTs with a minimum of
B T
3.6 Access to TST

available.[16] Reports from South Africa also found difficulty in keeping tuberculin for the test in stock. [22] However, it is possible that manufacturers could scale up production to meet the demand but that would largely depend on market forces.
efore TST can be used to screen for LTBI, it must first be made available to programmes providing HIV prevention, care and treatment services. The global supply may not be sufficient to meet the demand. According to a report from India, good-quality tuberculin in various strengths is seldom
3.7 Anergy
been recommended that other cutaneous allergens be used to differentiate between the presence of nonreactivity to tuberculin due to an absence of infection or due to anergy.[26] However, anergy testing is not an ideal field test as it requires a repeat visit 4872 hours after placement of the test and handling injectable materials (syringe, cold chain, etc.), as well as a cadre of trained personnel.[27] A study by Garcia et al. reported that the use of tetanus toxoid and Candida antigen were not useful in differentiating between the absence of reactivity to tuberculin caused by HIVassociated immunodeficiency and that due to absence of latent M. tuberculosis infection.[28] Tuberculin and anergy skin testing have a low predictive value
ST has been used for decades to detect LTBI, but is not entirely reliable due to its low specificity and sensitivity.[23] This is particularly important in persons with LTBI and immunosuppression. Because of their depressed cell-mediated immunity, people living with HIV may not be able to mount a response to TST. Thus, a negative TST in a person living with HIV may be due to true absence of infection with Mycobacterium tuberculosis or an inability to mount a response due to immunosuppression.[24] Anergy is defined as absence of induration in response to three antigens (PPD, Candida albicans and parotiditis antigen) applied by the Mantoux method.[25] It has
89
Annex 8
in detecting M. tuberculosis infection in HIV-infected persons, and therefore such testing has a limited role in identifying HIV-infected persons who may benefit from preventive therapy programmes for TB.[29] The usefulness of anergy testing has been questioned and
its use to guide administration of TB chemoprophylaxis has been discontinued in the United States.[30] Until well-designed studies of anergy testing as an adjunct to tuberculin testing are conducted, it is safest to discourage anergy testing.

TST is not a requirement for initiating IPT for people living with HIV. Where feasible, TST can be used as people with a positive test result benefit more from IPT than those with a negative test result.
Population: People living with HIV Intervention: TST Factor Decision High Explanation
Quality of evidence
Risks or undesired effects Values and preferences
Weak
Benefits: Early identification of LTBI Increase in numbers of people diagnosed with LTBI Minimizes number to be exposed to INH Correctly identifies people who will benefit from IPT
Provision of IPT reduces active TB incidence by: 33% (GRADE high) regardless of TST status 64% in TST-positive persons (GRADE high) 14% in TST-negative persons (GRADE high) 14% in those with unknown TST results (GRADE moderate) Therefore, TST is useful in identifying the population who would benefit most from IPT.
Strong
Risk of adverse reaction to PPD Risk of false-negative results (HIV anergy, reader error, poor product), false-positive results, interaction with BCG) Patients may not want to have TST Concerns regarding returning for results TST not a form of treatment Risk of adverse effects
Costs
Strong
Feasibility
Strong
Reduced by: Increased demand for TST may lower cost of PPD Reduction in the number of patients on INH (related toxicity management and follow up) Increased by: Procurement of PPD/need for storage Training of workers Time spent by health-care worker for test Travel cost for follow-up visits by the patient Cost associated with further screening for TB Not many studies in resource-poor settings examined the major issues Need for refrigeration and cold chain maintenance Considerable loss to follow up Need for training of health-care workers Increased workload of health-care workers Need for more training on the use of TST in clinical and public health practice
Strong
But: Cost of PPD could be offset by the TB cases prevented PPD could remain stable at room temperature
90
References
1. Akolo C et al. Treatment of latent tuberculosis infection in HIV infected persons. Cochrane Database of Systematic Reviews, 2010, (1):CD000171. 2. Landi S, Held HR. Effect of light on tuberculin purified protein derivative solutions. American Review of Respiratory Disease, 1975, 111:5261. 3. Medsafe. Tuberculin purified protein derivative (Mantoux) diagnostic antigen. 2007. Available from: http://www. medsafe.govt.nz/profs/datasheet/t/Tubersolinj.htm (accessed on 12 November 2010). 4. Landi S, Held HR. Stability of a dilute solution of tuberculin purified protein derivative at extreme temperatures. Journal of Biological Standardization, 1981, 9:195199. 5. Ladha A. Effect of refrigeration on Mantoux test result. Indian Pediatrics, 1995, 32:1036. 6. Landi S, Held HR. Stability of dilute solutions of tuberculin purified protein derivative. Tubercle, 1978, 59:121133. 7. Magnus K et al. Stability of purified tuberculin in high dilution. Bulletin of the World Health Organization, 1958, 19:765782. 8. Price LE, Rutala WA, Samsa GP. Tuberculosis in hospital personnel. Infection Control, 1987, 8:97101. 9. Raad I et al. Annual tuberculin skin testing of employees at a university hospital: a costbenefit analysis. Infection Control and Hospital Epidemiology, 1989, 10:465469. 10. Menzies D et al. Tuberculosis among health care workers. New England Journal of Medicine, 1995, 332:9298. 11. Hawken MP, Muhindi DW. Tuberculosis preventive therapy in HIV-infected persons: feasibility issues in developing countries. International Journal of Tuberculosis and Lung Disease, 1999, 3:646650. 12. Aisu T et al. Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS, 1995, 9:267273. 13. Antonucci G et al. Guidelines of tuberculosis preventive therapy for HIV-infected persons: a prospective, multicentre study. GISTA (Gruppo Italiano di Studio Tubercolosi e AIDS). European Respiratory Journal, 2001, 18:369375. 14. Sackoff JE et alB. Purified protein derivative testing and tuberculosis preventive therapy for HIV-infected patients in New York City. AIDS, 1998, 12:20172023. 15. DeRiemer K, Daley CL, Reingold AL. Preventing tuberculosis among HIV-infected persons: a survey of physicians knowledge and practices. Preventive Medicine, 1999, 28:437444. 16. Mohan A, Sharma SK. In search of a diagnostic test for tuberculosis infection: where do we stand? Indian Journal of Chest Disease and Allied Sciences, 2006, 48:56. 17. Syed J. TB diagnostics: a crisis for people living with HIV/AIDS worldwide. 2006. Available from: http://www.thebody. com/content/art1760.html (accessed on 12 November 2010). 18. Carter ER, Lee CM. Interpretation of the tuberculin skin test reaction by pediatric providers. Pediatric Infectious Disease Journal, 2002, 21:200203. 19. Ozuah PO et al. Assessing the validity of tuberculin skin test readings by trained professionals and patients. Chest, 1999, 116:104106. 20. Perez-Stable EJ, Slutkin G. A demonstration of lack of variability among six tuberculin skin test readers. American Journal of Public Health, 1985, 75:13411343. 21. Villarino ME et al. Comparable specificity of 2 commercial tuberculin reagents in persons at low risk for tuberculous infection. Journal of the American Medical Association, 1999, 281:169171. 22. Smart T. Continuous isoniazid preventive therapy (IPT) superior to short course in TST-positive HIV patients but only in those with a positive tuberculin skin test (TST). HIV & AIDS Treatment in Practice (HATIP), 2009, 151. Available from: http://www.stoptb.org/wg/tb_hiv/assets/documents/hatip151.pdf (accessed on 13 November 2010). 23. Farhat M, Greenaway C, Pai M, Menzies D. False-positive tuberculin skin tests: what is the absolute effect of BCG and non-tuberculous mycobacteria? International Journal of Tuberculosis and Lung Disease, 2006, 10:11921204. 24. Markowitz N et al. Tuberculin and anergy testing in HIV-seropositive and HIV-seronegative persons. Pulmonary Complications of HIV Infection Study Group. Annals of Internal Medicine, 1993, 119:185193. 25. Rivero A et al. [Randomized trial of three regimens to prevent tuberculosis in HIV-infected patients with anergy]. Enfermedades Infecciosas y Microbiologa Clnica, 2003, 21:287292. 26. Arnadottir T et al. Guidelines for conducting tuberculin skin test surveys in high prevalence countries. Tubercle and Lung Disease, 1996, 77 (Suppl 1):119. 27. Jones-Lopez EC et al. Delayed-type hypersensitivity skin test reactivity and survival in HIV-infected patients in Uganda: should anergy be a criterion to start antiretroviral therapy in low-income countries? American Journal of Tropical Medicine and Hygiene, 2006, 74:154161. 28. Garcia-Garcia ML et al. Underestimation of Mycobacterium tuberculosis infection in HIV-infected subjects using reactivity to tuberculin and anergy panel. International Journal of Epidemiology, 2000, 29:369375. 29. Yanai H et al. Utility of tuberculin and anergy skin testing in predicting tuberculosis infection in human immunodeficiency virus-infected persons in Thailand. International Journal of Tuberculosis and Lung Disease, 1997, 1:427434. 30. Centers for Disease Control and Prevention. Prevention and treatment of tuberculosis among patients infected with human immunodeficiency virus: principles of therapy and revised recommendations. Morbidity and Mortality Weekly Report, 1998, 47(RR-20):3641.
91

interferon gamma release assays (IGRA)
PICOT Question: What is the role of IGRA in identifying adults and children who would benefit from treatment for LTBI or active TB? Population: Adults and children living with HIV Intervention: Use of IGRA in identifying adults and children who would benefit from treatment for LTBI or active TB Comparison: No identification of LTBI; TST Outcomes: Negative predictive value, sensitivity, specificity, positive predictive value Timeline: One week during work-up to exclude TB before starting IPT
Outcomes
Predictive value for identification of persons at risk for developing TB

9
Comment
Critical Critical Important Important Important
Specificity for identification of persons at 9 risk for developing TB Sensitivity for identification of persons with active TB
Specificity for identification of persons at 4 low risk for TB
Degree of agreement between results of 5 IGRA and results of TST

Search criteria PubMed Search (Tuberculosis[Mesh]) AND HIV[Mesh]) AND Diagnosis[Mesh]) OR (IGRA, interferon gamma release assay, Quantiferon gamma Gold In-Tube, T-SPOT, ELISPOT, T-cell based assay, whole blood Quantiferon gamma release assay, TB, HIV). Leading researchers in the area were also contacted to provide any unpublished data fulfilling the search criteria.
92
Annex 9: Summary of findings and quality of evidence evaluation: interferon gamma release assays (IGRA)
articles
684
Limits Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase
IV, comparative study, controlled clinical trial, multicentre study.
225
By title
68
By abstract
58
Of interest
28
Studies of interest Twenty-six abstracts/articles [126] provided information about sensitivity/specificity of IGRA among the HIV-infected population or performance of IGRA in comparison to TST. Two articles [27,28] specifically provided information on the cost and feasibility of IGRA. Studies evaluating IGRA that are commercially available (T-SPOT, Quantiferon Gold In Tube [QFN G IT]) and their pre-commercial predecessors (ELISPOT based on antigens included in T-SPOT) were included but assays based on alternate antigens were eliminated.
Information was specifically abstracted on: predictive value of IGRA among persons who developed TB (incident TB) and persons with HIV and prevalent active TB (including information about the performance of IGRA in persons with active TB disaggregated by category of CD4 cell count, if available). We also extracted results of IGRA and TST among persons with HIV and no evidence of active TB, including information on the numbers of patients with positive, negative or indeterminate results of each test, and reported kappa values for the degree of agreement between TST and IGRA.
hree studies [2,6,8] considered the ability of IGRA to predict development of TB over time (IGRA Table 1.1). The results of one of these studies [8] were disregarded because only a CD4 cell count-corrected measure of interferon-gamma (IFN-) release was reported. One study [2] reported that among 822 HIV-positive persons in Austria without active TB who were tested with QFN G IT (results: 37 positive, 738 negative, 47 indeterminate), three developed active TB during follow up, all three

of whom had a positive QFN G IT test at baseline. The other study [6] followed 20 people living with HIV with no respiratory symptoms, but with positive ELISPOT; two developed active TB at three and 10 months of follow up. Eight studies [13,13,17,23,25,26] evaluated the performance of QFN G IT assay among adults with HIV in whom prevalent active TB was confirmed by a combination of signs, symptoms, AFB smear and
93
Annex 9
mycobacterial culture (IGRA Table 2.1). All studies were in settings with high TB and HIV prevalence. The reported sensitivity of QFN G IT ranged from 30% to 83% with a pooled sensitivity of 0.63 (95% CI: 0.570.68). The number of persons included in most studies was small (range 8105), and the total number of persons evaluated was 339. One study [22] evaluated the sensitivity of QFN G IT among HIV-infected children with active TB. Among 36 children with HIV and active TB, 17 had positive QFN G IT tests; the reported sensitivity for QFN G IT was 0.47 (0.310.64). Five studies [5,6,9,13,18] reported the sensitivity of a T-SPOT assay among HIV-infected adults in whom active TB was confirmed by a combination of signs, symptoms, AFB smear and mycobacterial culture (IGRA Table 2.2), although one of these [9] reported only disaggregated sensitivity for subgroups with lower and higher CD4 cell counts and was not included in the pooled estimate. In the remaining four studies [5,6,13,18] reported sensitivity ranged from 0.89 to 0.90 and the pooled sensitivity was 0.90 (0.840.96). The total number of persons included in these studies was 109. Two studies [7,14] reported the sensitivity of T-SPOT among HIV-infected children with active TB, confirmed by a combination of signs, symptoms, AFB smear and mycobacterial culture (IGRA Table 2.2). The pooled sensitivity was 0.69 (95% CI: 0.570.82). The total number of children studied was 52. One study [13] reported on rates of positive IGRA among 10 individuals considered to be at low risk for TB (IGRA Table 3). The number of persons with positive tests were 0 (QFN G IT) and one (with T-SPOT). Ten studies [4,10,12,13,15,16,19,20,22,24] compared the results of QFN G IT with those of TST in HIV-infected persons without evidence of active TB (IGRA Table 4).
In five studies [4,9,12,16,22], the number of persons with positive QFN G IT tests was greater than the number with positive TST and, in the remaining five [10,15,19,20,22], the number of persons with positive TST was greater than the number with positive QFN G IT. Eight studies (IGRA Table 4) compared the results of T-SPOT or ELISPOT with those of TST [9,11,13,16,19,20,22,24]; in six, a larger proportion of patients had positive T-SPOTs [9,11,13,16,19,22], in two, a larger proportion of patients had positive TST. [20,24] In addition to comparing the proportions of persons positive based on each type of test, many studies reported kappa values for correlation between tests (IGRA Table 4). Reported kappa values varied widely; from 0.02 to 0.43 for agreement between T-SPOT and TST, and between 0.23 and 0.59 for agreement between QFN G IT and TST. These values would be considered by some to be consistent with a range between no and moderate agreement. The performance of QFN G IT was compromised among HIV-infected people compared with HIVnegative persons. Rates of indeterminate test results for QFN G IT were significantly higher in persons with HIV compared with persons without HIV, and in persons with low CD4 cell counts compared with persons with higher CD4 cell counts. QFN G IT sensitivity was significantly reduced among patients with low CD4 counts. While most studies found no impact of low CD4 count on T-SPOT sensitivity, at least one study [24] found that low CD4 count was associated with an increased risk of indeterminate test results. Importantly, because there are few longitudinal studies [2,6,8] and no accurate gold standard for LTBI, it is difficult to assess the trade-offs that would occur between sensitivity and specificity using different levels of IFN- gamma release to define positive tests.
IGRA Table 1.1: Predictive value of IGRA among persons with HIV who developed TB during follow up (incident TB)
GRADE table
Reference Test Design Limitations Consistency Directness Precision Overall quality Total no. with positive results followed up 36 20 Predictive value No. of persons with active TB in whom test was + ve 0.8 (0.020.21) 0.1 (0.010.31) 3 2 - ve 0 0 indeterminate 0 0
1 study 1 study
QFN G IT T-SPOT
Observational Observational
1 1
1 1
1 1
1 1
Very low Very low
94
Summary table
Reference Test Design Limitations Consistency Directness Precision Total no. with positive results followed up 36 20 Predictive value 0.8 (0.020.21) 0.1 (0.010.31) No. of persons with active TB in whom test was + ve 3 2 - ve 0 0 indeterminate 0 0
Aichelburg et al. Clark et al.
QFN G IT T-SPOT
10 unclear 10 unclear
1 1
1 1
Aichelburg et al.: 18 months of follow up (radiology, culture for TB confirmation) Clark et al.: median follow up of 12 months
IGRA Table 2.1: Sensitivity of IGRA among persons with HIV and active TB at baseline (prevalent TB)
GRADE table for QFN G IT
Reference Test Design Limitations Consistency Directness Precision Overall quality Very low Very low N (with active TB) 339 36 Sensitivity (CI) 0.63 (0.570.68) 0.47 (0.310.64) No. of persons with active TB in whom test was + ve 8 studies (adults) 1 study (children) QFN G IT QFN G IT Observational Observational 1 1 1 1 OK 1 1 1 212 17 - ve 65 10 indeterminate 62 9
Summary table for QFN G IT

Reference Test Design Limitations (QUADAS) 4 and 10 10 not clear 10 not clear 10 not clear 10 not clear 10 not clear 10 not clear 10 not clear None 4,10 not clear Consistency Directness Precision N (with active TB) 68 8 12 12 105 19 59 26 30 36 Sensitivity (CI) 0.65 (0.530.76) 0.88 (0.641) 0.58 (0.30.9) 0.83 (0.621) 0.65 (0.560.74) 0.68 (0.470.89) 0.63 (0.50.78) 0.65 (0.460.81) 0.30 (0.140.46) 0.47 (0.310.64) No. of persons with active TB in whom test was + ve Aabye et al. Aichelburg et al. Baba et al. (confirmed) Baba et al. (probable) Kabeer et al. Leidl et al. Raby et al. Tsiouris et al. Veldsman et al. Stavri et al. (children) QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT QFN G IT Observational Observational Observational Observational Observational Observational Observational Observational Observational Observational OK 1 OK pleural OK pleural OK OK OK OK OK 1 OK 1 1 1 OK 1 OK 1 1 1 44 7 7 10 68 13 37 17 9 17 - ve 9 (13%) 1 (12%) 1 (6%) 0 19 (18%) 6 10 (17%) 4 15 10 indeterminate 15 (22%) 0 4 (33%) 2 (17%) 18 (17%) 0 12 (20%) 5 6 9
QUADAS Quality Assessment of Diagnostic Accuracy Studies
IGRA Table 2.2: Sensitivity of IGRA among persons with HIV with active TB at baseline (prevalent TB)
GRADE table for T-SPOT
Reference Test Design Limitations Consistency Directness Precision Overall quality Very low Very low N (with active TB) 109 52 Sensitivity (CI) 0.9 (0.840.96) 0.69 (0.570.82) No. of persons with active TB in whom test was + ve 4 (adults) * 2 (children) T-SPOT T-SPOT Observational Observational 1 OK 1 1 OK OK 1 1 98 36 - ve 7 15 indeterminate 4 1
95
Annex 9
Summary table for T-SPOT

Reference Test Design Limitations Consistency Directness Precision N (with active TB) 30 39 18* Sensitivity (CI) No. of persons with active TB in whom test was + ve Clark et al. Chapman et al. Jiang et al.* T-SPOT T-SPOT T-SPOT Observational Observational 10 not clear 10 No 1 OK OK 1 ? 1 0.90 (0.791) 0.9 (0.800.99) 66.7 (CD4 <200) 85.7 (CD4 200) Leidl et al. Rangaka et al. Davies et al. (children) Liebeschuetz et al.(children) T-SPOT T-SPOT T-SPOT T-SPOT Observational Observational Observational 10 not clear None OK OK OK OK OK OK 1 1 1 19 21 22 30 0.89 (0.751) 0.9 (0.781) 0.67 (0.430.85) 0.73 (0.540.88) 17 19 14 22 0 2 7 8 2 0 1 0 27 35 - ve 1 4 indeterminate 2 0
*Overall sensitivity/raw numbers not available for one study (Jiang et al.)
IGRA Table 3: Specificity of IGRA in HIV-positive persons at low risk for TB

Summary GRADE table
Reference Test Design Limitations Consistency Directness Precision Overall quality Very low Very low No. with no risk for TB 10 10 Sensitivity No. of persons with active TB in whom test was + ve 0 0.9 0 1 - ve 10 9 indeterminate 0 0
Leidl et al. Leidl et al.
QFN G IT T-SPOT
1 1
1? 1?
1 1
IGRA Table 4: Agreement between results of IGRA and TST

Study N Setting HIV/TB burden High High High High High Low Low Low Low Low Low Low High 12 52.2 0 0.0 TSPOT +ve (N) 46 125 59 13 38 4 66 14 TSPOT +ve (%) 67.6 50.6 54.1 65.0 52 3.4 24.0 4.2 TSPOT indeter (N) 0 0 4 2 1 0 8 5 3.7 10.0 1 0 2.9 1.5 TSPOT indeter (%) 0.0 TST +ve (N) 28 53 42 10 35 6 33 7 11 13 16 19 6 TST +ve (%) 41.2 21.5 47.2 50.0 52 5.1 12.0 9.3 10.1 6.5 9.6 9.3 26.1 74 11 32 5 52 9 17 11 32 25 2 67.9 55.0 43.2 4.3 18.9 2.7 15.6 5.5 19.2 8.4 16.7 4 3 7 7 0 5 0 10 0 15 0 3.7 15.0 5 6 0 1.5 0.0 5.0 0.0 5.1 0.0 -0.02 0.23 0.37 0.43 0.6 0.16 0.2 0.16 0.34 0.46 0.58 0.52 0.33 0.23 0.59 0.38 0.28 0.37 0.44 0.19 0.15 0.17 QFN G IT +ve (N) QFN G IT +ve (%) QFT N IT indeter (N) QFN G IT indeter (%) Kappa TSPOT vs TST Kappa QFN G IT vs TST Kappa T-SPOT QFN G IT
Jiang et al. Karam et al. Leidl et al. Mandalakas et al. (adults) Rangaka et al. Richeldi et al. Stephan et al. Talati et al. Balcells Jones et al. Kimura et al. Luetkemeyer et al. Mandalakas et al. (children)
68 247 109 20 74 116 275 336 109 201 167 296 23
96

Recommendation: There is currently insufficient evidence to support the use of IGRA to identify persons living with HIV (adults and children) who are eligible for IPT. Population: HIV-infected adults and children Factor Decision High Intervention: Use of IGRA to identify people eligible for treatment of LTBI Quality of evidence Explanation The quality of evidence on the role of IGRA in correctly identifying people eligible for treatment of LTBI is very high. While there is some information about test performance in persons with active TB or at low risk for TB, longitudinal data are very limited. Overall, the evidence does not support the use of IGRA at this time. Strong Values and preferences Costs Strong Sensitivity of T-SPOT may be higher than that of TST IGRA may require fewer return visits than TST Risk of adverse events with IGRA may be less than that of TST Potential for boosting (with TST) eliminated with IGRA
Risks or undesired effects
Insufficient information to determine whether IGRA can differentiate between LTBI and active TB disease Insufficient information to determine whether IGRA can identify persons likely to benefit from LTBI treatment IGRA performance is affected by immunosuppression (increased rate of indeterminate results in persons with low CD4 counts). IGRA requires a blood draw. There is a chance of failed phlebotomy, especially in children. Patients may prefer to avoid visible reaction to TST. Patients may prefer to avoid blood draw.
Feasibility
Strong
Increased by: Need to establish well-equipped laboratory Need to procure equipment and supplies for IGRA performance and quality assurance Need for staff training
Need for well-established laboratories Need for careful handling and processing of blood samples to ensure accuracy of tests Availability of well-trained staff or staff that can be trained
97
Annex 9
References
1. Aabye MG et al. The impact of HIV infection and CD4 cell count on the performance of an interferon gamma release assay in patients with pulmonary tuberculosis. PLoS One, 2009, 4:e4220. 2. Aichelburg MC et al. Detection and prediction of active tuberculosis disease by a whole-blood interferon-gamma release assay in HIV-1-infected individuals. Clinical Infectious Diseases, 2009, 48:954962. 3. Baba K et al. Evaluation of immune responses in HIV infected patients with pleural tuberculosis by the QuantiFERON TB-Gold interferon-gamma assay. BMC Infectious Diseases, 2008, 8:35. 4. Balcells ME et al. A comparative study of two different methods for the detection of latent tuberculosis in HIV-positive individuals in Chile. International Journal of Infectious Diseases, 2008, 12:645652. 5. Chapman AL et al. Rapid detection of active and latent tuberculosis infection in HIV-positive individuals by enumeration of Mycobacterium tuberculosis-specific T cells. AIDS, 2002, 16:22852293. 6. Clark SA et al. Tuberculosis antigen-specific immune responses can be detected using enzyme-linked immunospot technology in human immunodeficiency virus (HIV)-1 patients with advanced disease. Clinical and Experimental Immunology, 2007, 150:238244. 7. Davies MA et al. Detection of tuberculosis in HIV-infected children using an enzyme-linked immunospot assay. AIDS, 2009, 23:961969. 8. Elliott JH et al. Immunopathogenesis and diagnosis of tuberculosis and tuberculosis-associated immune reconstitution inflammatory syndrome during early antiretroviral therapy. Journal of Infectious Diseases, 2009, 200:17361745. 9. Jiang W et al. High-sensitive and rapid detection of Mycobacterium tuberculosis infection by IFN-gamma release assay among HIV-infected individuals in BCG-vaccinated area. BMC Immunology, 2009, 10:31. 10. Jones S et al. Utility of QuantiFERON-TB Gold in-tube testing for latent TB infection in HIV-infected individuals. International Journal of Tuberculosis and Lung Disease, 2007, 11:11901195. 11. Karam F et al. Sensitivity of IFN-gamma release assay to detect latent tuberculosis infection is retained in HIVinfected patients but dependent on HIV/AIDS progression. PLoS One, 2008, 3:e1441. 12. Kimura M et al. Comparison between a whole blood interferon-gamma release assay and tuberculin skin testing for the detection of tuberculosis infection among patients at risk for tuberculosis exposure. Journal of Infectious Diseases, 1999, 179:12971300. 13. Leidl L et al. Relationship of immunodiagnostic assays for tuberculosis and numbers of circulating CD4+ T-cells in HIV infection. European Respiratory Journal, 2009, 35:619626. 14. Liebeschuetz S et al. Diagnosis of tuberculosis in South African children with a T-cell-based assay: a prospective cohort study. Lancet, 2004, 364:21962203. 15. Luetkemeyer AF et al. Comparison of an interferon-gamma release assay with tuberculin skin testing in HIV-infected individuals. American Journal of Respiratory and Critical Care Medicine, 2007, 175:737742. 16. Mandalakas AM et al. High level of discordant IGRA results in HIV-infected adults and children. International Journal of Tuberculosis and Lung Disease, 2008, 12:417423. 17. Raby E et al. The effects of HIV on the sensitivity of a whole blood IFN-gamma release assay in Zambian adults with active tuberculosis. PLoS One, 2008, 3:e2489. 18. Rangaka MX et al. Clinical, immunological, and epidemiological importance of antituberculosis T cell responses in HIV-infected Africans. Clinical and Infectious Diseases, 2007, 44:16391646. 19. Rangaka MX et al. Effect of HIV-1 infection on T-cell-based and skin test detection of tuberculosis infection. American Journal of Respiratory and Critical Care Medicine, 2007, 175:514520. 20. Richeldi L et al. Performance of tests for latent tuberculosis in different groups of immunocompromised patients. Chest, 2009, 136:198204. 21. Stavri H et al. Comparison of tuberculin skin test with a whole-blood interferon gamma assay and ELISA, in HIV positive children and adolescents with TB. Roumanian Archives of Microbiology and Immunology, 2009, 68:1419. 22. Stephan C et al. Comparing QuantiFERON-tuberculosis gold, T-SPOT tuberculosis and tuberculin skin test in HIVinfected individuals from a low prevalence tuberculosis country. AIDS, 2008, 22:24712479. 23. Kabeer BSA et al. Role of interferon gamma release assay in active TB diagnosis among HIV infected individuals. PLoS One, 2009, 4:e5718. 24. Talati NJ et al. Poor concordance between interferon-gamma release assays and tuberculin skin tests in diagnosis of latent tuberculosis infection among HIV-infected individuals. BMC Infectious Diseases, 2009, 9:15. 25. Tsiouris SJ et al. Sensitivity analysis and potential uses of a novel gamma interferon release assay for diagnosis of tuberculosis. Journal of Clinical Microbiology, 2006, 44:28442850. 26. Veldsman C et al. QuantiFERON-TB GOLD ELISA assay for the detection of Mycobacterium tuberculosis-specific antigens in blood specimens of HIV-positive patients in a high-burden country. FEMS Immunology and Medical Microbiology, 2009, 57:269273. 27. Burgos JL et al. Targeted screening and treatment for latent tuberculosis infection using QuantiFERON-TB Gold is cost-effective in Mexico. International Journal of Tuberculosis and Lung Disease, 2009, 13:962968. 28. Dewan PK et al. Feasibility, acceptability, and cost of tuberculosis testing by whole-blood interferon-gamma assay. BMC Infectious Diseases, 2006, 6:47.
98
Annex 10: Summary of findings and quality of evidence:

INH resistance
PICOT Question: Does treatment for LTBI among people living with HIV lead to significant development of mono-resistance against the drugs used for LTBI?
oncerns regarding the development of isoniazid resistance during IPT in people who develop and/or have active TB while on preventive treatment is the dominant reason offered for not implementing IPT. Although the Population: Adults living with HIV who received treatment for LTBI Intervention: IPT (612 months INH) or any rifampicin-containing preventive treatment regimen Comparison: No IPT or any other regimen
study design of this issue is difficult, a review of the existing data suggests that there is little scientific evidence to support concerns regarding the development of mono-isoniazid resistance and/or multidrug resistance. Outcomes: Occurrence of mono-isoniazid resistance in TB isolates from people who have received the intervention compared with those who did not receive the intervention Timeline: Lifetime
ollowing the GRADE approach, the key outcomes that needed to be considered in making recommendations have been chosen and accordingly ranked in order to identify the data that had to be sought from the process of evidence retrieval.
Outcomes
Mono-resistance to IPT (IPT versus no treatment/placebo)

9
Comment
Critical Less critical
Mono-resistance to IPT vs rifampicin (IPT intervention vs rifampicin as control)

2.1 PubMed Search
(HIV[MeSH Terms]) AND (Tuberculosis[MeSH Terms]) AND (Antitubercular Agents[MeSH Terms]) OR (latent[Title/Abstract]) OR (isoniazid[Title/Abstract]) OR (prevent*[Title/Abstract]) OR (resist*[Title/ Abstract])
99
Annex 10
eading researchers in the field and WHO staff were contacted and provided with the articles reviewed in this study. They were requested to advise on any other unpublished data or further articles that were not identified by the search strategy.
The major HIV/AIDS and TB conference web sites were also reviewed for abstracts that satisfied the above search criteria.
articles
220
Limits activated Humans, clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial,
phase IV, comparative study, controlled clinical trial, evaluation studies, technical report, validation studies
studies (including 36 reviews)
95
Of interest
studies
17
abstracts from conferences
Studies of interest: Seventeen articles of original research including abstracts from conferences [117] provided some information about the incidence of resistance among
people living with HIV who had latent TB and were provided preventive therapy.
2.2 Inclusion and exclusion criteria
he patients in the studies had to meet specific criteria for being diagnosed with HIV and not suffering from active TB at the time of the study, or being on antitubercular medication prior
to the study. Latent TB or no previous history of TB was noted by the studies. Observational studies and trials were analysed separately for GRADE purposes.
Net studies reviewed for grading of evidence: 14 studies 7 IPT versus placebo, randomized trials 3 IPT versus rifampicin, randomized trials 4 IPT versus placebo, observational Mwinga et al. 1998 randomized trial excluded of 96 TB cases, only 12 tested for sensitivities Kawai et al. 2006 observational, no numerical data Mugisha et al. 2006 observational, no placebo group comparison
100
T
Study
Annex 10: Summary of findings and quality of evidence: INH resistance
he studies of interest include those that compared people living with HIV who had undergone previous treatment for TB and were
put on IPT, with no IPT, or a rifampicin-containing regimen for a duration of six to 12 months and/or a shorter duration of the rifampicin regimen.

Campos et al. 2003 Casecontrol retrospective observational Feb 1999Jan 2000 Lima and Callao, Peru Adults >18 years, 80/81 male Churchyard et al. 2003 Observational study Golub et al. 2008 Cohort retrospective Gordin et al. 2000 Randomized controlled trial Sept 1991May 1996 Outpatients in clinics, US/Mexico/Haiti/ Brazil, age >13 years Halsey et al. 1998 Randomized prospective unmasked trial Apr 1990July 1992 Haiti, age 1677 years, HIV-1-positive, TST-positive >5 mm, normal chest X-ray, median follow up 2.5 years 750 HIV-positive => 332 on IPT Hawken et al. 1997 Randomized doubleblind placebocontrolled trial Apr 1992Mar 1994 Kenya, HIV-positive, 1465 years, no past/current TB, clinic or voluntary counselling and testing (VCT) attendees 342 INH
Publication year Methods/design
Data collection Population
From 1998 to July 2001 South Africa, gold miners, HIV-positive, documented successful completion of TB treatment1 338 had received secondary preventive therapy
Mar 1990Mar 1998 Injecting drug users Baltimore, US, adults, TST-positive with active TB- (by symptom review and chest X-ray) 800 HIV-positive => 649 TST => 102 TST-positive => 60 IPT started => 36 completed IPT (35% of TST-positives completed IPT), 26week course 1576 HIV-negative => 1104 TST => 434 TST-positive => 239 IPT started => 129/434 IPT completed (30%) Latent IPT
Sample size
415 HIV-positive with new TB -> 258 reviewed -> 135 TB culture positive -> TB Drug susceptibility tested on 81 => 35 MDR TB-positive Vs 46 MDR TBnegative HIV-positive patients vs 38 MDR TB-positive of 965 HIV-negative patients New TB TB prophylaxis MDR TB-positive 6.75 months vs MDR TB-negative 4.43.3 months of prophylaxis (P=0.27)
1583 HIV-positive, TST-positive (>5 mm, non-active TB), no past treatment for TB >2 months 35.8% vs 36.8% on ART in the 2 treatment groups 636 completed RIF/ PZA vs 544 INH therapy, 36 months follow up
Comparison group
221 had no secondary preventive therapy
333 on RIF prophylactic therapy
342 controls
TB latent/nil Treatment regimen
History of TB INH 300 mg/daily
Latent INH 300 mg/day + pyridoxine HCl 12 months (N=792)
Latent IPT 24 weeks INH 800 mg/pyridoxine 25 mg bi-weekly
No TB history 6 months INH 300 mg daily
Control
No treatment
RIF 600 mg/day + PZA 20 mg/kg/day for 2 months (N=791) Culture-confirmed TB, adverse events No INH resistance detected among TB cases who had received any duration of IPT No IPT 20 TB cases/24 585 person-years (incidence ratio 0.81 [0.501.26]/1000 person-years) vs all starting IPT 2 TB cases/4185 personyears (IR 0.48 [0.061.72]/1000 person-years), IPT 30 days 1 TB/3358 person-years (IR 0.29 [0.011.66]), completing IPT 0/2385 person-years (IR 0 [01.55]) IPT group 24/26 TB cases susceptibility tested 2 with INH monoresistance, 1 MDR; RIF/PZA group 19/19 TB cases tested 0 RIF monoresistance, 2 MDR
RIF 450 mg/day + PZA 2000 mg for 2 months, bi-weekly Resistant TB 14 (3.8%) IPT and 19 (5.0%) RIF preventive therapy patients developed TB. Seven had suspectibility testing, 2 resistant to INH, not signficantly different from the 180 other MTB isolates from TB patients without preventive therapy.
Placebo
Outcomes Resistance information
MDR TB MDR TB in HIVinfected patients was not associated with previous TB therapy or prophylaxis.
Resistant TB 51 cases of TB were diagnosed, 28 (8.3%) among the IPT cohort and 23 (10.4%) among the control cohort. There was no significant difference in the prevalence of isoniazid resistance between the IPT and control cohorts (20% [2/10] and 23% [3/13], respectively, (P=1.0).
Culture-confirmed TB 19 INH culturepositive patients vs 22 culturepositive controls, 2 resistant/17 tested (INH) vs 0 resistant/21 tested (control) risk for resistant TB/1000 10.05 vs 1.46 (RR 6.88 [0.013882.85])
101
Annex 10
Study Publication year Methods/design
Johnson et al. 2001 Randomized placebo-controlled trial Mar 1993Apr 1995 Uganda, HIVpositive, 1850 years, TST, no history of TB
Kawai et al. 2006 Retrospective review of MDR TB in a prospective TB treatment trial Feb 1999July 2000 Peru, TB patients identified and treated. Risk factors for MDR TB retrospectively reviewed 224 patients diagnosed with TB, commenced on TB directly observed treatment, shortcourse (DOTS)
Mosimaneotsile et al. 2009 Cohort prospective no placebo group, part of ongoing trial 20042006 Botswana, HIVpositive
Moreno et al. 1997 Historical cohort, not blinded
Mugisha et al. 2006 Retrospective IPT review feasibility of service provision study Sep 2001Sep 2003 (25-month period) Uganda
Mwinga et al. 1998 Randomized doubleblind placebocontrolled trial Aug 1992Jul 1994 Zambia, HIV-positive, no TB treatment history, both TST-positive and -negative patients; 1.8 years median follow up Of 2063 HIV-positive => 1053 subjects => 350 INH
19851994 Spain, 121 HIVpositive and PPDpositive
Sample size
2736 TST-positive and -negative HIVpositive adults, 2018 TST-positive => 536 INH prophylaxis => 459 completed and evaluated Of 2018 TST-positive => 464 placebo => 318 completed and evaluated Latent INH 6 months, INF+RIF 3 months, INH+RIF+ PZA 3 months, or placebo 6 months
1995 (72% female, age 1870 years, no TST criteria
29 INH (median follow up 89 months)
6305 HIV-positive => 3366 TST => 894 TST-positive => 506 IPT => 335 completed
Comparison group
No control, part of ongoing trial
92 no treatment (median follow up 60 months) Latent 912 months of INH Latent 9 months in a 12-month period INH 300 mg/25 mg pyridoxine
351 on RIF/352 on placebo
TB latent/nil Treatment regimen
Current TB
No/latent TB (not treated in past) INH 6 months + pyridoxine 25 mg
No TB and latent TB 6 months INH twice a week
?? Outcomes
No placebo/no review 3 definite, 2 possible incident, 3 possible deaths TB cases INH vs control = 36 vs 46, resistant/total tested 5/20 vs 1/24 MDR TB is associated with previous TB treatment; found an association with previous IPT 0 out of 3 cultureconfirmed TB cases were resistant to INH
No treatment
RIF + PZA twice/ week, OR placebo
Resistance information
3 TB cases INH vs 39/43 tested in no treatment group. Resistance 2 out of 2 tested INH vs 0/12 tested no treatment. Risk for resistance 118.64 vs 5.41 (21.95 [0.0411 582.31]) Median survival was more than 111 months in patients who received isoniazid compared with 75 months in patients who did not receive isoniazid (P < 0.001).
5/533 on IPT developed TB; 4 had been on treatment for many months
96 TB cases in all three groups, culture results 26 => sensitivity results for 12 only. One placebo patient resistant to INH
Findings/comments
Risk for resistant TB/1000 13.69 vs 3.39 (RR 4.04 [0.5032.80]) INH vs placebo
At the infectious diseases unit where all patients were HIV-positive, 30 (47%) of them had MDR TB. Multiple regression analysis revealed that MDR TB was associated with previous TB preventive therapy (hazard ratio [HR] 16, 95% CI: 2.885, P < 0.002)
Eighty-six per cent (1710/1995) were adherent to the entire 6-month course of IPT
The prevalence of INH resistance in Uganda was 6.7% in previously untreated cases in surveillance from 1996 to 1997, --portion of IPT patients will develop disease IPT would not be expected to be effective.
102
Study Publication year Methods/design
Pape et al. 1993 Randomized clinical trial 19861989 Haiti, asymptomatic HIV-positive 1865 years, no TB history, 118 enrolled (77% female) 58 on INH+B6
Rivero et al. 2003 Randomized clinical trial Jun 1994Dec 1998 Madrid and Andalusia, Spain, HIV-positive patients 1865 years with TST anergy, no TB history/treatment Of 319 HIV-positive => 83 on 6 months INH 82 3 months RIF/ 77 in 2 months RIF+PZA / 77 in no treatment group (follow up 88 personyears INH for six months group / 96 person-years RIF 3 months / 81 RIF plus PZA 2 months / 126 person-years no treatment group) No TB history
Saenghirunvatta et al. 1996 Prospective trial
Zar et al. 2007 Prospective doubleblind placebocontrolled trial Jan 2003May 2004 South Africa, HIVpositive children >8 weeks (median age 24.7 months), average 5.7 months follow up 263 HIV-positive => 132 IPT 131 placebo
le Roux et al. 2009 Prospective trial
19941995 Thailand, HIVpositive
December 2002March 2008 South Africa, children >8 weeks, median age 25.9 months, average 25 months follow up
Sample size
46 HIV-positive => 10 IPT 36 controls no therapy
276 patients placebo 105 placebo
171 prophylaxis, 105
Comparison group
60 vitamin B6 alone
TB latent/nil
No TB history
If previous TB, placed on therapy, and once completed, placed on trial 12 months 300 mg INH daily No treatment INH three times/ week Placebo three times/ week
No TB history
Treatment regimen
12 months daily INH 300 mg+ pyridoxine (50 mg) 12 months pyridoxine only
INH 6 months
INH daily 43%, 3x per week 57%
Comparison
RIF + INH 3 months / RIF 2 months / no treatment 11 cases TB overall, 2/3 in INH group resistant to INH, 0/3 in RIF 3 months, 1/1 resistant to INH PZA+RIF in 2-month RIF group, 4/4 resistant to INH+Strep in no treatment group Small numbers
Placebo daily 49%, 3x per week 51%
Outcomes Resistance information Comments 11/60 vit B6 and 4/58 INH patients had TB; 0 out of 15 TB patients had any drug resistance (all responsive to TB therapy) 0/10 AND 1/36 cases TB, no resistance Small sample size2 TB incidence INH 5/132 vs placebo 13/131. Incidence of resistant M. tuberculosis infection did not increase in children on prophylaxis. At enrolment, 9% on HAART, by end 31% 22/41 children who were diagnosed with TB were taking daily prophylaxis (P = 0.53). Two diagnosed with MDR TB; both were on daily prophylaxis and had mean adherence below 90%.
103
Annex 10
Quality assessment
Indirectness Anti-TB medications
11/1255(0.9%) MODERATE 5/1069 (0.5%) RR 1.87 (0.65 5.38) 4 more per 1000 (from 2 fewer to 20 more) Critical
No. of studies No medications Relative risk (95% CI) Absolute

No serious indirectness Serious4 None
Design
Limitations
Inconsistency
Imprecision
Importance
Mono-resistance to IPT vs placebo (IPT vs placebo)
Randomized trials
Serious3
Mono-resistance to IPT vs rifampicin (IPT intervention vs rifampicin as control) No serious indirectness VERY LOW Very serious None 3/1469 (0.2%) 1/1469 (0.1%) RR 2 (0.1822.03) 1 more per 1000 (from 1 fewer to 14 fewer) Less critical
Randomized trials
Serious3
Recruitment bias Churchyard gold miners Few patients, few resistance measurements 3 Incomplete accounting of patients and outcomes 4 Small number of cases and patients
104

Providing IPT for people living with HIV who are unlikely to have active TB does not significantly increase the risk of developing INH-resistant TB. Therefore, concerns regarding the development of isoniazid resistance should not be a barrier to implementing IPT programmes.
Population: People living with HIV Intervention: Programmatic implementation of IPT for people living with HIV in resource-constrained settings Action: Concerns regarding the development of isoniazid resistance should not be a barrier to implementing IPT programmes. Quality of evidence Benefits or desired effects Factor Decision Moderate Explanation Risks or undesired effects Strong Values and preferences Costs Strong Strong Feasibility Overall ranking of recommendation Direct evidence from a number of studies and meta-analyses Comparison of those on IPT with those receiving placebo shows no significantly increased risk. IPT prevents active TB in people living with HIV. IPT can be given with no significant risk of developing INHresistant TB. The rare patients who do develop INH mono-resistance can be treated successfully with routine TB treatment. Treating mono-RIF resistant LTBI with IPT can prevent future MDR (INH- and RIF-resistant) TB Minimal risk that patients will develop TB while on IPT Minimal risk that those who develop TB will develop INHresistant TB Minimal risk that those who develop INH resistance will be untreatable Minimal risk of transmission of INH resistance Minimal risk of development of MDR TB due to using IPT in the face of pre-existing mono-RIF resistant TB High INH resistance may reduce efficacy of IPT
IPT is a valuable TB prevention intervention with considerable potential to prevent morbidity, mortality and TB transmission. Minimal risk of INH resistance not likely to increase cost as the majority of patients will not develop TB and those who do will likely be successfully treated with routine treatment MDR rare but potentially costly event (e.g. laboratory testing, second-line treatment, extended duration)
May be difficult to overcome health-care workers concerns regarding development of TB drug resistance Evidence and experience strongly supports implementation.
105
Annex 10
References
1. Campos PE et al. Multidrug-resistant Mycobacterium tuberculosis in HIV-infected persons, Peru. Emerging Infectious Diseases, 2003, 9:15711578. 2. Churchyard GJ et al. Efficacy of secondary isoniazid preventive therapy among HIV-infected Southern Africans: time to change policy? AIDS, 2003, 17:20632070. 3. Golub JE et al. Long-term effectiveness of diagnosing and treating latent tuberculosis infection in a cohort of HIVinfected and at-risk injection drug users. Journal of Acquired Immune Deficciency Syndromes, 2008, 49:532537. 4. Gordin F et al. Rifampin and pyrazinamide vs isoniazid for prevention of tuberculosis in HIV-infected persons: an international randomized trial. Terry Beirn Community Programs for Clinical Research on AIDS, the Adult AIDS Clinical Trials Group, the Pan American Health Organization, and the Centers for Disease Control and Prevention Study Group. Journal of the American Medical Association, 2000, 283:14451450. 5. Halsey NA et al. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet, 1998, 351:786792. 6. Hawken MP et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS, 1997, 11:875882. 7. Johnson JL et al. Duration of efficacy of treatment of latent tuberculosis infection in HIV-infected adults. AIDS, 2001, 15:21372147. 8. Kawai V et al. Tuberculosis mortality, drug resistance, and infectiousness in patients with and without HIV infection in Peru. American Journal of Tropical Medicine and Hygiene, 2006, 75:10271033. 9. le Roux SM et al. Adherence to isoniazid prophylaxis among HIV-infected children: a randomized controlled trial comparing two dosing schedules. BMC Medicine, 2009, 7:67. 10. Moreno S, et al. Isoniazid preventive therapy in human immunodeficiency virus-infected persons. Long-term effect on development of tuberculosis and survival. Archives of Internal Medicine, 1997, 157:17291734. 11. Mosimaneotsile B et al. Isoniazid tuberculosis preventive therapy in HIV-infected adults accessing antiretroviral therapy: a Botswana experience, 20042006. Journal of Acquired Immune Deficiency Syndromes, 2009, 54:7177. 12. Mugisha B et al. Tuberculosis case finding and preventive therapy in an HIV voluntary counseling and testing center in Uganda. International Journal of Tuberculosis and Lung Disease, 2006, 10:761767. 13. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 14. Pape JW et al. Effect of isoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection. Lancet, 1993, 342:268272. 15. Rivero A et al. [Randomized trial of three regimens to prevent tuberculosis in HIV-infected patients with anergy]. Enfermedades Infecciosas y Microbiologa Clnica, 2003, 21:287292. 16. Saenghirunvattana S. Effect of isoniazid prophylaxis on incidence of active tuberculosis among Thai HIV-infected individuals. Journal of the Medical Association of Thailand, 1996, 79:285287. 17. Zar HJ et al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomised controlled trial. British Medical Journal, 2007, 334:136.
106
adherence to preventive therapy
PICOT Question: Will low adherence rates to treatment for LTBI be a barrier to the implementation of LTBI treatment among people living with HIV? Adults Population: Adults living with HIV/AIDS in areas where TB prevalence is >30% Intervention: Good adherence to LTBI treatment (>80% medication taken) Comparison: Poor adherence, i.e. <80% Outcomes: Consider only outcomes graded critical in the table below Timeline: Lifetime (although no study has lifetime follow up) Children Population: Children (<15 years) living with HIV/AIDS in areas where TB prevalence is >30% Intervention: Good adherence to LTBI treatment (>80%) Comparison: Poor adherence, i.e. <80% Outcomes: Consider only outcomes graded critical in the table below Timeline: Lifetime (although no study has lifetime follow up)
Outcomes for individuals
Prevention of deaths from TB Reduction in incidence of TB Prevention of TB transmission Acceptably low rate of adverse effects
Relative score
8 8 9
Severity index
Outcomes for programmes

Resource allocation Time allocation
Critical
Service uptake justifies staff allocation
Critical
Critical
Less critical

Medline and Embase were used to locate appropriate articles. Keywords: latent TB infection treatment, adherence, HIV PLWHA, isonazid prophylaxis Inclusion criteria Trials (randomized controlled trials/observational studies) that involved LTBI treatment in an unselected HIV population (adult or child) and commented on adherence with reference to how it was monitored,
outcome with regard to adherence (if possible), and made qualitative and quantitative statements with regard to achieving good adherence.
107
Annex 11
Search strategy Articles identified:
63
Selected by title:
47
Selected by abstract:
24
After reading:
Additional material:
Reviewed:
15
nlike treatment for TB which, if unsuccessful, could result in severe morbidity, death and further TB transmission, IPT is a preventive measure, and whether an individual is adherent and the degree to which they are adherent primarily only directly impacts their personal risk for TB. There are a variety of perspectives on adherence including how many people are lost to follow up, or

how many complete treatment, and how many take most of their tablets. It has been suggested that taking >80% of a course of IPT can be considered successful completion of treatment.[1] This has been interpreted in a number of ways: taking six months treatment within nine months, collecting more than 80% of tablets, taking >80% of the regimen. We have specified how adherence has been measured
108
Annex 11: Summary of findings and quality of evidence: adherence to preventive therapy
in the studies reviewed for the guidelines. Studies examining the effect of individual adherence at a population level have not been performed; this is not surprising as any study designed to evaluate the effect of poor adherence on community levels of TB or latent TB would involve very large numbers with long-term follow up and, as yet, there are few places implementing IPT on such a large scale to allow this to be accurately and meaningfully studied. There are few studies that look at the effect of individual adherence on individual outcome. One study [2] attempts to link adherence with outcome (note, though, that the study [probably due to inadequate power] does not show the overall benefit of IPT over placebo in preventing TB, see GRADE proforma attached). If the definition of poor adherence is not finishing the course of treatment then those patients lost to follow up in the studies represent poor adherence; and the outcomes of patients who default would answer the question of the effect of adherence on individual outcomes there are no studies on this issue for patients with HIV. It is sometimes assumed that those who default are equivalent to those untreated, that the non-adherer returns to baseline TB risk and that the question of poor adherence affecting individual outcome is the same as asking whether treatment for LTBI is efficacious. However, this is by no means certain as it may depend on the amount of IPT takenit is biologically plausible that even some IPT may be better than none. There are no studies that show the graded response of adherence to individual outcome, e.g. outcome by adherent patient months, which would be
Study/author and year Population Location
interesting. For example, is adherence to five months of IPT equivalent to not taking the medication and/ or what is the difference between adherence to six months or nine months of treatment? If there is a graded response on an individual basis to varying levels of adherence, this may have a knock-on effect on population transmission, but again, this has not been studied. There has been some discussion in the literature suggesting that if adherence to an LTBI prevention programme is low it may reduce the enthusiasm of staff to run one and will potentially raise unit costs. However, this is not reported by the studies as an outcome measure and is often written from a TB programmes perspective where the traditional focus has been treatment of active TB. HIV services, having experience with long-term prophylaxis such as co-trimoxazole, may have a different chronic or preventive care perspective. The issue of adherence and drug resistance is often raised. Despite a thorough review, no study has been identified that assesses this theoretical risk. Although data are not available, the assumption is that good adherence will lead to better outcomes and the search brought up a number of studies that examined factors that affect adherence including social, personal, clinical and programmatic interventions. If IPT is offered as part of chronic care for people living with HIV then it may benefit from other efforts to improve adherence. The table below provides a summary of adherence and an approximation of how it has been measured by both the randomized controlled trials (RCTs) and observational studies (OS) that looked at adherence.
Regimen Method of adherence measurement (good adherence) Clinic visit attended (>80%) Adherence rates
Randomized controlled trial (RCT)/observational study (OS); Halsey et al. Lancet 1998 [3] RCT/OS; Hawken et al. AIDS 1997 [2]
784
Haiti
INH twice weekly 6/12 or PZA+RIF twice weekly 2/12 INH twice weekly 6/12 or RIF+PZA twice weekly for 2/12
74% patients taking RIF+PZA
1053
Uganda
Pill collection/self-report/ return within one month of last dose (collect 80 100% tablets, self-report taking >80% tablets) Self-report/pill collection/ urine (collect pills within 2 weeks of end of last prescription or positive urine) Not documented Pill counts, supervisor tick sheet (not given)
74%
OS; Lugada et al. Int J Tuberc Lung Dis 2002 [15]
98
Uganda
INH daily 6/12
88% by urine, 76% by pill collection
OS; Munseri et al. Int J Tuberc Lung Dis 2008 [8] OS; Mohammed et al. Int J Tuberc Lung Dis 2007 [6]
565 118
Tanzania Cape Town
INH daily 6/12 INH daily 12/12 if TSTpositive or randomized to INH despite TSTnegative or placebo if randomized to that and TST-negative
87% (NB: 4% physician terminated) Median percentage of doses taken per subject 81.291.7% (placebo open label INH)
109
Annex 11
Study/author and year
Population
Location
Regimen
Method of adherence measurement (good adherence) Collection of pills (collect 6 months medicines in 9/12 period) Urine dipstick one-off Monthly pill count 2-monthly pill count and clinic attendance Pill count Self-report and clinic attendance Self-report Pill count and self-report
Adherence rates
Data review; Rowe et al. Int J Tuberc Lung Dis 2005 [10] OS; Szakacs et al. BMC Infect Dis 2006 [12] OS; Ngamvithaypong et al. AIDS 1997 [9] OS: Mosimaneotsile et al. JAIDS 2009 [7] OS: De Souza et al. MIOC 2009 [16] RCT; Whalen et al. N Engl J Med 1997 [17] Antonnucci et al. Eur Respir J 2001 [18] RCT; le Roux et al. BMC Med 2009 [5]
87
South Africa
INH daily 6/12
47%
301 412 1995 138 2018
South Africa Thailand Botswana Brazil Uganda
INH daily long term 9/12 daily INH 6/12 INH 6/12 INH INH+RIF 3/12 or INH+RIF+PZA 3/12 or INH 6/12 INH 6/12 INH daily or 3 x week 6/12
72% 67.5% 86% 87.7%
40 324 children
Italy South Africa
34.5% 100% compliant 94.7%
Summary
here is no evidence available to accurately answer whether low adherence rates to treatment for LTBI should be a barrier to implementation of treatment for LTBI among people living with HIV. However, some useful information can be derived from the published studies that include adherence information. Firstly, adherence data vary between studies (4794%). However, the studies that achieved higher adherence rates are not always feasible models for scaling up service; they may provide transportation fees to attend clinic appointments [4,6] or incentives on arrival. There is also mention of ineligibility for IPT due to the possibility of poor adherence, which would mean adherence rates are higher than would be achieved in routine programme conditions. It is noticeable that adherence in observational studies that reviewed retrospective adherence in a non-trial setting have lower rates of adherence. [10,12,18] The only published data [2] that linked adherence to LTBI treatment in an HIV-infected population to risk of TB found no increased risk of TB in poor adherers compared with those who
adhered well.[2] This supports the proposal that even if adherence is less than desirable, individual outcomes are not affected. There are no published data suggesting a link between poor adherence and the development of mono-isoniazid resistance. The factors addressed in the research setting, however, are not those that subjects comment on as reasons for poor adherence in the qualitative assessments reported. Side-effects or toxicity are also not mentioned as a risk factor for poor adherence. These conclusions compare favourably with the community consultation document [see Annex 4] and represent opportunities for programme implementation to improve adherence in the design and running of an LTBI programme. Additional research is needed on operationalizing IPT in a patient-centred manner which can improve adherence; however, it must also be recognized that IPT is a preventive measure and non-adherence impacts the individuals future risk of developing TB. The Guidelines Group made a strong recommendation that concerns regarding adherence should not be a barrier to implementing treatment for LTBI.
110
Annex 11: Summary of findings and quality of evidence: adherence to preventive therapy

Concerns regarding adherence should not be considered a barrier to implementing treatment for LTBI.
Population: People living with HIV Intervention: Implementation of IPT regardless of adherence concerns Action: HIV programmes should implement IPT Factor Quality of evidence Decision Low Explanation Benefits or desired effects Risks or undesired effects Strong Strong
Adherence will: Reduce TB transmission Reduce morbidity and mortality secondary to TB Improve overall impact of the programme
Studies where adherence is not a measured outcome Observational studies with limited patient numbers Qualitative assessments by interviews with small numbers of patients (adherent and non-adherent) Wide variety of adherence-related factors suggesting that local context is important
Extra work for clinic staff Extra cost in terms of patient information materials
Costs
Strong
Feasibility
Strong
Increased by: Initiation of a programme that is not currently running will entail extra costs in terms of staff, equipment, drugs, storage space, clinic time, patient time Decreased by: Overall costs decrease by efficacy of LTBI treatment programme Community involvement in programme Less spending on treatment of active TB (initial or secondary cases)
People living with HIV will appreciate a complete package of HIV/TB care. People living with HIV will benefit from lack of active TB disease as a result of IPT.
Fits well within HIV care structure Drug staff are familiar with its use ART programmes are familiar with focusing on adherence issues
111
Annex 11
References
1. De Cock KM, Grant A, Porter JD. Preventive therapy for tuberculosis in HIV-infected persons: international recommendations, research, and practice. Lancet, 1995, 345:833836. 2. Hawken MP et al. Isoniazid preventive therapy for tuberculosis in HIV-1-infected adults: results of a randomized controlled trial. AIDS, 1997, 11:875882. 3. Halsey NA et al. Randomised trial of isoniazid versus rifampicin and pyrazinamide for prevention of tuberculosis in HIV-1 infection. Lancet, 1998, 351:786792. 4. Hiransuthikul N et al. INH preventive therapy among adult HIV-infected patients in Thailand. International Journal of Tuberculosis and Lung Disease, 2005, 9:270275. 5. le Roux SM et al. Adherence to isoniazid prophylaxis among HIV-infected children: a randomized controlled trial comparing two dosing schedules. BMC Medicine, 2009, 7:67. 6. Mohammed A et al. Randomised controlled trial of isoniazid preventive therapy in South African adults with advanced HIV disease. International Journal of Tuberculosis and Lung Disease, 2007, 11:11141120. 7. Mosimaneotsile B et al. Isoniazid tuberculosis preventive therapy in HIV-infected adults accessing antiretroviral therapy: a Botswana experience, 20042006. Journal of Acquired Immune Deficiency Syndromes, 2009, 54:7177. 8. Munseri PJ et al. Completion of isoniazid preventive therapy among HIV-infected patients in Tanzania. International Journal of Tuberculosis and Lung Disease, 2008, 12:10371041. 9. Ngamvithayapong J et al. Adherence to tuberculosis preventive therapy among HIV-infected persons in Chiang Rai, Thailand. AIDS, 1997, 11:107112. 10. Rowe KA et al. Adherence to TB preventive therapy for HIV-positive patients in rural South Africa: implications for antiretroviral delivery in resource-poor settings? International Journal of Tuberculosis and Lung Disease, 2005, 9:263269. 11. Scardigli A et al. Responding to the challenge of patients adherence to IPT: experience of an urban ART facility in Mozambique. PEPFAR Implementing Meeting, Namibia, 11 June 2009. 12. Szakacs TA et al. Adherence with isoniazid for prevention of tuberculosis among HIV-infected adults in South Africa. BMC Infectious Diseases, 2006, 6:97. 13. WHO, UNAIDS. Consultation for the revision of WHO/UNAIDS policy statement on Preventive Therapy against TB for PLWHIV. 2010. [Annex 4] 14. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 15. Lugada ES et al. Operational assessment of isoniazid prophylaxis in a community AIDS service organisation in Uganda. International Journal of Tuberculosis and Lung Disease, 2002, 6:326331. 16. Souza CT et al. Effectiveness and safety of isoniazid chemoprophylaxis for HIV-1 infected patients from Rio de Janeiro. Memrias do Instituto Oswaldo Cruz, 2009, 104:462467. 17. Whalen CC et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. UgandaCase Western Reserve University Research Collaboration. New England Journal of Medicine, 1997, 337:801808. 18. Antonucci G et al. Guidelines of tuberculosis preventive therapy for HIV-infected persons: a prospective, multicentre study. GISTA (Gruppo Italiano di Studio Tubercolosi e AIDS). European Respiratory Journal, 2001, 18:369375.
112

cost effectiveness
PICOT Question: Is the treatment of LTBI in people living with HIV cost effective?
revious PICOT discussions have shown LTBI treatment to be efficacious, safe and an important tool for preventing TB in people living
with HIV. However, there are significant demands on limited funds within HIV programmes, which raise the issue of the cost effectiveness of treatment for LTBI. Comparator: No treatment for LTBI Outcome: See below Timeline: Lifetime
Population: Adults living with HIV in areas where the prevalence of LTBI is >30% Intervention: Treatment for LTBI
Outcomes
Cost of screening for and treating LTBI related to the cost of treating active TB that would occur in the absence of treatment for LTBI Cost of treatment of LTBI related to the cost of treating active TB, i.e. does not take screening costs into account
Relative score
9
Severity status
Critical
Critical

The Medline and Embase databases were used. Keywords: IPT, latent TB infection treatment, PLWHA, HIV, latent TB treatment, preventive TB treatment, cost-efficacy, cost-effectiveness, cost analysis Inclusion criteria: LTBI programme in HIV-positive individuals, clearly defined cost inclusive of screening for eligibility for preventive therapy, and estimate made of savings, again clearly defining where costings estimated from and documenting assumptions with regard to TB and programme implementation
113
Annex 12
Articles identified:
113
Selected by title:
50
Selected by abstract:
14
After reading:
13
Additional articles supplied by Guidelines Committee or sourced from references in other articles
13
Reviewed:
There were no studies on cost efficacy specifically related to children. There were also no studies on cost efficacy in a population with HIV living in an area with low TB prevalence.
114
Annex 12: Summary of findings and quality of evidence: cost effectiveness
he GRADE system is designed to review interventions scored on a number of outcomes. Cost efficacy can be an outcome; however, the articles that deal directly with cost-efficacy do not specifically have an outcome or control group

and are therefore difficult to assess within the GRADE approach. Therefore, GRADE software and methodology has not been used to answer this particular question.[1]
Published literature on cost-efficacy of IPT in HIV-infected populations
lthough the GRADE software has not been used to review the published literature, the principles encompassed within GRADE (transparent assessment looking at directness, consistency, precision) can be applied. This allows for an assessment of the quality of evidence available and a strength of recommendation that is possible to make from the available studies. Cost considerations are very situation specific cost of human resources, transportation, training, inpatient care and investigations varies widely between locations; this means that any comparison has to be based on cost relative to an outcome. However, the limited literature does not use a standardized
Trial Foster et al. AIDS, 1997, 11:919925 Setting and set-up Zambia district hospital, 6 months IPT after initial HIV test, clinical examination, smear, chest X-ray, confirmatory HIV test. Cost of IPT vs cost of treating TB (2/12 in hospital, drugs, diagnostics, lost earnings, health service costs) 1000 patients, spreadsheet model Urban Uganda (VCT centre). No HAART. Cost of IPT includes screening with questionnaire, smear, chest X-ray, physical examination, TST or treat all, i.e. no TST. Also included staff, training, equipment, space costs with decline over years of programme, no start-up cost. IPT means either 6 or 9 months of INH. Cost discount rate of 3% per year. Markov cohort simulation Assumptions Efficacy of IPT 4575% Adherence 5080% Likelihood of developing TB 2540% Secondary cases 25 Costs taken from 1997 programmatic costs, no data on where assumptions come from, TB and HIV incidence from local data Costs supplied from programmatic costs with a range. TST sensitivity 0.20.9 TST specificity 0.650.98 Adherence 0.4450.99 Risk of developing TB in controls 0.0340.1 All-cause mortality per year 0.0470.141 Adverse events 0.0140.041 Protection provided by IPT 15 years For QALY, survival of HIV-infected individual estimated Outcomes
measure to assess cost. Some studies chose to measure cost per quality-adjusted life year (QALY), some cost per TB case averted, some cost of TB in the person receiving IPT, some cost per averted TB case and averted secondary cases. There is also the issue of economic perspective: cost to the health-care system, cost to the patient, cost inclusive of screening for TB prior to treatment (by a variety of means), and cost of screening for HIV (see tables below). This variability renders direct comparison between studies difficult. Each study is best assessed on its individual merits and the results accumulated in order to formulate a recommendation.
Directness Relevant setting BUT a model, i.e. not based on reality, and in terms of population assumed around 30% HIV seropositivity with costs inclusive of 2 HIV tests Precision Wide variability given assumptions. Note benefitcost ratio significantly >1 only when patient earnings taken into account Comment Initial population not all HIV-positive
Measured in benefit cost ratio 0.383.86 or including patients lost wages 0.866.12 (greatest being most efficacious with maximal infectivity)
Shrestha et al. Int J Tuberc Lung Dis, 2007, 11:747754
Expected QALY with no intervention 498, with targeted TST testing 509 and with a treat-all strategy 539. Cost per QALY of targeted testing averaged $102, cost per QALY gained for treat all $106 (NB: if TST sensitivity is less than 0.52 as might be expected in advanced HIV, then target testing becomes more expensive at $115 per QALY)
Modelled so by necessity not completely transferable
Wide range of results, given variable assumptions. Consistent
Authors comment on non-HAART population.
115
Annex 12
Trial Hausler et al. Bull WHO, 2006, 84:528536
Setting and set-up Part of ProTEST, Cape Town PHC, STI, CHC clinics. Costs from clinic data note cost of ICF/IPT included screening for active TB with smear, TST, chest X-ray, physical examination, also included personnel time and start-up, building, vehicle and coordination costs. Rated against number of cases of TB prevented.
Assumptions Each case of TB infects 114 cases per year. ICF decreases infectious period from 9.6 months by 30%, so for every 100 cases picked up by ICF, 25 cases of TB are prevented. For IPT, the assumed incidence of TB decreases by 60% for 2 years if TSTpositive (if adherence 6080%)1, annual incidence of TB in TST-positive HIVpositive individuals is 8% and each case causes one other. Therefore, for every 100 people completing IPT using TST, 19 cases of TB are averted. TB rate if TSTpositive 3.410% per annum Secondary cases 210 Efficacy of preventive therapy duration 1.5 years to lifetime Efficacy of INH 2386%, efficacy for INH +RIF 1482%, efficacy for RIF+PZA 3090%, adverse events mild/severe 0.25 per month0.99 per day
Outcomes Cost per case of TB prevented by IPT $486962 (PHCCHC) but a 36% reduction if chest X-ray removed from screening equation. Removing TST reduces cost by 60% but cost per case prevented by only 4% as IPT is less efficacious.
Directness The assumptions made are from generally well accepted data. The population studied is relevant. Specific costs are of course site- and country-specific (note significant variation from PHC to CHC in same location)
Precision Consistent and plausible variation in cost due to different site.
Comment Note: authors argue cost-effective with TST, and could be more so without chest X-ray.
Bell et al. AIDS, 1999, 13:15491556
Model based on reported costings in the literature and using assumptions for the literature. Modelled the effect of 3 preventive therapy strategies: INH 6/12, INH + RIF 3/12, twice weekly RIF+PZA for 2/12. Used a Markov model with 3% discount rate and costs inclusive of indirect social costs and well as cost of secondary cases of infection.2 Cost of preventive therapy included screening costs (with chest X-ray, smear, TST, drug cost, personnel in case of RIF+PZA includes cost of DOT and nutrition supplementation). Measured cost per case prevented and cost per QALY South Africa predominantly high TB prevalence area. Costs included TST in those from low-prevalence background, cost of drug (INH), clinical services, treatment of INH hepatitis, transportation and breakthrough TB compared with cost of treating TB and secondary TB cases. Costs estimated in 1993 SA Rand with 4% discount rate taken from consensus view of functional TB programme. Estimate 8-year follow up
Taking all extremes of assumptions and varying costs (halving or doubling) does little to change overall outcome. Compared with cost of treating TB, preventive therapy is cost saving only if social costs are included. However, all regimens are cost reducing in terms of cost of treating TB if incidence of secondary infection is >2.4
Translatable to answer the question
Wide variability in costings and efficacy data exists and are used; however, does not alter overall outcomes.
In a sense, a metaanalysis of costefficacy data
Masobe et al. S Afr Med J, 1995, 85:7581
Assumed 97.2% of patients from highrisk TB background, 2.8% from low risk. TST-positive in 5%. Assume risk of 4.68 adult and 3.2 child secondary cases per active TB case and 6% of contacts develop TB. Annual rate of TB in HIV-positive 58%, INH efficacy 90%, compliance 68.5% and INH hepatitis rate 0.48%.
Over 8 years cost saving of R 40 551 952 (assumed prevent 21 800 cases of active TB) If compliance only 41% then saving only R10.5 million
Translatable to answer the question Direct in terms of applicable setting
Some sensitivity analysis commented on in terms of compliance, 4168% and sensitivity analysis in terms of discount rate. Still cost-saving
Bucher et al. Isonazid prophylaxis for TB in HIV infection: a meta-analysis of randomized controlled trials. AIDS, 1999, 13:501-507. Cost taken from Saunderson. Social Science and Medicine, 1995, 40:1203-1212, Foster et al., AIDS, 1997,11:919-925, Aisu et al., AIDS, 1995, 9:267-273.
1 2
116
Annex 12 Summary of findings and quality of evidence: cost effectiveness Annex 12:
Trial Sutton et al. Int J Tuberc Lung Dis,, 2009, 13:713718.
Setting and set-up Cambodia analysis of figures from 2003 to 2006 of running programme. Cost of IPT includes drugs (INH 9/12) and transportation of drugs, cost of patient transportation, cost of screening X-ray and clinical time, capital cost, human resource costs including training and education. Number of patients from their centre, cost from the centre. Compliance (completion rate) 86% Zambia ProTEST initiative facilitating care programme. Cost of running programme (not clear exactly what included) from 3 sites, cost of IPT does not specify if includes screening. Costs from 2007 converted to USD.
Assumptions Assume annual TB incidence in HIVpositive population (where TST result not known) is 8%, assume that IPT reduces TB incidence by 60% for 2 years. Assume 17/100 patients taking IPT are the number of TB cases prevented. Estimate that there is 1 secondary case in an HIV-positive person per active case.
Outcomes Cost per TB case prevented $955 (if 13/100 cases prevented), cost per case prevented $1274 but if 25/100 then falls to $764
Directness Study direct, relative
Precision Costs similar to those reported in other studies
Comment Note 88% of patients without active TB were deemed ineligible for IPT (pregnant,alcoholic liver disease, poor adherence likely)
Observational studies of cost (not always compared with outcomes)
Terris-Prestholt et al. Cost Effectiveness and Resource Allocation, 2008, 6:2
All real-time costs, i.e. no estimate of efficacy or protection but sensitivity analysis performed to look at variation of compliance
Cost per person completing IPT $2938 (NB: cost per person starting IPT $718)overall though altering compliance has less than 10% effect on unit costs
Real-time costings
Similar in terms of cost to other studies
No attempt to analyse in terms of efficacy
he Guidelines Group made a strong recommendation that, while there is significant variation between the studies assessed, the literature suggests that treatment for LTBI is likely to be cost effective. In some studies, there are a number of factors that must be taken into consideration to justify this conclusion. For example, one study [2] argues that the intervention is cost efficacious only if the social cost of having TB and treating TB are taken into consideration, and another study [3] discovered that the intervention becomes cost effective only if each TB case creates five secondary cases, all of which need treatment. In some studies, there are interesting programmatic alterations that dramatically affect cost. A study [4] found that the inclusion of chest X-ray in screening reduced cost efficacy; another study [5] found that the use of
TST reduced the cost of IPT. Yet another study [6] showed that reduction in patient compliance reduced the overall cost savings. None of the studies have been conducted in a population that was also on ART, and this would potentially change costs as well as the additional efficacy of IPT, and the number of secondary active TB cases among people living with HIV. There is considerable room for further research in this area, particularly given the recent improvements in HIV services and the emphasis on an integrated approach to TB and HIV prevention, care and treatment. The strong recommendation supports the overall recommendation for the wide use of IPT within comprehensive HIV prevention, care and treatment services, both as a measure of good clinical practice and as a likely cost-effective measure.
117
Annex 12

Treatment of LTBI should be considered cost effective in people living with HIV.
Population: People living with HIV Intervention: Implementation of IPT programme Factor Decision Low Action: HIV programmes should not avoid the addition of LTBI treatment to their service on cost grounds Quality of evidence Explanation Limited number of studies with inconsistent outcomes means that cumulative evidence is weak. Studies are hypothetical with variable assumptions made with regard to TB incidence, transmission and efficacy of IPT. All studies report IPT to be cost effective.
Risks or undesired effects
Strong
Will allow programmes to start LTBI treatment programmes as part of a package of HIV care Will save money within HIV and TB programmes which can be re-allocated to other parts of the programme Considering LTBI treatment cost effective and therefore initiating IPT programmes will reduce TB transmission, active TB in the individual and potentially realize the benefits listed for recommendation of IPT Initial costs will be higher since all studies show cost efficacy over >2-year period and not in the first year of a programme. Given the limited data from studies, it is possible that a universal roll-out of IPT may have unforeseen costs (not accounted for on a smaller scale).
Strong
Costs
Strong
Feasibility
Strong
Increased by: Use of chest X-ray Poor adherence Longer regimen Use of RIF+PZA Decreased by: Use of TST Good adherence Symptom screening
For the individual, it is likely that cost efficacy is unimportant. For programmes, being able to justify IPT provision as cost effective will satisfy donors and financial backers.
The incremental cost in excess of HIV programme would be limited and feasibility improved as it would be an add-on to existing programmes and no capital investment would be needed.
118
Annex 12: Summary of findings and quality of evidence: cost effectiveness
References
1. GRADE Working Group. Grading quality of evidence and strength of recommendations. British Medical Journal, 2004, 328:14901499. 2. Aisu T et al. Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre. AIDS, 1995, 9:267273. 3. Foster S, Godfrey-Faussett P, Porter J. Modelling the economic benefits of tuberculosis preventive therapy for people with HIV: the example of Zambia. AIDS, 1997, 11:919925. 4. Shrestha RK et al. Cost-utility of tuberculosis prevention among HIV-infected adults in Kampala, Uganda. International Journal of Tuberculosis and Lung Disease, 2007, 11:747754. 5. Shrestha RK et al. Cost-effectiveness of including tuberculin skin testing in an IPT program for HIV-infected persons in Uganda. International Journal of Tuberculosis and Lung Disease, 2006, 10:656662. 6. Masobe P, Lee T, Price M. Isoniazid prophylactic therapy for tuberculosis in HIV-seropositive patientsa least-cost analysis. South African Medical Journal, 1995, 85:7581. 7. Hausler HP et al. Costs of measures to control tuberculosis/HIV in public primary care facilities in Cape Town, South Africa. Bulletin of the World Health Organization, 2006, 84:528536. 8. Bell JC, Rose DN, Sacks HS. Tuberculosis preventive therapy for HIV-infected people in sub-Saharan Africa is costeffective. AIDS, 1999, 13:15491556. 9. Bucher HC GLGGS. Isoniazid prophylaxis for tuberculosis in HIV infection: a meta-analysis of randomized controlled trials. AIDS, 1999, 13:501507. 10. Saunderson P. An economic evaluation of alternative programme designs for tuberculosis control in rural Uganda. Social Science and Medicine, 1995, 40:12031212. 11. Sutton BS et al. The cost of intensified case finding and isoniazid preventive therapy for HIV-infected patients in Battambang, Cambodia. International Journal of Tuberculosis and Lung Disease, 2009, 13:713718. 12. Terris-Prestholt F et al. Integrating tuberculosis and HIV services for people living with HIV: costs of the Zambian ProTEST Initiative. Cost Effectiveness and Resource Allocation, 2008, 6:2; doi:10.1186/1478-7547-6-2.
119

intensive TB case-finding for children living with HIV
PICOT Question: What is the best combination of signs and symptoms and diagnostic tools that can be used as a screening tool to identify children living with HIV who are eligible for treatment of LTBI and diagnosis of active TB? Population: People living with HIV, children living with HIV Intervention: Combination of signs and symptoms and diagnostic tools that can be used as a screening tool to identify children living with HIV who are eligible for treatment of LTBI and diagnosis of active TB Comparison: TB prevalence without intervention Outcomes: Negative predictive value, sensitivity, specificity, positive predictive value Timeline: 12 months during work-up for TB
Outcomes
Negative predictive value (to identify children eligible for treatment of LTBI)
Relative score (rank 1 9 most critical)

9
Severity status
Critical Critical Important
Sensitivity (to identify children for further 9 diagnostic work-up) Specificity Positive predictive value
Important
120
Annex 13: Summary of findings and quality of evidence evaluation: intensive TB case-finding for children living with HIV

Search criteria: PubMed Search (Child[Mesh] OR Child, Preschool[Mesh]) OR Infant[Mesh]) AND Tuberculosis[Mesh]) AND HIV Infections[Mesh]) AND Diagnosis[Mesh]) OR (TB screening, Children, HIV)
articles
546
Limits: Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial, evaluation studies, multicentre study, Medline, PubMed Central, all infant: birth23 months, newborn: birth1 month, infant: 123 months, preschool child: 25 years, child: 612 years
161
By title
57
(+ 17 added)
By abstract (+10 added from references)
20
Of interest
15
International conferences: CROI, IAS, International AIDS conference CROI 2008, 2009: 0 IAS 2009: 2 IAC 2008: 2 Abstracts of interest: 1
121
Annex 13
Studies of interest: A total of 16 articles/abstracts provided information on the clinical presentation of TB among HIVinfected children, or the utility of the TB scoring system in HIV-infected children, or a combination of signs and symptoms or diagnostic tests among HIV-infected children.[116] Information available:
TB among HIV-infected children.[17,18] The WHOCDC meta-analysis focused only on adults and adolescents and did not address children.[19] An important challenge was the identification of an appropriate gold standard for the confirmation of TB among HIV-infected children. Among children with active TB, both sample collection and the yield of positive culture are generally thought to be low. Reviewed studies [216] used bacteriological confirmation of Mycobacterium tuberculosis by culture alone or in combination with radiological findings, TST reaction or response to antituberculosis therapy as the standard for comparison of the yield of various tests or symptom combinations. However, the difficulties encountered in confirming TB, even with the best available tests and technologies, severely limit the ability to identify practical approaches to the identification of TB in children.
There was only one study by Song et al. that looked at the combination of signs, symptoms and diagnostic tests among HIVinfected children.[1] Most of the identified studies presented information on the clinical presentation of TB among HIV-infected children and gave the frequency of various symptoms or the results of tests such as TST and chest X-ray among HIV-infected children with TB.[214] Two other studies evaluated the potential role of scoring systems for identification of TB in HIVinfected children.[15,16] Some review articles have summarized the difficulty in diagnosing

of only 0.56 and thus would fail to identify a large proportion of HIV-infected children with TB if used as a screening tool for diagnosis of TB in this population. This combination of symptoms had a specificity of 0.62. A study by Song et al.[1] reported on the evaluation of TB screening among 303 HIV-infected children in Rwanda using a combination of signs and symptoms and diagnostic tests. The results of screening tests were compared to a gold standard of microbiological confirmation or chest radiograph findings characteristic of TB. Absence of cough for two weeks or more, failure to thrive and fever had a negative predictive value of 0.99. Half the children were found to have a negative symptom screen and would not need further evaluation. Only two children were found to have a false-negative test result when compared with the gold standard. The combination had a sensitivity of 0.90 and specificity of 0.65 for identifying children for further diagnostic evaluation of active TB. Addition of a positive TST to the combination increased the sensitivity to 0.95 with a slight reduction in specificity to 0.59. Thus, 95% of children likely to have TB would be identified by this combination and sent for further evaluation. However, 5% of children likely to have the disease would still be missed.
ost of the studies [214] identified reported on the clinical presentation of TB among HIVinfected children or positivity of different tests (such as TST or chest X-ray). However, they did not report on the best combination of symptoms or tests that could be used as a screening tool to identify TB among HIV-infected children. Some studies [15,16] assessed the utility of scoring systems or consistency of different scoring systems in HIV-infected children. These scoring systems are limited by the absence of standard symptom definitions and adequate validation. Therefore, in their current form, their utility among HIV-infected children is limited. The scoring system used in Brazil worked equally well as the clinical algorithm for diagnosis of TB among HIV-infected children; however, its utility in high HIV- and TB-prevalence settings has not been tested and is uncertain. A study by Marais et al.[7] suggested that an algorithm based only on signs and symptoms was a poor predictor of a diagnosis of TB among HIVinfected children. The presence any one symptom of persistent, non-remittent cough for more than two weeks, objective weight loss in the preceding three months, or reported fatigue reached a sensitivity
122
GRADE profile and summary of findings What is the best combination of symptoms and diagnostic tools that can be used as a screening tool to identify HIV-infected children who are eligible for treatment of LTBI?
Quality assessment
No. of studies Design/no. of participants Limitations Inconsistency Indirectness Imprecision Other considerations Quality
Any one of cough 2 weeks, fever, or failure to thrive Negative predictive value 0.99 1 Sensitivity 0.90 1 Specificity 0.65 1 Observational study/303 Observational study/303 Serious limitation* Serious limitation* No serious inconsistency No serious inconsistency No serious indirectness No serious indirectness No serious imprecision# No serious imprecision# Low Observational study/303 Serious limitation* No serious inconsistency No serious indirectness No serious imprecision# Low Observational study/303 Serious limitation* No serious inconsistency No serious indirectness No serious imprecision# Low
Positive predictive value 0.14 1 Low
Culture and radiological appearance were used as a gold standard, which is not a perfect gold standard. Being an observational study and not having a well-defined gold standard, the study did not qualify for the highest quality of evidence. * The reference standard used is unlikely to correctly classify all the children with disease as having the disease. Moreover, sputum was collected only from children with signs and symptoms suggestive of TB or abnormal chest X-rays. # Confidence intervals for the sensitivity and specificity were not reported.
What is the best combination of symptoms and diagnostic tools that can be used as a screening tool to identify children for further diagnostic evaluation of active TB? Cough 2 weeks failure to thrive fever positive TST Values and uncertainty around these Negative predictive value 0.99 Sensitivity 0.95 Specificity 0.59 Positive predictive value 0.14 303 (1 study) Low 303 (1 study) Low 303 (1 study) Low 303 (1 study) Low Number of participants (studies) Quality of evidence
123
Annex 13

Children living with HIV without poor weight gain,* fever and current cough are unlikely to have active TB and should be offered IPT. Children living with HIV who are more than 12 months of age and who are unlikely to have active TB should receive six months of INH preventive therapy (10 mg/kg/day) as part of a comprehensive package of HIV care services. After successful completion of treatment for TB disease, all children living with HIV who are more than 12 months of age should receive INH for an additional six months. All children with a history of contact with a TB case should receive six months of IPT.
* Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weight-for-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening.
Population: Children living with HIV Factor
Intervention: Careful history-taking and clinical assessment Quality of evidence Benefits or desired effects Risks or undesired effects Values and preferences Costs Strong Decision Moderate
Explanation
The quality of evidence is low and comes from one study by Song et al. Simplifies screening and limits the number of radiological and laboratory investigations Identifies children who can benefit from IPT (and thus reduces morbidity/mortality due to TB)
Strong Weak/conditional
A small number of children with active TB might be given monotherapy Increase in time spent by health-care workers in screening for TB Parents would like their children to be protected from TB by provision of IPT, especially in settings with a high TB burden.
Weak
Increased by: Training of clinicians and nurses to perform clinical assessment and correctly determine poor weight gain Additional staff required due to increase in time spent by existing staff for screening Reduced by: Limited resources needed for symptom screening among children already regularly attending clinical services for HIV Avoiding costs of additional diagnostic tests including chest X-ray Avoiding costs that would have been associated with treatment of TB (if LTBI is effectively treated) History-taking and clinical assessment would be feasible but would require additional training and time from already overburdened staff.
Strength of recommendation
Strong (initial IPT for six months) Conditional (post TB treatment)
124
Children living with HIV with any one of the following: poor weight gain, fever or current cough may have active TB and should be evaluated for TB and other diseases.
Population: Children living with HIV Factor Intervention: Careful history-taking and clinical assessment Quality of evidence Benefits or desired effects Risks or undesired effects Weak Decision Low Explanation
The quality of evidence is low and comes from one study by Song et al. Early identification of TB suspects followed by appropriate treatment of identified TB cases can reduce TB-associated morbidity and mortality among children living with HIV.
Values and preferences Costs Weak
Increased demand for clinical and laboratory investigations This approach would result in a larger number of children undergoing diagnostic testing, including possible risks (generally not high) associated with sample collection (e.g. from lymph nodes or gastric aspirates)
Weak
Increased by: Increase in number of children requiring diagnostic evaluation for TB will require resources (staff, reagents, transport, laboratory capacity) Increased need for quality assurance of diagnostic services Need for additional drugs due to increase in number of cases Decreased by: Decreased costs of managing severely ill or dying children if TB is recognized May require some additional training and time from overburdened health-care workers, laboratory workers and families of affected children
Patients/parents generally desire accurate diagnosis of disease and may be willing to undergo diagnostic evaluation or treatment in an effort to prevent morbidity/mortality.
125
Annex 13
Figure 2. Algorithm for TB screening in children more than one year of age and living with HIV
Child more than 12 months of age and living with HIV* Screen for TB with any one of the following: Poor weight gain Fever Current cough Contact with a TB case No Assess for contraindications to IPT No Give IPT Yes Defer IPT Yes Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Screen regularly for TB Not TB Follow up and consider IPT TB Treat for TB

* All children and infants less than one year of age should be provided with IPT if they have a household contact history with a TB case. Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weight-for-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening. Contraindications include: active hepatitis (acute or chronic) and symptoms of peripheral neuropathy. Past history of TB should not be a contraindication for starting IPT. Although not a requirement for initiating IPT, TST may be done as part of eligibility screening in some settings. Investigations for TB must be done in accordance with existing national guidelines.
126
References
1. Song R et al. Evaluation of TB screening approaches among HIV-infected children Rwanda, 2008. 5th IAS conference on HIV pathogenesis and treatment, 1922 July 2009 [Abstract no TUPEB132]. 2. Hesseling AC et al. Outcome of HIV infected children with culture confirmed tuberculosis. Archives of Disease in Childhood, 2005, 90:11711174. 3. Iriso R et al. The diagnosis of childhood tuberculosis in an HIV-endemic setting and the use of induced sputum. International Journal of Tuberculosis and Lung Disease, 2005, 9:716726. 4. Jeena PM et al. Impact of HIV-1 co-infection on presentation and hospital-related mortality in children with culture proven pulmonary tuberculosis in Durban, South Africa. International Journal of Tuberculosis and Lung Disease, 2002, 6:672678. 5. Kiwanuka J et al. Diagnosis of pulmonary tuberculosis in children in an HIV-endemic area, Malawi. Annals of Tropical Paediatrics, 2001, 21:514. 6. Madhi SA et al. HIV-1 co-infection in children hospitalised with tuberculosis in South Africa. International Journal of Tuberculosis and Lung Disease, 2000, 4:448454. 7. Marais BJ et al. A refined symptom-based approach to diagnose pulmonary tuberculosis in children. Pediatrics, 2006, 118:e13501359. 8. Mukadi YD et al. Impact of HIV infection on the development, clinical presentation, and outcome of tuberculosis among children in Abidjan, Cote dIvoire. AIDS, 1997, 11:11511158. 9. Palme IB et al. Impact of human immunodeficiency virus 1 infection on clinical presentation, treatment outcome and survival in a cohort of Ethiopian children with tuberculosis. Pediatric Infectious Disease Journal, 2002, 21:1053 1061. 10. Ramirez-Cardich ME et al. Clinical correlates of tuberculosis co-infection in HIV-infected children hospitalized in Peru. International Journal of Infectious Diseases, 2006, 10:278281. 11. Sassan-Morokro M et al. Tuberculosis and HIV infection in children in Abidjan, Cote dIvoire. Transactions of the Royal Society of Tropical Medicine and Hygiene, 1994, 88:178181. 12. Schaaf HS et al. Culture-confirmed childhood tuberculosis in Cape Town, South Africa: a review of 596 cases. BMC Infectious Diseases, 2007,7:140. 13. Van Rheenen P. The use of the paediatric tuberculosis score chart in an HIV-endemic area. Tropical Medicine and International Health, 2002, 7:435441. 14. Walters E et al. Clinical presentation and outcome of tuberculosis in human immunodeficiency virus infected children on anti-retroviral therapy. BMC Pediatrics, 2008, 8:1. 15. Edwards DJ, Kitetele F, Van Rie A. Agreement between clinical scoring systems used for the diagnosis of pediatric tuberculosis in the HIV era. International Journal of Tuberculosis and Lung Disease, 2007, 11:263269. 16. Pedrozo C et al. Clinical scoring system for paediatric tuberculosis in HIV-infected and non-infected children in Rio de Janeiro. International Journal of Tuberculosis and Lung Disease, 2009, 13:413415. 17. Marais BJ et al. Diagnostic and management challenges for childhood tuberculosis in the era of HIV. Journal of Infectious Diseases, 2007, 196 (Suppl 1):S76S85. 18. WHO. Guidance for national tuberculosis programmes on the management of tuberculosis in children. Geneva, World Health Organization, 2006 [WHO/HTM/TB/2006.37; WHO/FCH/CAH/2006.7]. 19. Getahun H et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource constrained settings: individual participant data meta-analysis of observational studies. PLoS Medicine, 2011, 8(1): e1000391. doi:10.1371/journal.pmed.1000391.
127

preventive therapy for children living with HIV
PICOT Question: What is the optimal duration and drug regimen (e.g. INH, RIF, etc.) for treatment of LTBI to reduce the risk of TB developing among children living with HIV/AIDS? Population: Children living with HIV Intervention: IPT (6 months INH) Comparison: No IPT Outcomes: Active TB incidence, progression of HIV disease, adverse events, adherence, TB drug resistance, interval to active TB, interval to death Timeline: Lifetime
mortality,
Outcomes
Active TB incidence (suspected, probable, confirmed) Confirmed TB Mortality
Relative score (rank 1 9 most critical)

9
Comment
Critical Critical
Progression of HIV disease Adverse events
Adherence TB drug resistance Cost effectiveness Interval to active TB Interval to death
8 7 7 7 6 6
Critical
Critical
Critical
Critical (addressed by Annex 11) Critical (addressed by Annex 10) Critical (addressed by Annex 12) Less critical Less critical
128
Annex 14: Summary of findings and quality of evidence: preventive therapy for children living with HIV

Pubmed Search (Child[Mesh] OR Child, Preschool[Mesh]) OR Infant[Mesh]) AND Tuberculosis[Mesh]) AND HIV Infections[Mesh]) AND Therapeutics[Mesh]
articles
187
Limits: Clinical trial, meta-analysis, randomized controlled trial, review, clinical trial, phase I, clinical trial, phase II, clinical trial, phase III, clinical trial, phase IV, comparative study, controlled clinical trial, evaluation studies, multicentre study, Medline, PubMed Central, all infant: birth23 months, newborn: birth1 month, infant: 123 months, preschool child: 25 years, child: 612 years
54
By title
By abstract
Of interest
International conferences: (CROI, IAS, ICAAC, World AIDS conference) excluding published papers: ICAAC 2009 and previous: 1 IAS conferences: 2 CROI 2009 and previous: 1
O
129
verall, two studies were considered for the grade analysis.[1,12] One study suggested significant benefits for children receiving INH for six months, in particular, with regard to significant reductions in mortality.[12] However, findings from a randomized controlled trial conducted in South Africa showed that when HIV-infected infants with no known exposure to a TB source case are identified in the first three to four months of life, given rapid access to ART and carefully monitored for new TB exposure or disease on a monthly basis, there is no benefit from IPT.[1]
Therefore, based on this, the Guidelines Group recommends that all children more than 12 months of age who are living with HIV and who are unlikely to have active TB should receive six months of IPT as part of a comprehensive package of HIV care. For those children less than 12 months of age, only those with a history of contact with a TB case should receive six months of IPT. In contrast to adults and adolescents, there is no evidence to support the use of INH for longer than six months in children. Therefore, the Guidelines Group concludes that until more data are available, INH for children could not
Annex 14
be recommended for more than six months. Similarly, there is no evidence on whether repeating a course of IPT is beneficial for children. INH should be given at a dose of 10 mg/kg body weight daily and it is desirable that vitamin B6 be supplied
with INH at a dose of 25 mg daily. All available data to date suggest that INH is not toxic for children, even in those receiving ART. The following table shows a simplified dosing schedule for INH 10 mg/kg/day.
Weight range (kg)

<5 5.19.9
Number of 100 mg tablets of INH to be administered per dose

tablet 1 tablet 1 tablet 2 tablets 2 tablets
Dose given (mg)

100 50
1013.9
1419.9 >25
150
2024.9
200
he Guidelines Group noted that there is no evidence on the use of IPT in children after successful completion of TB treatment. However, like adults, children living with HIV are exposed to reinfection and recurrence of TB. Therefore, the Group conditionally recommends that children who have been successfully treated for TB and are living in settings with high TB transmission should receive IPT for an additional six months. IPT can be started immediately following the last dose of antituberculosis therapy. Regardless of a history of TB treatment, TB screening should be carried out
3 tablets or one adult tablet
250
300
during each contact of the child with a health-care worker. The Guidelines Group also concludes that there are no data regarding the efficacy of IPT for children stratified by degree of immunosuppression. However, it noted that there is biological plausibility in extrapolating what is known for adults and adolescents to children. Therefore, the Group conditionally recommends the combined use of IPT with ART for all children. The Group emphasizes that ART should not be delayed while starting or completing a course of IPT.
130
131
Indirectness INH prophylaxis (6 months)
44/358 (12.3%) MODERATE 5% 1 fewer per 1000 (from 17 fewer to 22 more) RR 1.5 (0.258.89) 4 more per 1000 (from 7 fewer to 71 more) RR 0.84 (0.511.37) CRITICAL MODERATE 16 fewer per 1000 (from 49 fewer to 37 more) RR 0.62 (0.211.85) CRITICAL HIGH 0 fewer per 1000 (from 0 fewer to 0 more) RR 0 (00) 0% 0 fewer per 1000 (from 0 fewer to 0 fewer) 0 fewer per 1000 (from 0 fewer to 0 fewer) CRITICAL 23 fewer per 1000 (from 48 fewer to 52 more) 14 fewer per 1000 (from 43 fewer to 32 more) CRITICAL LOW 7 more per 1000 (from 10 fewer to 105 more) 45/357 (12.6%) RR 0.97 (0.66091.4384) 4 fewer per 1000 (from 43 fewer to 55 more)
Q3.1 Efficacy in children Question: Should INH prophylaxis (six months) vs placebo be used in HIV-infected children (PPD-positive or TB-exposed)? Settings: High HIV/TB prevalence country Bibliography: Zar et al. 2007; Madhi et al. 2008
No. of patients
Placebo Relative risk (95% CI) Absolute
Quality assessment
None
No. of studies
No serious indirectness2 No serious imprecision
Design
Limitations
Inconsistency
Summary of findings Effect Quality Importance

CRITICAL
Active TB (follow up 5.79 months; clinical algorithm criteria, chest X-ray, bacteriological isolates from any site)
Randomized trials
Serious1
Confirmed TB (follow up 5.79 months; culture-proven) No serious indirectness 3/226 (1.3%) 0.9% 3/226 (1.3%) Serious4 None
Randomized trial
Serious3
Mortality (all causes) (follow up 5.79 months; review of hospital records and patients files) No serious indirectness2 10% No serious imprecision None 26/358 (7.3%) 31/357 (8.7%)
Randomized trials
Serious1
Adverse reactions (grade 3 or 4 toxicity) (follow up 5.79 months; clinical and laboratory monitoring) No serious indirectness5 5/132 (3.8%) 0% 8/131 (6.1%) No serious imprecision None
Randomized trials
HIV disease progression None 0/0 (0%) 0/0 (0%)
3. GRADE profiles and summary of findings
Opposite direction of the effect. P1041 (Madhi et al.) represents an optimal HIV care setting, with capacity to diagnose infectious diseases and good facilities to rule out active TB: children were younger, healthier and presented in a less advanced stage of the disease. The study by Zar et al. represents the most common condition in rural areas with a later diagnosis, children presenting in a more advanced stage of the disease and challenging TB diagnosis. 3 One trial available 4 Wide confidence intervals 5 Raw data are missing for P1041 (but no significant difference was reported between the two groups).
Annex 14

HIV-infected children who are unlikely to have active TB should receive six months of IPT.
Population: HIV-infected children Intervention: 69 months of IPT (1020 mg/kg plus vitamin B6 25 mg daily) Action: Concerns regarding the development of isoniazid resistance should not be a barrier to implementing IPT programmes. Quality of evidence Factor Moderate Decision Explanation Benefits or desired effects Risks or undesired effects Strong The evidence for provision of IPT in children is still inconclusive (GRADE moderate) 69 months of INH prophylaxis is the only regimen that has been investigated so far in this specific population. No specific biological reasons to think that IPT in children would be less effective than in adults
Reduction in TB incidence Reduction in TB transmission (even if children are less likely to transmit) Reduction in mortality (Zar et al. GRADE moderate) especially in children with advanced disease in whom TB may remain undiagnosed Avoids side-effects of TB treatment Reduction in drugdrug interactions between drugs for TB treatment and ARVs in children on ART Improved TB infection control in health-care and community settings (particularly in HIV clinics) Increase in unnecessary treatment (especially when TST is not performed or in case of false-positive TST where TST is performed) Potential INH toxicity (even if both randomized controlled trials are reassuring on this) Potential development and transmission of INH-resistant strains by treating undiagnosed TB (especially in children with more advanced disease where ruling out TB might be even more challenging) Reduction in compliance with other treatment (ART, cotrimoxazole, etc.)
Values and preferences Strong
Avoids active TB disease as well as deaths and disease transmission to other family members Health-care workers would be less exposed to active TB cases and would feel more protected (less of a problem in children as they are considered to be less contagious) Infection control for patient (less of a problem in children as they are considered to be less contagious) Avoids all the daily implications of long TB treatment with high pill burden and significant side-effects Parents would feel their children were protected against TB, adding value to HIV programmes BuT May be worried about side-effects of INH Pill burden Stigma
132
Factor Costs
Decision Weak
Feasibility
Strong
Increased by: Costs of diagnosing LTBI (TST) Costs of INH (including storage, supply and transportation) Monitoring liver function tests (if standard monitoring or in case of toxicity) Additional staff in ART settings are already overloaded by routine activities Need for second-line TB treatment in case INH resistance occurs and is transmitted Costs for INH-related toxicity (blood tests, hospitalizations, parents loss of earning) Additional costs of providing vitamin B6 Decreased by: Avoiding active TB treatment costs Less cost of hospitalizations Reduction in social costs and loss of parents earnings Reduction in costs due to treatment of secondary cases including in health-care workers, improving staff retention Reduction in costs achieved by preventing TBART cotreatment including potential need for a more expensive regimen and possible ART failure Dose adjustment would require more frequent follow up
Explanation
Drug easily available and cheap Administration fits well with the schedule of HIV programmes Does not require strict monitoring (laboratory) Health-care workers would already be familiar with the drug BuT Lack of child-friendly formulation for infants and very small children Would mean adding an intervention to an overburdened system May increase clinic visits and require more staff Space required for storage and difficulties in procurement INH single formulation less easily available compared to fixed-dose combinations (FDCs) Difficult to overcome health-care workers reluctance to use IPT Health-care workers need to be trained in conducting and reading TST or IGRA Dose adjustment would require more frequent follow up
133
Annex 14
Figure 2. Algorithm for TB screening in children more than one year of age and living with HIV
Child more than 12 months of age and living with HIV* Screen for TB with any one of the following: Poor weight gain Fever Current cough Contact with a TB case No Assess for contraindications to IPT No Give IPT Yes Defer IPT Yes Investigate for TB and other diseases Other diagnosis Give appropriate treatment and consider IPT Screen for TB regularly Not TB Follow up and consider IPT TB Treat for TB

* All children and infants less than one year of age should be provided with IPT if they have a household contact history with a TB case. Poor weight gain is defined as reported weight loss, or very low weight (weight-for-age less than 3 z-score), or underweight (weight-for-age less than 2 z-score), or confirmed weight loss (>5%) since the last visit, or growth curve flattening Contraindications include: active hepatitis (acute or chronic) and symptoms of peripheral neuropathy. Past history of TB should not be a contraindication for starting IPT. Although not a requirement for initiating IPT, TST may be done as part of eligibility screening in some settings. Investigations for TB must be done in accordance with existing national guidelines.
References
1. Madhi SA et al.; and the P1041 Team. Lack of efficacy of primary isoniazid (INH) prophylaxis in increasing tuberculosis (TB) free survival in HIV-infected (HIV+) South African children. 48th ICAAC/IDSA 46th annual meeting, 2008 [G21346a]. 2. Martinez Alfaro EM, et al. [Evaluation of 2 tuberculosis chemoprophylaxis regimens in patients infected with human immunodeficiency virus. The GECMEI Group]. Medicina Clinica (Barc), 2000, 115:161165. 3. Martinson NB et al. Novel regimens for treating latent TB in HIV-infected adults in South Africa: a randomized clinical trial. 16th Conference on Retroviruses and Opportunistic Infections, Montreal, 811 February 2009 [Paper 36bLB]. 4. Mohammed A et al. Randomised controlled trial of isoniazid preventive therapy in South African adults with advanced HIV disease. International Journal of Tuberculosis and Lung Disease, 2007, 11:11141120. 5. Mwinga A et al. Twice weekly tuberculosis preventive therapy in HIV infection in Zambia. AIDS, 1998, 12:2447 2457. 6. Pape JW et al. Effect of isoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection. Lancet, 1993, 342:26872. 7. Quigley MA et al. Long-term effect of preventive therapy for tuberculosis in a cohort of HIV-infected Zambian adults. AIDS, 2001, 15:215222. 8. Rivero A et al. [A randomized trial of three regimens to prevent tuberculosis in HIV-infected patients with anergy]. Enfermedades Infecciosas y Microbiologia Clinica, 2003, 21:287292. 9. Rivero A et al. [Randomized clinical trial investigating three chemoprophylaxis regimens for latent tuberculosis infection in HIV-infected patients]. Enfermedades Infecciosas y Microbiologia Clinica, 2007, 25:305310. 10. Samandari T et al.; on behalf of BOTUSA IPT study. Preliminary results of the Botswana IPT Trial: 36 months vs. 6 months isoniazid for TB prevention in HIV-infected adults. 40th Union World Lung Conference, Cancun, 2009. 11. Whalen CC et al. A trial of three regimens to prevent tuberculosis in Ugandan adults infected with the human immunodeficiency virus. UgandaCase Western Reserve University Research Collaboration. New England Journal of Medicine, 1997, 337:801808. 12. Zar HJ et al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomised controlled trial. British Medical Journal, 2007, 334:136.
134
Annex 15: Summary of findings and quality of evidence

evaluation: selected research gaps
The optimal timing for starting IPT and whether/ when to combine it with ART should be further investigated, with particular attention to the possible benefits and risks for patients already on ART. In addition, the possibility of discontinuing INH once immunological recovery has been achieved would be of great interest. The efficacy, safety and tolerability of a longer regimen of INH prophylaxis should be addressed in the paediatric population. Potential co-formulation of INH with cotrimoxazole should be made available and longitudinal data on its effectiveness collected. The potential impact of INH monoresistance on IPT and possible efficacy of subsequent TB treatment should be assessed. Operational research should be conducted on scaling up the implementation of IPT. Programmes implementing IPT as part of HIV prevention and care should be monitored and evaluated. Mathematical modelling to evaluate the effectiveness of IPT should be planned, including examination of the risks and benefits with particular attention to benefits stratified by TST status. A proper cost-effectiveness analysis addressing different strategies for provision of IPT and TST should be performed. The optimal frequency of screening people with HIV for active TB with a symptom questionnaire and the timing of repeat TST in HIV-positive individuals receiving ART should be addressed. The algorithm for intensive case-finding in children should be validated. Limited studies have evaluated the clinical utility of IGRA in persons with HIV and, to date, no evidence is available to warrant its programmatic use in resource-limited settings as a screening test before IPT. There is a clear need for longitudinal studies that assess the ability of IGRA to identify people at high risk for developing active TB or those who would benefit from treatment of LTBI. Because of concerns about the impact of immunosuppression on the performance of IGRA, it would be most useful if such studies included persons with both high and low CD4 cell counts and provided results stratified by CD4 cell count. Longitudinal assessment with serial measures is required to elucidate the clinical significance of discordant IGRA and TST, and the predictive value of IGRA for M. tuberculosis infection among HIV-infected persons. Better diagnostics for children should be identified. Prospective studies using cases of bacteriologically confirmed TB and appropriate controls including children not suspected of having TB would be needed to evaluate the diagnostic value of TB symptoms. Although some studies have evaluated different aspects of the TST in the general population and among HIV-infected persons, very few studies have provided information on the feasibility of using TST in resource-poor settings. No studies have evaluated the actual costs involved in largescale TST screening among people with HIV infection, and the overall burden of performing TST on already overstretched health systems in resource-poor settings. The stability of tuberculin from the point of manufacture to the point of use should be assessed, especially under natural environmental conditions in developing countries where facilities for refrigeration are not readily available. The reliability of other categories of health-care professionals in reading TST results, particularly in the HIV-infected population, needs to be evaluated. Efforts are currently being made in the area of task-shifting with the overall goal of scaling up HIV prevention, care and treatment activities. Different strategies are also needed to reduce the proportion of patients who fail to return for the TST results to be read.
135

Guidelines For Intensified Tuberculosis Case-Finding and Isoniazid Preventive Therapy For People Living With HIV in Resource-Constrained Settings

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Guidelines For Intensified Tuberculosis Case-Finding and Isoniazid Preventive Therapy For People Living With HIV in Resource-Constrained Settings

Uploaded by

Copyright:

Available Formats

Guidelines for intensified tuberculosis case-finding and isoniazid preventive therapy for people living with HIV in resourceconstrained

Department of HIV/AIDS Stop TB Department World Health Organization, Geneva, Switzerland

(NLM classification: WF 220)

Abbreviations and acronyms

Background and process

1. Background and process

1.2 Target audience

1.4 Process of formulating the guidelines

Background and process

Evidence level High Moderate Low Very low

1.5 Strength of recommendations

Strength of recommendation Strong Conditional (weak)

1.6 Adaptation of the guidelines

2.2 Efficacy, regimen and duration

2.2.2 Regimen and duration

2.2.2.1 Table 1: Comparison of the efficacy of different drug regimens

2.2.3 Immune status and concomitant use of IPT with ART

2.2.4 Pregnant women

2.2.5 Patients previously treated for TB (secondary prophylaxis)

2.2.6 Special populations

Assess for contraindications to IPT No Give IPT Yes Defer IPT

FOOTNOTES TO ALGORITHM FOR ADULTS

2.3 Detecting latent TB infection in resource-constrained settings

2.3.2 Interferon-gamma release assays (IGRA)

2.4 Issues to consider for implementation of IPT

2.4.1 Primary ownership by HIV service providers

2.4.2 IPT and drug-resistant TB

2.4.3 Adherence and clinical follow up

2.4.4 Cost-effectiveness of IPT

Intensified tuberculosis case-finding and prevention in children living with HIV

3.2 Regimen and duration

Weight range (kg)

Dose given (mg)

3.3 Secondary prophylaxis and IPT with ART in children

3.3.1 Secondary prophylaxis

3.3.2 IPT with ART in children

Intensified tuberculosis case-finding and prevention in children living with HIV

FOOTNOTES TO ALGORITHM FOR CHILDREN

4.1 Screening for TB

4.2 Preventive treatment for TB

4.3 Operational research

6. Selected GRADE profiles

Selected GRADE profiles

GRADE profile table 2: TB screening for children

Positive predictive value 0.15 1 Low

Three out of 8 studies showed an opposite direction of the effect

Different direction of the effect across the studies

Selected GRADE profiles

Summary of findings Effect

123/1984 (6.2%) 2% 50%

419/1984 (21.1%) 5% 50%

43/171 (25.1%) 10% 50%

33/1873 (1.8%) 0% 20%

Selected GRADE profiles

Summary of findings Effect

46/618 (7.4%) 2% 50%

84/618 (13.6%) 2% 50%

38/25 (152%) 10% 50%

Selected GRADE profiles

Summary of findings Effect

54/1193 (4.5%) 2% 50%

298/1193 (25%) 2% 50%

32/146 (21.9%) 10%