Thesis

UvA-DARE (Digital Academic Repository)
Interventions for anaemia in children living in a resource-poor setting: Malawi
Esan, M.O.
Publication date
2012
Document Version
Final published version
Link to publication
Citation for published version (APA):

Esan, M. O. (2012). Interventions for anaemia in children living in a resource-poor setting:
Malawi. Rozenberg Publishers.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:11 Jul 2022
Interventions for anaemia in children
living in a resource-poor setting:
Michael Olusegun Esan

Malawi
21
Michael Olusegun Esan Interventions for anaemia in children living in a resource-poor setting: Malawi
Page |1
Interventions for anaemia in children living in a resource-
poor setting: Malawi

Page |2
All published papers and illustrations were reprinted with the permission of the copyright
owner.
© Copyright 2012, Michael O. Esan.
All rights reserved. No part of this publication may be reproduced, stored or transmitted in
any form or by any means without the written permission of the copyright owner, or where
applicable, of the publishers of the scientific journals.
Chapter 2 & 7: Copyright © 2012, Elsevier Ltd.
Chapter 5: Copyright © 2010, British Blood Transfusion Society.
Chapter 6: Copyright © 2011, Blackwell Publishing Ltd.
ISBN:
Printed by:
Rozenberg Publishers
Lindengraght 302 d+e
1015 KN, Amsterdam
The Netherlands
Tel: (+31) (0) 206255429
info@rozenbergps.com
www.rozenbergps.com
Financial support: The studies described in this thesis were supported financially by the
Netherlands-African partnership for capacity development and clinical interventions
against poverty-related diseases (NWO-NACCAP), under the College of Medicine, Malawi-
Amsterdam-Liverpool (COMMAL) partnership programme. The funders played no part in
the study design, data collection, analysis, and interpretation or writing of the scientific
papers.
Page |3
Interventions for anaemia in children living in
a resource-poor setting: Malawi
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam
op gezag van de Rector Magnificus
Prof. dr. D. C. van den Boom
ten overstaan van een door het college voor promoties ingestelde
commissie, in het openbaar te verdedigen in de Agnietenkapel
op vrijdag 16 november 2012, om 12:00 uur
door
Michael Olusegun Esan
geboren te Lagos, Nigeria

Page |4
Promotiecommissie
Promotor: Prof. dr. B. J. Brabin, University of Amsterdam
Co-promotors: Dr. K. S. Phiri, University of Malawi

Dr. M. Boele van Hensbroek, University of Amsterdam
Overige leden: Prof. dr. M. Grobush, University of Amsterdam

Prof. dr. C. C. John, University of Minnesota
Prof. dr. I. Bates, Liverpool School of Tropical Medicine
Dr. D. Pajkrt, University of Amsterdam
‘All things are possible to him who believes…’ Mark 9:23
This PhD thesis is dedicated to my mother, (late) Mrs. Margeret Titilola Esan and to my
aunt, (late) Prof. (Mrs.) Jadesola Akande.
Page |5
Contents:
Chapter 1. Introduction 7
Chapter 2. Iron deficiency in children with HIV-associated anaemia: a systematic 31

review and meta-analysis.
Esan M, Jonker F, Boele van Hensbroek M, Calis JJ, Phiri K.S
Trans R Soc Trop Med Hyg 2012; Oct 106(10): 579-587.
Chapter 3. Benefits and risks of iron supplementation in HIV-infected children: 48

results of a double-blind, randomized, controlled trial.
Esan M.O., Boele van Hensbroek M., Nkhoma E.C., Musicha C.,
White S.A., ter Kuile F.O., Phiri K.S. Submitted
Chapter 4. Iron supplementation in HIV-infected children improves iron status: 67

results of a double-blind, randomized, controlled trial.
Esan MO, Phiri KS, Truwah ZT, Musicha C, Kraaijenhagen RJ,
Boele van Hensbroek M. Submitted
Chapter 5. Development and evaluation of a new paediatric blood transfusion 78

protocol for Africa.
B. Cheema, E. M. Molyneux, J. C. Emmanuel, B. M’baya, M.
Esan, H. Kamwendo, L. Kalilani-Phiri & M. Boele van Hensbroek.
Transfusion Medicine, 2010 Jun; 20(3):140 – 151.
Chapter 6. High transfusion failure rates in Malawian children with severe 98

anaemia following a standard transfusion regimen.
Esan MO, Phiri KS, Molyneux EM, Mukaka M, Cheema B.,
and Boele van Hensbroek M.
British Journal of Haematology, 2011 Sept; 154(6); 783-785.
Chapter 7. Intermittent Preventive Therapy in the post-discharge management 105

of severe malarial anaemia in pre-school children; a multi-centre
randomized placebo controlled trial in Southern Malawi.
Kamija Phiri, Michael Esan, Michael Boele van Hensbroek,
Carole Khairallah, Brian Faragher and Feiko O. ter Kuile.
Lancet Infectious Diseases 2012 Mar 12(3); 191-200.
Chapter 8. Discussion & Conclusions. 127

Summary 139
Samenvatting 144
Acknowledgements 148
Author details 152
Page |6
Abbreviations
Hb Haemoglobin
WHO World Health Organisation
G6PD Glucose 6 phosphate dehydrogenase
HIV Human immunodeficiency virus
MDG Millenium Development Goals
HAART Highly active anti-retroviral therapy
IPTp Intermittent preventive therapy in pregnancy
IPTi Intermittent preventive therapy in infants
IPTc Intermittent preventive therapy in children
IPTpd Intermittent preventive therapy post-discharge
ACTs Artemisinin combination therapies
sTfR Soulble transferrin receptor
sTfR-F Soluble transferrin receptor-log ferritin index
MCV Mean corpuscular volume
MCHC Mean cell haemoglobin concentration
aMD Adjusted mean difference
aPR Adjusted prevalence ratio
Delta RR Change in respiratory rate
Delta HR Change in heart rate
Delta Hb Change in haemoglobin concentration
Delta CD4% Change in CD4 percentage
PC Packed red blood cells
WB Whole blood
Page |7
Chapter 1
Introduction
Page |8
Anaemia is defined as a state in which there are an insufficient number of red blood cells to
cater for the body’s physiologic demands.1 Haemoglobin (Hb), the oxygen-carrying protein
in red blood cells, is generally used to quantify the level of anaemia. Anaemia results in a
body oxygen deficit which manifests as a range of clinical symptoms and signs, from easy
fatigability to heart failure, depending on the severity. Anaemia can be classified into mild,
moderate and severe based on the Hb level and a combination of clinical signs and
symptoms1, 2 (Table 1). Anaemia is responsible for a diverse range of effects on growth,
work capacity, cognitive and behavioral development and contributes significantly to
maternal and child mortality.3-7 The definition of anaemia varies by age, sex, race, altitude,
geographic location, smoking and pregnancy status.2, 8-10 The intervention studies reported
in this thesis examined anaemia in children less than 5 years old; anaemia in this age group
is defined as a body Hb level of less than 11g/dl2, 11 (Table 1).
Anaemia affects nearly 2 billion people worldwide and about 50% of all children less than 5
years old.11-13 In Africa, the prevalence of anaemia in pre-school children (<5 years) is 67%
and anaemia is classified as a severe public health problem (defined as a prevalence of
anaemia >40%) in most parts of the African continent, including in Malawi where the
studies reported in this thesis were performed11 (Figure 1).
Iron deficiency is believed to be responsible for about half of the cases of anaemia seen in
women and children and is the most important single causal factor for the development of
anaemia. 2, 12 Other factors which have been found to contribute to the aetiology of anaemia
in resource-limited settings include poor dietary intake and intestinal absorption of
micronutrients (especially vitamins A, B12 and folate), hereditary conditions affecting red
blood cells (sickle cell anaemia, glucose-6-phosphate dehydrogenase (G6PD) deficiency,
thallasemia) and infections such as malaria, hookworm, schistosomiasis and HIV13-16
(Figure 2). Some of these aetiological factors (iron deficiency, malaria and HIV) will be
discussed in further detail in this thesis. Women of child-bearing age and young children
are at greatest risk for anaemia and are the target of most public health interventions.2
This PhD thesis focuses on studies looking at interventions for anaemia in pre-school
children (less than 5 years old) with different degrees of anaemia, living in a resource-
limited setting with a high pressure of infections, all done with a view towards achieving
Page |9
the WHO millennium development goal number 4 (MDG4) - a reduction in child

mortality.17
Table 1: Haemoglobin levels to diagnose anaemia at sea-level (g/l)

Anaemia
Population Non- Mild Moderate Severe
Anaemia
Children (6-59 months)* >110 100-109 70-99 <70
Children (5-11 years) >115 110-114 80-109 <80
Children (12-14 years) >120 110-119 80-109 <80
Non-pregnant women >120 110-119 80-109 <80
(>15years)
Pregnant women* >110 100-109 70-99 <70
Men (>15 years) >130 110-129 80-109 <80
Source: WHO 1, 2
*High-risk groups for anaemia
P a g e | 10
Malawi
Figure 1: Anaemia as a global public health problem in pre-school children (<5 years of age)
*Areas in red (including Malawi) are countries where the prevalence of anaemia in pre-school children is >40% (source: WHO11)
P a g e | 11
Figure 2: Aetiological factors associated with development of anemia in resource-limited settings
Iron deficiency
Iron deficiency is the most important nutritional deficiency worldwide, and is responsible
for up to 50% of anaemias seen in young children and pregnant women.2, 11 Iron deficiency
can be a result of reduced dietary intake, poor bioavailability and intestinal malabsorption,
acute or chronic infections, increased iron requirements (during pregnancy and growth)
and blood loss. Iron deficiency has been associated with poor growth, immunological
impairments, cognitive and behavioural deficits, and reduced work capacity.5, 18-24
Universal iron supplementation is recommended for the prevention and treatment of

anaemia in regions where the prevalence of anaemia in pre-school children is greater than
40%- this includes most of Africa, South America and South East Asia2, 11 (Figure 1).
However, iron deficiency was significantly less prevalent in severely anaemic Malawian
children when compared with non-severely anaemic controls.25 In addition, iron deficiency
has been found to be associated with a reduced risk of malaria in young children and
pregnant women;26-29 and was protective against bacterial infections.30 Subsequently, iron
supplementation has been found to be associated with an increased risk of infections,
malaria-associated morbidity and mortality in malaria-endemic regions.31-33 This has led
the WHO to advise that iron status be determined prior to the use of iron supplements in
P a g e | 12
young children living in regions with a high prevalence of malaria and other infectious
diseases.34 However this approach is impractical in resource-limited settings where simple,
affordable and reliable tests of iron status are not available and ferritin, which is commonly
used for the determination of iron status is also a mediator of the acute phase response and
may be difficult to interpret in the presence of inflammation.13, 35, 36 This challenge is even
greater in settings where iron deficiency, malaria and HIV are all highly prevalent, and the
benefits or risks of iron supplementation are unknown.37
HIV
Since it was first discovered in 1984, human immunodeficency virus (HIV) infection has
become a global epidemic, with an estimated 34 million people living with the disease and
1.8 million AIDS-related deaths worldwide in 2010.38 HIV (the virus that causes AIDS)
affects the body’s defenses, especially it’s ability to generate an immune response to
infection, thereby increases its susceptibility to opportunistic infections. The effect of the
epidemic has been particularly devastating in Africa where there were 1.9 million new HIV
infections in 2010 and 1.2 million HIV-related deaths, accounting for 69% of the global
total. Africa also has the highest regional prevalence of adults living with HIV (4.9%).38 In
Malawi, the prevalence of HIV-infection in children admitted to hospital is about 19%, and
HIV-infection accounts for about 40% of pediatric in-patient deaths.39 Anaemia is a major
complication of HIV-infection, even in the era of increased availability of highly active anti-
retroviral therapy (HAART)- and has been found to be independently associated with HIV
disease progression and mortality.40-42 In HIV-infected people who develop anaemia, the
treatment of anaemia has been found to be associated with improved survival when
compared with not giving any treatment.43 The benefits of treatment of HIV-associated
anaemia on quality of life and mortality have been well-described in previous studies.44, 45
Although iron deficiency is an important cause of anaemia, its role in HIV-associated
anaemia is unclear.46 Defining the role of iron deficiency in HIV is important as most public
health interventions for anaemia in resource-limited settings are based on iron
supplementation or fortification programmes, and iron supplementation is seen as a
relatively cost-effective strategy for preventing and controlling anaemia.2, 13 Most areas
that are highly endemic for anaemia are also endemic for malaria, HIV-infection and other
infectious diseases. In vitro studies indicate that HIV-infection alters iron metabolism
P a g e | 13
resulting in increased iron accumulation in tissues.47, 48 These result in anaemia of chronic

disease (or inflammation), a functional iron deficiency in which iron is withheld in tissues
and is not made available for red cell production.49-51 This hypothesis is important, as red
cell production failure is believed to be the most important patho-physiologic mechanism
associated with anaemia in HIV infection.52, 53 This readily available iron in tissues favours
replication of not only the HIV, but also other pathogens such as non-typhoid Salmonella.54
There is evidence from previous observational studies that ‘high’ iron status may adversely
affect the outcome of HIV infection by increasing the rate of disease progression and hence
decreasing survival times.55, 56 Thus, iron supplementation in this context may be
detrimental.
Malaria
About half of the world’s population are at risk of malaria.57 Malaria was responsible for
655,000 deaths in 2010, most of which were in African children.58 Malaria is a protozoan
infection caused by the bite of a female anopheles mosquito infected with the Plasmodium
parasite. Four of the five known Plasmodium species that commenly cause human
infections are: P. Falciparum, P. Vivax, P. Ovale and P. Malariae, but of these four, P.
Falciparum is the most important. Cerebral malaria and severe malaria anaemia from P.
falciparum infections are the two clinical malaria syndromes associated with the highest
risk of mortality in children.59, 60 Of these two deadly syndromes, severe malaria anaemia
was looked at in more detail in this thesis. Severe malaria anaemia is defined by the WHO
as a haemoglobin of <5 g/dl (or haematocrit of <15%) in the presence of P. Falciparum
asexual parasitemia.61 Severe malaria anaemia is associated with significant in-hospital and
post-discharge mortality.62-69 Deaths usually occur within the first 12 hours of hospital
admission, but a significant amount of mortality has been found to occur within 6 months
of discharge home, usually following another severe malaria episode.63, 64, 66, 70 Blood
transfusion as an intervention for symptomatic severe malaria anaemia has been found to
be associated with improved survival, but presence of malaria parasitemia on follow-up
after an episode of severe malaria anaemia negates the benefits of blood transfusion.63, 64,
71-73 Recovery following a severe malaria anaemia episode is slow; even following radical
clearance of parasitaemia, complete haematological recovery can take up to six weeks and
may be further delayed by co-morbidities such as multi-micronutrient deficiencies, repeat
P a g e | 14
or recrudescent malaria infections and helminth infections.74-77 In areas with moderate to

intense malaria transmission, repeated attacks of malaria infection following recent
hospitalization for severe malaria anaemia result in persistent bone marrow suppression
and dyserythropoiesis. 78 This results in a significantly increased risk of post-discharge
mortality and severe morbidity, as the bone marrow has not sufficiently recovered and
haemoglobin levels are still low. We hypothesized that if we are able to provide radical cure
for severe malaria and subsequently prevent repeated attacks of malaria by anti-malarial
chemo-prophylaxis, we could provide a time window that would allow for adequate bone
marrow recovery following blood transfusion, allowing haemoglobin concentrations to be
restored to normal levels and thus preventing readmission to hospital for a repeat severe
anaemia episode (Chapter 7).
Interventions for anaemia in resource-limited settings

Anaemia can be caused by a variety of aetiological factors, and this should be kept in mind
when planning for interventions. While iron deficiency is an important cause of anaemia in
resource-limited settings, it is not the only cause of anaemia and can exist alone
independently of anaemia. An integrated approach, taking into account the role of other
aetiological factors such as other micronutrient deficiencies (vitamin B12, folate) and
concomittant infections (malaria, HIV, schistosomiasis, hookworm); as well as improved
sanitation, health care and feeding practices needs to be adopted if such interventions are
to be effective2, 13 (Table 2). Provision of addtitional iron by food fortification and dietary
diversification has been used to prevent iron deficiency anaemia in populations at risk.2, 11
In malaria-endemic regions, the provision of malaria chemoprophylaxis and intermittent
preventive therapy in pregnant women (IPTp), infants (IPTi) and in children (IPTc) along
with iron and folic acid supplementation has been shown to be beneficial, even in HIV-
infected individuals.79-84 In regions with a high prevalence of hookworm and helminth
infestation, deworming exercises have been shown to increase haemoglobin
concentrations thus reducing anaemia prevalence.85-87
In the course of this thesis, we assessed the management of anaemia at different stages of
it’s development. Firstly, we examined the effects of oral iron supplementation for the
treatment of moderate anaemia in HIV-infected children. We then evaluated a new blood
transfusion protocol (adopted from the WHO transfusion guidelines) by assessing
P a g e | 15
adherence, and observing the clinical and haematologic responses of severely anaemic
children to a standard, WHO-recommended, blood transfusion. Finally, we investigated a
novel approach to the management of severe anaemia following discharge home from the
hospital, in an effort to reduce post-discharge morbidity and mortality, which has
previously been found to be high in this setting.66
P a g e | 16
Table 2: Interventions to reduce anaemia prevalence in children: a population-based

approach *
1. Prevention of iron deficiency anaemia
i. Dietary measures
- Cattle meat, poultry and fish-based diets; legumes and green leafy vegetables are rich in bio-
available iron
- Fruits and vegetables rich in vitamins A and C enhance iron absorption
- Phytates present in cereal bran, cereal grains and nuts; iron-binding tannins found in tea,
coffee and cocoa; and calcium from milk inhibit iron absorption.
- Has the potential for long term sustainability if successfully implemented.
ii. Iron fortification of foods

- Potential for multiple nutritional benefits
- Potential for long-term sustainability
- Efforts must be balanced against the prevalence of infections, especially in malaria-endemic
regions.
iii. Iron supplementation of groups at risk for 3 months (where prevalence of anaemia >40%)
- Can be implemented at the community level.
- At-risk groups include all women of child bearing age, adolescent girls and children younger
than 5 years old.
- Dose: 2mg/kg up to 30mg/day in children <2 years old; 2mg/kg up to 30mg/day in children
2-5 years old; 60mg/day given with 400µg/day of folic acid in women of child bearing age
and adolescent girls.
Challenges:
- Faeces may turn black (not harmful, iron supplements should not be stopped)
- Epigastric discomfort, nausea, diarrhea or constipation may occur with doses ≥60mg/day
- Iron inhibits the absorption of certain antibiotics (tetracyclines, sulphonamides,
trimethoprim)
2. Malaria prophylaxis (in malaria-endemic regions)

- Insecticide treated bed-nets.
- In-door residual spraying.
- Intermittent preventive therapy in children and pregnant women.
3. Micronutrient malnutrition control
4. Control of hookworm and helminth infections
5. Improved environmental sanitation
6. Childhood Immunizations
7. Improved community-based primary health care
*Source: WHO 2
P a g e | 17
Table 3: Interventions for anaemia: a health-facility based approach*

Type of anaemia Management
Non-severe1
Mild • Treatment dose of iron tablet/syrup (3mg/kg) to be given at home
Moderate for 2 weeks initially, then review.
• Give oral iron for up to 3 months (2-4 weeks to correct anaemia;

then 1-3 months to build normal iron stores).
• If child has not been treated for hookworm infection in preceding 6

months, give one dose of mebendazole 500mg.
• Advise mother about good feeding practices.
Severe2 • Blood transfusion as soon as possible, preferably with packed cells

10mls/kg body weight over 3-4 hours. If packed cells are not
available, give fresh whole blood 20mls/kg over 3-4 hours.
• Check pulse rate and respiratory rate every 15 minutes. If either

rises, transfuse more slowly.
• If there is evidence of fluid overload due to blood transfusion, give

IV furosemide 1-2mg/kg body weight up to a maximum of 20mg.
• If after transfusion Hb remains as before, repeat transfusion.

1Defined as Hb <11 g/dl in children <6 years old
2Defined as Hb of <4 g/dl or <6 g/dl with signs of severe anaemia (deep, labored breathing, heart
failure, shock, clinical dehydration)
*Source: WHO 88
P a g e | 18
Methodology
Study setting: Malawi

Malawi, popularly known as the warm heart of Africa, is a land-locked country located in
South-east Africa. It is bordered by Mozambique to the south, southwest and east; Zambia
to the west and north west; and Tanzania to the north and north east. It is 901 km long and
between 80 to 161 km wide, with a land mass of 118, 484 sq km. Lake Malawi, Africa’s
third largest lake, makes up about 20% of the country’s land mass and runs along the
eastern part of the country, bordering Mozambique. The population of Malawi is rapidly
growing, with 13,077,160 people reported from the 2008 National Population and Housing
Census, up 32% from the 1998 census.89 Approximately 85% of the population live in the
rural areas, mostly existing as subsitence farmers. The country is made up of 3 regions
(Northern, Central and Southern) which are further made up of 28 districts, 13 of which are
in the Southern region. The official languages are English and Chichewa, and the population
is made up of several ethnic groups- the major ones are Chewa, Ngoni, Tumbuka, Yao, and
Lomwe. The climate is generally subtropical, with rainfall from December to April every
year, then a cool period from May till August and finally hot weather is experienced
between September and November.
Malawi is one of the world’s least-developed countries, with a gross domestic product
(GDP) per capita of US $ 870 and an agriculture-based economy, with the export of tobacco,
tea, sugar and coffee accounting for more than 80% of total export revenue.90
Healthwise, the infant mortality rate is 66 deaths per 1000 live births; the under-5
mortality rate is 112 deaths per 1000 live births. The life expectancy is 47 years. The
prevalence of anaemia in children 6 – 59 months old in Malawi is 63%, with the prevalence
higher in children from rural areas and of lower socio-economic status. The prevalence of
HIV in adults aged 15-49 years in Malawi is 10.6%.91
The studies discussed in this thesis were conducted in 4 different districts in the Southern
region of Malawi- Thyolo, Zomba, Blantyre and Chikhwawa (Figure 3).
P a g e | 19
Figure 3: Map of Malawi (Source: MDHS 2010 91)

P a g e | 20
Study sites
1. Thyolo
Thyolo district is located within the Shire Highlands, and is famous for its numerous tea
plantations. The district has a total land area of 1,715 sq km which is 2% and 19% of total
land area for Malawi and the Southern Region respectively. Thyolo district has a population
of 587,053 persons; it is estimated that 44% of the population derive their livelihood from
agriculture.89
The prevalence of anaemia in children aged 6-59 months in Thyolo district is 49%, and the
under-five mortality rate is 122 deaths per 1000 live births.91 The district has 2 referral
hospitals, which serve the whole population of Thyolo:
i. Thyolo district hospital, a government-owned facility run by the Ministry of Health
ii. Malamulo hospital, owned by Christian Health Association of Malawi (CHAM).
Thyolo district hospital was selected as a study site because it has well established
pediatric out-patient HIV-clinics organized by doctors without borders (MSF) based at the
hospital. This provided a good platform for recruiting participants for the iron
supplementation trial reported in this thesis (chapters 3 and 4). Thyolo district has
seasonal malaria transmission, with increased hospital visits during the rainy season
mostly due to malaria infections, also making it a suitable location for recruiting
participants for the the intermittent preventive therapy post-discharge (IPTpd) trial
discussed in detail in chapter 7.
2. Zomba
Zomba, Malawi’s first capital, is located in the South-eastern part of Malawi, on the lower
slopes of the Zomba mountain 60 km northeast of Blantyre (Figure 3). It covers an area of
2,580 square kilometres and has a population of 667,953 people. Majority of the
economically active population in Zomba are subsistence farmers (57.2%).89
Zomba Central Hospital is the main referral centre in the district. The prevalence of
anaemia in children aged 6-59 months in Zomba district is 64%, and Zomba has one of the
highest under-five mortality rates in the country, at 132 deaths per 1,000 live births.91
Zomba was chosen as a study site because it has well-organised paediatric out-patient HIV
clinics being run by Digitas organisation within the hospital, which provides free highly
P a g e | 21
active ant-retroviral therapy (HAART) initiation and monitoring services to HIV-infected

children in the district. This provided an ideal platform for recruiting participants for the
iron supplementation trial (chapters 3 and 4). The paediatric emergency department at
Zomba Central Hospital has a heavy patient flow especially during the months with heavy
rainfall (December to April), mostly due to malaria infections- thus also making it a good
location for recruiting participants for the IPTpd trial (chapter 7).
3. Blantyre:
Blantyre is the commercial capital of Malawi and with a population of just over 1 million
people, is Malawi’s second most populous district, after Lilongwe. The district lies 800m-
1600m above sea level. Only 13% of its population derive their livelihood from agriculture.
Blantyre has one of the highest literacy rates in the country, with 89% of men and 85% of
women able to read a sentence or part of a sentence.91 Blantyre has the lowest prevalence
of anaemia in children aged 6-59 months in Malawi (44%) and one of the lowest under-five
mortality rates in the Southern region, at 110 deaths per 1000 live births.91 Queen
Elizabeth Central Hospital is a government-run, 1000-bed facility which serves as the
district’s main referral centre and also as the teaching hospital for the country’s only
College of Medicine, based at the University of Malawi. Blantyre has seasonal malaria
transmission (usually during the rainy season from December to April every year), with an
entomological innoculation rate (EIR) estimated at just one infectious bite per person per
year.25 Blantyre was chosen as a study site because of the high patient flow seen in the
pediatric emergency department, especially during the rainy season. It is estimated that
there are 80,000 sick-child visits per year to the pediatric emergency department, and
23,000 hospital admissions.92 In addition, a research clinic had been established at the
pediatric accident and emergency centre for a case-control study previously conducted by
our study group.25 This research clinic was utilized for recruiting participants for IPTpd
trial reported in this thesis. The paediatric accident and emergency centre was also used to
recruit participants for the blood transfusion studies reported in chapters 5 and 6.
4. Chikhwawa:
Chikhwawa is located in the lower Shire Valley, about 50km south-west of Blantyre (Figure
3). It is bordered by Blantyre to the north, Nsanje to the south, Thyolo to the east and
P a g e | 22
Mwanza to the west. It has a population of 434,648 people, with 56% of the population
working as subsistence farmers.89, 91 The prevalence of anaemia in children aged 6-59
months in Chikhwawa district is 75%, the second highest in the country. The under-five
mortality rate is 139 deaths per 1000 live births.91 Chikhwawa District Hospital serves as
the main referral centre in the area. Chikhwawa district has year-round malaria
transmission with an EIR of approximately 170 infectious bites per person per year.25 It
also has relatively warm weather compared with other parts of the country, with
temperatures rising as high as 420C during the dry, hot season (September to November).
Chikhwawa was chosen as a study site for the IPTpd study because of its contrasting
malaria transmission pattern compared with other study sites. There is also a well-
established research clinic and study team already set-up in the paediatric ward of
Chikhwawa District Hospital from a previous large case-control study carried out by our
research group,25 making the task of recruiting patients for a new trial at this site easier.
Studies:
Three separate studies were conducted during the course of this PhD; one descriptive,
cross-sectional study and two randomised, double-blinded, controlled trials. Firstly, we
reviewed the literature on iron deficiency in HIV-infected children to determine the
prevalence of iron deficiency in HIV-infected children from high and low-income settings
and compared it with that of HIV-uninfected controls (chapter 2). The research studies are
then described in detail in chapters’ three to seven. The three main studies described in
this thesis are:
1. Iron supplementation in HIV-infected Malawian children: Is it safe and beneficial?
The World Health Organization (WHO) recently advised that iron status be determined
prior to the use of iron supplements in young children living in regions with a high
prevalence of malaria and infectious diseases due to an increased risk of mortality.34
However, this recommendation is impractical in resource-limited settings, as simple,
affordable and reliable tests for the determination of iron status are not available and
ferritin, which is also a marker of the acute phase response, is difficult to interpret due to
inflammation.13, 35, 36 Iron supplementation is still used to treat anaemia in children living in
P a g e | 23
resource-limited settings, including HIV-infected children, but the potential benefits or

risks of iron supplementation in HIV-infected children are not known.37 We carried out a
randomised, double-blind, controlled trial of multi-vitamins with iron supplementation
versus multivitamins alone in anaemic HIV–infected children aged 6-59 months to
determine the effect of iron on haemoglobin, HIV disease progression and morbidity
(chapter 3). The objectives of this trial were:
Primary Objective:
To determine whether iron supplementation in HIV-infected children is associated with an

increased risk of morbidity.
Secondary objectives:
To investigate in HIV-infected children with anaemia:
- The erythropoietic responses to iron supplementation

- The effect of iron supplementation on HIV disease progression
- The effect of iron supplementation on the prevalence of infections
Other specific outcomes that were examined include the effect of iron on:
1. All-cause sick out-patient clinic visits

2. All-cause hospital admissions
3. Incidence of malaria and respiratory infections during trial follow-up.
4. Change in haemoglobin levels during the follow-up period
5. Prevalence of anaemia and iron deficiency at recruitment, 3 and 6 months follow-up
6. Immune response (CD4 count and percentage) over the trial follow-up period.
7. Changes in different serological iron markers observed during follow-up period
(chapter 4).
2. Development and evaluation of a new paediatic blood transfusion protocol

i. Development and evaluation of a new paediatric blood transfusion protocol for Africa
(chapter 5).
Severe malaria anaemia contributes significantly to child mortality during the rainy season
in Malawi.65 Previous studies have shown that transfusion protocols when strictly adhered
to can significantly reduce the number of transfusions without increasing mortality.93, 94
P a g e | 24
The WHO have published a set of guidelines for the management of severe anaemia in
resource-limited settings,95 however this has not been evaluated in a busy African hospital
setting.
Primary Objective
To develop a new paediatric blood transfusion protocol (adopted from the WHO
transfusion guidelines) for use in under-resourced environments and to evaluate its
usability in a busy African hospital setting.
ii. High transfusion failure rates in Malawian children with severe anaemia following the
standard, WHO-recommended blood transfusion regimen (chapter 6).
In order to better understand the clinical and haematologic responses of severely anaemic
children to a standard, WHO-recommended blood transfusion, we carried out a descriptive,
cross-sectional study as part of a new blood transfusion protocol assessment at Queen
Elizabeth Central Hospital, Blantyre. We aimed to evaluate the current WHO guidelines for
blood transfusion in children admitted to hospital for severe anaemia (defined as Hb <6
g/dl with signs and symptoms of anaemia) due to any cause, focussing on transfusion
failures (defined as Hb <6 g/dl 24 hours after blood transfusion) and the association
between the resolution of clinical symptoms and haematological recovery in severely
anaemic Malawian children.
Primary objective:
To evaluate the current WHO transfusion guidelines by observing the clinical and
haematologic responses of severely anaemic children to blood transfusion.
- To determine the effect of blood transfusion on haemoglobin levels.

- To assess risk factors associated with transfusion failure.
P a g e | 25
3. Intermittent preventive therapy with monthly artemether-lumefantrine for the post-

discharge management of severe anaemia in children aged 4-59 months in Southern
Malawi (chapter 7).
Severe malaria anaemia is a significant cause of in-hospital and post-discharge mortality.

As much as 17% of children admitted to hospital for severe malaria anaemia will die within
6 months following in-patient treatment.66 In the design of this study, we hypothesized that
by giving full treatment doses of an effective anti-malaria drug at monthly intervals
following discharge home from hospital, we would give the bone marrow adequate time to
recover and build up haemoglobin levels, and thus reduce the risk of death following a
subsequent attack of malaria. Thus we carried out a randomised, double-blind controlled
trial of monthly treatment with artemether-lumefantrine versus placebo in the
management of young children (aged 4-59 months) with severe malaria anaemia (defined
as Hb <5g/dl with P.falciparum asexual parasitemia) following discharge from the hospital
to determine the effect on post-discharge morbidity and mortality.
Primary Objective:
To determine in pre-school children the effect of intermittent preventive therapy post-

discharge with monthly artemether-lumefantrine on all-cause mortality and hospital re-
admission for severe anaemia or severe malaria between 1 and 6 months.
To determine in pre-school children the effect of intermittent preventive therapy post-

discharge with monthly artemether-lumefantrine on:
- All-cause mortality
- Hospital re-admission for all-cause severe anaemia or severe malaria
- All-cause hospital admissions
- All-cause sick-child clinic visits
- Clinic visits because of microscopically confirmed non-severe malaria.
P a g e | 26
References
1. WHO. Haemoglobin concentrations for the diagnosis of anaemia and assessment of

severity. Vitamin and Mineral Nutrition Information System. Geneva, World Health
Organization, 2011 (WHO/NMH/NHD/MNM/11.1) 2011.
2. WHO. Iron Deficiency Anaemia Assessment, prevention and control: A guide for
programme managers. 2001.
3. Chang SC, O'Brien KO, Nathanson MS, Mancini J, Witter FR. Hemoglobin
concentrations influence birth outcomes in pregnant African-American adolescents. J Nutr
2003;133(7):2348-55.
4. Ai Y, Zhao SR, Zhou G, Ma X, Liu J. Hemoglobin status associated with performance
IQ but not verbal IQ in Chinese preschool children. Pediatr Int.
5. Carter RC, Jacobson JL, Burden MJ, et al. Iron deficiency anemia and cognitive
function in infancy. Pediatrics;126(2):e427-34.
6. Scholz BD, Gross R, Schultink W, Sastroamidjojo S. Anaemia is associated with
reduced productivity of women workers even in less-physically-strenuous tasks. Br J Nutr
1997;77(1):47-57.
7. Scholl TO, Hediger ML. Anemia and iron-deficiency anemia: compilation of data on
pregnancy outcome. Am J Clin Nutr 1994;59(2 Suppl):492S-500S discussion S-1S.
8. Milman N, Byg KE, Agger AO. Hemoglobin and erythrocyte indices during normal
pregnancy and postpartum in 206 women with and without iron supplementation. Acta
Obstet Gynecol Scand 2000;79(2):89-98.
9. Johnson-Spear MA, Yip R. Hemoglobin difference between black and white women
with comparable iron status: justification for race-specific anemia criteria. Am J Clin Nutr
1994;60(1):117-21.
10. Smith JR, Landaw SA. Smokers' polycythemia. N Engl J Med 1978;298(1):6-10.
11. WHO. Worldwide prevalence of anaemia 1993 - 2005: WHO Global Database on
Anaemia. 2008.
12. WHO. Micronutrient deficiencies: iron deficiency anaemia.
13. WHO. Assessing the iron status of populations. 2007.
14. Semba RD, Bloem MW. The anemia of vitamin A deficiency: epidemiology and
pathogenesis. Eur J Clin Nutr 2002;56(4):271-81.
15. Sim JJ, Lac PT, Liu IL, et al. Vitamin D deficiency and anemia: a cross-sectional study.
Ann Hematol;89(5):447-52.
16. Milman N. Anemia--still a major health problem in many parts of the world! Ann
Hematol;90(4):369-77.
17. WHO. World Health Statistics 2010. Part 1:Health-related Millenium Development
Goals. 2011.
18. Lawless JW, Latham MC, Stephenson LS, Kinoti SN, Pertet AM. Iron supplementation
improves appetite and growth in anemic Kenyan primary school children. J Nutr
1994;124(5):645-54.
19. Oppenheimer SJ. Iron and its relation to immunity and infectious disease. J Nutr
2001;131(2S-2):616S-33S; discussion 33S-35S.
20. Haas JD, Brownlie Tt. Iron deficiency and reduced work capacity: a critical review of
the research to determine a causal relationship. J Nutr 2001;131(2S-2):676S-88S;
discussion 88S-90S.
P a g e | 27
21. Thibault H, Galan P, Selz F, et al. The immune response in iron-deficient young
children: effect of iron supplementation on cell-mediated immunity. Eur J Pediatr
1993;152(2):120-4.
22. Soewondo S, Husaini M, Pollitt E. Effects of iron deficiency on attention and learning
processes in preschool children: Bandung, Indonesia. Am J Clin Nutr 1989;50(3 Suppl):667-
73; discussion 73-4.
23. Latham MC, Stephenson LS, Kinoti SN, Zaman MS, Kurz KM. Improvements in
growth following iron supplementation in young Kenyan school children. Nutrition
1990;6(2):159-65.
24. Chandra RK. Reduced bactericidal capacity of polymorphs in iron deficiency. Arch
Dis Child 1973;48(11):864-6.
25. Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J
Med 2008;358(9):888-99.
26. Gwamaka M, Kurtis JD, Sorensen BE, et al. Iron deficiency protects against severe
Plasmodium falciparum malaria and death in young children. Clin Infect Dis;54(8):1137-44.
27. Nyakeriga AM, Troye-Blomberg M, Dorfman JR, et al. Iron deficiency and malaria
among children living on the coast of Kenya. J Infect Dis 2004;190(3):439-47.
28. Jonker FA, Calis JC, van Hensbroek MB, et al. Iron status predicts malaria risk in
malawian preschool children. PLoS One;7(8):e42670.
29. Kabyemela ER, Fried M, Kurtis JD, Mutabingwa TK, Duffy PE. Decreased
susceptibility to Plasmodium falciparum infection in pregnant women with iron deficiency.
J Infect Dis 2008;198(2):163-6.
30. Paradkar PN, De Domenico I, Durchfort N, Zohn I, Kaplan J, Ward DM. Iron depletion
limits intracellular bacterial growth in macrophages. Blood 2008;112(3):866-74.
31. Oppenheimer SJ, Gibson FD, Macfarlane SB, et al. Iron supplementation increases
prevalence and effects of malaria: report on clinical studies in Papua New Guinea. Trans R
Soc Trop Med Hyg 1986;80(4):603-12.
32. Sazawal S, Black RE, Ramsan M, et al. Effects of routine prophylactic
supplementation with iron and folic acid on admission to hospital and mortality in
preschool children in a high malaria transmission setting: community-based, randomised,
placebo-controlled trial. Lancet 2006;367(9505):133-43.
33. Oppenheimer SJ, Macfarlane SB, Moody JB, Bunari O, Hendrickse RG. Effect of iron
prophylaxis on morbidity due to infectious disease: report on clinical studies in Papua New
Guinea. Trans R Soc Trop Med Hyg 1986;80(4):596-602.
34. WHO, UNICEF. Iron supplementation of young children in regions where malaria
transmission is intense and infectious disease is highly prevalent. 2006:2.
35. Ojukwu JU, Okebe JU, Yahav D, Paul M. Oral iron supplementation for preventing or
treating anaemia among children in malaria-endemic areas. Cochrane Database Syst Rev
2009(3):CD006589.
36. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to
inflammation. N Engl J Med 1999;340(6):448-54.
37. Adetifa I, Okomo U. Iron supplementation for reducing morbidity and mortality in
children with HIV. Cochrane Database Syst Rev 2009(1):CD006736.
38. WHO. Global Health Observatory (GHO): HIV/AIDS. 2012.
39. Rogerson SR, Gladstone M, Callaghan M, et al. HIV infection among paediatric in-
patients in Blantyre, Malawi. Trans R Soc Trop Med Hyg 2004;98(9):544-52.
P a g e | 28
40. Ellaurie M, Burns ER, Rubinstein A. Hematologic manifestations in pediatric HIV

infection: severe anemia as a prognostic factor. Am J Pediatr Hematol Oncol
1990;12(4):449-53.
41. Mocroft A, Kirk O, Barton SE, et al. Anaemia is an independent predictive marker for
clinical prognosis in HIV-infected patients from across Europe. EuroSIDA study group. Aids
1999;13(8):943-50.
42. Tovo PA, de Martino M, Gabiano C, et al. Prognostic factors and survival in children
with perinatal HIV-1 infection. The Italian Register for HIV Infections in Children. Lancet
1992;339(8804):1249-53.
43. Sullivan PS, Hanson DL, Chu SY, Jones JL, Ward JW. Epidemiology of anemia in
human immunodeficiency virus (HIV)-infected persons: results from the multistate adult
and adolescent spectrum of HIV disease surveillance project. Blood 1998;91(1):301-8.
44. Saah AJ, Hoover DR, He Y, Kingsley LA, Phair JP. Factors influencing survival after
AIDS: report from the Multicenter AIDS Cohort Study (MACS). J Acquir Immune Defic Syndr
1994;7(3):287-95.
45. Moore RD, Keruly J, Richman DD, Creagh-Kirk T, Chaisson RE. Natural history of
advanced HIV disease in patients treated with zidovudine. The Zidovudine Epidemiology
Study Group. Aids 1992;6(7):671-7.
46. Kosalaraksa P, Bunupuradah T, Vonthanak S, et al. Prevalence of Anemia and
Underlying Iron Status in Naive Antiretroviral Therapy HIV-Infected Children with
Moderate Immune Suppression. AIDS Res Hum Retroviruses.
47. Savarino A, Calosso L, Piragino A, et al. Modulation of surface transferrin receptors
in lymphoid cells de novo infected with human immunodeficiency virus type-1. Cell
Biochem Funct 1999;17(1):47-55.
48. Pugliese A, Gennero L, Pescarmona GP, et al. Serum citrate levels, haptoglobin
haplotypes and transferrin receptor (CD71) in patients with HIV-1 infection. Infection
2002;30(2):86-9.
49. Fuchs D, Zangerle R, Artner-Dworzak E, et al. Association between immune
activation, changes of iron metabolism and anaemia in patients with HIV infection. Eur J
Haematol 1993;50(2):90-4.
50. Butensky James E, Harmatz P, Lee M, et al. Altered iron metabolism in children with
human immunodeficiency virus disease. Pediatr Hematol Oncol 2009;26(2):69-84.
51. Collins HL. Withholding iron as a cellular defence mechanism--friend or foe? Eur J
Immunol 2008;38(7):1803-6.
52. Boele van Hensbroek M, Calis JC, Phiri KS, et al. Pathophysiological mechanisms of
severe anaemia in Malawian children. PLoS One 2010;5(9):e12589.
53. Calis JC, Phiri KS, Vet RJ, et al. Erythropoiesis in HIV-infected and uninfected
Malawian children with severe anemia. Aids 2010;24(18):2883-7.
54. Chlosta S, Fishman DS, Harrington L, et al. The iron efflux protein ferroportin
regulates the intracellular growth of Salmonella enterica. Infect Immun 2006;74(5):3065-7.
55. Costagliola DG, de Montalembert M, Lefrere JJ, et al. Dose of desferrioxamine and
evolution of HIV-1 infection in thalassaemic patients. Br J Haematol 1994;87(4):849-52.
56. McDermid JM, Jaye A, Schim van der Loeff MF, et al. Elevated iron status strongly
predicts mortality in West African adults with HIV infection. J Acquir Immune Defic Syndr
2007;46(4):498-507.
57. WHO. TDR Diseases and Topics: Malaria.
58. WHO. Malaria: fact sheet. 2012.
P a g e | 29
59. Maitland K, Marsh K. Pathophysiology of severe malaria in children. Acta Trop

2004;90(2):131-40.
60. Marsh K, English M, Crawley J, Peshu N. The pathogenesis of severe malaria in
African children. Ann Trop Med Parasitol 1996;90(4):395-402.
61. WHO. Guidelines for the treatment of malaria-- 2nd edition. 2010:35.
62. Zucker JR, Lackritz EM, Ruebush TK, 2nd, et al. Childhood mortality during and after
hospitalization in western Kenya: effect of malaria treatment regimens. Am J Trop Med Hyg
1996;55(6):655-60.
63. Lackritz EM, Campbell CC, Ruebush TK, 2nd, et al. Effect of blood transfusion on
survival among children in a Kenyan hospital. Lancet 1992;340(8818):524-8.
64. Lackritz EM, Hightower AW, Zucker JR, et al. Longitudinal evaluation of severely
anemic children in Kenya: the effect of transfusion on mortality and hematologic recovery.
Aids 1997;11(12):1487-94.
65. Slutsker L, Taylor TE, Wirima JJ, Steketee RW. In-hospital morbidity and mortality
due to malaria-associated severe anaemia in two areas of Malawi with different patterns of
malaria infection. Trans R Soc Trop Med Hyg 1994;88(5):548-51.
66. Phiri KS, Calis JC, Faragher B, et al. Long term outcome of severe anaemia in
Malawian children. PLoS One 2008;3(8):e2903.
67. Obonyo CO, Vulule J, Akhwale WS, Grobbee DE. In-hospital morbidity and mortality
due to severe malarial anemia in western Kenya. Am J Trop Med Hyg 2007;77(6 Suppl):23-
8.
68. Biemba G, Dolmans D, Thuma PE, Weiss G, Gordeuk VR. Severe anaemia in Zambian
children with Plasmodium falciparum malaria. Trop Med Int Health 2000;5(1):9-16.
69. Newton CR, Warn PA, Winstanley PA, et al. Severe anaemia in children living in a
malaria endemic area of Kenya. Trop Med Int Health 1997;2(2):165-78.
70. English M, Ahmed M, Ngando C, Berkley J, Ross A. Blood transfusion for severe
anaemia in children in a Kenyan hospital. Lancet 2002;359(9305):494-5.
71. Brewster DR. Blood transfusions for severe anaemia in African children. Lancet
1992;340(8824):917.
72. Obonyo CO, Steyerberg EW, Oloo AJ, Habbema JD. Blood transfusions for severe
malaria-related anemia in Africa: a decision analysis. Am J Trop Med Hyg 1998;59(5):808-
12.
73. Akech SO, Hassall O, Pamba A, et al. Survival and haematological recovery of
children with severe malaria transfused in accordance to WHO guidelines in Kilifi, Kenya.
Malar J 2008;7:256.
74. Price RN, Simpson JA, Nosten F, et al. Factors contributing to anemia after
uncomplicated falciparum malaria. Am J Trop Med Hyg 2001;65(5):614-22.
75. Bloland PB, Lackritz EM, Kazembe PN, Were JB, Steketee R, Campbell CC. Beyond
chloroquine: implications of drug resistance for evaluating malaria therapy efficacy and
treatment policy in Africa. J Infect Dis 1993;167(4):932-7.
76. Looareesuwan S, Merry AH, Phillips RE, et al. Reduced erythrocyte survival
following clearance of malarial parasitaemia in Thai patients. Br J Haematol
1987;67(4):473-8.
77. Sowunmi A, Gbotosho GO, Happi CT, Fateye BA. Factors contributing to anaemia
after uncomplicated Plasmodiumfalciparum malaria in children. Acta Trop;113(2):155-61.
78. Abdalla S, Weatherall DJ, Wickramasinghe SN, Hughes M. The anaemia of P.
falciparum malaria. Br J Haematol 1980;46(2):171-83.
P a g e | 30
79. Wilson NO, Ceesay FK, Obed SA, et al. Intermittent preventive treatment with
sulfadoxine-pyrimethamine against malaria and anemia in pregnant women. Am J Trop
Med Hyg;85(1):12-21.
80. Filler SJ, Kazembe P, Thigpen M, et al. Randomized trial of 2-dose versus monthly
sulfadoxine-pyrimethamine intermittent preventive treatment for malaria in HIV-positive
and HIV-negative pregnant women in Malawi. J Infect Dis 2006;194(3):286-93.
81. Kapito-Tembo A, Meshnick SR, van Hensbroek MB, Phiri K, Fitzgerald M, Mwapasa V.
Marked reduction in prevalence of malaria parasitemia and anemia in HIV-infected
pregnant women taking cotrimoxazole with or without sulfadoxine-pyrimethamine
intermittent preventive therapy during pregnancy in Malawi. J Infect Dis;203(4):464-72.
82. Barger B, Maiga H, Traore OB, et al. Intermittent preventive treatment using
artemisinin-based combination therapy reduces malaria morbidity among school-aged
children in Mali. Trop Med Int Health 2009;14(7):784-91.
83. Senn N, Rarau P, Stanisic DI, et al. Intermittent preventive treatment for malaria in
Papua New Guinean infants exposed to Plasmodium falciparum and P. vivax: a randomized
controlled trial. PLoS Med;9(3):e1001195.
84. Meremikwu MM, Donegan S, Sinclair D, Esu E, Oringanje C. Intermittent preventive
treatment for malaria in children living in areas with seasonal transmission. Cochrane
Database Syst Rev;2:CD003756.
85. Watthanakulpanich D, Maipanich W, Pubampen S, et al. Impact of hookworm
deworming on anemia and nutritional status among children in Thailand. Southeast Asian J
Trop Med Public Health;42(4):782-92.
86. Smith JL, Brooker S. Impact of hookworm infection and deworming on anaemia in
non-pregnant populations: a systematic review. Trop Med Int Health;15(7):776-95.
87. Bhargava A, Jukes M, Lambo J, et al. Anthelmintic treatment improves the
hemoglobin and serum ferritin concentrations of Tanzanian schoolchildren. Food Nutr Bull
2003;24(4):332-42.
88. WHO. Management of the child with a serious infection or malnutrition: Guidelines
for care at the first referral level in developing countries. 2000.
89. N.S.O. 2008 Population and Housing Census Main Report. National Statistics Office,
Malawi. 2008.
90. World-Bank. Malawi: Country data profile. The World Bank Group. 2011.
91. N.S.O. Malawi Demographic and Health Survey 2010. Zomba, Malawi and Calverton,
Maryland, USA: NSO and ICF Macro. 2011.
92. Cheema B, Molyneux EM, Emmanuel JC, et al. Development and evaluation of a new
paediatric blood transfusion protocol for Africa. Transfus Med;20(3):140-51.
93. Vos J, Gumodoka B, Ng'weshemi JZ, Kigadye FC, Dolmans WM, Borgdorff MW. Are
some blood transfusions avoidable? A hospital record analysis in Mwanza Region,
Tanzania. Trop Geogr Med 1993;45(6):301-3.
94. Craighead IB, Knowles JK. Prevention of transfusion-associated HIV transmission
with the use of a transfusion protocol for under 5s. Trop Doct 1993;23(2):59-61.
95. WHO. The clinical use of blood in medicine, obstetrics, paediatrics, surgery and
anaesthesia, trauma and burns. 2001:337.
P a g e | 31
Chapter 2
Iron deficiency in children with HIV-associated anaemia: a

systematic review and meta-analysis
Esan M.O1, 2, 3, Jonker F.A.M2, Boele van Hensbroek M2, Calis J.C.J2, Phiri K.S1, 3
1Wellcome Trust Research Laboratories, P. O. Box 30096, Blantyre, Malawi

2 Global Child Health Group- AIGHD, Emma Children’s Hospital/Academic Medical Centre,
Amsterdam, the Netherlands
3Community Health Department, College of Medicine, University of Malawi, P. O. Box 360,
Blantyre, Malawi
Trans R Soc Trop Med Hyg 2012; Oct 106(10): 579-587.

P a g e | 32
Abstract
We conducted a systematic review and meta-analysis to determine the prevalence of iron
deficiency in HIV-infected children from high and low-income settings and compared it
with that of HIV-uninfected controls.
We searched five major databases for primary studies reporting on anaemia and iron
markers in HIV-infected children. A pooled analysis was done using random-effects models,
with Forest plots and heterogeneity test estimates provided.
Fifteen articles (2778 children) met the inclusion criteria. In the pooled analysis, mean
overall prevalence of iron deficiency in HIV-infected children was 34% (95%CI 19-50%).
Prevalence rates were similar in high-income (31%; 95%CI 2-61%) and low-income
settings (36%; 95%CI 17-54%) (p=0.14). Studies that included a HIV-uninfected control
population (n=4) were only available from low-income settings and showed less iron
deficiency in HIV-infected children (28%) compared with HIV-uninfected controls (43%);
OR 0.50 (0.27-0.94); p=0.03.
The findings suggest that HIV-infected children are less likely to be iron deficient when
compared with HIV-uninfected children. Possible explanations for this include HIV-induced
haematosuppression and associated hypoferraemia, with adequate iron stores.
Nevertheless iron deficiency is a common co-morbidity in HIV. Studies are needed to
determine the role of iron deficiency in HIV-associated anaemia and the effects of iron
supplementation in this population.
P a g e | 33
Introduction
Anaemia is a common haematological complication of HIV infection and has been
consistently found to be independently associated with HIV-disease progression and
mortality.1-5 Mild to moderate anaemia is the most common presentation in HIV-infected
children from both high and low-income settings, with a prevalence of 3-82% and 22-94%
respectively.6 The aetiology of HIV-associated anaemia is thought to be multi-factorial with
HIV-associated infections, neoplasms, drug-related side effects, and micronutrient
deficiencies being the most important aetiological (sub) groups involved.7-9 Red cell
production failure is often the most important underlying pathogenic mechanism.10
Iron deficiency is considered to be the most important micronutrient deficiency causing
anaemia globally. In low-income settings, iron deficiency is estimated to be responsible for
up to 50% of cases of anaemia seen in pregnant women and children.11 However, the
contribution of iron deficiency to anaemia in HIV-infection is unclear. Poor dietary iron
intake and bioavailability, as well as reduced intestinal absorption due to repeated
infections could result in a diminished iron status.11, 12 On the other hand, there is evidence
that iron metabolism is altered in HIV-infection, resulting in an immune-mediated relative
increase in iron stores.13 Increased iron levels have been associated with advanced stages
of HIV disease and mortality;14 and iron supplementation in this context may be
detrimental. The purpose of this review is to determine the prevalence of iron deficiency in
HIV-infected children from high and low-income settings and compare it with that of HIV-
uninfected controls. The results are then used to discuss the role of iron deficiency, the use
of iron markers and inflammation in HIV-associated anaemia.
Methods
Search Strategy
Primary studies reporting on haemoglobin (Hb) and markers of iron status in HIV-infected
children were searched for in the following databases (excluding Web of Science) in
November 2009, with an updated search done in January 2012: PubMed (1950-2012),
Embase (1980-2012), Africa Index Medicus (1960-2009), Africa Journals On-line (1998-
2009) and Web of Science (1975-2012). Conference abstracts were searched via
conference proceedings citation index database available on Web of Science. A
P a g e | 34
standardized search protocol was developed based on the Cochrane Collaboration

guidelines15 using the following key words: ‘HIV’, ‘children’, ‘iron status’ and ‘anaemia’. The
search strategy aimed to identify all relevant papers and conference abstracts regardless of
language or publication status (published, unpublished, in press or in progress). Relevant
papers were translated where necessary. Finally, references lists of all selected articles
were reviewed for relevant articles. Selection of papers and data extraction was done
independently by two of the reviewers (MOE and FAMJ); discrepancies were resolved by
discussion.
Selection Criteria
All studies that met the following criteria were included: presented data in HIV-infected
children (<18 years) on mean haemoglobin/haematocrit levels and one or more of the
following markers of iron status: serum ferritin, soluble transferrin receptor (STfR), soluble
transferrin receptor-log ferritin index (STfR-F Index), zinc protoporphyrin (ZPP), serum
iron, serum transferrin (TrF), total iron binding capacity (TIBC), mean cell haemoglobin
concentration (MCHC), mean corpuscular volume (MCV), haemosiderin or bone marrow
iron. If both haemoglobin and haematocrit values were provided, the former was used.
Where more than one marker of iron status was used, bone marrow iron was considered
first, but where this was not available, peripheral blood iron markers (ferritin, TrF, STfR,
TIBC, TfR-F Index) were preferred over indices of the full blood count (MCV, MCHC).16, 17 All
observational and interventional studies that met the inclusion criteria were included. The
following studies were excluded: individual case reports and case reviews, studies
assessing restricted populations such as children with haemoglobinopathies; or studies
done in children with specific HIV-related morbidities only. All other relevant non-primary
articles on anaemia and iron deficiency in HIV-infected children were used for the
discussion section.
Definitions
For the purpose of this review, anaemia was defined as haemoglobin of less than 11.5 g/dl
or haematocrit of less than 33%. This value was derived from the cut-off values for
anaemia used by the selected studies, and is the WHO cut-off for anaemia for children >5
years.11 Different markers and cut-offs for iron deficiency were used in the included studies
P a g e | 35
and included ferritin (6 -40 µg/dl),2, 9, 12, 18-24 soluble transferrin receptor-log ferritin >5.625
and >0.75, with a CRP <1.0 mg/L;26 serum iron <40 µmol/l;19 mean corpuscular volume
<70fl and mean cell haemoglobin concentration <32.4g/dl.1
After careful consideration, studies were classified into 2 broad groups: low-income and
high-income settings, taking into account several factors which included socio-economic
factors, accessibility/availability of health care and infection pressure. Studies done in
Africa, Asia, and South America were classified as from low-income settings. Studies done
in Europe and North America were classified as from high-income settings.
Statistical analysis
Comprehensive meta-analysis software version 2 (Biostat Inc., Englewood NJ, USA) and
STATA version 10 (StataCorp, College Station TX, USA) were used for data analysis.
Descriptive data of all identified studies are presented in Table 1. Pooled mean prevalence
estimates (95% Confidence Intervals) using a random-effects model were generated for
studies done in HIV-infected children using Forest plots stratified by setting (high versus
low-income settings) and presented in Figure 2. A pooled analysis of available data from
controlled studies (Figures 3 & 4) and of iron deficiency prevalence stratified by highly
active anti-retroviral therapy (HAART) use (Figure 5) was done, with Forest plots and
heterogeneity test estimates provided. A two-sample z-test was used to compare the
prevalence of iron deficiency in HIV-infected children in high-income versus low-income
settings, and the prevalence of iron deficiency in HIV-infected children on HAART with
those not on HAART. We visually assessed for publication bias using funnel plots. Two-
sided p-values of <0.05 were considered statistically significant.
Results
We followed the preferred reporting items for systematic reviews and meta-analyses
(PRISMA) guidelines in presenting the findings of this review.27
Selection of articles
A combined search of PubMed, Web of Science and Embase databases retrieved 507 hits,
which included 13 conference abstracts. Thirty-eight articles qualified using the selection
criteria described above (Figure 1). Africa Index Medicus and Africa Journals On-line were
P a g e | 36
excluded from the updated search as the initial search of these databases gave a low return
of articles and did not provide any additional literature to that which was available on
PubMed or Embase. From the 38 articles left, 23 were excluded either because they did not
present relevant findings exclusively in the study target population (children <18 years) or
they did not present data on iron deficiency or anaemia.
Figure 1: Systematic review flow diagram

242 hits from Web of
95 hits from 170 hits from
Science, including 13
Pubmed Medline
conference abstracts
469 articles failed to meet study

inclusion criteria after abstracts were
screened
38 full-text articles were
assessed for eligibility 8 studies were not done in
the study target population 1
23 articles
were excluded
15 did not present data on iron
deficiency and/or anemia in study
15 articles were eligible and used target population 2
for the systematic review and meta-
analysis
1 Study results were either not reported exclusively for children <18 years or exclusively for the paediatric
HIV-infected sub-population
2 Did not present results of iron markers for determination of iron status or haemoglobin cut-off levels for
definition of anaemia in their respective study populations
Description of selected studies

Fifteen articles representing 2778 children met the inclusion criteria. Eleven studies were
from low-income settings: Brazil,9, 18, 24 Thailand,19 South Africa,2 Uganda,1, 21-23 Malawi25
and India;26 and four were from high-income settings: USA3, 28, 29 and Italy.12 Twelve of the
studies were cross-sectional studies1, 2, 9, 12, 18, 19, 21, 22, 25, 26, 28, 29 and three were cohort
studies.3, 23, 24 Four of the fifteen studies presented data on iron deficiency in an HIV-
uninfected control group, all from low-income settings (Figure 4).22-25
P a g e | 37
Three of the four controlled studies were cohorts of children born to HIV-infected and
uninfected mothers followed from birth with blood parameters assessed periodically; HIV
PCR testing was used determine which HIV-exposed children made up the HIV-infected
cohort, all other children (HIV-exposed and unexposed) were used as controls.22-24 The
fourth study was a case-control study in which healthy, non-severely anaemic
(haemoglobin >5g/dl) community and hospital controls were recruited for HIV-infected
children admitted for severe anaemia.25
All HIV-infected children from two of the studies presented were on HAART;18, 29 four
studies had some of their HIV-infected cohorts on HAART;24 9, 26, 28 seven studies were done
before HAART use was adopted as part of the standard of care in the countries where they
were carried out1-3, 21-23, 25 and two studies did not provide any information on HAART use.
12, 19 Four of the five studies which provided prevalence estimates for iron deficient HIV-
infected children on HAART were made up of cohorts that were either followed from birth
or started on HAART at an early age.9, 18, 24, 29 The fifth study cohort was made up of HIV-
infected children initiated on HAART using national guidelines, adapted from WHO
criteria.26 Only children on daily maintenance medications with HAART were considered
for inclusion the HAART subgroup, children who received HAART only as part of peri-natal
prophylaxis protocols were excluded.
P a g e | 38
Table 1: Prevalence of anaemia and iron deficiency in HIV-infected children and HIV-uninfected controls as reported in 15 studies meeting the inclusion criteria for
this systematic review and meta-analysis
Prevalence Prevalence
N Anaemia Definition of iron
Age in of Anaemia of Iron def.
Country Authors Period Description cut-off a
months deficiency (%) (%)
Hb (g/dl)
HIV+ HIV- HIV+ HIV- HIV+ HIV-
Low-income settings
Review looking at nutritional problems
Thailand Tienboon 19 1994 3-46 45 - <10 Serum Iron <40 65.0 - 29.0 -
in hospitalized children in Thailand
Cross-sectional study on usefulness of
Uganda Ray, et al 21 1995- 1998 9 134 - <11 Ferritin <12 90.7 - 48.5 -
sTfR as a marker of ID in HIV anaemia
Study on contribution of iron deficiency
Uganda Totin, et al 22 1995 - 1998 9 165 39 <11 Ferritin <12 90.9 76.9
& chr. disease to anaemia 47.1 59.1
Longitudinal study characterizing risk <11
Uganda Clark, et al 1 1995 - 1998 9 225 - MCHC <32.4 91.7 - 73.8 -
factors for anaemia in HIV
Longitudinal study on haematological
Brazil Silva, et al 18 1996 – 1997 3-18 37 - <11.5 Ferritin <6 100.0 - 2.7 -
parameters in HIV infected children
Longitudinal study on haematological
Brazil Silva, et al 24 1996 – 1997 <1-93 26 53 <11.5 Ferritin <6 73.1 41.5 17.9 16.2
aspects of vertical exposure to HIV-1
Nested cohort study on haematological 99 460 <10.5
Zimbabwe Miller, et al 23 1997 – 2000 0–24 Ferritin <12 75.9 36.6 2.5b 33.0
changes in HIV + children & controls
Descriptive study on haematological Haemosiderin:
Brazil Meira, et al 9 2000 - 2001 4–84 8 - <10 Decreased 62.5 - 42.9 -
alterations in HIV
South Descriptive study on peripheral iron
Eley, et al 2 Prior to 2002 13-38 60 - <11 Ferritin <10 72.0 - 52 -
Africa status in clinically stable HIV + children
Cross-sectional (case-control) study on TfR-F index >5.6
Malawi Calis, et al 25 2002-2004 6 – 60 85 958 <11 60.0 43.3 41.7 64.0
risk factors for severe anaemia.
Cross-sectional study on prevalence & <11.5 STfR/F index >0.75
India Shet, et al 26 2007-2008 24-144 80 - 52.5 - 36.3 -
aetiology of anaemia in HIV + children + CRP ≤ 1.0
High-income settings
Retrospective review of bone marrow 60 Haemosiderin: - -
USA Mueller, et al 28 1987 - 1994 4–199 - ≤10 Decreased 64.0 44.0
aspirates
Haematological profile of symptomatic Haemosiderin: -
USA Ellaurie, et al 3 Prior to 1989 2-96 100 47 <11 Decreased 94.0 32.0 0.0
HIV + versus HIV - children
c Study assessing contribution of ID to <11.5(>24) Serum iron <30
Italy Castaldo, et al 12 Prior to 1995 3–156 71 - 66.0 - 48.0 -
anaemia in HIV <10.5(6-24) (≤12);<53 (>12)
TfR-F index of 27.8-
d Butensky James Observational study classifying anaemia Below ref.
USA Prior to 2009 <216 26 - 46.3 15.0 - 35.0 -
et al 29 as either due to ID or chronic disease lab values
Values for serum ferritin presented in ug/L; serum iron in umol/L; MCHC in g/dL & MCV in fL b Estimate based on mean (SD) ferritin values for HIV-infected cohort at 6 months c Age specific
cut-offs for anaemia and serum iron are presented. Anaemia defined as Hb <11.5g/dl for children >24months, and < 10.5g/dl for children 6-24months. ID defined as serum iron <30 for
children <12months; serum iron <53 for children >12months. d Laboratory cut-off value for Hb not referenced by the authors.
P a g e | 39
Figure 2: Pooled analysis of prevalence of iron deficiency in HIV-infected children
Studies Mean prevalence & 95% CI
Ellaurie, et al 1990 0.01 (-0.01 – 0.03)

Costaldo, et al 1996 0.48 (0.36 – 0.60)
Mueller, et al 1996 0.43 (0.31 – 0.55)
Butensky, et al 2009 0.35 (0.17 – 0.53)
Sub-total, high-income settings1 0.31 (0.02– 0.61)
Silva, et al 1999 0.03 (-0.03 – 0.09)
Silva, et al 2001 0.19 (0.03 – 0.35)
Clark, et al 2002 0.74 (0.68 – 0.80)
Eley, et al 2002 0.52 (0.40 – 0.64)
Tienboon 2002 0.29 (0.15 – 0.43)
Totin, et al 2002 0.47 (0.39 – 0.56)
Miera, et al 2005 0.38 (0.05 – 0.71)
Miller, et al 2006 0.03 (-0.001 – 0.06)
Ray, et al 2007 0.49 (0.41 – 0.57)
Calis, et al 2008a 0.41 (0.31 – 0.51)
Shet, et al 2011 0.36 (0.20 – 0.52)
Sub-total, low-income settings2 0.36 (0.17 – 0.54)
Overall 3 0.34 (0.19 – 0.50)
-1.00 -0.50 0.00 0.50 1.00
Not iron deficient Iron deficient
1 N=257; test for heterogeneity (random): Q=1.52 ; d. f.=3 (p=0.02) ; I2=0
2 N=964; test for heterogeneity (random ): Q=4.75 ; d. f.=10 (p=0.01) ; I2=0
3 Test for overall effect (random): Z= 4.28 (p<0.001)
Figure 3a: Prevalence of iron deficiency in HIV-infected children from controlled studies
Studies1 Mean and 95% CI
Silva, et al 2001 0.19 (0.03 – 0.35)

Totin, et al 2002 0.47 (0.39 – 0.56)
Miller, et al 2006 0.03 (-0.001 – 0.06)
Calis, et al 2008a 0.42 (0.32 – 0.52)
Overall2 0.28 (0.01 – 0.54)
-1.00 -0.50 0.00 0.50 1.00

Figure 3b: Prevalence of iron deficiency in HIV-uninfected controls

Studies Mean and 95% CI
Silva, et al 2001 0.17 (0.07 – 0.27)
Totin, et al 2002 0.59 (0.43 – 0.75)
Miller, et al 2006 0.33 (0.29 – 0.37)
Calis, et al 2008a 0.64 (0.60 – 0.68)
Overall3 0.43 (0.21 – 0.65)
-1.00 -0.50 0.00 0.50 1.00
1 N= 291 (HIV-infected cohort), 1,510 (HIV-uninfected controls); data obtained from studies done in low-income settings, no data available for
controlled studies from high-income settings.
2 Pooled estimates; Test of heterogeneity (random effects model): Q= 1.75, d.f.=3 (p=0.04), I2=0
3 Pooled estimates; Test of heterogeneity (random effects model): Q=2.94, d.f.=3 (p<0.001), I2=0
P a g e | 40
Figure 4: Pooled analysis of iron deficiency in controlled studies
Controlled studies1 Odds ratio and 95% CI

Silva, et al 2001 1.10 (0.47 – 2.57)
Totin, et al 2002 0.65 (0.35 – 1.23)
Miller, et al 2006 0.23 (0.15 – 0.34)
Calis, et al 2008a 0.50 (0.34 – 0.75)
Overall2 0.50 (0.27 – 0.94)
0.01 0.1 1 10 100
Favours HIV -infected Favours HIV-uninfected
1 N=1,801 (HIV positive :291); compares prevalence rates of iron deficiency in HIV-infected children with
controls in the 4 controlled studies; only studies done in low-income settings had a control group included.
2 Test of overall effect : Z= -2.17 (p=0.03)
Anaemia
The prevalence of anaemia in HIV-infected children from high and low-income settings ranged
from 15-94% and 63-100% respectively (Table 1). Prevalence of anaemia in uninfected
controls from high and low-income settings ranged from 5-31% and 36-100% respectively
(Table 1).
Iron deficiency
The iron markers and their cut-off values used to define iron deficiency for each study are
presented in Table 1. In the pooled analysis, the mean overall prevalence of iron deficiency in
HIV-infected children was 34% (95% CI 19-50%). Prevalence rates were similar in high-
income settings (31%; 95% CI 2-61%) and low-income settings (36%; 95% CI 17-54%)
(p=0.14; Figure 2). Studies that included a HIV-uninfected control population (n=4) were only
available in low-income settings and showed that HIV-infected children were less likely to be
iron deficient when compared with HIV-uninfected controls [28% vs. 43%; OR 0.50 (0.27-
0.94); p=0.03]; (Figures 3 and 4).
Pooled estimates of the prevalence of iron deficiency in HIV-infected children stratified by
HAART use are provided in Figure 5. The prevalence of iron deficiency was significantly lower
in HIV-infected children on HAART (24%; 95% CI 7-41%) when compared with those not on
HAART (37%; 95% CI 17-58%) (p=0.005).
Visual assessment of funnel plots for the prevalence of iron deficiency in HIV-infected children
(not shown) revealed some asymmetry, with 3 outlying studies detected.3, 18, 23
P a g e | 41
Intervention studies
Only one of the studies reviewed was an intervention study, done in a high-income setting
looking at the effect of oral iron therapy on iron deficiency and intestinal malabsorption in
HIV-infected children.12 Iron deficiency was present in 48% of the study population, and was
significantly associated with intestinal malabsorption.
Figure 5: Prevalence of iron deficiency in HIV-infected children by HAART use
Mean prevalence & 95% CI

Studies
Silva, et al 1999 0.03 (-0.03 – 0.09)
Silva, et al 2001 0.19 (0.03 – 0.35)
Miera, et al 2005 1 0.67 (0.14 – 1.20)
Butensky, et al 2009 0.35 (0.17 – 0.53)
Shet, et al, 2011 1 0.28 (0.14 – 0.42)
Sub-total, HAART 2 0.24 (0.07 – 0.41)
Ellaurie, et al 1990 0.01 (-0.03 – 0.05)
Clark, et al 2002 0.74 (0.68 – 0.80)
Eley, et al 2002 0.52 (0.40 – 0.64)
Totin, et al 2002 0.47 (0.39 – 0.56)
Miera, et al (2), 2005 1 0.20 (-0.15 – 0.55)
Miller, et al 2006 0.03 (-0.001 – 0.06)
Ray, et al 2007 0.49 (0.41 – 0.57)
Calis, et al 2008a 0.42 (0.32 – 0.52)
Shet, et al (2), 20111 0.44 (0.28 – 0.60)
Sub-total, (non-HAART) 2
0.37 (0.17 – 0.58)
-1.00 -0.50 0.00 0.50 1.00
1Studies provided data on iron def iciency f or sub-populations on HAART

2 N=131 (HAART), N= 555 ( non-HAART). Only children on maintenance HAART regimens were included in the HAART group,
children who received HAART only as part of peri-natal prophylaxis protocols were excluded.
Discussion
In this first review on iron deficiency in paediatric HIV-infection, we observed that there is
limited data on iron states in HIV-infected children. However, we identified that HIV-infected
children are less likely to be iron deficient when compared with HIV-uninfected children. This
comparison could only be made for children from low-income settings, since there were no
studies from high-income settings reporting on the prevalence of iron deficiency in HIV-
uninfected controls. We also observed that prevalence rates of iron deficiency in HIV-infected
children are significantly modified by HAART use.
P a g e | 42
True versus functional iron deficiency in HIV-associated anemia

Our results suggest that iron deficiency in HIV-infected children is a problem in both high and
low-income settings, despite the increasing availability of HAART- with studies from both
areas reporting similar mean prevalence rates. Our results also suggest that iron deficiency
may be less common in HIV-infected children than in uninfected children. HIV-infected
children are commonly malnourished due to macronutrient and multi-micronutrient deficient
diets, as well as increased metabolic demands during infections.30 Iron deficiency however
does not appear to be more common. One of the studies presented in this review which looked
at bone marrow aspirates in a subset of their HIV cohort found normal iron content in all 20
specimens examined- although other bone marrow studies done in children reported higher
iron deficiency prevalence rates.3, 9, 28 Pathophysiologic mechanisms that may explain the
lower prevalence of iron deficiency seen in HIV-infected children compared with controls
include a cytokine-mediated inhibition of erythroid progenitor cells and an inflammation-
induced hypoferraemia, with haemoglobin levels closely correlated with low serum iron
levels.31 These mechanisms may result in functional iron deficiency, where indices of body iron
stores are relatively normal while indices of peripheral body iron are reduced. This is in
contrast with true iron deficiency where iron stores (notably ferritin levels) are depleted.
Functional iron deficiency, which may be seen in HIV and other chronic inflammatory
disorders, may render iron unavailable for erythropoiesis resulting in anaemia but also makes
it unavailable for infective organisms that require iron for growth and proliferation32, 33 - thus
may be protective against severe bacterial infections and malaria in regions with a high
pressure of infection.34
The use of iron markers in infection: the role of inflammation

The interpretation of serological markers of body iron status is unclear in areas with a high
pressure of infection.35 Ferritin, which is commonly used to estimate body iron is also a
mediator of the acute phase response and is raised in HIV-infection while serum iron and
blood transferrin levels are decreased.36 Some of the studies identified in our search did not
adequately adjust their ferritin cut-off values for inflammation- thus true iron deficiency may
be present even when the indices of iron status exceed the recommended normal cut-off value.
This could have contributed to the lower prevalence of iron deficiency seen in the HIV-infected
cohorts of studies that used ferritin alone to assess iron status.35 However, two of the three
P a g e | 43
controlled studies (both done in regions of high infection pressure) that used ferritin alone to
assess iron status did not report a significant difference in mean ferritin values or the
prevalence of iron deficiency between their HIV-infected and control populations.22, 24 More
importantly, the only controlled study which used soluble transferrin receptor-log ferritin
index to determine iron status- a marker which is less influenced by infection showed that iron
deficiency was less prevalent in HIV-infected children compared with controls, further adding
weight to our findings.25 In the absence of bone marrow iron values, ideally studies using a
combination of two or more serological iron markers, preferably including a marker of
inflammation in regions with a high pressure of infections should have been used but this was
not possible as the number of eligible studies presenting this data was too small.
Iron states in HIV-associated anaemia: is intervention needed?

The results of this review casts some doubt as to the role of iron deficiency in the aetiology of
anaemia of HIV-infection, and raises important questions on how HIV-associated anaemia
should be managed. It is known that anaemia in HIV disease is associated with an increased
risk of death3, 4 and anaemia treatment is associated with an improved survival and quality of
life in HIV-infected persons.37 However, it is unknown if iron deficiency plays a major role in
the aetiology of anaemia in HIV-infected children, with the results suggesting that it is less
common when compared with uninfected children. An important question remains- if the
WHO policy of iron supplementation for the prevention and treatment of anaemia is
favourable in HIV-infected children, especially in the era of increased availability of HAART.
Castaldo et al demonstrated that iron supplementation is beneficial in HIV-infected children;
however the six patients that received iron in this study were neither randomised nor blinded,
and there was no control group with which to compare their findings- thus their results were
subject to selection and measurement bias.12 They also did not report if there were any
untoward effects seen as a result of iron supplementation. The detrimental effects of iron
supplementation for the treatment of anaemia in iron-replete children have been clearly
demonstrated in previous studies, with the results showing an increased risk of severe
infections and death in populations with a high prevalence of malaria.38 Given the increased
risk of susceptibility to infections in HIV-infected children, this evidence is needed to inform
management, especially in regions with a high prevalence of infections, including malaria.39
P a g e | 44
In trying to assess the impact of important factors that could have affected our results, we
examined the effect of HAART on our prevalence estimates. We observed that routine HAART
use in HIV-infected children was associated with a lower prevalence of iron deficiency when
compared with HIV-infected children not on HAART. The explanation for this observation is
unclear, but it may be related to the beneficial effect of HAART on anemia and haematopoiesis,
where a reduction in viral load results in reversal of cytokine-mediated haematosuppression.40
A recent study carried out in India by Shet et al in which a cohort of HIV-infected children
managed according to national guidelines were followed for 6 months reported that HAART
when given with iron supplementation was associated with a better haemoglobin response
than either HAART or iron given alone, without any serious adverse effects seen.26 These
observations make a strong case for the conduct of studies examining the effects of HAART on
iron status in HIV-infected children.
Other factors that could potentially have had an impact on our findings include poor dietary
intake and bioavailability of iron, inadequate HIV disease monitoring and management in
resource-limited settings, prevention and treatment of opportunistic infections, and
management of co-morbidities such as malnutrition and parasitic infections- especially in
regions with a high pressure of infection. We have tried to limit the impact of these factors by
classification of the studies included in this review; however their importance cannot be
understated, especially in low-income settings.11, 19
Limitations of this review

The number of studies identified by our search was small and heterogenous, restricting the
differential analytic power of the meta-analysis. We partly corrected for this by doing sub-
group analyses using high and low income settings. However, as we found some asymmetry
on visual assessment of funnel plots (not presented), we cannot rule out publication bias
which is not uncommon with diagnostic studies.
None of the studies from high-income settings provided data on iron deficiency in HIV-
uninfected controls, thus it was not possible to compare findings in controlled studies from
different geographic (socio-economic) settings. There were no randomized clinical trials on
iron supplementation in HIV-infected children with anaemia identified in our search.
P a g e | 45
Clinical impact and areas for future research

Our findings suggest that iron deficiency may be less common in HIV-infected children
compared with HIV-uninfected children. Studies examining the role of iron deficiency in the
aetiology of HIV-associated anaemia are needed. There is a need for more studies in which the
iron status of HIV-infected children is correctly defined, ideally using bone marrow iron
estimates, and also for studies examining the effect of HAART on iron status. There is a need
for randomized clinical trials evaluating the safety and benefit of iron supplementation in the
treatment of HIV-associated anaemia, especially in low-income settings.
Conclusions
Though the data available was limited, the findings suggest that HIV-infected children are less
likely to be iron deficient when compared with HIV-uninfected children. Iron deficiency
however remains a common co-morbidity. Studies are needed to determine the role of iron
deficiency in HIV-associated anaemia and to assess the safety and benefit of iron
supplementation. This information is required to guide management, especially in low-income
settings with a high prevalence of infections, including malaria.
Contributions and acknowledgements: The manuscript was prepared by MOE and FAMJ. JCJC,
MBVH and KSP provided editorial support. Statistical analysis was done by MOE and JCJC. The
authors would like to thank Caroline Veraart and Mw. Dyserinck of Emma Children’s Hospital
AMC for logistics and help with the database search strategies respectively.
Funding: Netherlands-African partnership for capacity development and clinical interventions
against poverty-related diseases (NACCAP).
Competing interests: None declared.
Ethical approval: None required.
P a g e | 46
References
1. Clark TD, Mmiro F, Ndugwa C, et al. Risk factors and cumulative incidence of anaemia among
human immunodeficiency virus-infected children in Uganda. Ann Trop Paediatr 2002;22(1):11-7.
2. Eley BS, Sive AA, Shuttleworth M, Hussey GD. A prospective, cross-sectional study of anaemia
and peripheral iron status in antiretroviral naive, HIV-1 infected children in Cape Town, South Africa.
BMC Infect Dis 2002;2:3.
3. Ellaurie M, Burns ER, Rubinstein A. Hematologic manifestations in pediatric HIV infection:
severe anemia as a prognostic factor. Am J Pediatr Hematol Oncol 1990;12(4):449-53.
4. Tovo PA, de Martino M, Gabiano C, et al. Prognostic factors and survival in children with
perinatal HIV-1 infection. The Italian Register for HIV Infections in Children. Lancet
1992;339(8804):1249-53.
5. Mocroft A, Kirk O, Barton SE, et al. Anaemia is an independent predictive marker for clinical
prognosis in HIV-infected patients from across Europe. EuroSIDA study group. Aids 1999;13(8):943-50.
6. Calis JC, van Hensbroek MB, de Haan RJ, Moons P, Brabin BJ, Bates I. HIV-associated anemia in
children: a systematic review from a global perspective. Aids 2008;22(10):1099-112.
7. Ellaurie M, Rubinstein A. Elevated tumor necrosis factor-alpha in association with severe
anemia in human immunodeficiency virus infection and Mycobacterium avium intracellulare infection.
Pediatr Hematol Oncol 1995;12(3):221-30.
8. Shah I. Adverse effects of antiretroviral therapy in HIV-1 infected children. J Trop Pediatr
2006;52(4):244-8.
9. Meira DG, Lorand-Metze I, Toro AD, Silva MT, Vilela MM. Bone marrow features in children with
HIV infection and peripheral blood cytopenias. J Trop Pediatr 2005;51(2):114-9.
10. Boele van Hensbroek M, Calis JC, Phiri KS, et al. Pathophysiological mechanisms of severe
anaemia in Malawian children. PLoS One 2010;5(9):e12589.
11. WHO. Iron Deficiency Anaemia Assessment, prevention and control: A guide for programme
managers. 2001.
12. Castaldo A, Tarallo L, Palomba E, et al. Iron deficiency and intestinal malabsorption in HIV
disease. J Pediatr Gastroenterol Nutr 1996;22(4):359-63.
13. Fuchs D, Zangerle R, Artner-Dworzak E, et al. Association between immune activation, changes
of iron metabolism and anaemia in patients with HIV infection. Eur J Haematol 1993;50(2):90-4.
14. de Monye C, Karcher DS, Boelaert JR, Gordeuk VR. Bone marrow macrophage iron grade and
survival of HIV-seropositive patients. Aids 1999;13(3):375-80.
15. Higgins PT, Green S. Cochrane Handbook for Systematic Reviews of Interventions 2011;Version
5.1.0:6.4.1.
17. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptor and its ratio to serum ferritin in
the diagnosis of iron deficiency. Blood 1997;89(3):1052-7.
18. Silva EB, Silva MT, Vilela MM. [Evolution of hematological parameters in a group of children
with human immunodeficiency virus infection - HIV 1]. J Pediatr (Rio J) 1999;75(6):442-8.
19. Tienboon P. Nutrition problems of hospitalised children in a developing country: Thailand. Asia
Pac J Clin Nutr 2002;11(4):258-62.
20. Miller MF, Stoltzfus RJ, Mbuya NV, Malaba LC, Iliff PJ, Humphrey JH. Total body iron in HIV-
positive and HIV-negative Zimbabwean newborns strongly predicts anemia throughout infancy and is
predicted by maternal hemoglobin concentration. J Nutr 2003;133(11):3461-8.
21. Ray A, Ndugwa C, Mmirot F, Ricks MO, Semba RD. Soluble transferrin receptor as an indicator of
iron deficiency in HIV-infected infants. Ann Trop Paediatr 2007;27(1):11-6.
P a g e | 47
22. Totin D, Ndugwa C, Mmiro F, Perry RT, Jackson JB, Semba RD. Iron deficiency anemia is highly
prevalent among human immunodeficiency virus-infected and uninfected infants in Uganda. J Nutr
2002;132(3):423-9.
23. Miller MF, Humphrey JH, Iliff PJ, Malaba LC, Mbuya NV, Stoltzfus RJ. Neonatal erythropoiesis and
subsequent anemia in HIV-positive and HIV-negative Zimbabwean babies during the first year of life: a
longitudinal study. BMC Infect Dis 2006;6:1.
24. Silva EB, Grotto HZ, Vilela MM. [Clinical aspects and complete blood counts in children exposed
to HIV-1: comparison between infected patients and seroreverters]. J Pediatr (Rio J) 2001;77(6):503-
11.
25. Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J Med
2008;358(9):888-99.
26. Shet A, Arumugam K, Rajagopalan N, et al. The prevalence and etiology of anemia among HIV-
infected children in India. Eur J Pediatr 2011;171(3):531-40.
27. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. Bmj
2009;339:b2700.
28. Mueller BU, Tannenbaum S, Pizzo PA. Bone marrow aspirates and biopsies in children with
human immunodeficiency virus infection. J Pediatr Hematol Oncol 1996;18(3):266-71.
29. Butensky James E, Harmatz P, Lee M, et al. Altered iron metabolism in children with human
immunodeficiency virus disease. Pediatr Hematol Oncol 2009;26(2):69-84.
30. Friis H. Micronutrients and HIV infection. CRC Press, Florida; 1st Edition 2002.
31. Fuchs D, Hausen A, Reibnegger G, et al. Immune activation and the anaemia associated with
chronic inflammatory disorders. Eur J Haematol 1991;46(2):65-70.
32. Paradkar PN, De Domenico I, Durchfort N, Zohn I, Kaplan J, Ward DM. Iron depletion limits
intracellular bacterial growth in macrophages. Blood 2008;112(3):866-74.
33. Collins HL. Withholding iron as a cellular defence mechanism--friend or foe? Eur J Immunol
2008;38(7):1803-6.
34. Gwamaka M, Kurtis JD, Sorensen BE, et al. Iron deficiency protects against severe Plasmodium
falciparum malaria and death in young children. Clin Infect Dis;54(8):1137-44.
35. Lewis DK, Whitty CJ, Epino H, Letsky EA, Mukiibi JM, van den Broek NR. Interpreting tests for
iron deficiency among adults in a high HIV prevalence African setting: routine tests may lead to
misdiagnosis. Trans R Soc Trop Med Hyg 2007;101(6):613-7.
36. Kreuzer KA, Rockstroh JK. Pathogenesis and pathophysiology of anemia in HIV infection. Ann
Hematol 1997;75(5-6):179-87.
37. Sullivan PS, Hanson DL, Chu SY, Jones JL, Ward JW. Epidemiology of anemia in human
immunodeficiency virus (HIV)-infected persons: results from the multistate adult and adolescent
spectrum of HIV disease surveillance project. Blood 1998;91(1):301-8.
38. Sazawal S, Black RE, Ramsan M, et al. Effects of routine prophylactic supplementation with iron
and folic acid on admission to hospital and mortality in preschool children in a high malaria
transmission setting: community-based, randomised, placebo-controlled trial. Lancet
2006;367(9505):133-43.
39. Adetifa I, Okomo U. Iron supplementation for reducing morbidity and mortality in children with
HIV. Cochrane Database Syst Rev 2009(1):CD006736.
40. Huang SS, Barbour JD, Deeks SG, et al. Reversal of human immunodeficiency virus type 1-
associated hematosuppression by effective antiretroviral therapy. Clin Infect Dis 2000;30(3):504-10.
P a g e | 48
Chapter 3
Benefits and risks of iron supplementation in HIV-infected Malawian

children: results of a double-blind, randomized controlled trial
Esan M.O.1, 2, 4, Boele van Hensbroek M. 2, Nkhoma E.C.1, 4, Musicha C.1, White S.A.1, 3, ter Kuile
F.O.5, 6, Phiri K.S.1, 3, 4
1 Malawi-Liverpool-Wellcome Trust Research Laboratories, P. O. Box 30096, Chichiri, Blantyre 3,

Malawi
2 Global Child Health Group- AIGHD, Emma Children’s Hospital/Academic Medical Centre, Amsterdam,
the Netherlands
3 Community Health Department, College of Medicine, University of Malawi, P. O. Box 360, Chichiri,
Blantyre 3, Malawi
4Tropical Haematology Research Unit, College of Medicine, University of Malawi, P. O. Box 360, Chichiri,
Blantyre 3, Malawi
5 Liverpool School of Tropical Medicine, Liverpool, United Kingdom
6 Department of Infectious Diseases, Tropical Medicine & AIDS, Academic Medical Centre, University of
Amsterdam, Amsterdam, Netherlands
Submitted
P a g e | 49
Abstract
Background: It is unknown whether iron supplementation in HIV-infected children, living in

regions with high infection pressure is safe or beneficial. A 2-arm double-blind randomized,
controlled trial was conducted to examine the effects of iron supplementation on haemoglobin,
HIV-disease progression and morbidity.
Methods: HIV-infected Malawian children aged 6-59 months with moderate anaemia (Hb 7.0-
9.9 g/dl) were randomized to 3mg/kg/day of elemental iron and multi-vitamins (Vitamins A, C
and D) or multi-vitamins alone for 3 months. Participants were followed for 6 months. Clinical
trial number: ISRCTN-62947977.
Results: Two-hundred and nine children were randomized and 196 (93.8%) completed 6
months follow-up. Iron supplementation was associated with greater increases in
haemoglobin concentrations (g/dL): adjusted mean difference (aMD) (95% CI) 0.60 (0.06-
1.13); p=0.03, and reduced the risk of anaemia persisting up to 6 months follow-up: adjusted
prevalence-ratio (95% CI) 0.59 (0.38-0.92); p=0.02. Children who received iron had a better
CD4 percentage response at 3-months: aMD (95%CI) 6.00 (1.84-10.16); p=0.005, but an
increased incidence of clinical malaria at 6 months: incidence rate (IR) 120.2 vs. 71.7; adjusted
incidence rate ratio (aIRR) (95%CI) 1.81 (1.04–3.16); p=0.04, especially during the first 3
months: IR 78.1 vs. 36.0; aIRR (95%CI) 2.68 (1.08-6.63); p=0.03.
Conclusions: Iron supplementation for the treatment of anaemia in HIV-infected children has
beneficial effects on haemoglobin, anaemia and immunity, but increases the risk of malaria.
Iron supplementation in similar paediatric populations in malaria-endemic areas should only
be provided in combination with protection from malaria.
P a g e | 50
Introduction:
Anaemia is a global health problem with approximately 25% of the world’s population affected
[1, 2]. The greatest burden is in Africa and South-East Asia, where more than 60% of pre-
school children are anaemic [1]. Iron deficiency is responsible for approximately 50% of cases
of anaemia seen in women and pre-school children, the two groups most at risk of the
condition, and for 2.3 million disability adjusted life years (DALYs) [3]. In developing
countries, an overlap is often found between anaemia and HIV in areas where both conditions
are prevalent. This is not surprising since anaemia is an important hematologic complication of
HIV-infection, associated with poor survival and reduced quality of life [4-6]. Recent studies
suggest that iron supplementation may be harmful to iron-replete children living in areas with
a high prevalence of malaria and other infectious diseases [7-10]. This prompted the WHO to
recommend targeted iron supplementation of children at risk of anaemia and iron deficiency
[7]. However, this strategy is impractical in resource-limited countries where simple, reliable,
and affordable tests for the assessment of iron status are not routinely available; and ferritin
findings may be difficult to interpret where infections are prevalent [11-13]. Blanket iron
supplementation is still widely used for children living in areas with a high prevalence of
infections, including HIV. However, it is not known if giving iron to this specific sub-population
is either safe or beneficial, given their increased susceptibility to infections and the
suppressive effects of HIV-1 infection on bone marrow haematopoiesis, which may not readily
respond to iron [14, 15]. We have conducted a randomized, controlled trial of supplementation
of multi-vitamins with iron versus multi-vitamins without iron in anaemic HIV-infected
children investigating the effect of iron supplementation on haemoglobin, HIV disease
progression and morbidity.
Methods:
Participants
We conducted a two-arm, double-blind, randomized, placebo-controlled trial of oral iron
supplementation in HIV-infected children aged 6-59 months with moderate anaemia (defined
as Hb 7 to 9.9 g/dl) between January 2009 and August 2010 in Southern Malawi, a region of
moderate to intense perennial malaria transmission. Eligible participants were recruited from
Thyolo District Hospital and Zomba Central Hospital. HIV-positive children
P a g e | 51
identified by universal HIV counseling and testing within the hospital and who had clinical
pallor on physical examination were screened at paediatric out-patient clinics for moderate
anaemia using Hemocue 301 analyser (HemoCue AB, Angleholm, Sweden). Children were
excluded from participation if: 1) they had evidence of severe malnutrition (defined as weight-
for-height less than 75% of expected for age or bi-pedal oedema [16]); 2) they were already on
micronutrient supplements or fortified diets; 3) they had a history of hypersensitivity to study
medications; 4) they were enrolled in another study involving treatment with medicinal
products at the time of recruitment; or 5) had a recognized specific cause of chronic anaemia
(such as sickle cell disease or suspected malignancy). All children received co-trimoxazole
prophylaxis as it is standard practice for management of HIV-infection in Malawi.
Children whose parent/guardian consented were randomized to two arms: (iron) arm
received 3mg/kg of elemental iron and multivitamins (Vitamins A 1500 I.U/ml, C 35mg/ml
and D 400 I.U/ml) as once-daily concentrated drops for 3 months (commercially available as
Tri-Vi-Sol with iron©, Mead Johnson Nutritionals, USA) and the second (placebo) arm received
multi-vitamins alone (commercially available as Tri-Vi-Sol©, Mead Johnson Nutritionals, USA)
once daily for 3 months. Both medicinal products were identical in appearance and content,
with the exception of elemental iron in the multi-vitamins with iron preparation. The primary
end point for the study was the incidence of all-cause out-patient sick visits. All participants
were followed up for 6 months.
Procedures
Recruitment: On enrolment a blood sample was taken for full blood count (ABX Micros 60,
Horiba ABX S.A.S, Montpellier, France), malaria microscopy, blood cultures, and basic
biochemistry assays, including C-reactive protein (CRP) and serum ferritin. Serum samples
collected for ferritin and CRP assessment were stored at -800C until study completion and then
shipped to the Netherlands where the assays were run as a fully automated single-in-one
series using Beckman LX20 chemistry analyzer (Beckman Coulter, Brea, CA, USA). In addition,
urine and stool analysis were carried out to exclude bacterial and parasitic infections. Positive
results for infection were treated according to standard hospital protocols.
Follow up: Study participants were seen routinely at 1, 2, 3 and 6-months after recruitment in
the study clinic, at which time clients were assessed for adherence to study medication,
general well-being and nutritional status. At recruitment, 1 and 2 month visits, guardians were
P a g e | 52
given a month’s supply of study medications. At each scheduled hospital visit a blood and urine
sample was collected for haemoglobin measurements, malaria microscopy (on clinical
suspicion of malaria) and urine analysis. Serum samples were also collected at 3 and 6 months
for CRP and ferritin measurements. Children who presented to hospital due to an illness before
the scheduled hospital visits were examined and treated according to standard hospital
protocols. Children who did not show up for scheduled follow-up after a week were traced to
their homes by research staff. Details of hospital sick visits by trial participants to other health
care facilities prior to their scheduled hospital visit were collected from their health passport
books and parent/guardian interviews at the time they presented for a routine follow-up or
hospital sick visit.
Case definitions: Malaria was defined by the presence of Plasmodium falciparum asexual
parasites on Giemsa-stained thick blood smear microscopy. The parasite density was
documented as the number of parasites per 200 white blood cells in the thick smear. Malaria
slides were independently read by two readers; and further by a third reader if the initial
results differed by more than 25%. A Rapid diagnostic test for P. falciparum histidine-rich
protein-2 (Paracheck-Pf®, Orchid Biomedical Systems, Goa, India) was done if the microscopy
was negative or if a trial participant who did not show up for a scheduled hospital visit was
traced and found to be ill at home.
A respiratory infection was defined using WHO clinical criteria [17] as a child presenting with
a history of cough or runny nose, along with fast breathing (defined as a respiratory rate over
50 breaths/minute if less than one year and over 40 breaths/minute if more than one year old)
and signs of respiratory distress (chest in-drawing, labored breathing, grunting).
The number of children who progressed to AIDS in each arm was determined by using a
combination of clinical (WHO clinical staging [18]) and immunological markers (CD4
percentage <20%) taken for each trial participant at 3 and 6 months follow-up visits. Iron
deficiency was defined as serum ferritin <30 µg/dl where CRP level was ≥5.0 (evidence of
inflammation) and serum ferritin <12 µg/dl where CRP was <5.0 [19, 20].
Randomization and blinding: Block randomization (random block sizes ranging from two to
eight) with 1:1 allocation was used and stratified by study site. The blinded participant-
specific study drugs were prepared in accordance with a computer generated randomization
list by a designated person who otherwise was not involved in the study. Study drugs were
P a g e | 53
pre-packed in sealed envelopes and labeled with the study numbers for each study site. For
each patient a copy of the study code was kept in sealed envelopes by the study safety monitor.
Ethical approvals were obtained from the College of Medicine Research Ethics Committee,
Malawi and the Liverpool School of Tropical Medicine Research Ethics Committee, UK. The
clinical trial registration number was ISRCTN 62947977.
Sample size calculation and statistical analysis

A sample size of 1260 was calculated based on data from a large case-control study [21], in
which we assumed that the mean (SD) number of sick-child out-patient clinic visits over 6
months follow-up would be 2.2 (2.2), with iron supplementation regarded as safe if the
incidence of sick visits was not increased by more than 15% compared with placebo, with 90%
power to infer non-inferiority, assuming 10% loss to follow-up. However, it became clear after
12 months of recruitment that we would be unable to achieve our projected sample size within
the approved time period, due to unforeseen changes in national policy on HAART use in
children and pregnant women, which took effect soon after study commencement [22]. We
informed the local ethics committee of our decision to stop recruitment in February 2010. All
participants were followed till August 2010.
A statistical analysis plan was agreed before un-blinding the data. Data were analyzed by
intention-to-treat using Stata 10 (StataCorp, College Station, Texas, USA). Multiple linear
regression models were used to test the intervention effect on changes in haemoglobin (delta
Hb) and CD4 percentage (delta CD4%) over the follow-up period, with treatment effects
presented as the mean difference (MD). Prevalence endpoints (anaemia, iron deficiency) were
summarized as percentages with treatment effects expressed as prevalence ratios (PR); and as
the number of events per 100 person-years with the treatment effect expressed as incidence
rate ratios (IRR) for morbidity outcomes, using negative binomial regression with robust
estimates of the standard error to account for multiple events per-child. All treatment effects
were adjusted for important baseline variables (previous blood transfusion in the past 3
months, bed-net use, CD4 percentage at recruitment and baseline haemoglobin). These were
selected after exploring the association between baseline variables and outcome variables in
univariate models and variables with a p-value of <0.10 entered into the initial full multi-
variable model. Exploratory analyses was done to determine to what extent the magnitude of
P a g e | 54
treatment effects varied with time period (active intervention vs. passive post-intervention
period), age (<24 months or >24 months), baseline iron status (iron deficient vs. iron replete),
bed net use and HAART use at recruitment using interaction term models [23]. As the study
lacked power to test for treatment effect modification, we used the magnitude of the treatment
effect ratio (interaction term) between subgroups and its corresponding p-value to guide our
conclusions. For all other statistical analyses, p-values of <0.05 were taken as statistically
significant and each estimate is reported with a 95% confidence interval.
Study monitoring
The study was conducted according to Good Clinical Practice (ICH-GCP) standards. A study
initiation visit was performed prior to the start of patient recruitment, and a close-out visit
was performed at the end of the study. Regular study monitoring visits were made by clinical
research associates (CRA) to both trial study sites, with detailed assessments of trial
recruitment procedures, documentation and standard operating procedures (SOP) done.
Monitoring reports and recommendations made available to investigators and the trial
sponsor at the end of each visit.
Role of funding source

The study was funded by the Netherlands-African partnership for capacity development and
clinical interventions against poverty-related diseases (NWO-NACCAP), under the College of
Medicine, Malawi-Amsterdam-Liverpool (COMMAL) partnership programme. The funders
played no part in the study design, data collection, analysis, and interpretation or writing of the
final report.
P a g e | 55
Figure 1: HIV-Iron trial profile
284 HIV-positive children with clinical

anaemia were assessed for study eligibility
75 were excluded:
• 69 did not meet eligibility criteria
• 8 did not consent to the study
209 children were randomized assigned to

receive study drugs
105 received multivitamins 104 received multivitamins

with iron at baseline without iron at baseline 1 moved out of study catchment area
1 Died 1 missed follow-up for unknown reason
with iron at 1 month without iron at 1 month
1 Died 1 died
1 lost to follow-up 2 withdrew consent
with iron at 2 months without iron at 2 months
1 died
1 lost to follow-up
102 completed 3 months 97 completed 3 months
follow-up follow-up
1 Died
1 Died
1 lost to follow-up
100 completed 6 months 96 completed 6 months
Follow-up as per-protocol Follow-up as per-protocol
P a g e | 56
Results:
A total of 209 HIV-infected children (78 from Thyolo District Hospital, 131 from Zomba Central
Hospital) were recruited between January 2009 and February 2010. One hundred and ninety
six patients (93.8%) completed 6 months follow-up (figure 1). Baseline characteristics of study
participants are summarized in Table 1. The mean (SD) age was 25.8 (15.9) months and 47.4%
(99/209) of the study population were male. Median (IQR) Hb at baseline was 9.4 (8.6–
9.8)g/dl and the prevalence of iron deficiency was 34.5% (72/209; Table 1).
Table 1: Demographic and baseline characteristics by study arm

Household characteristics Iron (n=105) Placebo (n=104) Total (n=209)
Maternal age (years) 29.5 (5.1) 29.4 (5.4) 29.5 (5.2)
Maternal educational level (Primary) 66 (62.9%) 59 (56.7%) 125 (59.8%)
Guardian/Father employment status (Employed) 85 (81.0%) 82 (78.9%) 167 (79.9%)
Household has electric power supply 13 (12.4%) 17 (16.4%) 30 (14.4%)
Nutrition (≥3 meals with fish/meat per week) 68 (64.8%) 70 (67.4%) 138 (66.0%)
Study site (Thyolo) 39 (37.1%) 39 (37.5%) 78 (37.3%)
Use of a mosquito bed net in the past week1 54 (53.5%) 55 (55.6%) 109 (54.5%)
Child parameters
Sex (Male) 42 (40.0%) 57 (54.8%) 99 (47.4%)
Age at recruitment (months) 25.2 (15.8) 26.4 (16.1) 25.8 (15.9)
Weight (kg) 10.2 (2.8) 10.1 (2.9) 10.2 (2.8)
Height (cm) 76.4 (11.2) 76.3 (11.3) 76.4 (11.2)
MUAC (cm) 14.2 (1.8) 14.0 (1.7) 14.1 (1.8)
HAART use at recruitment 36 (34.3%) 39 (37.5%) 75 (35.9%)
Hospital admission in past 3 months2 21 (21.9%) 27 (27.3%) 48 (24.6%)
Previous use of anti-malarials in past month 28 (27.2%) 27 (27%) 55 (27.1%)
History of blood transfusion in past 3 months3 6 (6.1%) 4 (4.0%) 10 (5.1%)
Fever5 at recruitment 16 (15.2%) 6 (5.8%) 22 (10.5%)
Respiratory symptoms at recruitment 18 (17.1%) 22 (21.2%) 40 (19.1%)
Haemoglobin level at recruitment 9.4 (8.8-9.8) 9.4 (8.5-9.8) 9.4 (8.6-9.8)
Prevalence of iron deficiency4 38 (36.2%) 34 (32.7%) 72 (34.5%)
Positive malaria test6
at recruitment 6 (6.0%) 4 (3.9%) 10 (4.9%)
CD4 percentage at recruitment 28.7 (8.7) 31.1 (11.2) 29.9 (10.1)
Data presented as n (%), mean (s. d.) or median (IQR); 1N=200 (four missing values for iron arm & five for placebo); 2N=195 (nine
missing values for iron arm & five for placebo); 3N=198 (six missing values for iron arm & five for placebo); 4iron deficiency defined as
serum ferritin<12 µg/dl when CRP is <5.0 & ferritin <30 µg/dl when CRP >5.0; 5 Defined as axillary temperature ≥37.50C; 6 Asexual
parasites seen on blood smear microscopy.
P a g e | 57
Changes in haemoglobin, CD4 percentage, the prevalence of anaemia (Hb<11g/dl) and iron
deficiency over the study follow-up period are summarized in Table 2. Iron supplementation
was associated with a better haemoglobin (delta Hb) response compared with placebo at 3
months (adjusted mean difference (aMD) 0.62 g/dL 95%CI 0.20-1.04, p=0.004) and 6 months
follow-up (aMD 0.60 g/dL, 95%CI 0.06-1.13, p=0.03). Iron supplementation was also
associated with a reduced risk of anaemia persisting at 3 months (adjusted prevalence ratio
(aPR) 0.71 95%CI 0.51-0.97; p=0.03) and 6 months follow-up (aPR 0.59 95%CI 0.38-0.92;
p=0.02). In addition, iron supplementation was associated with a reduced risk of iron
deficiency (aPR 0.28 95%CI 0.15-0.49; p<0.001), and a better CD4 percentage response at 3
months (aMD 6.00 95%CI 1.84-10.16; p=0.005), but not at 6 months follow-up (p>0.05).
Morbidity outcomes by intervention period are presented in Table 3. Iron supplementation

was associated with an increased risk of clinical malaria infections over the 6 months follow-
up period (incidence rate (IR) 120.2 vs. 71.7; adjusted IRR 1.81 95%CI 1.04–3.16; p=0.04),
which was more pronounced in the first 3 months (IR 78.1 vs. 36.0; adjusted IRR 2.68 95%CI
1.08-6.63; p=0.03). In contrast, iron supplementation was associated with a lower incidence of
respiratory infections (IR 28.1 vs. 102.8; adjusted IRR 0.28 95%CI 0.09-0.91; p=0.04) in the
post-intervention period (3-6 months). There was no significant difference by study arm in the
incidence of out-patient sick visits (IR 350.0 vs. 383.0; adjusted IRR 0.98 95%CI 0.67 – 1.43;
p=0.91), hospital admissions (IR 28.6 vs. 42.3; adjusted IRR 0.62 95%CI 0.27-1.42; p=0.26) and
risk of progression to AIDS by 6 months (IR 31.9 vs. 50.8; adjusted IRR 0.61 95%CI 0.35–1.07;
p=0.08).
Treatment effect modification of morbidity outcomes by age, iron status, bed-net use and
HAART use are presented in table 4. Iron supplementation was associated with an increased
risk of clinical malaria infections in children >24 months old (IRR 3.30 95%CI 1.28-8.49;
p=0.01, p interaction=0.10), and in children who did not use bed nets (IRR 4.12 95%CI 1.89-
8.98; p=<0.001, p interaction=0.003), In addition, iron supplementation was associated with a
reduced risk of progression to AIDS in iron deficient children (IRR 0.25 95%CI 0.08-0.83;
p=0.02, p interaction=0.07; Table 4).
P a g e | 58
Table 2: Changes in laboratory values during the follow-up period by study arm
Laboratory parameters Iron 1 Placebo1 Unadjusted MD/PR2 P-value Adjusted MD/PR2 P-value
Delta Hb3 in g/dL at 3 months 2.08 (1.42) 1.53 (1.60) 0.55 (0.13-0.97) 0.01 0.62 (0.20-1.04) 0.004
Delta Hb in g/dL at 6 months 2.28 (1.54) 1.69 (1.84) 0.59 (0.11-1.06) 0.02 0.60 (0.06-1.13) 0.03
Prevalence of anaemia4 at 3 months 41 (39.1%) 60 (57.7%) 0.68 (0.51-0.91) 0.01 0.71 (0.51-0.97) 0.03
Prevalence of anaemia at 6 months 29 (27.6%) 48 (46.2%) 0.59 (0.41-0.86) 0.005 0.59 (0.38-0.92) 0.02
Delta CD4%5 at 3 months 3.86 (12.15) -1.45 (14.28) 5.32 (1.46-9.17) 0.007 6.00 (1.84-10.16) 0.005
Delta CD4% at 6 months 3.3 (12.07) -1.24 (14.52) 4.54 (0.59-8.49) 0.02 0.80 (-2.90-4.50) 0.67
Prevalence of ID6 at 3 months 19 (18.1%) 51 (49.0%) 0.27 (0.15-0.49) <0.001 0.28 (0.15-0.49) <0.001
Prevalence of ID at 6 months 30 (28.6%) 41 (39.4%) 0.60 (0.38-0.97) 0.04 0.86 (0.49-1.50) 0.60
1Data presented as mean change (95%CI) or N (%)
2 Presented as mean difference (MD) (95%CI) or prevalence ratio (PR) (95%CI); adjusted MD/PR values were adjusted for baseline haemoglobin,
bed net use, previous blood transfusions, HAART use at recruitment and CD4 percentage at recruitment.
3Delta Hb is the change in Hb from baseline
4Anaemia defined as Hb <11g/dl
5Delta CD4 percentage is the change in CD4 percentage from baseline
6Iron deficiency (ID) defined as ferritin <12 µg/dl when CRP <5 & ferritin <30 µg/dl when CRP≥5
P a g e | 59
Table 3: Morbidity outcomes by study arm and time period

Morbidity Outcomes Iron Placebo Unadjusted IRR1 P- Adjusted IRR1, 2 P-
Events (N) Incidence3 Events (N) Incidence (95%CI) value (95%CI) value
All-cause out-patient sick visits
All events over 6 months follow-up 133 (58) 350.0 135 (67) 383.0 0.95 (0.68 – 1.33) 0.76 0.98 (0.67 – 1.43) 0.91
Intervention period (0-3 months) 63 (27) 275.4 47 (26) 206.6 1.35 (0.83 – 2.19) 0.22 1.48 (0.81 – 2.68) 0.20
Post-intervention period (>3-6 months) 70 (31) 323.7 88 (41) 460.1 0.79 (0.50 – 1.26) 0.32 0.83 (0.51 – 1.35) 0.45
All-cause hospital admissions
Clinical malaria episodes
Respiratory infections
Progression to AIDS
1Treatment effects presented as incidence rate ratios (IRR) (95% confidence intervals)
2All adjusted treatment effects (excluding progression to AIDS) adjusted for baseline haemoglobin, bed net use, baseline cd4 percentage HAART use at
recruitment and recent blood transfusion prior to recruitment. Progression to AIDS adjusted for the same variables with the exception of HAART use at
recruitment.
3 Expressed as per 100 person-years
P a g e | 60
Table 4: Treatment effect modification with age, iron status, bed-net use and HAART use at recruitment over 6 months follow-up
Baseline Out-patient hospital sick Hospital admissions Progression to AIDS
Malaria
parameters visits
IRR1 (95% CI) p-value IRR1 (95% CI) p-value IRR1 (95% CI) p-value IRR1 (95% CI) p-value
Age (months)
≤24 1.11 (0.68 – 1.83) 0.67 1.20 (0.58 – 2.49) 0.62 0.41 (0.15 – 1.14) 0.09 0.30 (0.10 – 0.97) 0.04
>24 0.74 (0.45 – 1.23) 0.25 3.30 (1.28 – 8.49) 0.01 1.42 (0.36 – 5.53) 0.61 0.81 (0.44 – 1.50) 0.51
Interaction term 0.26 0.10 0.14 0.14
Iron status2
Iron deficient 1.33 (0.70 – 2.53) 0.39 2.67 (0.68 – 10.52) 0.16 0.33 (0.06 – 1.73) 0.19 0.25 (0.08 – 0.83) 0.02
Iron replete 0.87 (0.54 – 1.38) 0.55 1.79 (0.98 – 3.25) 0.06 0.86 (0.32 – 2.29) 0.76 0.88 (0.46 – 1.70) 0.71
Bed net use
Yes 1.25 (0.74 – 2.11) 0.41 0.74 (0.33 – 1.67) 0.47 1.13 (0.37 – 3.48) 0.83 0.79 (0.36 – 1.73) 0.56
No 0.78 (0.47 – 1.29) 0.33 4.12 (1.89 – 8.98) <0.001 0.34 (0.11 – 1.06) 0.06 0.56 (0.28 – 1.12) 0.10
HAART use at recruitment
On HAART 0.77 (0.40 – 1.45) 0.41 1.03 (0.39 – 2.70) 0.96 0.53 (0.17 – 1.66) 0.28
Not on HAART 1.14 (0.72 – 1.81) 0.58 2.39 (1.17 – 4.86) 0.02 0.74 (0.24 – 2.31) 0.60
Interaction term 0.31 0.17 0.68
1Treatment effects presented as incidence rate ratios (95% confidence intervals); All treatment effects (excluding progression to AIDS) adjusted for
baseline haemoglobin, bed net use, baseline cd4 percentage HAART use at recruitment and recent blood transfusion prior to recruitment.
2Defined as ferritin <12 µg/ml when CRP <5 & ferritin <30 µg/ml when CRP ≥5
P a g e | 61
Discussion:
To our knowledge, this is the first double-blind, randomized placebo-controlled trial of iron
supplementation for the treatment of anaemia in HIV-infected children living in an area with a
high pressure of infections (including malaria). This is not surprising, given the logistical and
ethical challenges we encountered in the conduct of this study. Mild-to-moderate anaemia is
still commonly seen in HIV infection despite increasing HAART use, and anaemia remains
strongly associated with HIV-disease progression [24, 25]. We have tried to conduct our study
to mirror common practice in resource-limited settings, where iron supplements are routinely
given for the treatment of anaemia without prior determination of iron status- in this way, our
study findings can be generalized for similar settings. The decision to treat anaemia in HIV-
infection, especially in resource-limited settings is often based on clinicians’ opinions and the
results of observational studies, as the effects (beneficial or detrimental) of iron
supplementation in this population are unknown [26]. Intervention trials for anaemia in HIV-
infected individuals are urgently needed, especially in resource-limited settings where the
pressure of infections is high.
Iron supplementation in our study cohort increased haemoglobin levels and reduced the
prevalence of anaemia persisting at 6 months follow-up by 40%, but increased the risk of
clinical malaria infections by 6 months. Our study was not powered to show that iron
supplementation did not have a deleterious effect on all-cause sick visits, all-cause hospital
admissions and progression to AIDS by 6 months.
Effect of iron supplementation on haemoglobin, immune function and HIV disease progression
Previous studies on iron supplementation in HIV-infection have shown varying results.
Castaldo et al [27] found an improvement in haemoglobin levels in Italian HIV-infected
children following daily oral iron, similar to the findings of a study in Indian HIV-infected
children following iron supplementation [28]. However, Olsen et al [29] did not observe any
haemoglobin improvement following iron supplementation of HIV-infected Kenyan adults over
a 12-month period [29]. This lack of effect may have been due to under-treatment as iron was
given only twice weekly. It could also have been due to differences in anaemia-associated co-
factors like the high incidence of helminth infections observed in the Kenyan cohort. These
findings have resulted in investigators questioning the role of iron deficiency in the aetiology
of HIV-associated anaemia [30, 31]. However the reduction in the prevalence of anaemia and
P a g e | 62
iron deficiency observed following iron supplementation in our study cohort suggests that iron
deficiency plays an important role in the aetiology of anaemia in this population [32].
Iron supplementation in our study was also associated with a better CD4 percentage response
at 3 months follow-up. This observation is in line with the findings of a US-based study in a
cohort of HIV- and Hepatitis C co-infected women following 6 months of iron supplementation
[33]. Studies from India, Togo and France have previously found iron deficiency to be
associated with reduced CD4 T-cell subsets and impaired cell-mediated immunity in children
without HIV-infection [34-36], with some studies demonstrating a beneficial effect on CD 4
with iron supplementation [34, 35] , and others failing to do so [36]. We have been able to
show in our study cohort that CD4 percentages respond positively to iron supplementation in
HIV-infected children.
Effect of iron supplementation on susceptibility to infections in HIV-infected children

Iron supplementation in our study appeared to increase the risk of malaria particularly during
the active intervention period, but was protective against respiratory infections during the
post-intervention period which are rather contrasting effects. Perhaps the differential effects
of iron on malaria and respiratory infections may be related to the mechanisms by which
different pathogens derive iron from the body. While Plasmodium species make use of the labile
intracellular iron pool which can be significantly increased by iron supplementation, many
bacteria derive iron either by the use of high-affinity siderophores or by direct interaction with
circulating iron-transport protein complexes [37, 38]. It may also be related to the observed
beneficial effect of iron supplementation on immune function, which may have been
responsible for the reduced risk of respiratory infections. The association between iron
supplementation and malaria morbidity has been highlighted in previous trials done in the
Gambia, Papa New Guinea and Tanzania [8, 9, 39, 40]. Iron supplementation in pre-school
Tanzanian children was associated with an increased risk for serious adverse events, hospital
admissions and deaths mainly due to malaria and other infectious diseases [8]. The increased
risk for adverse outcomes with iron appears to be mostly limited to regions with a high
prevalence of malaria, as a large trial of iron and folic acid supplementation of pre-school
children carried out in southern Nepal (an area with low malaria risk) did not show significant
differences in morbidity and mortality but appeared to be protective against acute respiratory
infections [41]. The significant interaction between iron supplementation and age resulting in
P a g e | 63
a higher incidence of malaria infections in older children (>24 months) is also interesting, as
children aged 6-24 months are at the greatest risk of severe disease and death following
malaria infection due to an immature immune response [42]. This observation may be a result
of younger children being more likely to be iron deficient due to nutrient-poor weaning diets,
hence are protected from malaria [43, 44]. However, further studies are needed to elucidate
the mechanisms by which iron supplementation resulted in its beneficial effects on immunity.
A recent meta-analysis on iron supplementation in the treatment of anaemia in children living

in malaria-endemic regions concluded that iron supplementation is safe when combined with
regular malaria surveillance and treatment [13]. Given the benefits of iron supplementation,
the increased risk of malaria and the protective effect of bed nets observed in our study,
perhaps a combination of interventions providing iron supplementation with malaria
prevention (by surveillance, treatment and prophylaxis) may be considered for the
management of HIV-associated anaemia in malaria-endemic regions [45].
The main limitation of our study was that we were unable to attain the projected sample size
and thus could not make any inferences on the incidence of all-cause sick child visits. However,
the effect of iron on malaria was very pronounced and was statistically significant despite the
reduced sample size. Another limitation of the study was that we were unable to examine the
effect of iron supplementation on mortality. A much larger sample size is needed if mortality
would be the primary outcome, which was not feasible in our setting. However these studies in
HIV-infected children are needed, since studies in HIV-infected adults suggest that iron
supplementation may be associated with an increased risk of mortality [46, 47].
Conclusions:
Our findings suggest that iron supplementation in anaemic HIV-infected children has beneficial
effects on haemoglobin; anaemia and immune function but increases the risk of malaria. Based
on these findings one may consider combining interventions of iron supplementation with
malaria prevention, in regions with a high prevalence of malaria. It is clear that further
mechanistic studies are needed to unravel this complex, but interesting interaction between
iron, immunity and infections in HIV-infected children.
P a g e | 64
Acknowledgements, Contributions and Declarations:

The investigators would like to thank the Netherlands-African partnership for capacity
development and clinical interventions against poverty-related diseases (NACCAP) for funding
support, Professors Malcolm E. Molyneux, Neil French, Bernard Brabin, Brian Faragher, Sian
Roberts and Alison Reynolds of the Liverpool School of Tropical Medicine; Queen Dube,
Macpherson Mallewa and Limangeni Mankhambo of the Department of Paediatrics and Child
Health, College of Medicine, University of Malawi; Professors Exnervia Gomo, Victor Mwapasa
and the entire staff of the Research Support Centre, College of Medicine; Felanji Simukonda
and Richard Kagomo of the Malawi-Liverpool-Wellcome Trust; Charlotte Adamczick and Kevin
Clarke of Zomba Central Hospital; the hospital directors and entire paediatric departments of
Thyolo District Hospital and Zomba Central Hospital as well as the families of all the children
who participated in this study.
MOE was responsible for preparing the manuscript, with support from MBvH, FOtK and KSP.
MOE and ECN were responsible for implementing study protocols, patient recruitment and
clinical trial site co-ordination under the supervision of KSP. MOE, KSP, CM, SW all had access
to the trial database and were all involved with statistical analysis. All the authors reviewed
the manuscript and agreed to submit the study findings for publication. The authors declare
that there are no conflicts of interest.
P a g e | 65
References
1. WHO, Worldwide prevalence of anaemia 1993 - 2005: WHO Global Database on Anaemia. 2008.
2. McLean, E., et al., Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition
Information System, 1993-2005. Public Health Nutr, 2009. 12(4): p. 444-54.
3. Ezzati M, L.A., Rodgers AA, Murray CJL. , Comparative quantification of health risks: global and
regional burden of disease attributable to selected major risk factors. WHO, 2004: p. Chapter 26;
page 2151.
4. Saah, A.J., et al., Factors influencing survival after AIDS: report from the Multicenter AIDS Cohort
Study (MACS). J Acquir Immune Defic Syndr, 1994. 7(3): p. 287-95.
5. Steinberg, J.P., et al., Predictors of outcome in AIDS patients receiving zidovudine. J Acquir
Immune Defic Syndr, 1989. 2(3): p. 229-34.
6. Moore, R.D., et al., Natural history of advanced HIV disease in patients treated with zidovudine.
The Zidovudine Epidemiology Study Group. Aids, 1992. 6(7): p. 671-7.
7. WHO and UNICEF, Iron supplementation of young children in regions where malaria transmission
is intense and infectious disease is highly prevalent. 2006: p. 2.
8. Sazawal, S., et al., Effects of routine prophylactic supplementation with iron and folic acid on
admission to hospital and mortality in preschool children in a high malaria transmission setting:
community-based, randomised, placebo-controlled trial. Lancet, 2006. 367(9505): p. 133-43.
9. Oppenheimer, S.J., et al., Iron supplementation increases prevalence and effects of malaria: report
on clinical studies in Papua New Guinea. Trans R Soc Trop Med Hyg, 1986. 80(4): p. 603-12.
10. Oppenheimer, S.J., et al., Effect of iron prophylaxis on morbidity due to infectious disease: report
on clinical studies in Papua New Guinea. Trans R Soc Trop Med Hyg, 1986. 80(4): p. 596-602.
11. Lewis, D.K., et al., Interpreting tests for iron deficiency among adults in a high HIV prevalence
African setting: routine tests may lead to misdiagnosis. Trans R Soc Trop Med Hyg, 2007. 101(6):
p. 613-7.
12. WHO, Assessing the iron status of populations. 2007.
13. Ojukwu, J.U., et al., Oral iron supplementation for preventing or treating anaemia among children
in malaria-endemic areas. Cochrane Database Syst Rev, 2009(3): p. CD006589.
14. Calis, J.C., et al., Erythropoiesis in HIV-infected and uninfected Malawian children with severe
anemia. Aids, 2010. 24(18): p. 2883-7.
15. Moses, A., J. Nelson, and G.C. Bagby, Jr., The influence of human immunodeficiency virus-1 on
hematopoiesis. Blood, 1998. 91(5): p. 1479-95.
16. Q.E.C.H Blantyre, M., Protocols for the management of common childhood illnesses in Malawi.
2004: p. 109.
17. WHO, Management of the child with a serious infection or malnutrition: Guidelines for care at the
first referral level in developing countries. 2000.
18. WHO. Antiretroviral therapy of HIV infection in infants and children: towards universal access.
Recommendations for a public health approach. 2010 [cited; 2nd:[93, Annex C]. Available from:
http://whqlibdoc.who.int/publications/2010/9789241599801_eng.pdf.
19. Thurnham, D.I., et al., Effects of subclinical infection on plasma retinol concentrations and
assessment of prevalence of vitamin A deficiency: meta-analysis. Lancet, 2003. 362(9401): p.
2052-8.
20. WHO, Iron Deficiency Anaemia Assessment, prevention and control: A guide for programme
managers. 2001.
21. Calis, J.C., et al., Severe anemia in Malawian children. N Engl J Med, 2008. 358(9): p. 888-99.
22. Malawi.N.A.C., Treatment of AIDS: Guidelines for the use of antiretroviral therapy in Malawi, third
edition. 2008: p. 1.
23. Wang, R., et al., Statistics in medicine--reporting of subgroup analyses in clinical trials. N Engl J
Med, 2007. 357(21): p. 2189-94.
24. Moore, R.D. and D. Forney, Anemia in HIV-infected patients receiving highly active antiretroviral
therapy. J Acquir Immune Defic Syndr, 2002. 29(1): p. 54-7.
P a g e | 66
25. Mocroft, A., et al., Anaemia is an independent predictive marker for clinical prognosis in HIV-
infected patients from across Europe. EuroSIDA study group. Aids, 1999. 13(8): p. 943-50.
26. Adetifa, I. and U. Okomo, Iron supplementation for reducing morbidity and mortality in children
with HIV. Cochrane Database Syst Rev, 2009(1): p. CD006736.
27. Castaldo, A., et al., Iron deficiency and intestinal malabsorption in HIV disease. J Pediatr
Gastroenterol Nutr, 1996. 22(4): p. 359-63.
28. Shet, A., et al., The prevalence and etiology of anemia among HIV-infected children in India. Eur J
Pediatr, 2011. 171(3): p. 531-40.
29. Olsen, A., et al., Failure of twice-weekly iron supplementation to increase blood haemoglobin and
serum ferritin concentrations: results of a randomized controlled trial. Ann Trop Med Parasitol,
2006. 100(3): p. 251-63.
30. Kosalaraksa, P., et al., Prevalence of Anemia and Underlying Iron Status in Naive Antiretroviral
Therapy HIV-Infected Children with Moderate Immune Suppression. AIDS Res Hum Retroviruses.
31. Esan, M.O., et al., Iron deficiency in children with HIV-associated anaemia: a systematic review and
meta-analysis. Trans R Soc Trop Med Hyg.
32. WHO/UNAIDS, Towards Universal Access: Scaling up priority HIV/AIDS interventions in the
health sector. 2010: p. pp7.
33. Semba, R.D., et al., Effect of micronutrients and iron supplementation on hemoglobin, iron status,
and plasma hepatitis C and HIV RNA levels in female injection drug users: a controlled clinical
trial. J Acquir Immune Defic Syndr, 2007. 45(3): p. 298-303.
34. Mullick, S., et al., Impact of iron deficiency anaemia on T lymphocytes & their subsets in children.
Indian J Med Res, 2006. 124(6): p. 647-54.
35. Berger, J., et al., Effect of daily iron supplementation on iron status, cell-mediated immunity, and
incidence of infections in 6-36 month old Togolese children. Eur J Clin Nutr, 2000. 54(1): p. 29-35.
36. Thibault, H., et al., The immune response in iron-deficient young children: effect of iron
supplementation on cell-mediated immunity. Eur J Pediatr, 1993. 152(2): p. 120-4.
37. Loyevsky, M., et al., Chelation of iron within the erythrocytic Plasmodium falciparum parasite by
iron chelators. Mol Biochem Parasitol, 1999. 101(1-2): p. 43-59.
38. Wandersman, C. and P. Delepelaire, Bacterial iron sources: from siderophores to hemophores.
Annu Rev Microbiol, 2004. 58: p. 611-47.
39. Smith, A.W., et al., The effects on malaria of treatment of iron-deficiency anaemia with oral iron in
Gambian children. Ann Trop Paediatr, 1989. 9(1): p. 17-23.
40. van den Hombergh, J., E. Dalderop, and Y. Smit, Does iron therapy benefit children with severe
malaria-associated anaemia? A clinical trial with 12 weeks supplementation of oral iron in young
children from the Turiani Division, Tanzania. J Trop Pediatr, 1996. 42(4): p. 220-7.
41. Tielsch, J.M., et al., Effect of routine prophylactic supplementation with iron and folic acid on
preschool child mortality in southern Nepal: community-based, cluster-randomised, placebo-
controlled trial. Lancet, 2006. 367(9505): p. 144-52.
42. Doolan, D.L., C. Dobano, and J.K. Baird, Acquired immunity to malaria. Clin Microbiol Rev, 2009.
22(1): p. 13-36, Table of Contents.
43. Parker, M.E., et al., The health of HIV-exposed children after early weaning. Matern Child Nutr.
44. Gwamaka, M., et al., Iron deficiency protects against severe Plasmodium falciparum malaria and
death in young children. Clin Infect Dis. 54(8): p. 1137-44.
45. Phiri, K., et al., Intermittent preventive therapy for malaria with monthly artemether-lumefantrine
for the post-discharge management of severe anaemia in children aged 4-59 months in southern
Malawi: a multicentre, randomised, placebo-controlled trial. Lancet Infect Dis. 12(3): p. 191-200.
46. Salmon-Ceron, D., et al., Lower survival in AIDS patients receiving dapsone compared with
aerosolized pentamidine for secondary prophylaxis of Pneumocystis carinii pneumonia. Study
Group. J Infect Dis, 1995. 172(3): p. 656-64.
47. McDermid, J.M., et al., Elevated iron status strongly predicts mortality in West African adults with
HIV infection. J Acquir Immune Defic Syndr, 2007. 46(4): p. 498-507.
P a g e | 67
Chapter 4
Iron supplementation improves the iron status of HIV-infected

Malawian children
Esan MO1, 2, 4, Phiri KS3, 4, Truwah ZT4, Musicha C1, Kraaijenhagen RJ5 , Boele van Hensbroek
M2.
1 Malawi-Liverpool-Wellcome Trust Research Laboratories, P. O. Box 30096, Blantyre, Malawi

2 Global Child Health Group- AIGHD, Emma Children’s Hospital/Academic Medical Centre, Meibergdreef
9. 1105 AZ, Amsterdam, the Netherlands
3 Community Health Department, College of Medicine, University of Malawi, P. O. Box 360, Blantyre,
Malawi
4Tropical Haematology Research Unit, College of Medicine, University of Malawi, P. O. Box 360,
Blantyre, Malawi
5Meander Medical Centre, Amersfoort, P. O. Box 1502 3800BM, the Netherlands
Submitted
P a g e | 68
Abstract:
Objective: To determine the iron status of HIV-infected Malawian children and the responses
of iron markers to iron supplementation.
Design: The study was part of a double-blind, randomized controlled trial of iron
supplementation in anemic HIV-infected children (Clinical trial number: ISRCTN-62947977).
Methods: HIV-infected children aged 6-59 months with moderate anemia (hemoglobin 7.0-9.9
g/dl) were randomized to receive 3mg/kg/day of elemental iron and multi-vitamins (Vitamins
A, C and D) or multi-vitamins alone for 3 months. Iron status was assessed at recruitment, 3
and 6 months. All participants were followed for 6 months.
Results: One thousand seven hundred and ninety serological iron tests were conducted in 209
HIV-infected children over 6 months follow-up. The baseline prevalence of iron deficiency
ranged from 8.1-48.3% when different serological iron markers were applied. Iron
supplementation was associated with significant reductions in soluble transferrin receptor
(sTfR) levels (adjusted mean difference (aMD) -1.00; 95%CI -1.47 to -0.54; p<0.001) and sTfR-
log ferritin (sTfR-F) index (aMD -0.65; 95%CI -0.90 to -0.40; p<0.001) at 3 months, and in
sTfR-F index at 6 months (aMD 0.21; 95%CI 0.03-0.39; p=0.02). Children who received iron
were less likely to be iron-deficient at 3 months when assessed using serum ferritin with C-
reactive protein (adjusted prevalence ratio (aPR) 0.29; 95%CI 0.16-0.51; p<0.001); sTfR (aPR
0.35; 95%CI 0.17-0.73; p=0.005) and MCV (aPR 0.49; 95%CI 0.24-1.00; p=0.05).
Conclusion: These findings suggest that iron supplementation is effective in improving iron
status in HIV-infected children, even in resource-limited settings with a high pressure of
infections.
P a g e | 69
Background:
The importance of knowing iron status prior to giving iron supplements has recently become
more significant, with studies suggesting that iron supplementation may increase the risk of
infections and mortality, especially in malaria-endemic regions [1-3] and in HIV-infected
individuals [4, 5]. However, the assessment of iron status is difficult in areas with a high
pressure of infections, where markers of iron status are also markers of the acute phase
response and may be difficult to interpret in the presence of inflammation [6, 7]. The
contribution of iron deficiency to anemia in HIV-infection is unknown, but observational
studies suggest that it may play an important role [8-10]. The use of iron supplements for the
management of HIV-associated anemia is an on-going discussion among clinicians in resource-
limited settings, where different opinions exist on the effects of iron on iron status in this
population, and no clear evidence-based guidelines are available [11]. We have assessed iron
status and the changes in iron markers over time as part of a randomized, double-blinded,
placebo-controlled trial of iron supplementation in HIV-infected children with moderate
anemia.
Methods:
HIV-infected children with moderate anemia (defined as hemoglobin of 7 to 9.9 g/dl) were
recruited sequentially from pediatric out-patient clinics in 2 hospitals in Southern Malawi
(Zomba Central Hospital and Thyolo District Hospital) as part of a randomized, controlled trial
from January 2009 to August 2010. After obtaining parental consent, eligible children were
randomized into two arms with the first (iron) arm receiving 3mg/kg of elemental iron and
multivitamins (Vitamins A 1500 I.U/ml, C 35mg/ml and D 400 I.U/ml) as once-daily
concentrated drops for 3 months (commercially available as Tri-Vi-Sol with iron©, Mead
Johnson Nutritionals, USA) and the second (placebo) arm receiving multi-vitamins (Vitamin A,
C and D) alone (Tri-Vi-Sol©) once daily for 3 months. Both supplements were identical in
content and appearance with the exception of iron. All participants were followed for 6
months. Ethical approval was obtained from the College of Medicine Research Ethics
Committee, Malawi.
P a g e | 70
Blood samples were collected for a full blood count using a Coulter counter (ABX Micros 60,
Horiba ABX S.A.S, Montpellier, France), the assessment of iron status (ferritin and soluble
transferrin receptor (sTfR)) and inflammation (C-reactive protein (CRP)) at recruitment, 3
months and 6 months follow-up. Samples were stored at -80 0C till the study was completed
and then shipped to Meander Medical Center, Amersfoort, the Netherlands where assays were
done using a Beckman LX20 analyzer (Beckman Coulter, Brea, CA, USA). The assays were
carried out according the instructions of the manufacturer, where samples were run in a fully
automated sequence as a single-in-one series with calibrators and control material as per the
manufacturer’s specifications.
Iron deficiency was defined as serum ferritin of <12 µg/l in participants without significant
inflammation (defined as CRP <5.0 mg/l) or <30 µg/l in participants with on-going
inflammation (defined as CRP was >5.0 mg/l) [12, 13]. Other definitions used were mean cell
hemoglobin concentration (MCHC) of <32 g/l, mean corpuscular volume (MCV) <67 fl (in
children <2 years old) and <73 fl (in children 2-5 years old) [13]; sTfR of >3.50 mg/l
(recommended cut-off for kit specific assay) and sTfR-F index (defined as sTfR / log10 ferritin)
[14] of >2.38 when CRP ≥5 mg/l or >3.24 when CRP <5 mg/l.
Stata 10 (StataCorp, College Station, Texas, USA) was used for data analysis. Multiple linear
regression models were used to test the intervention effect on continuous outcomes, with data
presented as mean (SD) and treatment effects expressed as mean difference (MD) and 95%
confidence intervals. Prevalence endpoints were summarized as percentages, with the
treatment effects expressed as prevalence ratios (PR) and 95% confidence intervals. All
adjusted treatment effects were adjusted for baseline variables (hemoglobin, bed net use, CD4
percentage, highly active anti-retroviral therapy (HAART) use at recruitment, and a history of
blood transfusion in the 3 months preceding recruitment) using negative binomial regression.
These variables were selected after exploring the association between baseline variables and
treatment outcomes in uni-variate models and variables with a p-value of <0.10 entered into
the final multi-variate model. P-values of <0.05 were taken as statistically significant.
P a g e | 71
Results:
One thousand seven hundred and ninety serological iron tests were conducted in 209 HIV-
infected children over 6 months follow-up. Baseline characteristics of study participants are
provided in Table 1. Ninety-nine children (47.4%) were male and the mean (SD) age was 25.8
(15.9) months. The baseline prevalence of iron deficiency ranged from 8.1-48.3% when
different serological iron markers were applied (Table 1).
Table 1: Baseline characteristics of study participants1

Child parameters N (%)/Mean (SD) Iron Placebo
Sex (Male) 99 (47.4%) 42 (40.0%) 57 (54.8%)
Age at recruitment (months) 25.8 (15.9) 25.2 (15.8) 26.4 (16.1)
Weight (kg) 10.2 (2.8) 10.2 (2.8) 10.1 (2.9)
Mid Upper Arm Circumference (cm) 14.1 (1.8) 14.2 (1.8) 14.0 (1.7)
Length (cm) 76.4 (11.2) 76.4 (11.2) 76.3 (11.3)
Laboratory parameters
Positive malaria test at enrolment 10 (4.9%) 6 (6.0%) 4 (3.9%)
C-reactive protein (CRP) (mg/l) 17.5 (38.5) 21.8 (39.2) 13.1 (37.5)
CD4 count at enrolment (cells/µl) 1418.1 (793.1) 1423.3 (806.8) 1412.8 (783.1)
Prevalence of iron deficiency based on different iron markers 2
Ferritin (F) & CRP 72 (34.5%) 38 (36.2%) 34 (32.7%)
Soluble transferrin receptor (sTfR) 59 (28.2%) 28 (26.7%) 31 (29.8%)
Mean Cell Haemoglobin Concentration (MCHC) 101 (48.3%) 49 (46.7%) 52 (50.0%)
Mean Corpuscular Volume (MCV) 67 (32.1%) 35 (33.3%) 32 (30.8%)
sTfR-F index 17 (8.1%) 9 (8.6%) 8 (7.7%)
1Data presented as N (%) or mean (s. d.)
2Defined as Ferritin of <12 µg/l where CRP is <5.0 mg/l & ferritin <30 µg/l where CRP is ≥5.0 mg/l; sTfR >3.5 mg/l;
sTfR-F >3.24 when CRP <5.0 mg/l & sTfR-F >2.38 when CRP is ≥5.0 mg/l; MCV <69 fl when age <24 months & MCV
<73 fl when age >24 months; MCHC <32 g/l.
The changes in iron markers over the study follow-up period are presented in Table 2. Iron
supplementation was associated with significant reductions in sTfR levels (adjusted mean
difference (aMD) -1.00; 95%CI -1.47- -0.54; p<0.001) and sTfR-F index (aMD -0.65; 95%CI -
0.90- -0.40; p<0.001) at 3 months, and in sTfR-F index at 6 months (aMD 0.21; 95%CI 0.03-
0.39; p=0.02). Iron supplementation was also associated with an increase in MCV at 3 months
(aMD 4.45; 95%CI 1.76-7.15; p=0.001).
P a g e | 72
Children who received iron were less likely to be iron deficient at 3 months follow-up when
assessed using serum ferritin with CRP (adjusted prevalence ratio (aPR) 0.29; 95%CI 0.16-
0.51; p<0.001), sTfR (aPR 0.35; 95%CI 0.17-0.73; p=0.005) and MCV (aPR 0.49; 95%CI 0.24-
1.00; p=0.05) when compared with placebo; but this effect was not sustained at 6 months
follow-up (p>0.05; Table 2).
P a g e | 73
Table 2: Changes in iron markers in HIV-infected children over a 6-month follow-up period
Changes in iron Iron1 Placebo1 Unadjusted MD/PR2 P- Adjusted MD/PR2 P-
markers (95%CI) value (95%CI) value
Ferritin
3 months 18.25 (93.15) -0.49 (39.14) 18.74 (-1.88-39.36) 0.08 17.49 (-4.76-39.74) 0.12
ID at 3 months3 12 (12.6%) 42 (42.4%) 0.28 (0.16-0.51) <0.001 0.29 (0.16-0.51) <0.001
6 months 25.19 (178.18) 21.06 (131.66) 4.13 (-41.54-49.79) 0.86 -2.56 (-55.00-49.88) 0.92
ID at 6 months3 21 (21.9%) 34 (34.7%) 0.61 (0.38-0.98) 0.04 0.88 (0.50-1.54) 0.65
sTfR
3 months -1.18 (1.69) -0.33 (1.59) -0.85 (-1.31- -0.39) <0.001 -1.00 (-1.47- -0.54) <0.001
ID at 3 months4 11 (10.5%) 27 (26.0%) 0.40 (0.21-0.77) 0.006 0.35 (0.17-0.73) 0.005
6 months -0.88 (1.85) -0.66 (1.56) -0.22 (-0.70-0.26) 0.37 0.37 (-0.02-0.76) 0.06
ID at 6 months4 18 (17.1%) 21 (20.2%) 0.85 (0.48-1.50) 0.57 1.01 (0.50-2.05) 0.97
MCHC
3 months 0.89 (6.65) 0.83 (2.59) 0.06 (-1.37-1.49) 0.93 -0.07 (-1.73-1.60) 0.94
ID at 3 months5 37 (35.2%) 37 (35.6%) 0.99 (0.69-1.43) 0.96 1.08 (0.75-1.57) 0.68
6 months 4.77 (13.99) 2.42 (7.50) 2.36 (-0.87-5.58) 0.15 3.10 (-0.71-6.90) 0.11
ID at 6 months5 18 (17.1%) 24 (23.1%) 0.74 (0.43-1.29) 0.29 0.79 (0.44-1.41) 0.43
MCV
3 months 4.70 (8.29) 0.35 (9.71) 4.35 (1.82-6.88) 0.001 4.45 (1.76-7.15) 0.001
ID at 3 months6 11 (10.5%) 24 (23.1%) 0.45 (0.23-0.88) 0.02 0.49 (0.24-1.00) 0.05
6 months 6.47 (9.91) 4.57 (9.48) 1.89 (-0.87-4.65) 0.18 -0.87 (-3.03-1.29) 0.43
ID at 6 months6 13 (12.3%) 12 (11.5%) 1.07 (0.51-2.25) 0.85 2.33 (0.96-5.67) 0.06
sTfR-F index
3 months -0.63 (1.07) -0.13 (0.73) -0.51 (-0.76- -0.25) <0.001 -0.65 (-0.90- -0.40) <0.001
ID at 3 months7 6 (5.7%) 11 (10.6%) 0.54 (0.21-1.41) 0.21 0.30 (0.07-1.24) 0.10
6 months -0.50 (1.08) -0.25 (0.98) -0.25 (-0.54- -0.04) 0.10 0.21 (0.03-0.39) 0.02
ID at 6 months7 9 (8.6%) 16 (15.4%) 0.56 (0.26-1.21) 0.14 0.67 (0.27-1.67) 0.39
1Datapresented as mean difference from baseline (SD), or as N (%)
2Treatment effects presented as mean difference (MD) or prevalence ratios (PR) with 95% confidence intervals; values adjusted for baseline iron
marker results, baseline haemoglobin level, bed net use, cd4 percentage, HAART use & blood transfusion in 3 months preceding recruitment.
3 Values adjusted for inflammation. Iron deficiency: ferritin <12 µg/l when CRP <5 mg/l & ferritin <30 µg/l when CRP ≥5 mg/l
4 Irondeficiency (ID): sTfR >3.5 mg/l; 5 MCHC<32 g/l; 6 MCV <67fl when age <24 months & MCV< 73fl when age >24 months
7ID: sTfR-F >3.24 when CRP>5 mg/l & sTfR-F >2.38 when CRP <5 mg/l
P a g e | 74
Discussion:
Determination of iron status in HIV-infection is difficult due to associated inflammation, and is
even more challenging in resource-limited settings where there is a high pressure of infections
[7]. However information on iron status is crucial as it may impact the outcomes of
interventions for anemia in resource-limited settings, and the determination of iron status
prior to giving iron supplements is recommended by the WHO in areas with a high prevalence
of malaria and infectious diseases [1]. Bone marrow iron assessment, which is the gold
standard for the determination of iron status, is invasive and hence not ideal for use as a
screening tool. Studies from Sub-Saharan Africa which have previously investigated iron
deficiency in HIV-infected children using ferritin as an iron marker were limited by the fact
that they did not take into account the effect of inflammation; hence there was a risk of
misclassification bias [9, 10]. We have shown in a recently conducted meta-analysis that iron
deficiency is equally prevalent in HIV-infected children from high and low income settings
[15]. We have now confirmed, using different serological iron markers, that iron deficiency is
prevalent in HIV-infected children, and plays an important role in the multi-factorial etiology
of anemia in this population.
Previous studies done in adults have suggested that iron supplementation may be harmful in
HIV-infection due to pre-existing high iron states [5, 16]. However iron deficiency was
prevalent in our study cohort prior to iron supplementation. Possible explanations for this
could be that children are a higher risk group for iron deficiency, as other factors such as poor
dietary intake and reduced intestinal absorption of iron may contribute to its prevalence [13,
17]. Also, our HIV-infected cohort was relatively healthy at recruitment with good baseline
CD4 counts- high iron states are often associated with late stage HIV-infection and mortality
[5].
A consensus is yet to be reached on a single iron marker for the determination of iron status in
regions with a high pressure of infections [18]. Ferritin, which is commonly used for this
purpose is also a marker of the acute phase response [6]. Serum iron is abnormally low in
chronic disease and varies diurnally and with meals [18]. Recently, soluble transferrin
receptor (sTfR) and sTfR-log ferritin index (sTfR-F index) have been put forward for the
determination of iron status in regions with a high pressure of infections as they are less
affected by inflammation [14, 19]. Iron supplementation in this study was associated with an
improvement in iron status as evidenced by a reduction in sTfR and sTfR-F index. There was
P a g e | 75
also a reduction in the prevalence of iron deficiency after 3 months of oral iron use. These
findings suggest that iron supplementation is effective in the treatment of iron deficiency
anemia in HIV-infected children living in regions with a high pressure of infections. They also
suggest that inadequate dietary intake (probably from micro-nutrient deficient diets), rather
than intestinal malabsorption, is important in the etiology of iron deficiency anemia in HIV-
infected children from resource-limited settings [17].
The results of this study further strengthen the case for the use of iron supplementation in
resource-limited settings with a high pressure of infections, including malaria. It is hoped that
the beneficial effect of iron supplementation on iron status seen in our study would inform
policy in such settings. However, iron supplementation should be provided with adequate
protection from malaria in malaria-endemic regions.
Acknowledgements, contributions and declarations

MOE was responsible for the preparation of the manuscript, with editorial advice from RJK,
KSP and MBvH. MOE and CM performed the statistical analysis. RJK supervised the laboratory
component of the study. MOE and ZTT were responsible for the field work under the
supervision of KSP.
The authors would like to thank Brigette Denis and Mavis Manyere of Malawi-Liverpool-
Wellcome Trust Laboratories for arranging for sample shipments and the Netherlands-African
partnership for capacity development and clinical interventions against poverty-related
diseases (NWO-NACCAP) for funding the study.
The authors declare that there are no conflicts of interest.
P a g e | 76
References
1. WHO, UNICEF. Iron supplementation of young children in regions where malaria

transmission is intense and infectious disease is highly prevalent. 2006:2.
2. Sazawal S, Black RE, Ramsan M, Chwaya HM, Stoltzfus RJ, Dutta A, et al. Effects of
routine prophylactic supplementation with iron and folic acid on admission to
hospital and mortality in preschool children in a high malaria transmission
setting: community-based, randomised, placebo-controlled trial. Lancet
2006,367:133-143.
3. Oppenheimer SJ, Gibson FD, Macfarlane SB, Moody JB, Harrison C, Spencer A, Bunari O.
Iron supplementation increases prevalence and effects of malaria: report on
clinical studies in Papua New Guinea. Trans R Soc Trop Med Hyg 1986,80:603-612.
4. Salmon-Ceron D, Fontbonne A, Saba J, May T, Raffi F, Chidiac C, et al. Lower survival in
AIDS patients receiving dapsone compared with aerosolized pentamidine for
secondary prophylaxis of Pneumocystis carinii pneumonia. Study Group. J Infect
Dis 1995,172:656-664.
5. McDermid JM, Jaye A, Schim van der Loeff MF, Todd J, Bates C, Austin S, et al. Elevated
iron status strongly predicts mortality in West African adults with HIV infection. J
Acquir Immune Defic Syndr 2007,46:498-507.
6. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to
inflammation. N Engl J Med 1999,340:448-454.
7. Lewis DK, Whitty CJ, Epino H, Letsky EA, Mukiibi JM, van den Broek NR. Interpreting
tests for iron deficiency among adults in a high HIV prevalence African setting:
routine tests may lead to misdiagnosis. Trans R Soc Trop Med Hyg 2007,101:613-617.
8. Calis JC, Phiri KS, Vet RJ, de Haan RJ, Munthali F, Kraaijenhagen RJ, et al. Erythropoiesis
in HIV-infected and uninfected Malawian children with severe anemia. Aids
2010,24:2883-2887.
9. Eley BS, Sive AA, Shuttleworth M, Hussey GD. A prospective, cross-sectional study of
anaemia and peripheral iron status in antiretroviral naive, HIV-1 infected
children in Cape Town, South Africa. BMC Infect Dis 2002,2:3.
10. Totin D, Ndugwa C, Mmiro F, Perry RT, Jackson JB, Semba RD. Iron deficiency anemia
is highly prevalent among human immunodeficiency virus-infected and
uninfected infants in Uganda. J Nutr 2002,132:423-429.
11. Adetifa I, Okomo U. Iron supplementation for reducing morbidity and mortality in
children with HIV. Cochrane Database Syst Rev 2009:CD006736.
12. Thurnham DI, McCabe GP, Northrop-Clewes CA, Nestel P. Effects of subclinical
infection on plasma retinol concentrations and assessment of prevalence of
vitamin A deficiency: meta-analysis. Lancet 2003,362:2052-2058.
13. WHO. Iron Deficiency Anaemia Assessment, prevention and control: A guide for
programme managers. 2001.
14. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptor and its ratio to serum
ferritin in the diagnosis of iron deficiency. Blood 1997,89:1052-1057.
P a g e | 77
15. Esan MO JF, Boele van Hensbroek, Calis JCJ, Phiri KS. Iron deficiency in children with
HIV-associated anaemia: a systematic review and meta-analysis Trans R Soc Trop
Med Hyg 2012.
16. de Monye C, Karcher DS, Boelaert JR, Gordeuk VR. Bone marrow macrophage iron
grade and survival of HIV-seropositive patients. Aids 1999,13:375-380.
17. Castaldo A, Tarallo L, Palomba E, Albano F, Russo S, Zuin G, et al. Iron deficiency and
intestinal malabsorption in HIV disease. J Pediatr Gastroenterol Nutr 1996,22:359-
363.
19. Ray A, Ndugwa C, Mmirot F, Ricks MO, Semba RD. Soluble transferrin receptor as an
indicator of iron deficiency in HIV-infected infants. Ann Trop Paediatr 2007,27:11-
16.
P a g e | 78
Chapter 5
Development and evaluation of a new paediatric blood

transfusion protocol for Africa
1 2 3 4 5 6
B. Cheema, E. M. Molyneux, J. C. Emmanuel, B. M’baya, M. Esan, H. Kamwendo,
2 7
L. Kalilani-Phiri & M. Boele van Hensbroek
1
Division of Emergency Medicine, University of Cape Town, Bellville, Cape Town 7535, South Africa,
2 3
College of Medicine, University of Malawi, Blantyre, Malawi, Blood Transfusion Medicine Specialist
Consultant, Zimbabwe,
4
Malawi Blood Transfusion Service, Blantyre, Malawi,
5
Malawi-Liverpool-Wellcome Trust Clinical Research Programme, Blantyre, Malawi,
6
Queen Elizabeth Central Hospital, Blantyre, Malawi,
7
Global Child Health Group, Emma Children’s Hospital, AMC, University of Amsterdam, The
Netherlands
Transfusion Medicine, 2010 Jun; 20(3):140 – 151.

P a g e | 79
Abstract
Severe anaemia is a common childhood emergency in developing countries. Practical evidence-

based guidance on when to transfuse, volume of transfusion and ideal duration of transfusion
is lacking. The aim of this study is to develop a paediatric transfusion protocol for use in under-
resourced environments and evaluate its usability in a busy African hospital setting. A
paediatric transfusion protocol based on the WHO Guidelines was developed for the Queen
Elizabeth Central Hospital (QECH), Blantyre, Malawi. On the basis of simple bedside clinical
features of respiratory, cardiovascular and neurological compromise, the protocol allocates
children with severe anaemia (haemoglobin ≤6g dL-1) to one of the three groups: complicated
anaemia, uncomplicated anaemia and anaemia with severe malnutrition. Data were collected
to monitor protocol adherence, delays to transfusion, post-transfusion haemoglobin and need
for repeat transfusion. Two-hundred and fifteen severely anaemic children were enrolled: 180
complicated, 25 uncomplicated and 10 severely malnourished. With respect to protocol
adherence, all children were allocated to the correct transfusion group; correct volume
(±10%) was given in 89·3%; correct duration (±30 min) in 86·2% and correct overall rate
(±10%) in 78·6%. Comparing old and new transfusion guidelines, a potential avoidable
transfusion rate of 29% was found. This study demonstrates that clear and detailed
transfusion guidelines based on simple bedside clinical features can be used in a very busy
children’s hospital in sub-Saharan Africa. With minimal additional equipment, volume and
duration of transfusion can be well controlled. Furthermore, having a protocol in place results
in a significant reduction of avoidable transfusions.
P a g e | 80
INTRODUCTION
In some parts of sub-Saharan Africa, 20–47% of all hospitalised children are given blood
transfusions (Greenberg et al., 1988; Obonyo et al., 2007). Global blood shortages and the risk
of transfusion-transmissible infections (TTIs) make the rational use of blood in children a high
priority. Whilst some research has been directed towards delineating the causes and effects of
anaemia, little work has been done on identifying which children need transfusion, what level
of haemoglobin (Hb) to transfuse at and what is the best way to give the blood.
Malawi is one of the world’s poorest countries and has significant challenges in health care
provision. According to the WHO Global database on anaemia, 73·2% of Malawian children
−1
aged 0·5–5 years have a haemoglobin level (Hb) less than 11 g dL and 4·8% have an Hb less
−1
than 5 g dL (the most commonly used definition of severe anaemia) [World Health
Organization (WHO), 2005]. Malawi is a malaria-endemic country with a peak of malaria and
severe anaemia patients seen during the rainy season (December to May) at which time there
is a marked rise in demand for blood transfusions. Previous studies have shown that having
transfusion protocols in place can markedly reduce the number of unnecessary transfusions
given to children (Craighead & Knowles, 1993; Vos et al., 1993).
However, in Malawi, as in many other African countries, a detailed paediatric transfusion
protocol is not available for doctors working in the health system.
Therefore, at the Queen Elizabeth Central Hospital (QECH), Blantyre, a new paediatric
transfusion protocol was developed as part of a Hospital Transfusion Committee initiative to
produce hospital-wide transfusion guidelines. The new protocol is based on WHO Guidelines:
clinical use of blood in medicine, obstetrics, paediatrics, surgery and anaesthesia, trauma and
burns (WHO, 2001) but is adapted to give more detailed guidance on which patients to
transfuse and speed and volume of transfusion for different patient groups. Baseline
information on transfusion practices prior to the introduction of the new protocol was
collected, followed by a prospective evaluation of the usability of and adherence to the new
protocol.
MATERIALS AND METHODS

This prospective cohort study was conducted at the Paediatric Emergency Department (ED) of
QECH in Blantyre, where an estimated 80 000 children are seen every year and 23 000 of these
P a g e | 81
children are admitted to the paediatric wards.

Subjects studied
Children attending the ED with suspected anaemia had a capillary blood sample taken for Hb
and packed cell volume (PCV) measurement. Children were eligible for the study if they were
−1
aged between 2 months and 15 years and had an Hb ≤6g dL (or PCV <18%).
At the time of enrolment, a medical history was taken and a standardised physical
examination was performed by a clinician. The clinicians working in this area at the time of the
study comprised three clinical officers and one consultant (B. C.). The three clinical officers
underwent a training session when the protocol was introduced to the paediatric department
(January 2007), and they had an extra training session at the start of the study. The study was
supervised on a daily basis by B. C. and H. K. Feedback on the study was given to the clinical
officers on a weekly basis and to the nursing staff on a monthly basis. All four ED clinicians
were involved in enrolling patients and completing study forms.
Hb was assessed using a HemoCue®Hb301 haemoglobin analyser (Angelholm, Sweden). In
addition, the blood sample was used for malaria parasite screen (MPS), which was done by
microscopic examination of thick blood films stained with Field’s stain. Parasite density was
graded from 0 to 5 (0 to more than 250 000 malaria parasites per microlitre). For children
requiring transfusion grouping and cross-matching were done and the time between taking
the blood sample and receiving the blood from the blood bank was documented, as was the
donor blood Hb (measured by HemoCue®). Either whole blood or red cell suspension
(hereafter referred to as ‘red cells’) was administered depending on availability from the
hospital blood bank.
Each unit of red cells from the Malawi Blood Transfusion Service (MBTS) was produced by
the following method. Blood is collected in triple/quadruple multiple (interconnected) bags;
the primary blood collection bag contains Citrate Phosphate Dextrose Adenine (CPDA-1) as the
preservative-anticoagulant solution (the donor’s blood is collected into this bag); the second
bag contains Saline, Adenine, Glucose, Mannitol (SAGM) and the third bag is empty. After
centrifugation, plasma is allowed to flow into the empty bag and the SAGM into the red cells
bag. This is done in a closed system (as in most European countries) so there is no risk of
laboratory contamination. Paediatric packs are made by sharing the contents of this adult unit
into two bags, again in a closed system. If smaller volumes are required, a sterile docking
device is then used to connect a set of bags known as transfer packs. Smaller paediatric packs
P a g e | 82
are then made, again in a closed system, hence sterile. Terumo Sterile Connecting Devices are
used. The plasma is used to make fresh frozen plasma and cryoprecipitate; it is not sent for
fractionation.
During transfusion, the rate was controlled by manually adjusting the number of drops
infused per minute, making use of burettes to measure the volume of blood. No infusion pumps
or other flow control devices were used. Vital signs were recorded hourly by the nursing staff.
The clinician reviewed the patients at 1 h (complicated group) or 2 h (malnutrition and
uncomplicated group). After completion of the transfusion, all children were admitted to the
paediatric ward. Decisions regarding the need for re-transfusion were made by the ward
clinician based on clinical indications and repeat Hb measurement.
Old transfusion guidelines
The existing paediatric departmental guidelines for blood transfusion advised transfusion for
−1
all children with PCV <15% (≈Hb 5·0g dL ) and transfusion in unwell children if PCV <18%
−1 −1
(≈Hb 6·0g dL ). The recommended volume of transfusion was 20 mL kg of whole blood or 10
−1
mL kg of red cells to run over 3–4 h. In malnourished children, it was recommended that half
this volume be given. In December 2006, over a 4-week period, data were gathered to assess
transfusion practice and adherence to the ‘old transfusion guidelines.’
New transfusion protocol
The ‘new’ paediatric blood transfusion protocol (hereafter referred to as the ‘new QECH
protocol’) was introduced in January 2007 (Table 1 and Fig. 1). Protocol evaluation was
carried out between January and May 2007, after approval by the College of Medicine Research
Ethics Committee.
The new QECH protocol was developed by adapting the WHO blood transfusion guidelines
(WHO, 2001). The specific adaptations are described under the following sub-headings: (i)
patient grouping, (ii) clinical signs, (iii) indications for transfusion, (iv) volume and (v)
duration of transfusion.
Patient grouping
The WHO guideline groups anaemic patients into two categories: chronic/compensated
anaemia and acute/decompensated anaemia. The new QECH protocol has three categories:
uncomplicated anaemia, complicated anaemia and severe malnutrition.
Justification: The first two categories are comparable with the two WHO categories
(compensated and decompensated). The third category was added to ensure that severely
P a g e | 83
malnourished children, who are at increased risk of volume overload, receive a smaller volume
and slower transfusion than well-nourished children. Severe malnutrition was defined as the
presence of either pitting oedema of both feet or severe visible wasting.
Clinical signs.
As in the WHO guideline, the terms ‘respiratory distress’, ‘neurological changes’ and
‘circulatory changes’ were retained in the new QECH protocol. However, the definitions for
these terms were slightly adjusted: for ‘respiratory distress,’ the criterion ‘increased use of
abdominal muscles for breathing’ was removed. ‘Neurological changes’ in the new QECH
protocol were defined by ‘impaired consciousness’ [Blantyre coma score of 4 or less (Molyneux
et al., 1989)] or ‘prostration’ (alert but unable to sit unaided if over 1 year or unable to
drink/breastfeed if less than 1 year of age). In the new QECH protocol, ‘circulatory changes’ in
a severely anaemic child is defined as cold extremities with prolonged capillary refill time
(over 2 s) and/or fine bi-basal crepitating.
Justification: ‘Increased use of abdominal muscles for breathing’ was removed as this is a
poorly defined, not commonly used, criterion for ‘respiratory distress’ in children.
‘Neurological changes’ were defined in more detail in the new QECH protocol, since the WHO
guideline only states ‘mental status changes’ without any further clarification. The definition
for ‘circulatory changes’ was adjusted by defining signs of possible shock and ‘congestive
cardiac failure’ (CCF) in more detail. Importantly, detection of fine bi-basal crepitations is
added since this may be an early sign of CCF.
P a g e | 84
Indications for transfusion: The new QECH protocol is in keeping with the WHO guideline
−1
recommending transfusion for all children with Hb of 4 g dL or less (irrespective of clinical
−1
condition). For the children with an Hb of 4–6 g dL , the new QECH protocol recommendation
is to transfuse all well-nourished children if any of the following signs are present: respiratory
distress, bi-basal crepitations, impaired consciousness, prostration or prolonged capillary refill
time (for definitions see above). The WHO guideline recommends transfusion for this group
only if showing ‘features of hypoxia, acidosis, impaired consciousness or hyperparasitaemia’.
Justification: Hypoxia, acidosis and hyperparasitaemia are not included in the new protocol
because these are often difficult to interpret clinically.
Volume of transfusion: The new QECH protocol advises a transfusion volume of 10 mL kg of red
−1
cells or 20 mL kg of whole blood for well-nourished children and half the standard volume
for children with severe malnutrition.
P a g e | 85
Fig.1: New QECH transfusion protocol outline.

P a g e | 86
Justification: The WHO guidance on volume of transfusion is slightly confusing. On one hand,
it states that if transfusion is needed, give sufficient blood to make the child clinically stable. On
−1 −1
the other hand, it recommends 20 mL kg of whole blood or 10 mL kg of red cells. Although,
−1 −1
in children, 5 mL kg of red cells or 10 mL kg of whole blood is usually sufficient to relieve
acute shortage of oxygen-carrying capacity (Esan, unpublished observations), the new QECH
protocol recommends the full WHO transfusion volume, in order to reduce the chance of a
rebound severe anaemia.
Duration of transfusion: In keeping with the WHO recommendation, the new QECH protocol
advises to give the initial part of the transfusion more quickly (first half over 1 h). The child is
then re-assessed and if the clinical condition has not worsened, the second half is given over
the following 2 h. If the child’s condition has deteriorated and bi-basal crepitations are present
(as a new finding), further transfusion needs to be stopped and frusemide considered. If,
however, the child is worsening without bi-basal crepitations then the second half of the
transfusion is also given over 1 h on the basis that the child is likely to still be critically
anaemic and requires further urgent blood (Fig. 1).
When calculating the correct volume, correct duration and correct overall rate of
transfusion, a 10% margin of error was allowed for volume and overall rate and a 30-min
margin of error for duration, giving upper and lower limits as follows:
−1
1. Correct volume of transfusion (±10%): complicated and uncomplicated 9–11 mL kg for red
−1 −1
cells; 18–22 mL kg for whole blood. Severe malnutrition: 4·5–5·5mLkg for red cells; 9–11
−1
mL kg for whole blood.
2. Correct duration of transfusion (±30 min): 150– 210 min – complicated; 210–270 min–
uncomplicated. Severe malnutrition: 210–270 min.
−1 −1
3. Correct overall rate of transfusion (±10%): Complicated 3·0–3·6mLkg h red cells; 6·0–
−1 −1 −1 −1 −1 −1
7·4mL kg h whole blood. Uncomplicated: 2·3– 2·8mLkg h red cells; 4·5–5·5mLkg h
−1 −1 −1 −1
whole blood. Severe malnutrition: 1·1–1·4mL kg h red cells and 2·3–2·8mLkg h whole
blood.
P a g e | 87
RESULTS
In order to get a baseline for comparison, prior to the introduction of the new QECH protocol,
−1
information was collected on patients with Hb <6·0g dL who were managed according to the
old transfusion guideline. Data were gathered on a total of 29 patients, of these 15 were female
and the median age of this group was 18 months (inter-quartile range (IQR) 12–29 months). Of
these 29 patients, 11 (38%) had complicated anaemia (one of whom died awaiting
transfusion); 7 (24%) had severe malnutrition and 11 (38%) had uncomplicated anaemia, all
of whom were transfused. Enough blood for transfusion was received for 19 of these 28
transfusions (67·9%). Children for whom an insufficient volume of blood was received from
the blood bank were excluded from analysis. Of the 19 patients evaluated, correct blood
volume was given in 11 cases (57·8%); transfusion duration was correct in 9 cases (47·4%)
and overall rate of transfusion was correct in 2 cases (10·5%).
As part of the new QECH protocol evaluation, 215 children were enrolled between January
and May 2007. The median age of the study population was 22 months and there were 107
females (49·8%). Table 2 summarises the main characteristics of the study group and Fig. 2
summarises the group allocation and study flow.
Table2: Patient demographics, laboratory findings and diagnoses

P a g e | 88
Protocol adherence
All 215 patients (100%) were allocated to the correct group and followed the correct
transfusion pathway. In 159 of 184 transfused cases (86·4%), enough blood was sent from the
blood bank to allow the transfusion protocol to be followed. Analysis of correct adherence to
the new QECH protocol is therefore restricted to these 159 patients; by anaemia group these
comprised 154 of 176 children with complicated anaemia; 3 of the 6 children with
uncomplicated anaemia and both severely malnourished children, respectively.
The correct total volume of blood (+/−10%) was given in 142 cases (89·3%); correct
duration of transfusion (+/−30 minutes) was kept to in 137 (86·2%) and the correct overall
volume per kilogram per hour (±10%) was given in 125 (78·6%) cases. Due to small numbers
in the pre-protocol group, statistical significance testing which compares correct adherence to
protocol in the two groups could not be performed.
Haemoglobin rise
The Hb rise was estimated in the 144 well-nourished children with complicated anaemia in
whom an Hb check was carried out within 24 h of transfusion. In these 144 children, the mean
−1 −1
Hb rise (SD) was 2·5g dL (1·3). Following transfusion, Hb remained less than 5·0g dL in 9 of
−1
these 144 children (6·3%) and less than 6·0g dL in 33 children (22·9%).
Repeat transfusion
Five of the 159 children (3·1%) had sufficient blood available for full transfusion in the ED
received a second transfusion in the ward.
Not transfused in ED but transfused in ward
In accordance with the new QECH protocol, 27 of the 215 patients (12·6%) were not
transfused in the ED – these comprised 8 of 10 children (80%) with severe malnutrition and
19 of 25 children (76·0%) with uncomplicated anaemia. Subsequently 10 of these 27 cases
(37·0%) received transfusion in the ward during their admission period (3 of 8 with severe
malnutrition and 7 of 19 with uncomplicated anaemia).
Not transfused
Seventeen patients (12 of 25 with uncomplicated anaemia and 5 of 10 with severe
malnutrition) were not transfused at all during admission. Of these 17 patients, 14 (82·4%)
had a repeat Hb done on the next day in the ward – the mean Hb of these 14 patients on
−1 −1
admission was 5·4g dL and the following day, without transfusion, it was 5·3g dL .
P a g e | 89
Delays to transfusion
184 patients in this study received a transfusion in the ED; the median (IQR) time taken from
sending a cross-match sample to blood bank to receiving blood for transfusion in ED was 55
min (40–85 min). In 30 cases (16·3%) the time to receive blood was ≤30 min, however, in 20
cases (10·9%) it was ≥120 min and a delay of ≥180 min was experienced in 7 cases (3·8%).
Donor Hb
Of the 184 patients transfused in ED, 96 (52·2%) were transfused with red cell concentrate
−1
and 88 (47·8%) with whole blood. The median donor blood Hb (IQR) was 19·7g dL (17·2–
−1 −1 −1
20·8g dL ) for red cell concentrate and 14·2g dL (12·9–17·0g dL ) for whole blood.
Outcome
Eleven of the 215 patients (5·1%) died in hospital (see Table 3 for details). Four children died
before transfusion – the time interval between arrival and death in these children was 20, 30,
100 and 105 min. Three of these four children had malaria parasites on blood film and
received a presumed diagnosis of severe malarial anaemia.
Two anaemic children who were not transfused at presentation in the ED subsequently died.
One was a jaundiced, hypoglycaemic child with marasmic kwashiorkor and a presenting Hb of
−1
5·6g dL . Despite treatment with dextrose, antibiotics and quinine, this child died 22 h after
−1
admission. The other was a 13-year old girl who presented with an Hb of 5·0g dL and a 4-
week history of abdominal pain. She was found to have splenomegaly and multiple abdominal
masses and was diagnosed with Burkitt’s lymphoma. Following bone marrow biopsy, she
received multiple blood transfusions and was treated with steroids and antibiotics. She
developed signs of bowel obstruction and died 24 days after admission.
P a g e | 90
NOTE: only those with sufficient blood sent from blood bank for full transfusion (N=159) included in final analysis of transfusion data
Figure 2: Flow of patients.

P a g e | 91
DISCUSSION
In a recent article exploring strategies for reducing the mortality and morbidity of anaemia in
African infants, Crawley concluded that ‘there is a need to develop transfusion guidelines that
are based on clinical criteria and not solely on the level of haemoglobin’ and also that ‘the
optimal speed and volume of transfusion remain to be determined’ (Crawley, 2004). This study
was carried out to look at whether a transfusion protocol adapted from the WHO transfusion
guidelines (2001) to give more detailed guidance on both clinical features of severe anaemia
and volume and speed of transfusion could be followed in a busy children’s ED in Malawi. The
study was not designed to look at clinical effectiveness of transfusion itself or at the longer
term morbidity or mortality resulting from transfusion or non-transfusion in children with
severe anaemia. Following the independent development of the new QECH protocol, it was
noted that this protocol has much in common with the guidelines for transfusion in the WHO
pocket book of hospital care for sick children (WHO, 2005) (e.g. separate guidance for children
−1
with severe malnutrition; only transfuse if Hb <4g dL unless signs of complication and more
appropriate signs of complicated anaemia such as deep and laboured breathing).
This study confirms that detailed transfusion protocols for children can be adhered to in a
busy hospital in a resource-poor setting with minimal additional equipment. The simple
bedside clinical criteria for allocation to the three groups of severe anaemia were easily
followed with all patients allocated to the correct group. The majority of patients (83·7%)
were allocated to the complicated group, having presented with respiratory, cardiovascular
and/or neurological signs of decompensated severe anaemia.
The 100% correct group allocation may have been influenced by the method of data
collection. During the study, the clinical officers completed a study data collection form for
each patient. This form asked if there were any signs of complicated anaemia and/or any signs
of severe malnutrition – with boxes for ticking the respective defining features. It is possible
that the act of recording the clinical features in this manner may have prompted the clinician
to assess more carefully and also to allocate the patient to the correct transfusion pathway.
With regards to volume, duration and overall rate of transfusion, this was given correctly in
89·3%, 86·2% and 78·2%, respectively, of the 159 patients who received sufficient blood for
full transfusion. This was an improvement from the old transfusion guideline group when 57·8,
47·4 and 10·5% of the 19 patients who received sufficient blood for full transfusion were given
the correct volume, duration and rate, respectively. (Due to small numbers in the old
P a g e | 92
transfusion guideline group, statistical testing of the significance of these differences was not
possible.)
−1
The mean (SD) Hb rise in this study was 2·5g dL (1·3) which is comparable with previous
studies in similar settings: English et al. (2002) documented a median (IQR) Hb rise of 3·0g
−1
dL (1·6–4·3) in 984 transfused children in Kenya and Holzer et al. (1993) found a mean (SD)
−1
Hb rise equivalent to 3·3g dL (1·5) in 60 children who received blood transfusion in
Tanzania. However, English et al. (2002) found that 24% of children in their study still had an
−1
Hb of less than 5 g dL following transfusion whereas in our study post-transfusion Hb of less
−1
than 5 g dL was found in only 6·3% of patients. Possible explanation for this difference
−1
includes the fact that our patients had higher median (IQR) pre-transfusion Hb of 4·2g dL
−1
(3·4–4·8) compared to a median (IQR) pre-transfusion Hb of 3·5g dL (2·8–4·2) in the Kenya
study. Another potential contributing factor could be that only patients who have had an
adequate volume of blood transfused were analysed for Hb rise in our study whereas it is not
clear if this was the case in the Kenyan study. Furthermore, our post-transfusion Hb
measurements were done within 24 h of transfusion compared to 18–48-h post-transfusion in
the Kenya study – this could have resulted in further drop in Hb in some of their patients.
Other studies have shown that transfusion appears to benefit only those children with an Hb
−1 −1
<4g dL or those with Hb <5g dL and exhibiting signs of decompensation as long as the
transfusion is given within the first 2 days of hospitalisation (Lackritz et al., 1992; Lackritz et
−1
al., 1997). In this study, a total of 27 children with an Hb between 4·1 and 6 g dL were not
transfused in the ED (19 in the uncomplicated group and 8 in the severe malnutrition group).
Ten of these 27 children were subsequently transfused in the paediatric ward on clinical
grounds.
Reductions in the number of transfusions after introduction of transfusion guidelines for
children have previously been documented. Craighead & Knowles (1993) found that strict
enforcement of a transfusion protocol in Malawi resulted in a drop in the percentage from 44
to 11% of severely anaemic under-5-year children who were transfused. Vos et al. (1993)
analysed patient records of 497 transfused children in Tanzania and found that 75% of all
avoidable transfusions were in those under the age of 5 and that applying a transfusion
−1 −1
threshold of 4 gdL vs. 5 g dL would have resulted in avoidable transfusions being reduced
from 62 to 35%.
P a g e | 93
In the present study, the combined number of children with uncomplicated anaemia (Hb <6g
−1
dL ) in both the pre-and post-new protocol groups was 35. If the old transfusion guideline of 5
−1
g dL had been applied to these 35 children, 22 (63%) would be transfused; however, using
−1
the new QECH protocol, with a threshold of 4 g dL , only 12 (34%) would be transfused. This
study had small numbers of children with uncomplicated anaemia and was not designed to
specifically address avoidable transfusion rates but with these limitations in mind, our results
indicate that a potential 29% avoidable transfusion rate amongst children with uncomplicated
−1
anaemia may be possible by bringing the threshold for transfusion from 5 to 4 g dL . Given the
dire shortages of blood supplies in sub-Saharan Africa and the danger of TTIs, this potential
reduction is important.
There were 11 deaths amongst the 215 patients in this study giving a mortality rate of 5·1%.
This compares favourably with mortality rates of 6–13% documented in previous studies in
African settings. Berkley et al. (2003) reported a 9% mortality amongst 1160 Kenyan children
−1
admitted with Hb <5·0g dL .Bojang et al. (1997) found a mortality rate of 13% amongst 173
children with severe malarial anaemia and a PCV of less than 12% in those who were assigned
to receive blood transfusion. Using data from eight prospective hospital-based studies in
African settings between 1990 and 1997, Obonyo et al. (1998) estimated an average risk of in-
hospital mortality from severe malarial anaemia of 12%. A recent prospective case-controlled
study reported a 6·3% mortality amongst 382 Malawian children presenting to the clinic with
−1
an Hb <5·0g dL (Calis et al., 2008).
In our study 4 of the 11 children who died – did so before transfusion could be commenced
(36·4%) and this is comparable with reported literature from similar settings. Calis et al.
(2008) reported that 9 of 24 deaths (37·5%) in anaemic children in Malawi occurred before
transfusion; English et al. (2002) found that 27 of 123 deaths (22·0%) in children with severe
malarial anaemia in Kenya occurred before transfusion and Bojang et al. (1997) reported that
15 of 23 Gambian children (65·2%) with severe malarial anaemia who died did so before blood
transfusion was given.
Previous researchers have observed bimodal patterns of presentation and diagnoses
between children who died soon after admission and those who died later – postulating that
children with a clinical picture of severe malarial anaemia tend to die early whilst those with
other diagnoses generally die later (English et al., 2002; Berkley et al., 2003). These
P a g e | 94
researchers have categorised deaths from severe anaemia into early (<6 h) and late deaths (>6
h) (English et al., 2002) or immediate (<4 h), early (4–48 h) and late deaths (>48 h) (Berkley et
al., 2003). Using the definitions of Berkley, in our study 4 of 11 (36·4%) children who died
would be classified as immediate, 4 (36·4%) as early and 3 (27·2%) as late deaths. This
compares with the reported percentages of 26·0% immediate, 41·3% early and 32·7% late
deaths amongst 1160 severely anaemic children in the Berkley study (2003). In keeping with
the above theory – three of four (75%) of immediate deaths and four of eight (50%) of the
deaths in first 48 h had malaria parasites on initial blood screen.
Blood shortages and delays in receiving blood for transfusion are widespread in the
developing world with consequent increased morbidity and mortality. Few studies have
documented the length of time from arrival in ED to death whilst awaiting blood for
transfusion (Lackritz et al., 1992). In this study, the time between arrival into the study and
death, in those who died before blood transfusion, was between 20 and 105 minutes. Although
this study does not formally address issues of blood supply, it is worth noting that the median
(IQR) time to receipt of blood for the 184 patients who received transfusion in this study was
55 min (40–85 min). Given the enormous constraints and difficulties in blood supply in the
developing world, this figure is impressive.
This protocol has options for transfusing either whole blood or red cells (also known as ‘red
cell suspension’ or ‘red cell concentrate’). However, in many sub-Saharan countries (with less
well developed blood transfusion services), red cells are not available and ‘packed cells’ are
produced by the removal of most of the supernatant plasma from a unit of whole blood –
leaving just sufficient plasma to allow adequate viscosity and blood flow through
administration sets.
QECH is very fortunate to have close links with the MBTS, which provides ongoing support of
the QECH blood bank, has membership on the Hospital Transfusion Committee and provides
effective administrative support. A recent paper (Hassall et al., 2009) has described the high
incidence of bacterial contamination of paediatric whole blood transfusions in Kenya and
elsewhere in sub-Saharan Africa. The authors cite needle puncture of blood bags to decant
smaller amounts for paediatric transfusion as one of the main factors for the high
contamination rates. The MBTS uses a sterile multiple blood bag system and a docking device
as is the system in Europe and provides QECH with pre-prepared paediatric units of red cells
and whole blood. Undoubtedly, the presence of MBTS results in QECH having a much swifter
P a g e | 95
and safer supply of blood than is available in many parts of the developing world.
Neither burettes nor any other form of flow control device (such as infusion pumps) was
routinely available at QECH at the time of the study. The existing method for volume control
involved looking at the volume in the blood pack and then drawing a line on the blood bag to
indicate the estimated point at which the transfusion should cease. Burettes were therefore
deemed necessary for the study – even though they may well have contributed significantly to
the correct volume of blood given – as otherwise we would not have been able to document
with any accuracy the volumes of blood given. Although burettes are not routinely available in
poorer African countries, we feel that they should be advocated for as they are an essential
item to prevent over-infusion of blood or intravenous fluids to small children and
malnourished patients, who are particularly at risk from fluid overload.
This study demonstrates that clear and detailed transfusion guidelines based on simple
bedside clinical features, aimed at identifying children at most risk from severe anaemia, can
be used in a busy children’s ED in sub-Saharan Africa. With minimal additional equipment,
volume and duration of transfusion can be well controlled. Furthermore, having a protocol in
place with clear indications on which anaemic children do not require transfusion results in a
reduction of avoidable transfusions.
DECLARATIONS OF INTEREST: None declared.
ACKNOWLEDGMENTS
This project was supported by a grant of £5500 from the British Blood Transfusion Society.
Two Hb301 haemoglobin analysers used in the study were donated by HemoCue® Angelholm,
Sweden. The MBTS supported this project by managing the financial administration of the
project and providing technical information and advice.
P a g e | 96
REFERENCES
Berkley, J.A., Ross, A., Mwangi, I. et al. (2003) Prognostic indicators of early and late deaths in children
admitted to district hospital in Kenya: cohort study. British Medical Journal , 326, 361–366.
Bojang, K.A.,Van Hensbroek, M.B., Palmer, A., Banya, W.A., Jaffar, S. & Greenwood, B.M. (1997)
Predictors of mortality in Gambian children with severe malaria anaemia. Annals of Tropical
Paediatrics, 17, 355–359.
Calis, J.C., Phiri, K.S., Faragher, E.B. et al. (2008) Severe anemia in Malawian children. The New England
Journal of Medicine, 358, 888–899.
Craighead, I.B. & Knowles, J.K. (1993) Prevention of trans-fusion-associated HIV transmission with the
use of a transfusion protocol for under 5s. Tropical Doctor , 23, 59–61.
Crawley, J. (2004) Reducing the burden of anaemia in infants and young children in malaria-endemic
countries of Africa: from evidence to action. The American Journal of Tropical Medicine and Hygiene,
71(Suppl. 2), 25–34.
English, M., Ahmed, M., Ngando, C., Berkley, J. & Ross, A. (2002) Blood transfusion for severe anaemia in
children in a Kenyan hospital. The Lancet , 359, 494–495.
Greenberg, A.E., Nguyen-Dinh, P., Mann, J.M. et al. (1988) The association between malaria, blood
transfusions, and HIV seropositivity in a pediatric population in Kinshasa, Zaire. The Journal Of the
American Medical Association, 259, 545–549.
Hassall, O., Maitland, K., Pole, L. et al. (2009) Bacterial contamination of pediatric whole blood
transfusions in a Kenyan hospital. Transfusion, 49, 2594–2598.
Holzer, B.R., Egger M., Teuscher T., Koch S., Mboya D. & Davey Smith G. (1993) Childhood anaemia in
Africa: to transfuse or not transfuse. Acta Tropica, 55, 47–51.
Lackritz, E.M., Campbell, C.C., Ruebush, T.K. II, Hightower A.W., Wakube W., Steketee R.W. & Were J.B.
(1992) Effect of blood transfusion on survival among children in a Kenyan hospital. The Lancet , 340,
524–528.
Lackritz, E.M., Hightower, A.W., Zucker, J.R. et al. (1997) Longitudinal evaluation of severely anemic
children in Kenya: the effect of transfusion on mortality and hematologic recovery. AIDS , 11, 1487–
1494.
Molyneux, M.E., Taylor, T.E., Wirima, J.J., Borgstein, A. (1989) Clinical features and prognostic indicators
in paediatric cerebral malaria: a study of 131 comatose Malawian children. Quarterly Journal of
Medicine, 71, 441–459.
Obonyo, C.O., Steyerberg, E.W., Oloo, A.J. & Habbema, J.D.F. (1998) Blood transfusion for severe malaria-
related anemia in Africa: a decision analysis. The American Journal of Tropical Medicine and Hygiene,
59, 808–812.
Obonyo, C.O., Vulule, J., Akhwale, W.S. & Grobbee, D.E. (2007) In-hospital morbidity and mortality due
P a g e | 97
to severe malarial anemia in western Kenya. The American Journal of Tropical Medicine and Hygiene,
77(Suppl. 6), 23–28.
Vos, J., Gumodoka, B., Ng’weshemi, J.Z., Kigadye, F.C., Dolmans, W.M. & Borgdorff, M.W. (1993) Are some
blood transfusions avoidable? A hospital record analysis in Mwanza Region, Tanzania. Tropical and
Geographical Medicine, 45, 301–303.
World Health Organization. (2001) Clinical use of Blood in Medicine, Obstetrics, Paediatrics, Surgery and
Anaesthesia, Trauma and Burns. World Health Organization, Geneva. 337 pp. ISBN 92 4 154538 0.
World Health Organization. Vitamin and Mineral Nutrition Information Service Global Database on
Anaemia Malawi 2004–2005. URL:http://www.who.int/vmnis/anaemia/data/database/countries/mwi
ida.pdf (last accessed on 8 February 2010).
World Health Organization. (2005) Pocket Book of Hospital Care for Sick Children Guidelines for the
Management of Common Illnesses with Limited Resources. World Health Organization, Geneva. 378 pp.
ISBN 9241546700.
P a g e | 98
Chapter 6
High transfusion failures rates in Malawian children with severe

anaemia following a standard blood transfusion regimen
Esan MO1 ,3, 4, Phiri KS1 ,2, Molyneux EM2, Mukaka M1, Cheema B5 and Boele van Hensbroek M3, 4
1 MLW Clinical Research Programme, Blantyre, Malawi

2 College of Medicine, University of Malawi
3 Global Child Health Group, Emma Children’s Hospital, AMC, Amsterdam, the Netherlands
4 Liverpool School of Tropical Medicine, UK
5 Division of Emergency Medicine, University of Cape Town, South Africa
British Journal of Haematology, 2011 Sept; 154(6); 783-785.

P a g e | 99
Abstract
Background: Severe anaemia is a major cause of childhood morbidity and mortality. It has
been suggested that post-discharge mortality may be partly explained by an inadequate
haematological response to blood transfusion.
Methods: A prospective cohort study was conducted at Queen Elizabeth Hospital, Blantyre in
2007. All children meeting the WHO criteria for severe anaemia were recruited and transfused
according to WHO guidelines. Regular clinical assessments were made and post-transfusion
haemoglobin (Hb) measured after 24 hours. Transfusion failure was defined as a post-
transfusion Hb ≤6 g/dl.
Results: 128 children aged 3-60 months were recruited. Malaria was seen in 73.4% of
children, and 23.4% of children had transfusion failure. In the multivariate analysis, failure
was predicted by pre-transfusion Hb (Adjusted OR 0.4; 95% CI 0.25 – 0.68; p=0.001) and little
reduction in respiratory rate during transfusion (Adjusted OR 0.93; 95% CI 0.88 – 0.98;
p=0.008), but not by malaria infection. Malaria parasitaemia was associated with a poor Hb
response to blood transfusion (mean (SD) Hb change=2.6(1.1) versus 3.1(1.2) g/dl; p=0.02).
Conclusion: Blood transfusion using the current WHO guidelines results in a high transfusion
failure rate. Introducing post-transfusion Hb and malaria tests in malaria-endemic settings
may help reduce post-discharge mortality.
P a g e | 100
Background
Severe anaemia is major cause of childhood morbidity and mortality in Africa. In Malawi,
severe anaemia accounts for up to 54% of malaria-related deaths (Slutsker, et al 1994). Most
severe anaemia-related deaths occur in children with signs of respiratory distress and/or
cardiac failure, often within the first 24 hours of admission (Marsh, et al 1995). The WHO
advises to transfuse severely anaemic children with 10mls/kg of packed cells (PC) or 20mls/kg
of whole blood (WB), if PC are not available (Organisation 2000). We have assessed the clinical
and early haematological responses of severely anaemic children transfused according to WHO
guidelines. In addition we have tried to identify clinical and laboratory markers that would
predict a transfusion failure.
Methods
A prospective cohort study was conducted at the Paediatric Emergency of Queen Elizabeth
Central Hospital (QECH), Blantyre, Malawi in March-April 2007. Children <5 years presenting
with signs and symptoms suggestive of severe anaemia (defined as haemoglobin (Hb) of <4
g/dl or Hb <6g/dl with evidence of hypoxia or hyperparasitaemia) were screened, and eligible
children enrolled after obtaining informed consent. A clinical history was taken, physical
examination performed and a sample of venous blood collected for Hb check (Hemocue Hb
301; Angelholm, Sweden), thick film for malaria parasites, blood glucose (Hemocue B-glucose;
Angelholm, Sweden), blood group and cross-match. When a diagnosis other than malaria was
suspected, children were managed according to hospital protocol.
Blood for transfusion was obtained from the Malawi Blood Transfusion Service (MBTS) and
screened according to MBTS standard operating procedures. Severely anaemic children were
transfused with either 10mls/kg of PC (or 20mls/kg of WB) over 3 hours based on hospital
availability. Clinical assessments were done manually every 15 minutes for the first two hours
and half-hourly for the last hour. Post-transfusion Hb was measured after 24 hours. Ethical
approvals were obtained from the College of Medicine Research Ethics Committee and the
Liverpool School of Tropical Medicine (LSTM) Ethics Committee.
Statistical analysis was done using Stata version 10 (StataCorp, College Station, TX USA).
Continuous data were analysed by independent samples Student t-tests; categorical data were
analyzed using the chi-squared or Fisher’s exact test. The change in Hb (delta Hb) was defined
as the difference between the post-transfusion Hb and the pre-transfusion Hb. Delta
P a g e | 101
respiratory rate (RR) and delta heart rate (HR) were defined as the difference in RR/HR
between the start of transfusion and the RR/HR at 60, 120 and 180 minutes after transfusion
was commenced. Transfusion failure was defined by a post-transfusion Hb ≤6 g/dl. Univariate
and multivariate logistic regression using transfusion failure as the outcome was done looking
at different clinical and laboratory markers. Important clinical and baseline parameters were
analysed after patients were categorised into groups based on presence or absence of malaria
parasitaemia. P-values of <0.05 were taken as statistically significant.
Results
128 children were recruited with a mean (SD) age of 22.6(13.8) months. Fever was the most
common presenting symptom seen in 124 children (96.9%). Malaria parasites were found in
73.4% of children, suspected septicaemia in 14.1% and suspected pneumonia in 7%. The in-
patient mortality rate was 4.7% (6/128).
Table I: Baseline characteristics of study participants

Baseline Parameter Mean (SD) or N (%)
Age (months) 22.6 (13.8)
Sex (male) 63/128 (49.2%)
Weight (kg) 9.5 (2.5)
Symptoms (history)
Fever 124/128 (96.9%)
Convulsions 6/128 (5.4%)
Clinical Signs
Fast breathing1 59/128 (46.1%)
Bi-basal crackles 1/128 (0.8%)
Impaired consciousness2 24/128 (18.8%)
Enlarged liver 79/128 (61.7%)
Enlarged spleen 72/128 (56.3%)
Prolonged CRT3 3/128 (2.3%)
Laboratory
Malaria Parasitaemia 94/128 (73.4%)
Pre-transfusion Hb 4 (1.0)
Pre-transfusion Hb ≤4 55/128 (43%)
Hypoglycaemia4 12/128 (10.8%)
Donor Hb (PC) in g/dl 18.9 (3.5)
1>50 breaths/minute; 2Blantyre Coma Score ≤ 4/5;
3Capillary re-fill time ≥ 3 seconds; 4Blood sugar ≤ 2.2mMol/L (40 g/dl)
Thirty children (23.4%) were classified as transfusion failures (post-transfusion Hb ≤6 g/dl).

Multivariate analysis identified high pre-transfusion Hb (Adjusted Odds Ratio [adj. OR] 0.4;
95% CI 0.25 – 0.68; p=0.001) and a reduction in respiratory rate (delta RR) (Adj. OR 0.93; 95%
P a g e | 102
CI 0.88 – 0.98; p=0.008) to be significantly associated with a reduced risk of transfusion failure
(table II).
Table II: Logistic regression of potential predictors of a low post-transfusion Hb (Hb ≤ 6g/dl)
Outcome1 Uni-variate OR2 (95% P-value Multi-variate OR2 (95% C.I.) Adj. P-value
Malaria 1.25(0.48 – 3.25) 0.65 1.08(0.33 – 3.60) 0.90
Age 1.00(0.97 – 1.03) 0.96 1.00(0.97 – 1.04) 0.89
Weight 0.95(0.80 – 1.13) 0.58 0.95(0.74 – 1.20) 0.65
Gender 1.56(0.68 – 3.59) 0.29 1.80(0.68 – 4.74) 0.24
Glucose 0.96(0.84 – 1.11) 0.60 0.92(0.77 – 1.09) 0.33
Blood type (PC or WB) 1.24(0.53 – 2.89) 0.61 3.11(0.95 – 10.2) 0.06
Donor Hb 0.94(0.86 – 1.01) 0.13 0.89(0.78 – 1.01) 0.07
Pre-transfusion Hb 0.47(0.30 – 0.73) 0.001 0.41(0.25 – 0.68) 0.001
Delta RR3 0.92(0.88 – 0.97) 0.002 0.93(0.88 – 0.98) 0.008
Delta HR4 0.98(0.95 – 1.00) 0.034 0.99(0.96 – 1.01) 0.420
1 Post transfusion Hb ≤ 6 g/dl; N=30 (23.4%); 2 Odds Ratio; Change in RR3 and HR4 during transfusion period.
Children without malaria had better early Hb responses (within 24hours) to blood transfusion
compared with malaria-infected children (mean (SD) 3.1(1.2) to 2.6(1.1); p=0.02), but
transfusion failures were not predicted by malaria infection (table II).
Table III: Changes in clinical parameters and haemoglobin by malaria parasitaemia

Parameters Malaria negative (N = 34) Malaria parasitaemia (N = 94) P-value
Change in RR (cycles/min)1
After 60 minutes 8 (0 - 12) 4 (-4 - 12) 0.88
After 120 minutes 4 (-4 - 12) 4 (-4 - 8) 0.27
After 180 minutes 7 (0 - 12) 4 (0 - 12) 0.60
Change in HR (beats/min)1
After 60 minutes 11 (4 - 20) 10 (4 - 18) 0.91
After 120 minutes 18 (8 - 34) 18 (8 - 26) 0.61
After 180 minutes 25 (12 - 36) 21 (8 - 32) 0.73
Haemoglobin (g/dl)2
Pre-transfusion Hb 3.9 (1.2) 4.1 (0.9) 0.22
Post-transfusion Hb 7.1 (1.6) 6.7 (1.0) 0.12
Delta Hb 3.1 (1.2) 2.6 (1.1) 0.02
Other
Age (months)2 24.4 (14) 22.5 (14.2) 0.50
Weight (kg)2 8.8 (2.6) 9.7 (2.4) 0.07
Blood transfused (PC or WB)3 0.7 (0.54 – 0.86) 0.55 (0.45 – 0.65) 0.14
1 Median (IQR); 2 Mean (SD); 3 Proportion (95% CI), Chi2 with one degree of freedom = 2.22
Discussion
The high transfusion failure rate (23.4%) observed in this study, comparable with the 24%
transfusion failure rate (defined as post-transfusion Hb ≤5 g/dl) observed in Kenya, raises
P a g e | 103
concern as to whether the current WHO guidelines adequately treat severe anaemia in
resource-limited settings (English, et al 2002). In many clinics across sub-Saharan Africa,
access to laboratory investigations is limited and post-transfusion Hb checks are not routine.
Children inadequately transfused, but clinically improved, are discharged home without a
follow-up appointment- putting them at risk of rebound severe anaemia. When this occurs,
parents may chose not to return to hospital but instead keep the child at home or seek advice
from a traditional healer (Mota, et al 2009). In this way transfusion failures may contribute
significantly to the high post-discharge mortality observed in Malawi, Kenya and Gambia
following hospital admission for severe anaemia (Bojang, et al 1997, Lackritz, et al 1992,
Lackritz, et al 1997, Phiri, et al 2008, Zucker, et al 1996). Our results suggest that early
identification of children at high risk of transfusion failure may be possible. A low pre-
transfusion Hb level and the absence of a significant reduction in respiratory rate (delta RR)
during transfusion were reliable predictors of transfusion failure.
Malaria infection was associated with poor early Hb response to blood transfusion. Adequate
emphasis should be placed on parasitological clearance and close clinical monitoring during
hospitalization in areas endemic for Falciparum malaria.
A limitation of the study was that the type of transfusion (either PC or WB) given was based on
availability and not random allocation. All clinical assessments were done manually so were
subject to observer bias. HIV testing was not part of routine practice at the paediatric
department during our study thus was a missed opportunity.
Conclusions
Blood transfusion in resource-limited settings using the current WHO guidelines results in a
high transfusion failure rate. A low pre-transfusion Hb level and little reduction in respiratory
rate during transfusion may help to detect early those at high risk of failing. Introducing post-
transfusion Hb checks in resource-limited settings and reviewing the ‘one-size-fits-all’
approach on transfusion policy may need to be considered. There is a need to evaluate the
safety and efficacy of alternative transfusion strategies.
Acknowledgements
Special thanks to Professors Bernard Brabin, Feiko ter Kuile and Dr Imelda Bates of LSTM, Dr
Bridon M’baya of MBTS and the staff of the Paediatric Emergency Department, QECH, Blantyre
Malawi. The study was funded by the LSTM Master of Tropical Paediatrics programme.
P a g e | 104
References
Bojang, K.A., Palmer, A., Boele van Hensbroek, M., Banya, W.A. & Greenwood, B.M. (1997)
Management of severe malarial anaemia in Gambian children. Trans R Soc Trop Med
Hyg, 91, 557-561.
English, M., Ahmed, M., Ngando, C., Berkley, J. & Ross, A. (2002) Blood transfusion for severe
anaemia in children in a Kenyan hospital. Lancet, 359, 494-495.
Lackritz, E.M., Campbell, C.C., Ruebush, T.K., 2nd, Hightower, A.W., Wakube, W., Steketee, R.W.
& Were, J.B. (1992) Effect of blood transfusion on survival among children in a Kenyan
hospital. Lancet, 340, 524-528.
Lackritz, E.M., Hightower, A.W., Zucker, J.R., Ruebush, T.K., 2nd, Onudi, C.O., Steketee, R.W.,
Were, J.B., Patrick, E. & Campbell, C.C. (1997) Longitudinal evaluation of severely
anemic children in Kenya: the effect of transfusion on mortality and hematologic
recovery. Aids, 11, 1487-1494.
Marsh, K., Forster, D., Waruiru, C., Mwangi, I., Winstanley, M., Marsh, V., Newton, C., Winstanley,
P., Warn, P., Peshu, N. & et al. (1995) Indicators of life-threatening malaria in African
children. N Engl J Med, 332, 1399-1404.
Mota, R.E., Lara, A.M., Kunkwenzu, E.D. & Lalloo, D.G. (2009) Health seeking behavior after
fever onset in a malaria-endemic area of Malawi. Am J Trop Med Hyg, 81, 935-943.
Organisation, W.H. (2000) Management of the child with a serious infection or severe
malnutrition. p. 162.
Phiri, K.S., Calis, J.C., Faragher, B., Nkhoma, E., Ng'oma, K., Mangochi, B., Molyneux, M.E. & van
Hensbroek, M.B. (2008) Long term outcome of severe anaemia in Malawian children.
PLoS One, 3, e2903.
Slutsker, L., Taylor, T.E., Wirima, J.J. & Steketee, R.W. (1994) In-hospital morbidity and
mortality due to malaria-associated severe anaemia in two areas of Malawi with
different patterns of malaria infection. Trans R Soc Trop Med Hyg, 88, 548-551.
Zucker, J.R., Lackritz, E.M., Ruebush, T.K., 2nd, Hightower, A.W., Adungosi, J.E., Were, J.B.,
Metchock, B., Patrick, E. & Campbell, C.C. (1996) Childhood mortality during and after
hospitalization in western Kenya: effect of malaria treatment regimens. Am J Trop Med
Hyg, 55, 655-660.
P a g e | 105
Chapter 7
Intermittent preventive therapy for malaria with monthly

artemether–lumefantrine for the post-discharge management
of severe anaemia in children aged 4–59 months in southern
Malawi: a multicentre, randomised, placebo-controlled trial.
Kamija Phiri 1, 2, Michael Esan 1, 2, 3, Michael Boele van Hensbroek 3 , Carole

Khairallah 4 , Brian Faragher 4 and Feiko O. ter Kuile 5
1 Community Health Department, College of Medicine, University of Malawi

2 MLW Clinical Research Programme, College of Medicine, University of
Malawi
3 The Amsterdam Institute for Global Health and Development, Emma
Children’s Hospital AMC, University of Amsterdam, the Netherlands
4 Liverpool School of Tropical Medicine, Liverpool, UK.
5 Department of Infectious Diseases, Tropical Medicine and AIDS, Academic
Medical Center, University of Amsterdam, the Netherlands
Lancet Infectious Diseases 2012 Mar 12(3); 191-200.

P a g e | 106
Abstract
Background: Young children with severe malarial anaemia in Africa are at high risk of
readmittance to hospital or death within 6 months of discharge. We aimed to assess whether 3
months of chemoprevention with artemether– lumefantrine reduced this risk.
Methods: We did a randomised, placebo-controlled, multicentre trial in four hospitals in Malawi
testing the efficacy and safety of intermittent preventive therapy post-discharge (IPTpd) in
children aged 4–59 months admitted for severe malarial anaemia. All convalescent children who
had completed a blood transfusion received artemether–lumefantrine at discharge and were
randomly assigned by a computer-generated sequence to receive placebo or artemether–
lumefantrine at 1 month and 2 months after discharge, providing about 1 month and 3 months of
protection, respectively. Patients and study staff were masked throughout the study. The
primary endpoint was a composite of all-cause mortality or hospital readmittance because of all-
cause severe anaemia or severe malaria between 1 and 6 months after enrolment. This trial is
registered, number ISRCTN89727873.
Results: Of 1414 children enrolled, 708 were assigned to receive placebo and 706 the
intervention. By 6 months, 192 children (14%) had died or were readmitted with severe
malaria or severe anaemia. 1–6 months after randomisation, 109 primary events occurred in 85
children in the placebo group and 86 in 74 children in the intervention group (adjusted
protective efficacy [PE] 31%, 95% CI 5–50; absolute rate reduction 11·7 per 100 children years,
95% CI 1·8–18·9; p=0·024). The protective effect was greatest during the IPTpd period (1–3
months), when 58 primary events occurred in 49 children in the placebo group and 37 in 34
children in the intervention group (PE 41%, 10–62; p=0·01), but was not sustained after the
third month (4–6 months, PE 17%, –27 to 45; p=0·395). When episodes in the first month were
included—i.e., before the first dose of IPTpd, when both groups benefited from the post-
treatment prophylactic effect of artemether–lumefantrine provided at discharge—the overall
cumulative PE by 6 months was 26% (–2 to 46; p=0·06).
Interpretation: In areas with intense malaria transmission, chemoprevention with IPTpd
given to children with severe malarial anaemia might reduce rates of readmittance to hospital
for severe anaemia or malaria. Studies to confirm these findings and to investigate different
delivery mechanisms and cost-effectiveness are needed.
Funding: The Netherlands African Partnership for Capacity Development and Clinical
Interventions Against Poverty Related Diseases, the UBS-Optimus Foundation, and the Gates
Malaria Partnership.
P a g e | 107
Introduction
Severe anaemia is a major cause of hospital admissions in sub-Saharan Africa and contributes
substantially to paediatric mortality, particularly among young children in areas of high
malaria transmission. Previous studies in western Kenya and southern Malawi indicate that
not only are young children admitted to hospital for severe anaemia at high risk of in-hospital
mortality, but also an additional 10–16% of patients die or are readmitted in the first 3–6
months after discharge.1–6
In children who have recovered from severe malarial anaemia, the initial rise in haemoglobin
resulting from blood transfusions is likely to be negated by subsequent episodes of new or
recrudescent malaria infections after discharge. Recovery from malaria-associated anaemia is
slow—full haematological recovery takes at least 6 weeks and takes substantially longer in
patients with recrudescence or reinfection,7,8 with infections with other pathogens, or with
prolonged nutritional deficiencies.9 Although continued destruction of unparasitised erythrocytes
after radical clearance of parasitaemia is a contributing factor, persistent dyserythropoiesis and
bone-marrow suppression can persist for much longer.9 We postulated that interventions that
result in radical cure and prevention of subsequent malaria episodes could provide a time-
window during which the bone marrow of recently transfused children can recover, allowing
haemoglobin to be restored and reducing the risk of re-admittance to hospital because of severe
malaria or recurrence of severe anaemia. The standard treatment for severe malaria anaemia
in Malawi and many other countries in sub-Saharan Africa is a blood transfusion and
intravenous quinine, then artemether–lumefantrine when children can switch to oral
medication, but no policy exists to address the high risk of morbidity and mortality after
discharge. The lumefantrine component in artemether–lumefantrine can provide several
weeks of post-treatment prophylaxis.10
Intermittent preventive therapy is the administration of full treatment courses of long-acting
antimalarial drugs at pre-defined time intervals, irrespective of whether patients are known to
have malaria. Such therapy clears existing infections and provides prolonged periods of
prophylaxis against new infections after each treatment course.11 This strategy can prevent
clinical malaria and malaria-associated severe anaemia in pregnant women,12 infants,13 and
children,14 and could be cost-effective for the management of children with severe malarial
anaemia after discharge.15 We designed a randomised, double-blind, placebo-controlled trial to
test the efficacy and safety of intermittent preventive therapy post-discharge (IPTpd) in the
P a g e | 108
management of young children admitted for severe malarial anaemia requiring a blood
transfusion.
Methods
Participants
We recruited children aged 4–59 months from four hospitals in southern Malawi, the Queen
Elizabeth Central Hospital (Blantyre) and three hospitals within an hour drive of Blantyre:
Chikwawa and Thyolo District Hospitals, and Zomba Central Hospital. This area has moderate to
intense perennial malaria transmission. All children had been admitted with severe malarial
anaemia, had received a blood transfusion, and had completed the in-hospital course of
intravenous quinine. Convalescent children surviving this initial in-hospital phase were eligible for
inclusion if after transfusion they had haemoglobin concentrations of more than 5 g/dL, weighed
more than 5 kg, were able to switch to oral medication, and could sit unaided. Children with blood
loss due to trauma, haematological malignancy, a known bleeding disorder, or known sickle-cell
disease were excluded. Other exclusion criteria were known hypersensitivity to artemether–
lumefantrine, treatment with artemether–lumefantrine within the week before admission, non-
residency in the study area, previous participation in the study, participation in another clinical
trial, known need for medication prohibited during the intervention period, and surgery scheduled
during the study. Written, informed consent was obtained. The study was approved by the
research ethics committees of the College of Medicine (Malawi) and the Liverpool School of
Tropical Medicine (UK).
Procedures
Eligible children in both groups received six doses as part of the standard 3 day course of
artemether–lumefantrine (Novartis, 20 mg artemether, 120 mg lumefantrine) in hospital.
Children weighing less than 15 kg received one tablet and those weighing 15 kg or more
received two tablets, about once every 12 h for 3 days. They were then randomly assigned
during convalescence to receive either IPTpd with the same 3 day course of artemether–
lumefantrine or placebo (Lab-Allied, Nairobi, Kenya) at 1 month and 2 months after discharge
(figure 1).
The first daily doses of IPTpd or placebo at 1 month and 2 months after discharge were
provided in the community by study team members who visited each home in the morning for
3 days; the second daily dose was left with the parents or guardian to give in the evening.
P a g e | 109
Adherence was assessed the next morning. Children were followed up for 6 months by passive
case detection.
The observation time was divided into three periods: the first month after discharge before
IPTpd, the IPTpd period starting at 1 month after discharge when children received the first
course of IPT or placebo and ended at 3 months (1 month after the second IPTpd dose), and the
extended follow-up period (4–6 months) to assess whether any initial beneficial effect of IPTpd
was sustained beyond the intervention period. Parents and guardians were asked to take children
to a study clinic if they had fever or were unwell. At these visits clinical information was recorded
on standardised forms, the axillary temperature and haemoglobin concentrations measured, and
a malaria blood smear taken. Children with a positive malaria smear were treated with 5 days of
quinine during the first 3 months after discharge or with a standard 3 day course of artemether–
lumefantrine during the extended follow-up period. Children with severe disease were admitted
to hospital. Children with recurrent severe anaemia received a blood transfusion. Bacterial and
other infections were treated at the discretion of clinicians. All children were seen at 6 months
for assessment of haemoglobin concentrations and malaria parasitaemia.
Haemoglobin was measured at point-of-care with the HemoCue B-haemoglobin analyser. Malaria
infection was defined as the presence of asexual Plasmodium falciparum in a blood smear.
HIV testing was done according to WHO guidelines with two rapid tests (Determine and Uni-
Gold). Discordant results and reactive results in children younger than 18 months were resolved
by PCR.
Figure 1: Study treatment and follow-up.

A cross-sectional survey was done 6 months after randomization. AL=artemether-lumefantrine. IPTpd=intermittent
preventive therapy post-discharge.
P a g e | 110
The primary outcome was a composite of all-cause mortality and hospital readmission because
of all-cause severe anaemia or severe malaria between 1 and 6 months. Severe anaemia was
defined as a haemoglobin concentration of less than 5 g/dL or a clinical indication for blood
transfusion. Severe malaria was defined as readmittance to hospital because of confirmed
malaria treated with parenteral quinine. Secondary endpoints were all-cause mortality,
hospital readmission because of all-cause severe anaemia or severe malaria, all-cause hospital
admission, all-cause sick-child clinic visits, and clinic visit because of microscopically
confirmed non-severe malaria.
Randomisation and masking
The trial statistician in Liverpool generated the random numbers by computer to allocate
children to groups. The randomisation sequence was stratified by hospital and weight group
(<15kg and 15 kg or more) in randomly varying block sizes of two, four, or six. An independent
statistician in Blantyre assigned the labels A and B to either the active drug or placebo and
oversaw the packaging and coding of sequentially numbered drug envelopes by a pharmacist.
When children meeting the enrolment criteria were able to switch to oral medication they were
allocated to one of the two study groups by the coordinating clinician in each hospital in order of
their study identification number by drawing successive envelopes from the box corresponding to
each weight stratum.
Statistical analysis
A composite endpoint, rather than all-cause mortality only, was used to decrease the sample
size needed for a given power by increasing the observed event rate. The individual
components were judged to have similar patho-physiological pathways with substantial
potential for overlap.16
The study was designed to detect a 40% reduction in the cumulative incidence of the primary
endpoint from 9·75% in the placebo group to 5·85% in the IPTpd group, with 80% power and a
two-sided significance threshold of 5%. The 40% reduction in primary endpoint was based on
the average protective efficacy (43%) from four completed studies of intermittent preventive
therapy in infants.17 The anticipated risk of 9·75% in the control group was chosen on the basis
of preliminary results of observational studies in Malawi, in which the researchers reported a
risk of about 16%.1 During these studies sulfadoxine–pyrimethamine was the treatment of
choice to complete intravenous quinine in hospitals. Malawi changed to artemether–
lumefantrine in 2006 because of sulfadoxine–pyrimethamine resistance. We anticipated that
P a g e | 111
recrudescent infection had contributed substantially to morbidity after discharge, therefore we

assumed that replacing sulfadoxine–pyrimethamine with artemether– lumefantrine would
reduce the risk of morbidity after discharge by 40%, from 16% to 9·75%. Assuming 10% loss to
follow-up, the study aimed to recruit 1650 participants (825 patients per group) or until 126
primary events had occurred.
Analysis was stratified by time period (figure 1). We anticipated that artemether–lumefantrine
would provide prophylaxis for about 1 month after each course and that children in the IPTpd
group would therefore be protected from malaria for the first 3 months after discharge, whereas
children in the placebo group, who only received artemether–lumefantrine at discharge, would
be protected for 1 month. Because both groups received artemether–lumefantrine at discharge,
no effect was expected in the period 0–1 month after discharge. The main treatment effect was
anticipated at 1–3 months. However, an extended follow-up period between 4 and 6 months was
included in the primary time period of interest, because any potential rebound effect resulting
from a delayed acquisition of protective immunity to malaria once the direct pharmacological
protective effect has waned, would decrease the public health significance of the intervention.
The 0–6 month period after discharge was also analysed because an IPTpd strategy is likely to be
considered for policy only if it substantially adds benefit to the effect of artemether–
lumefantrine at discharge. Furthermore, any intervention starting 1 month after discharge is
likely to be arranged with carers around the time of discharge.
Analysis was by intention to treat and done with SAS (version 9.2) and STATA (version 11). The
predictive analytics software missing values module was used to create five datasets in which any
missing values were imputed with the fully conditional specification, which is an iterative Markov
chain Monte Carlo method (ten iterations for each of the five imputations; done with PASW version
18). 13% of parents withheld consent to test their child for HIV; therefore, no HIV status was
imputed and a separate category for missing HIV status was used. The primary analyses included
all events.18 Endpoints with overlap (e.g., a child with severe malaria who also had severe
anaemia) were counted as a single event. We counted events as distinct if the child had been
discharged before the next event and a minimum of 3 days had passed since the diagnosis of the
earlier event. We calculated protective efficacy adjusted for prognostic factors at baseline,
calculated as 100 multiplied by (1–hazard ratio). Hazard ratios were calculated by Cox regression
for repeated events with robust standard error estimation methods to account for correlation
between episodes within children. All covariates with a p-value less than 0·2 in the univariate
P a g e | 112
Cox-regression models and other potential predictors identified in our previous observational
studies1 were entered into the initial multivariate model.
Figure 2: Trial profile

IPTpd=intermittent preventive therapy post-discharge. *Four patients diagnosed with sickle cell disease after
randomisation. †The original design included a third group (436 patients) of sulfadoxine–pyrimethamine at
discharge, then placebo at 1 and 2 months (according to the national policy in Malawi in 2006) and used mean
haemoglobin concentrations at 3 months as the primary endpoint. However, shortly after the trial started, the
national policy changed to artemether– lumefantrine, so the sulfadoxine–pyrimethamine group was discontinued.
By that time 40 patients had been recruited. Recruitment in the other two groups was continued with a new
randomisation list. The primary endpoint was changed to the composite endpoint by 6 months. Masking was
maintained throughout trial and the 13 patients in the sulfadoxine–pyrimethamine group were identified after
closure of all datasets and breaking of the study code.
P a g e | 113
These covariates were maternal education, presence of electricity in the household, number of
siblings in household, the child’s age, sex, HIV status, use of a bednet, history of previous hospital
admissions and blood transfusion, disease severity at admission, presence of malaria
parasitaemia and haemoglobin concentrations at randomisation (i.e., post-transfusion), and study
dose (mg/kg). The unadjusted absolute rate reduction was calculated as incidence in the placebo
group minus incidence in the IPTpd group. The adjusted absolute rate reduction was calculated
as the adjusted hazard ratio multiplied by incidence per person-year in the placebo group. We
also calculated unadjusted incidence rates per person-year. Poisson regression was used to
assess the effect of the intervention on anaemia and malaria at the end of the 6 months
observation period and results expressed as adjusted prevalence ratios.
Pre-specified subgroup analyses were done to test to what extent the magnitude of treatment
effects depended on age, HIV status, and use of insecticide-treated nets, and exploratory subgroup
analysis by history of previous hospital admissions, baseline haemoglobin concentration,
presence of malaria parasites at randomisation, season, and study site. Because the study was not
powered to test effect modification, we used the magnitude of the difference in treatment effect
between subgroups as well as the corresponding p value of the interaction terms to analyse
interactions. The trial is registered with Current Controlled Trials, number ISRCTN89727873.
Role of the funding source
The sponsors of the study had no role in the study design, data collection, data analysis, data
interpretation, or writing of the report. KP, FOtK, CK, and BF had full access to all the data in
the study. All authors reviewed the manuscript and had final responsibility for the decision to
submit for publication.
Results
Between June, 2006, and August, 2009 (when the required number of events was reached), 1414
eligible children aged 4–59 months were randomly assigned to receive placebo or intervention
(figure 2); 1310 children (93%) completed the 6 months follow-up, and 34 (2%) died during this
period. These percentages were much the same in the two groups (figure 2). 15 deaths (44%)
occurred within 1 month of discharge. Demographic characteristics were similar between groups,
but the occurrence of HIV, cerebral malaria, and some other markers of disease severity were
slightly higher in the IPTpd group than in the placebo group (table 1). Only six HIV-infected
patients (5%) were receiving co-trimoxazole prophylaxis.
P a g e | 114
Table 1: Demographic and health baseline characteristics

Data are n (%) or mean (SD); IPTpd=intermittent preventive therapy post-discharge
P a g e | 115
Table 2: Crude incidence per person-year and treatment effect by follow-up

Follow-up ranges are months since randomisation. PE computed by univariate Cox regression for repeated events
with the exception of the effect on mortality, which included time to first event only. Total number of children at the
start of each follow-up period and corresponding follow-up time (in person-days) for 0–1, 1–3, and 4–6 months:
placebo, 708 (21 057), 687 (40 490), 669 (64 122); IPTpd, 706 (21 213), 685 (40 396), 669 (64 352). PE=unadjusted
protective efficacy. ARR=unadjusted absolute rate reduction per 100 person-years. IPTpd=intermittent preventive
therapy post-discharge.
P a g e | 116
Overall, 192 (14%) of 1414 children had 235 primary events. Between 1 and 6 months after
randomisation more primary events had occurred in the placebo group than had in the
intervention group (table 2). Adjusted protective efficacy [PE] was 31% (95% CI 5–50); the
absolute rate reduction was 11·7 per 100 children years (95% CI 1·8–18·9; p=0·024; figure 3).
The protective effect was greatest during the IPTpd period (1–3 months; PE 41%, 95% CI 10–
62; p=0·01), but was not sustained after the third month (17%, –27 to 45; p=0–395; figure 3).
Inclusion of the pre-IPTpd period showed that IPTpd prevented about one in four primary
events up to 6 months (figure 3). The effect did not differ between age groups, by HIV status,
bednet use, baseline haemoglobin concentration, or the presence of active malaria infection at
enrolment (figure 4).
IPTpd reduced the number of all-cause hospital admissions (figure 3), a result of the reduced
admissions because of severe malaria and severe anaemia; mortality was not affected
significantly (figure 3). IPTpd also reduced the number of clinic visits resulting from non-severe
clinical malaria, which was most evident during the IPTpd period (figure 3). Pooling of
artemether–lumefantrine courses showed that episodes of clinical malaria became evident
from day 19 after the start of artemether–lumefantrine, with a substantial increase from day 23
onwards. At the final 6 months follow-up visit, treatment groups did not differ for the
prevalence of malaria parasitaemia (151 of 653 patients [23%; IPTpd] vs 162 of 657 patients
[25%; placebo]; prevalence ratio [PR] 0·96 [95% CI 0·79–1·16]; p=0·654), moderate anaemia
(haemoglobin concentration <8 g/dL; 79 of 653 patients [12%] vs 80 of 657 patients [12%]; PR
0·98 [0·73– 1·30]; p=0·870), and mean haemoglobin (10·6 g/dL vs 10·6 g/dL; adjusted mean
difference 0·12 g/dL [95% CI 0·09–0·39]; p=0·291). No drug-related serious adverse events
were reported.
Discussion
Compared with a standard single course of artemether–lumefantrine at discharge, which
provides a maximum of 1 month of post-treatment prophylaxis, provision of an additional 2
months of chemoprevention by two full treatment courses of artemether–lumefantrine at 1
month and 2 months after discharge prevented 40% of deaths or hospital admissions because of
recurrence of severe anaemia or severe malaria 1–3 months after discharge. IPTpd also halved
the number of clinic visits needed because of uncomplicated malaria.
P a g e | 117
Figure 3: Adjusted treatment effect by endpoint and time period.

PE=adjusted protective efficacy. ARR=adjusted absolute rate reduction per 100 children years.
*Protective efficacy (1–6 months) for severe anaemia was 25% (95% CI –13 to 50, p 0·165) and for severe
malaria 28% (–3 to 50, p=0·072).
P a g e | 118
Figure 4: Subgroup analyses for primary outcome (death or re-admittance to hospital for
severe malaria or severe anaemia between 1 and 6 months)
Interaction p-values calculated from interaction term Cox regression models. For HIV status, the p value of the
interaction term was calculated for the model that excluded children with missing HIV status. *Numbers in
parentheses are based on the incidence of clinical malaria recorded in the placebo group in each site, which
serves as a proxy indicator of malaria transmission intensity. †Malaria transmission season based on the
monthly incidence of clinical malaria in the placebo group.
P a g e | 119
The protective effect was not sustained after 3 months, when the direct pharmacodynamic effect
of the drugs had waned.19
No rebound effect occurred in the intervention group in the 4–6 months period, and the overall
cumulative protective efficacy up to 6 months, although smaller than that seen at 3 months, was
still in favour of IPTpd. Importantly, the beneficial effect of chemoprevention occurred in
addition to the effect of untreated or treated nets, consistent with the effect of seasonal IPT in
children14 and in addition to the initial effect obtained from the first course of artemether–
lumefantrine provided at discharge.
The protective effect was much the same irrespective of HIV infection, despite the potentially
different aetiology of anaemia, which in HIV-infected children tends to be more related to
persistent bone-marrow suppression associated with a chronic state of inflammation.20, 21 HIV-
infected and HIV-exposed children were included because antiretroviral therapy and co-
trimoxazole prophylaxis in this population were only scaled up to a national level at the end of
the recruitment period. Therefore, our data are insufficient to test whether co-trimoxazole
prophylaxis had a modifying effect. IPTpd might provide less benefit in children already
protected by co-trimoxazole than in those in our study because co-trimoxazole has anti-malarial
properties and is highly effective for prevention of malaria in HIV-infected adults, children, and
pregnant women including in areas with sulfadoxine–pyrimethamine resistance.22, 23
Intermittent preventive therapy for pregnant women and infants differs from seasonal
intermittent preventive therapy in children and IPTpd. The current regimens are two or three
courses given over 3–6 months for intermittent preventive therapy in pregnant women and three
courses given over 7 months for intermittent preventive therapy in infants, leaving individuals
unprotected between doses, which allows re-infection and the acquisition of protective immunity.
By contrast, seasonal intermittent preventive therapy for children and IPTpd aim to maintain
therapeutic drug concentrations for several months throughout the period of greatest malaria risk.
The prophylactic effect is a key component of all these strategies and requires the use of long-acting
drugs. Artemether–lumefantrine provided roughly 3 weeks of post-treatment prophylaxis, after
which episodes of clinical malaria became apparent. The reduced protection at the end of the month
could account for the 50% protective efficacy against clinical malaria during the IPTpd period, which
is more modest than the 80–90% reported for a combination of amodiaquine and sulfadoxine–
pyrimethamine when used for monthly intermittent preventive therapy in children in the same age
group in west Africa,14 where sulfadoxine–pyrimethamine is still highly effective.24 The slightly
P a g e | 120
shorter period of post-treatment prophylaxis with artemether–lumefantrine than with sulfadoxine–

pyrimethamine is also consistent with the results of comparative trials of amodiaquine–sulfadoxine–
pyrimethamine or dihydroartemisinin–piperaquine.10,25,26 We used artemether–lumefantrine
because it was the only long-acting drug produced according to good manufacturing standards that
was widely available in Malawi and other parts of sub-Saharan Africa, and because high-grade
resistance to sulfadoxine–pyrimethamine was widespread in Malawi in 2006. Mefloquine, another
long-acting anti-malarial, was not used because it is associated with high rates of vomiting in young
children.19, 27, 28 Dihydroartemisinin–piperaquine had not been assessed for any IPT strategy when
the study was designed, but trials of piperaquine have since shown the drug to be very effective for
intermittent preventive therapy in children, either as monotherapy or combined with dihydro-
artemisinin. 22, 23
Five previous trials, two including children with severe anaemia15, 31 and three including
children with mild anaemia, 32–34 provide some guidance on the generalisability of our findings
(panel). The three trials providing monthly IPT with sulfadoxine–pyrimethamine for mild
anaemia each showed that IPT roughly halved the frequency of clinical malaria, as in our trial,
but the effect on haematological recovery beyond the effect of a single course of sulfadoxine-
pyrimethamine or iron supplementation was small. Thus, IPTpd is likely to be particularly
beneficial in severely anaemic children and in settings where malaria is an important
contributing factor to severe morbidity after discharge. Although further confirmatory studies
are required, our results might also apply beyond areas with moderate to intense perennial
malaria transmission and include areas with highly seasonal malaria transmission, such as the
Sahel region in west Africa. Two trials in The Gambia (panel) showed that in children with
severe malarial anaemia, chemoprevention, as intermittent preventive therapy15 or
prophylaxis,31 targeted during the malaria transmission season also halved the rate of clinical
malaria and in one trial reduced all-cause hospital readmittance by 78%, and in the other trial
reduced recurrence of severe anaemia by 78%.
A potential limitation of our study is the focus on children who had severe anaemia and
malaria. Although most children admitted to hospital with severe anaemia had malaria, they
may have a higher risk of exposure to malaria after discharge than severely anaemic children
who do not have concomitant malaria. Thus, whether our findings can be generalised to the
latter group is unclear.
P a g e | 121
Overall, 14% of children had a

primary event, confirming the high
morbidity and mortality after
discharge previously reported.1
However, in our trial mortality in the
placebo group was much lower than
that reported for 2002–06 (2% vs 8%),
likely reflecting the switch from use of
sulfadoxine–pyrimethamine to
artemether–lumefantrine at discharge
and the scale-up of anti-retroviral
therapies for children with HIV at the
end of the study.
Another factor is the intensive care
provided to children in this study.
Therefore, our results might
underestimate the true potential effect
of IPTpd, because artemether–
lumefantrine provided prophylaxis for
just less than a month, and because of
the high degree of care provided to
participants.
The lower than anticipated mortality
and the absence of a consistent effect
on all-cause mortality also shows the
potential limitation of the use of
composite endpoints in clinical trials
when they include components of
unequal clinical importance (e.g.,
mortality vs. non-fatal hospital
admissions) and heterogeneity in
treatment effect between these
P a g e | 122
components is observed.16,35 Analysis of the individual components of our primary endpoint

showed that the beneficial effect was largely due to a reduction in hospital admissions because
of severe malaria and anaemia. We used the composite endpoints because we anticipated that
this would increase the power of the study by increasing the event rates. However, our
findings do not show that this occurred, which is consistent with a systematic review35 of the
role of endpoint selection, which reported that trials using composite endpoints are less likely
to generate positive results than are those that do not. A possible explanation proposed by the
authors is that a disproportionately high percentage (81%) of trials using a composite
endpoint include mortality as part of the endpoint, and the mortality endpoints in many trials
were associated with negative or neutral results.
Although IPTpd is a relatively simple intervention, its implementation requires appropriate
delivery mechanisms. By contrast with intermittent preventive therapies in infants and
pregnant women, which are delivered through the expanded programme on immunisation and
antenatal clinics, systems for the delivery of IPTpd would need to be established, as was done for
intermittent preventive therapy in children. Studies in rural areas in West Africa have shown that
delivery of intermittent preventive therapy to children through community health workers is
feasible and well accepted.36, 37 Several countries, including Malawi, have systems of community
health workers that could be used to deliver IPTpd or to arrange scheduled visits after discharge to
clinics or hospital outpatient departments. One advantage of IPTpd is that it is targeted at a very
high-risk group of hospitalised children who have already had contact with the health-care system,
providing a potential point of entry. The optimum delivery mechanism could vary with the
available health infrastructure and should be explored in future studies, ideally using mobile
telephone technology where appropriate.
IPTpd should be assessed in other settings with either regimens of artemether–lumefantrine
given every 3 weeks or less frequent regimens of other long-acting artemisinin-based
combination therapies. Data are also needed to compare different delivery mechanisms and to
test the cost-effectiveness of IPTpd. Although malaria was a major contributor, severe anaemia
is often caused by a combination of aetiological factors that result in persistent failure to
produce red blood cells. Interventions that target a broader range of aetiologies, including
bacterial infections (a major cause of severe anaemia in this group),11 micronutrient
deficiencies, and hookworm,5 might provide an even greater benefit that can be sustained for
longer.21 Our study clearly shows that a pro-active approach to the management of transfused
P a g e | 123
children after discharge could provide important public health benefits. Urgent attention should
be given to this neglected issue.
Contributors
KP, MBvH, and FOtK conceived the concept of intermittent preventive therapy post-discharge.
FOtK, KP, MBvH, and BF designed the study. KP coordinated the field work with support from
ME. CK did the analysis with statistical support from BF. KP, MBvH, BF, and FOtK interpreted
the data. FOtK and KP wrote the first draft of the manuscript; all authors reviewed and revised
the final version.
Conflicts of interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The study was co-funded by a grant from the Netherlands African Partnership for Capacity
Development and Clinical Interventions against Poverty Related Diseases, and part of the College
of Medicine Malawi Amsterdam Liverpool programme, a collaborative research and training
grant to the College of Medicine in Blantyre, the University of Amsterdam (Netherlands), and the
Liverpool School of Tropical Medicine (UK). Funding was also provided by the UBS Optimus
Foundation and by a re-entry grant to KP from the Gates Malaria Partnership, London School of
Hygiene & Tropical Medicine (UK), which received support from the Bill and Melinda Gates
Foundation. FOtK and CK thank the US Centers for Disease Control and Prevention for salary
support through a cooperative agreement between the Division of Parasitic Diseases and Malaria
(Centers for Disease Control and Prevention, USA) and the Malaria Epidemiology Unit of the
Child and Reproductive Health group, Liverpool School of Tropical Medicine held by FOtK. We
thank Novartis (Basel, Switzerland) for provision of the study drug and Lab-Allied (Nairobi,
Kenya) for provision of the placebo. We also thank the parents and guardians of the children
who participated in this study. We thank Sarah White from the College of Medicine who was the
independent statistician and held the study code and coordinated the preparation of the study
drugs and envelopes. We thank Charlotte Adamczick for her helpful contribution to the
coordination and clinical care of patients in the paediatric ward of Zomba Central Hospital. We
also thank Elizabeth Molyneux for being the independent safety monitor of the trial, and the
members of the data safety monitoring committee: Geoffrey Targett, Paul Milligan, and Enitan
Carrol for their valuable comments and independent review of the safety data, protocol, and
P a g e | 124
analytical plan. We would also thank Victor Mwapasa, Neil French, and Malcolm Molyneux who
were the independent chair and members of the trial steering committee, respectively. Lastly,
we thank Malcolm Molyneux and Robert Heyderman, the previous and current directors of the
Malawi–Liverpool Wellcome Clinical Research programme, and Peter Winstanley, the previous
director of the Wellcome Trust Tropical Centre, for hosting the study and for providing
invaluable logistic and administrative support.
References
1. Phiri KS, Calis JC, Faragher B, et al. Long term outcome of severe anaemia in Malawian
children. PLoS One 2008; 3: e2903.
2. Lackritz EM, Campbell CC, Ruebush TK 2nd, et al. Effect of blood transfusion on survival
among children in a Kenyan hospital. Lancet 1992; 340: 524–28.
3. Marsh K, Forster D, Waruiru C, et al. Indicators of life-threatening malaria in African
children. N Engl J Med 1995; 332: 1399–404.
4. Biemba G, Dolmans D, Thuma PE, Weiss G, Gordeuk VR. Severe anaemia in Zambian
children with Plasmodium falciparum malaria. Trop Med Int Health 2000; 5: 9–16.
5. Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J Med 2008;
358: 888–99.
6. Zucker JR, Lackritz EM, Ruebush TK 2nd, et al. Childhood mortality during and after
hospitalization in western Kenya: effect of malaria treatment regimens. Am J Trop Med
Hyg 1996; 55: 655–60.
7. Price RN, Simpson JA, Nosten F, et al. Factors contributing to anemia after uncomplicated
falciparum malaria. Am J Trop Med Hyg 2001; 65: 614–22.
8. Bloland PB, Lackritz EM, Kazembe PN, Were JB, Steketee R, Campbell CC. Beyond
chloroquine: implications of drug resistance for evaluating malaria therapy efficacy and
treatment policy in Africa. J Infect Dis 1993; 167: 932–37.
9. Looareesuwan S, Merry AH, Phillips RE, et al. Reduced erythrocyte survival following
clearance of malarial parasitaemia in Thai patients. Br J Haematol 1987; 67: 473–78.
10. Sinclair D, Zani B, Donegan S, Olliaro P, Garner P. Artemisinin-based combination therapy
for treating uncomplicated malaria. Cochrane Database Syst Rev 2009; 3: CD007483.
11. White NJ. Intermittent presumptive treatment for malaria. PLoS Med 2005; 2: e3.
12. ter Kuile FO, van Eijk AM, Filler SJ. Effect of sulfadoxine– pyrimethamine resistance on the
efficacy of intermittent preventive therapy for malaria control during pregnancy: a systematic
review. JAMA 2007; 297: 2603–16.
P a g e | 125
13. Aponte JJ, Schellenberg D, Egan A, et al. Efficacy and safety of intermittent preventive
treatment with sulfadoxine–pyrimethamine for malaria in African infants: a pooled analysis of
six randomised, placebo-controlled trials. Lancet 2009; 374: 1533–42.
14. Wilson AL. A systematic review and meta-analysis of the efficacy and safety of
intermittent preventive treatment of malaria in children (IPTc). PLoS One 2011; 6:
e16976.
15. Bojang KA, Milligan PJ, Conway DJ, et al. Prevention of the recurrence of anaemia in
Gambian children following discharge from hospital. PLoS One 2010; 5: e11227.
16. Ferreira-Gonzalez I, Permanyer-Miralda G, Busse JW, et al. Methodologic discussions for
using and interpreting composite endpoints are limited, but still identify major concerns.
J Clin Epidemiol 2007; 60: 651–57.
17. ter Kuile FO, Steketee RW. Intermittent preventive treatment in infants—adjusting
expectations and seeing opportunity. J Infect Dis 2006; 194: 269–72.
18. Cheung YB, Xu Y, Tan SH, Cutts F, Milligan P. Estimation of intervention effects using first or
multiple episodes in clinical trials: the Andersen-Gill model re-examined. Stat Med 2010; 29:
328–36.
19. Cairns M, Gosling R, Carneiro I, et al. Duration of protection against clinical malaria provided
by three regimens of intermittent preventive treatment in Tanzanian infants. PLoS One 2010;
5: e9467.
20. Calis JC, Phiri KS, Vet RJ, et al. Erythropoiesis in HIV-infected and uninfected Malawian
children with severe anemia. Aids 2010; 24: 2883–87.
anaemia in Malawian children. PLoS One 2010; 5: e12589.
22. Mermin J, Ekwaru JP, Liechty CA, et al. Effect of co-trimoxazole prophylaxis, antiretroviral
therapy, and insecticide-treated bednets on the frequency of malaria in HIV-1-infected
adults in Uganda: a prospective cohort study. Lancet 2006; 367: 1256–61.
23. Malamba S, Sandison T, Lule J, et al. Plasmodium falciparum dihydrofolate reductase and
dihyropteroate synthase mutations and the use of trimethoprim–sulfamethoxazole
prophylaxis among persons infected with human immunodeficiency virus. Am J Trop Med
Hyg 2010; 82: 766–71.
24. Naidoo I, Roper C. Drug resistance maps to guide intermittent preventive treatment of
malaria in African infants. Parasitology 2011; 138: 1469–79.
25. Zongo I, Dorsey G, Rouamba N, et al. Randomized comparison of amodiaquine plus
sulfadoxine–pyrimethamine, artemether– lumefantrine, and dihydroartemisinin–
piperaquine for the treatment of uncomplicated Plasmodium falciparum malaria in Burkina
Faso. Clin Infect Dis 2007; 45: 1453–61.
P a g e | 126
26. Zongo I, Dorsey G, Rouamba N, et al. Artemether–lumefantrine versus amodiaquine plus

sulfadoxine–pyrimethamine for uncomplicated falciparum malaria in Burkina Faso: a
randomised non-inferiority trial. Lancet 2007; 369: 491–98.
27. Gosling RD, Gesase S, Mosha JF, et al. Protective efficacy and safety of three antimalarial
regimens for intermittent preventive treatment for malaria in infants: a randomised,
double-blind, placebo-controlled trial. Lancet 2009; 374: 1521–32.
28. Slutsker LM, Khoromana CO, Payne D, et al. Mefloquine therapy for Plasmodium falciparum
malaria in children under 5 years of age in Malawi: in vivo/in vitro efficacy and correlation
of drug concentration with parasitological outcome. Bull World Health Organ 1990; 68: 53–
59.
29. Cisse B, Cairns M, Faye E, et al. Randomized trial of piperaquine with sulfadoxine–
pyrimethamine or dihydroartemisinin for malaria intermittent preventive treatment in
children. PLoS One 2009; 4: e7164.
30. Bojang K, Akor F, Bittaye O, et al. A randomised trial to compare the safety, tolerability and
efficacy of three drug combinations for intermittent preventive treatment in children. PLoS
One 2010; 5: e11225.
31. Bojang KA, Palmer A, Boele van Hensbroek M, Banya WA, Greenwood BM. Management of
severe malarial anaemia in Gambian children. Trans R Soc Trop Med Hyg 1997; 91: 557–61.
32. Desai MR, Mei JV, Kariuki SK, et al. Randomized, controlled trial of daily iron
supplementation and intermittent sulfadoxine– pyrimethamine for the treatment of mild
childhood anemia in western Kenya. J Infect Dis 2003; 187: 658–66.
33. Tomashek KM, Woodruff BA, Gotway CA, Bloland P, Mbaruku G. Randomized intervention
study comparing several regimens for the treatment of moderate anemia among refugee
children in Kigoma Region, Tanzania. Am J Trop Med Hyg 2001; 64: 164–71.
34. Verhoef H, West CE, Nzyuko SM, et al. Intermittent administration of iron and sulfadoxine-
pyrimethamine to control anaemia in Kenyan children: a randomised controlled trial.
Lancet 2002; 360: 908–14.
35. Hochman M, McCormick D. Endpoint selection and relative (versus absolute) risk reporting
in published medication trials. J Gen Intern Med 2011; 26: 1246–52.
36. Kweku M, Webster J, Adjuik M, Abudey S, Greenwood B, Chandramohan D. Options for
the delivery of intermittent preventive treatment for malaria to children: a community
randomised trial. PLoS One 2009; 4: e7256.
37. Bojang KA, Akor F, Conteh L, et al. Two strategies for the delivery of IPTc in an area of
seasonal malaria transmission in the Gambia: a randomised controlled trial. PLoS Med 2011;
8: e1000409.
P a g e | 127
Chapter 8
Discussion and Conclusions

P a g e | 128
Given the magnitude of the public health problem posed by anaemia in most resouce-limited
settings where it’s prevalence is over 40%,1 innovative approaches to controlling the epidemic
are needed. Iron deficiency is estimated to account for up to 50% of cases of anaemia in
pregnant women and children, thus understandably the prevention of iron deficiency anaemia
in populations at risk is the focus of many public health interventions.2 Although the
contribution of iron deficiency to the development of anaemia is important, the role of other
important aetiological factors needs to be considered, particularly in resource-limited settings
where the prevalence of infections (including malaria) is high. An integrated approach may be
required, where aetiological factors of anaemia are identified for each unique setting and
incorporated into anaemia control strategies. Such interventions need to be feasible, affordable
and sustainable if they are to have the desired impact.
Each chapter in this thesis examines anaemia at different stages of its development, to
determine which interventions are likely to have the greatest impact in resource-limited
settings, and which ones should potentially be avoided. As anaemia contributes significantly to
child mortality, successful interventions will further help towards achieving the WHO
millenium development goal (MDG) number 4: a two-thirds reduction in child mortality by
2015.3
i. Iron deficiency in children with HIV-associated anaemia
In chapter 2, the objective of the review was to determine the prevalence of iron deficiency in
HIV-infected children, and compare it with that of HIV uninfected children. Although the data
available was limited, we demostrated from relevant studies that the prevalence of iron
deficiency in HIV-infected children was 34% and that it was equally prevalent in HIV-infected
children from high and low-income settings (31% vs 36% respectively; p=0.14). Studies that
included a HIV-uninfected control population (n=4) were only available from low-income
settings and showed less iron deficiency in HIV-infected children (28%) compared with HIV-
uninfected controls (43%); p=0.03.
The findings further question the role of iron deficiency in the aetiology of HIV-associated
anaemia, with the results suggesting that it is less common in HIV-infected children than in
uninfected children. Previous studies have suggested that red cell production failure is the
most important aetiological mechanism, with HIV-infected children demonstrating fewer
P a g e | 129
erythroid progenitor cells in the bone marrow than in uninfected children.4, 5 However further
mechanistic studies are needed if intervention strategies for the management of anaemia in
this population are to be succcessful. Such interventions are required in HIV-infected children,
as the successful management of anaemia has been associated with improved survival
outcomes and quality of life in HIV-infected adults.6, 7 However, it is unknown if iron
supplementation, the recommended intervention for treating anaemia in children living in
resource-limited settings,8 is either beneficial or harmful in HIV-infected children.9 Given the
increased risk of infections in this population due to immunosuppression, studies evaluating
the safety and benefit of iron supplementation are needed to inform evidence-based guidelines
for the management of anaemia in HIV-infected children.
ii. Iron supplementation in HIV-infected children: benefits and risks
In chapter 3, the main objective of the study was to determine whether iron supplementation
in HIV-infected children is associated with an increased risk of morbidity, with the primary
end-point being the incidence of all-cause sick child clinic visits over 6 months follow-up. Other
specific endpoints that were assessed included the effect of iron on haemoglobin, CD4
percentages, and morbidity such as HIV-disease progression, all-cause hospital admissions,
respiratory infections and malaria.
We found that 3-months of iron supplementation was associated with greater increases in
haemoglobin concentrations and reduced the risk of anaemia at 6 months follow-up. Children
who received iron had a better CD4 percentage response at 3-months, but an increased
incidence of clinical malaria by 6 months, especially during the first 3 months.
This study is the first randomized, controlled trial examining the effects of iron
supplementation in HIV-infected children living in a malaria-endemic setting, thus the findings
have important implications for the management of anaemia in resource-limited settings with
a high prevalence of infections. In the conduct of this study, we tried to mirror the routine
practice in Malawi and many other countries in Sub-Saharan Africa, of regularly prescribing
iron supplements for the treatment of anaemia without prior determination of iron status to
improve generalisability. Though we were underpowered to detect a difference between the
intervention arms in our primary outcome (incidence of all-cause sick-child clinic visits), the
effect of iron on malaria was very apparent and statistically significant. Iron supplementation
P a g e | 130
in HIV-infected children with moderate anaemia significantly improved haemoglobin levels

and reduced the prevalence of anaemia, thus is an effective intervention for the management
of anaemia in this setting. However, the higher incidence of clinical malaria infections observed
in children who received iron confirmed what is already known from iron supplementation
studies performed in otherwise healthy children from malaria-endemic settings,10-13 which is
that the use of iron supplements is associated with a significant risk of malaria-associated
morbidity. Thus when considering the use of iron in HIV-infected children with anaemia, the
beneficial effects of iron supplementation on haemoglobin, anaemia prevalence and T-cell
immunity observed in this study will have to be weighed carefully against the increased risk of
infections, particularly of malaria. One thing is certain: iron supplementation must be given
with adequate protection from malaria when used in malaria-endemic settings.
In chapter 4, we set out to determine the iron status of HIV-infected children and the
responses of serological iron markers to iron supplementation.
We found that the baseline prevalence of iron deficiency in our HIV-infected cohort ranged
from 8.1-48.3% when different serological iron markers were applied. Iron supplementation
was associated with significant reductions in soluble transferrin receptor (sTfR) levels and
sTfR-log ferritin (sTfR-F) index at 3 months, and in sTfR-F index at 6 months. Children who
received iron were less likely to be iron-deficient at 3 months when assessed using serum
ferritin with C-reactive protein, sTfR and mean corpuscular volume (MCV) but these effects
were not sustained at 6 months.
The findings confirm that iron deficiency is prevalent in HIV-infected children when assessed
using different serological iron markers, in line with the findings of previous observational
studies from Sub-Saharan Africa.14-16 The findings are even more significant, given that
previously very few studies in African HIV-infected children have either taken into account the
effect of inflammation when reporting iron deficiency,16 or presented bone marrow findings.4,
17 The improvements seen in iron status following 3 months of iron supplementation in our
study suggests that iron supplementation is effective in the treatment of iron deficiency
anaemia in HIV-infected children. The added benefit of iron supplementation on iron status
further strengthens the case for the use of iron supplements in settings where anaemia, HIV
and malaria are highly prevalent. However, as was highlighted in chapter 3, iron
supplementation in this population must be provided with adequate protection from malaria.
P a g e | 131
iii. Development and evaluation of a new paediatric blood transfusion protocol for
Africa
In chapter 5, the objective of the study was to develop a paediatric transfusion protocol based
on the WHO transfusion guidelines for use in resource-limited settings, and evaluate its
usability in a busy African-hospital setting.
Based on the new protocol, severely anaemic children were classified into 3 groups:
complicated anaemia, uncomplicated anaemia and anaemia with severe malnutrition using
simple bedside clinical criteria. We found that all children were correctly assigned to
transfusion groups and that the correct volume of blood was given in 89.3%; the correct
duration was used in 86.2% and the correct rate of transfusion in 78.6% of participants. When
transfusion practices by clinicians using the old hospital transfusion guidelines were compared
with the new protocol, we found that 29% of all transfusions done could have been avoided.
The findings of this study are in agreement with the findings of other studies from Sub-
Saharan Africa, where the use of simple protocols reduced the number of blood transfusions
without increasing mortality18, 19 and by so doing avoid its associated risks.20-25 The protocol
was easy to use in a busy African hospital setting, with high protocol adherence rates by
clinicians and little additional equipment. In addition, it allowed clinicians to quickly identify
children at greatest risk of mortality from severe anaemia, and in the most urgent need of a
blood transfusion.
In the Malawi context, the potential impact of this intervention is huge, with 22% of the
country’s population of 13 million people aged 5 years or younger, and the prevalence of
severe anaemia (proportion of people that may require a blood transfusion) in this age group
estimated at 4%;26, 27 over 30,000 blood transfusions could be avoided annually during the
rainy season, when most malaria infections occur. This would take a significant strain off the
local blood transfusion service (MBTS), which like in other malaria-endemic settings struggles
to keep up with the high demand for blood during the peak malaria transmission season.28, 29
In chapter 6, the objective of the study was to evaluate the current WHO guidelines for blood
transfusion, by looking at the clinical and early haematological responses of severely anaemic
children to a standard blood transfusion regimen.
We found that out of 128 children recruited for the study, 30 children (23.4%) had a post-
transfusion Hb of 6 g/dl or less, and thus were classified as transfusion failures. In the
multivariate analysis it was observed that a high pre-transfusion Hb and a downward change
P a g e | 132
in respiratory rate (delta RR) were significantly associated with a reduced risk of transfusion
failure after adjusting for the presence of malaria parasites, age, weight, gender, glucose, type
of blood given (packed cells (PC) or whole blood (WB)), donor Hb and change in heart rate
(delta HR). In addition, children without malaria infection had better early Hb responses to
blood transfusion (at 24 hours) when compared with malaria-infected children, but
transfusion failures were not predicted by malaria, nor was there a difference in the clinical
responses to transfusion between children with and without malaria.
The high transfusion failure rate observed in this study is similar to that reported previously in
other studies from Sub-Saharan Africa30, 31 and raises concerns about whether the WHO
transfusion guidelines32 adequately treat severe anaemia, especially in malaria-endemic
settings. Though blood transfusion has been shown to improve survival outcomes in
symptomatic severely anaemic children where the risk of mortality is high,33, 34 the presence of
malaria parasitaemia during follow-up has been found to negate the benefit of blood
transfusion on haematologic recovery31, 35 and is an important determinant of treatment
outcome in malaria-endemic settings.36, 37 In this study, we found that the presence of malaria
parasitaemia was associated with a smaller haemoglobin response to transfusion after 24
hours. This was most likely due to on-going destruction and clearance of parasitized and non-
parasitized red blood cells by the spleen, and impaired reticulocytosis.38 In Malawi, post-
transfusion Hb or post-treatment malaria checks following treatment for severe anaemia are
not routinely done. Children are discharged home mostly based on clinical response to
treatment. These children who are clinically improved, but still moderately (or severely)
anaemic, are discharged home, and are at a high risk of a repeat severe anaemia episode
following a repeat or recrudescent malaria infection.
Perhaps incorporating a post-transfusion Hb check (as was done in this study) and a post-
treatment malaria test (either malaria microscopy or a rapid diagnostic test) into management
protocols after completing hospital treatment for severe anaemia may help identify children at
greatest risk of a repeat severe anaemia episode, and thus may help reduce the high associated
post-discharge mortality.36 39
P a g e | 133
iv. Intermittent preventive therapy with monthly artemether-lumefantrine for the post-
discharge management of severe anaemia in children aged 4-59 months in Southern
Malawi
In chapter 7, the objective of the study was to determine in pre-school children the effect of
intermittent preventive therapy post-discharge with monthly artemether-lumefantrine on all-
cause mortality and hospital re-admission for severe anaemia or severe malaria between 1 and
6 months.
We found that by 6 months, 14% of 1,414 study participants had either died or were re-
admitted to hospital with severe malaria or severe anaemia (primary endpoint). By 1–6 months
after randomization, intermittent preventive therapy post-discharge (IPTpd) with artemether-
lumefantrine was found to be protective against death or re-admission with severe malaria or
severe anaemia when compared with placebo (adjusted protective efficacy 31%; absolute rate
reduction 11.7 per 100 children years). The protective effect was greatest during the IPTpd
period (1–3 months; 41%), but was not sustained after the third month. When events that
occurred in the first month were included (before the first dose of IPTpd), when both groups
benefited from the post-treatment prophylactic effect of artemether–lumefantrine provided at
discharge—the overall cumulative protective efficacy by 6 months was 26%.
This trial was designed to test a novel approach to the challenge posed by the high post-
discharge morbidity and mortality observed following a severe malaria anaemia episode in
malaria-endemic settings.31, 36, 39 Intermittent preventive treatment with anti-malarials as an
intervention for malaria-associated anaemia has previously been used in children (IPTc)40, 41
and in infants (IPTi),42-44 and both strategies have been found to be very effective in reducing
malaria-associated morbidity and mortality. In IPTc, treatment doses of anti-malarials are
given to children in regions with seasonal malaria transmission at regular intervals usually for
3-4 months, regardless of the presence of malaria infection. However IPTpd differs from IPTc
in that IPTc is recommended for children living in regions with short, seasonal malaria
transmission, not for regions with intense, year-round malaria transmission where severe
malaria episodes tend to be more common. We found IPTpd to be effective against malaria-
associated morbidity and mortality in Malawi, a region with moderate to intense perennial
malaria transmission. In addition, IPTc advocates universal coverage of all children during the
peak malaria transmission season, which may raise concerns about the development of drug
P a g e | 134
resistance and impairment of naturally acquired immunity to malaria. IPTpd offers a more
targeted approach where only children less than 5 years old who are at the greatest risk of
malaria-associated morbidity and mortality receive the intervention, having already been in
contact with the health system following hospitalization for a severe malaria anaemia episode.
Intermittent preventive therapy post-discharge by giving two additional treatment courses of

artemether-lumefantrine at one-monthly intervals prevented 40% of deaths or hospital re-
admissions because of severe malaria or severe anaemia at 3 months, thus implementation of
this strategy could potentially have a huge public health impact on under-five mortality
(MDG4) during the peak malaria transmission season. However, the safety and efficacy of this
strategy using other artemisinin combination therapies (ACTs) will need to be assessed, to see
if they can offer similar or better protection. In addition, the feasibility of implementation of
this strategy would need to be determined, as potential mechanisms for delivery of the
intervention following discharge from hospital and its associated cost implications will have to
be assessed in the framework of existing health systems.
Conclusions
Iron deficiency is prevalent in HIV-infected children and plays a role in the multi-factorial
aetiology of anaemia in this population. The use of iron supplements in HIV-infected children
with anaemia has beneficial effects on haemoglobin, T-cell immunity and iron status, but
increases the risk of clinical malaria infections. Thus iron supplementation in this population,
as in other paediatric populations in malaria-endemic settings, has to be given with adequate
protection from malaria.
The use of simple, yet detailed protocols for blood transfusion is feasible in busy African
hospital settings and can significantly reduce the number of blood transfusions performed
during the malaria transmission season, with minimal additional equipment. This can take a
considerable strain off existing national blood transfusion services. The provision of post-
transfusion Hb checks and a malaria test following in-hospital management of severe anaemia
may help in identifying children at greatest risk of a repeat severe anaemia episode following a
repeat or recrudescent malaria infection, and in this way may contribute towards reducing
post-discharge mortality.
P a g e | 135
Intermittent preventive therapy post-discharge with monthly artemether-lumefantrine for 3

months following a severe malaria anaemia episode is an effective strategy for reducing post-
discharge morbidity and mortality, and could potentially have a huge impact on reduction of
under-five child mortality (MDG4). Studies are needed to assess the safety and efficacy of other
artemisinin-combination therapy regimens in this setting, and the feasibility of
implementation of this strategy in the framework of existing national health systems.
P a g e | 136
References
1. WHO. Worldwide prevalence of anaemia 1993 - 2005: WHO Global Database on Anaemia. 2008.
2. WHO. Iron Deficiency Anaemia Assessment, prevention and control: A guide for programme
managers. 2001.
3. WHO. World Health Statistics 2010. Part 1:Health-related Millenium Development Goals. 2011.
4. Calis JC, Phiri KS, Vet RJ, et al. Erythropoiesis in HIV-infected and uninfected Malawian children
with severe anemia. Aids 2010;24(18):2883-7.
anaemia in Malawian children. PLoS One 2010;5(9):e12589.
6. Saah AJ, Hoover DR, He Y, Kingsley LA, Phair JP. Factors influencing survival after AIDS: report
from the Multicenter AIDS Cohort Study (MACS). J Acquir Immune Defic Syndr 1994;7(3):287-95.
7. Sullivan PS, Hanson DL, Chu SY, Jones JL, Ward JW. Epidemiology of anemia in human
immunodeficiency virus (HIV)-infected persons: results from the multistate adult and adolescent
spectrum of HIV disease surveillance project. Blood 1998;91(1):301-8.
8. WHO. Management of the child with a serious infection or malnutrition: Guidelines for care at
the first referral level in developing countries. 2000.
9. Adetifa I, Okomo U. Iron supplementation for reducing morbidity and mortality in children with
HIV. Cochrane Database Syst Rev 2009(1):CD006736.
10. Sazawal S, Black RE, Ramsan M, et al. Effects of routine prophylactic supplementation with iron
and folic acid on admission to hospital and mortality in preschool children in a high malaria
transmission setting: community-based, randomised, placebo-controlled trial. Lancet
2006;367(9505):133-43.
11. Oppenheimer SJ, Gibson FD, Macfarlane SB, et al. Iron supplementation increases prevalence
and effects of malaria: report on clinical studies in Papua New Guinea. Trans R Soc Trop Med Hyg
1986;80(4):603-12.
12. Smith AW, Hendrickse RG, Harrison C, Hayes RJ, Greenwood BM. The effects on malaria of
treatment of iron-deficiency anaemia with oral iron in Gambian children. Ann Trop Paediatr
1989;9(1):17-23.
13. van den Hombergh J, Dalderop E, Smit Y. Does iron therapy benefit children with severe
malaria-associated anaemia? A clinical trial with 12 weeks supplementation of oral iron in young
children from the Turiani Division, Tanzania. J Trop Pediatr 1996;42(4):220-7.
14. Eley BS, Sive AA, Shuttleworth M, Hussey GD. A prospective, cross-sectional study of anaemia
and peripheral iron status in antiretroviral naive, HIV-1 infected children in Cape Town, South Africa.
BMC Infect Dis 2002;2:3.
15. Totin D, Ndugwa C, Mmiro F, Perry RT, Jackson JB, Semba RD. Iron deficiency anemia is highly
prevalent among human immunodeficiency virus-infected and uninfected infants in Uganda. J Nutr
2002;132(3):423-9.
16. Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J Med
2008;358(9):888-99.
17. Phiri KS. Assessing iron deficiency in Malawian children living in an area of high malaria and
bacterial infection morbidity. University of Liverpool 2006.
18. Craighead IB, Knowles JK. Prevention of transfusion-associated HIV transmission with the use
of a transfusion protocol for under 5s. Trop Doct 1993;23(2):59-61.
P a g e | 137
19. Vos J, Gumodoka B, Ng'weshemi JZ, Kigadye FC, Dolmans WM, Borgdorff MW. Are some blood
transfusions avoidable? A hospital record analysis in Mwanza Region, Tanzania. Trop Geogr Med
1993;45(6):301-3.
20. Lackritz EM, Ruebush TK, 2nd, Zucker JR, Adungosi JE, Were JB, Campbell CC. Blood transfusion
practices and blood-banking services in a Kenyan hospital. Aids 1993;7(7):995-9.
21. Blaser MJ, Cohn DL. Opportunistic infections in patients with AIDS: clues to the epidemiology of
AIDS and the relative virulence of pathogens. Rev Infect Dis 1986;8(1):21-30.
22. Greenberg AE, Nguyen-Dinh P, Mann JM, et al. The association between malaria, blood
transfusions, and HIV seropositivity in a pediatric population in Kinshasa, Zaire. Jama
1988;259(4):545-9.
23. N'Tita I, Mulanga K, Dulat C, et al. Risk of transfusion-associated HIV transmission in Kinshasa,
Zaire. Aids 1991;5(4):437-9.
24. Hassall O, Maitland K, Pole L, et al. Bacterial contamination of pediatric whole blood
transfusions in a Kenyan hospital. Transfusion 2009;49(12):2594-8.
25. Nebie K, Ouattara S, Sanou M, et al. Poor procedures and quality control among nonaffiliated
blood centers in Burkina Faso: an argument for expanding the reach of the national blood transfusion
center. Transfusion;51(7 Pt 2):1613-8.
26. N.S.O. 2008 Population and Housing Census Main Report. National Statistics Office, Malawi.
2008.
27. N.S.O. Malawi Demographic and Health Survey 2010. Zomba, Malawi and Calverton, Maryland,
USA: NSO and ICF Macro. 2011.
28. Obonyo CO, Steyerberg EW, Oloo AJ, Habbema JD. Blood transfusions for severe malaria-related
anemia in Africa: a decision analysis. Am J Trop Med Hyg 1998;59(5):808-12.
29. Ryder RW. Difficulties associated with providing an HIV-free blood supply in tropical Africa.
Aids 1992;6(11):1395-7.
30. English M, Ahmed M, Ngando C, Berkley J, Ross A. Blood transfusion for severe anaemia in
children in a Kenyan hospital. Lancet 2002;359(9305):494-5.
31. Akech SO, Hassall O, Pamba A, et al. Survival and haematological recovery of children with
severe malaria transfused in accordance to WHO guidelines in Kilifi, Kenya. Malar J 2008;7:256.
32. WHO. The clinical use of blood in medicine, obstetrics, paediatrics, surgery and anaesthesia,
trauma and burns. 2001:337.
33. Lackritz EM, Campbell CC, Ruebush TK, 2nd, et al. Effect of blood transfusion on survival among
children in a Kenyan hospital. Lancet 1992;340(8818):524-8.
34. Brewster DR. Blood transfusions for severe anaemia in African children. Lancet
1992;340(8824):917.
35. Lackritz EM, Hightower AW, Zucker JR, et al. Longitudinal evaluation of severely anemic
children in Kenya: the effect of transfusion on mortality and hematologic recovery. Aids
1997;11(12):1487-94.
36. Zucker JR, Lackritz EM, Ruebush TK, 2nd, et al. Childhood mortality during and after
hospitalization in western Kenya: effect of malaria treatment regimens. Am J Trop Med Hyg
1996;55(6):655-60.
37. Bojang KA, Palmer A, Boele van Hensbroek M, Banya WA, Greenwood BM. Management of
severe malarial anaemia in Gambian children. Trans R Soc Trop Med Hyg 1997;91(5):557-61.
38. Maitland K, Marsh K. Pathophysiology of severe malaria in children. Acta Trop 2004;90(2):131-
40.
P a g e | 138
39. Phiri KS, Calis JC, Faragher B, et al. Long term outcome of severe anaemia in Malawian children.
PLoS One 2008;3(8):e2903.
40. Cisse B, Sokhna C, Boulanger D, et al. Seasonal intermittent preventive treatment with
artesunate and sulfadoxine-pyrimethamine for prevention of malaria in Senegalese children: a
randomised, placebo-controlled, double-blind trial. Lancet 2006;367(9511):659-67.
41. Dicko A, Sagara I, Sissoko MS, et al. Impact of intermittent preventive treatment with
sulphadoxine-pyrimethamine targeting the transmission season on the incidence of clinical malaria in
children in Mali. Malar J 2008;7:123.
42. Gosling RD, Gesase S, Mosha JF, et al. Protective efficacy and safety of three antimalarial
regimens for intermittent preventive treatment for malaria in infants: a randomised, double-blind,
placebo-controlled trial. Lancet 2009;374(9700):1521-32.
43. Macete E, Aide P, Aponte JJ, et al. Intermittent preventive treatment for malaria control
administered at the time of routine vaccinations in Mozambican infants: a randomized, placebo-
controlled trial. J Infect Dis 2006;194(3):276-85.
44. Schellenberg D, Menendez C, Kahigwa E, et al. Intermittent treatment for malaria and anaemia
control at time of routine vaccinations in Tanzanian infants: a randomised, placebo-controlled trial.
Lancet 2001;357(9267):1471-7.
45. Konate AT, Yaro JB, Ouedraogo AZ, et al. Intermittent preventive treatment of malaria provides
substantial protection against malaria in children already protected by an insecticide-treated bednet in
Burkina Faso: a randomised, double-blind, placebo-controlled trial. PLoS Med;8(2):e1000408.
46. Dicko A, Diallo AI, Tembine I, et al. Intermittent preventive treatment of malaria provides
substantial protection against malaria in children already protected by an insecticide-treated bednet in
Mali: a randomised, double-blind, placebo-controlled trial. PLoS Med;8(2):e1000407.
P a g e | 139
Summary
P a g e | 140
Anaemia is a severe public health problem in Africa, with its prevalence in children less than 5
years old greater than 40% in most countries on the continent, including Malawi. Anaemia can
be defined as a state in which there is an insufficient number of red blood cells to meet the
body’s physiologic demands. It can present in different ways, from having no symptoms to
pesenting with symptoms of difficulty breathing and heart failure. It can be classified as mild,
moderate or severe depending on the severity. Anaemia can result in a wide range of
untoward effects, from poor growth to impaired thinking and work capacity. This PhD thesis
examines interventions for anaemia at different stages of it’s development.
In chapter 1, the purpose of the chapter was to introduce the problem of anaemia, the scope of
the problem globally, the different interventions currently being recommended by the WHO in
resource-poor hospital settings and communities, to describe the study setting (Malawi), and
to briefly state the objectives of the different studies performed in this thesis.
In chapter 2, we reviewed the literature on iron deficiency in HIV-infected children with

anaemia. Although the data available was limited, we demostrated from relevant studies that
the prevalence of iron deficiency in HIV-infected children was 34% and that it was equally
prevalent in HIV-infected children from high and low-income settings (31% vs 36%
respectively; p=0.14). Studies that included a HIV-uninfected control population (n=4) were
only available from low-income settings and showed less iron deficiency in HIV-infected
children (28%) compared with HIV-uninfected controls (43%); p=0.03. The findings further
question the role of iron deficiency in the development of anaemia in HIV-infected children,
with the results suggesting that it is less common in HIV-infected children than in uninfected
children. Given the increased risk of infections in this population due to a depressed immune
system, and the known beneficial effects of the treatment of anaemia on survival and quality of
life in HIV-infected adults, studies evaluating the safety and benefit of iron supplementation
are needed to inform evidence-based guidelines for the management of anaemia in HIV-
infected children.
In chapter 3, the objective of the study was to determine whether iron supplementation in
HIV-infected children is associated with an increased risk of morbidity, with the primary end-
P a g e | 141
point being the incidence of all-cause sick child clinic visits over 6 months follow-up. Other
specific endpoints that were assessed included the effect of iron on haemoglobin, CD4
percentages, and morbidity such as HIV-disease progression, all-cause hospital admissions,
respiratory infections and malaria.
We found that 3-months of iron supplementation was associated with greater increases in
haemoglobin concentrations and reduced the risk of anaemia at 6 months follow-up. Children
who received iron had a better CD4 percentage response at 3-months, but an increased
incidence of clinical malaria by 6 months, especially during the first 3 months.
This study is the first randomized, controlled trial examining the effects of iron
supplementation in HIV-infected children living in a malaria-endemic setting, and
demonstrated that iron supplementation has beneficial effects on haemoglobin levels, iron
status and immune function but increases the risk of clinical malaria infections. We concluded
that iron supplementation of HIV-infected children in malaria-endemic settings should only be
given with adequate protection from malaria.
In chapter 4, the objective of the study was to determine the iron status of HIV-infected
children and the responses of serological iron markers to iron supplementation.
We found that the baseline prevalence of iron deficiency in our HIV-infected cohort ranged
from 8.1-48.3% when different serological iron markers were applied. Iron supplementation
was associated with significant reductions (improvements) in soluble transferrin receptor
(sTfR) levels and sTfR-log ferritin (sTfR-F) index at 3 months, and in sTfR-F index at 6 months.
Children who received iron were less likely to be iron-deficient at 3 months when assessed
using serum ferritin with C-reactive protein, sTfR and mean corpuscular volume (MCV) but
these effects were not sustained at 6 months.
The study demonstrated that iron deficiency is prevalent in HIV-infected children with
anaemia and that iron supplementation is an effective treatment option for iron deficiency
anaemia in this population, even in a setting with a high prevalence of infections. However, as
was highlighted in chapter 3, iron supplementation in this population should only be provided
with adequate protection from malaria.
P a g e | 142
In chapter 5, the objective of the study was to develop a paediatric transfusion protocol based
on the WHO transfusion guidelines for use in resource-limited settings, and evaluate its
usability in a busy African-hospital setting.
Based on the new protocol, severely anaemic children were classified into 3 groups:
complicated anaemia, uncomplicated anaemia and anaemia with severe malnutrition using
simple bedside clinical criteria. We found that all children were correctly assigned to
transfusion groups and that the correct volume of blood was given in 89.3%; the correct
duration was used in 86.2% and the correct rate of transfusion in 78.6% of participants. When
transfusion practices by clinicians using the old hospital transfusion guidelines were compared
with the new protocol, we found that 29% of all transfusions done could have been avoided.
The study demonstrated that transfusion protocols can be implemented effectively in busy
African hospital settings with minimal additional equipment, and can significantly reduce the
number of blood transfusions given during the peak malaria transmission season. This can
take a significant strain off existing national blood transfusion services.
In chapter 6, the objective of the study was to evaluate the current WHO guidelines for blood
transfusion, by looking at the clinical and early haematological responses of severely anaemic
children to a standard blood transfusion regimen.
We found that out of 128 children recruited for the study, 30 children (23.4%) had a post-
transfusion haemoglobin of 6 g/dl or less, and thus were classified as transfusion failures. In
the multivariate analysis it was observed that a high pre-transfusion Hb and a downward
change in respiratory rate were significantly associated with a reduced risk of transfusion
failure after adjusting for several baseline factors. In addition, children without malaria
infection had better haemoglobin responses to blood transfusion (at 24 hours) when
compared with malaria-infected children.
The study suggests that performing post-transfusion Hb checks and a malaria test following in-
hospital management of severe anaemia may help in identifying children at greatest risk of a
repeat severe anaemia episode following a repeat or recrudescent malaria infection, and in this
way may contribute towards reducing post-discharge mortality in malaria-endemic settings.
In chapter 7, the objective of the study was to determine in pre-school children the effect of
intermittent preventive therapy post-discharge with monthly artemether-lumefantrine on all-
P a g e | 143
cause mortality and hospital re-admission for severe anaemia or severe malaria between 1 and
6 months.
We found that by 6 months, 14% of 1,414 study participants had either died or were re-
admitted to hospital with severe malaria or severe anaemia (primary endpoint). By 1–6 months
after randomization, intermittent preventive therapy post-discharge (IPTpd) with artemether-
lumefantrine was found to be protective against death or re-admission with severe malaria or
severe anaemia when compared with placebo (adjusted protective efficacy 31%). The
protective effect was greatest during the IPTpd period (1–3 months; 41%), but was not
sustained after the third month. When events that occurred in the first month were included
(before the first dose of IPTpd), the cumulative protective efficacy by 6 months was 26%.
The study demonstrated that intermittent preventive therapy post-discharge by giving two
additional treatment courses of artemether-lumefantrine at one-monthly intervals following a
severe malaria anaemia episode is an effective strategy for reducing post-discharge morbidity
and mortality in a setting of moderate to intense malaria transmission, and could potentially
have a huge impact on reduction of under-five child mortality (MDG4) during the peak malaria
season. However, studies are needed to assess the safety and efficacy of other artemisinin-
combination therapy regimens in the context of IPTpd, and the feasibility of implementation of
this strategy in the framework of existing national health systems.
In chapter 8, all the studies performed in this thesis were discussed, and general
recommendations/conclusions were made.
P a g e | 144
Samenvatting
P a g e | 145
Bloedarmoede (anemie) is een groot gezondheidsprobleem in Afrika. In de meeste Afrikaanse

landen is rond de 40% van de kinderen anemisch. Bij anemie is het aantal rode bloedcellen
onvoldoende waardoor normaal functioneren onmogelijk wordt. Dit kan zich uiten in moeite
met ademhalen, versnelde hartfrequentie, een verminderde inspanningstolerantie en, op
langere termijn, een verminderde groei en ontwikkeling van het kind. Vaak wordt anemie
onderverdeeld in ‘mild’, ‘matig’ of ‘ernstig’. In dit proefschrift worden resultaten gepresenteerd
van verschillende studies naar preventie en behandeling van anemie bij Malawiaanse
kinderen.
In Hoodstuk 1 wordt een introductie gegeven in het probleem van anemie, dat besproken
wordt in een globaal kader. Daarnaast worden de verschillende, door de Wereld Gezondheids
Organisatie (WHO) aangeraden, behandelingen ter preventie en behandeling van anemie
besproken. In dit hoofdstuk worden eveneens de Malawiaanse onderzoekslocaties beschreven
en komen de doelen van de verschillende onderzoeken van dit proefschrift aan de orde.
In Hoofdstuk 2 wordt de literatuur besproken die in het verleden gepubliceerd is op het

gebied van HIV-geïnfecteerde kinderen met bloedarmoede en een tekort aan ijzer. De
hoeveelheid beschikbare studies bleek zeer beperkt, maar toonde aan dat ijzertekort even
vaak voorkomt bij HIV-geïnfecteerde kinderen in ontwikkelingslanden als in dezelfde groep
kinderen in ontwikkelde landen (respectievelijk 31% en 36%, p=0.14). Vier studies,
uitgevoerd in ontwikkelingslanden, lieten zien dat ijzertekort vaker voorkomt bij anemische
kinderen die niet HIV geïnfecteerd zijn dan bij met HIV geïnfecteerde anemische kinderen
(43% tegen 28%; p=0.03). Dit roept twijfels op over het belang van ijzertekort als oorzaak van
anemie bij HIV geïnfecteerde kinderen. Dit is met name van belang aangezien er aanwijzingen
zijn voor een verhoogde gevoeligheid voor infecties tijdens ijzerbehandeling. Het is dus van
belang om goed inzicht te krijgen in het voorkomen (prevalentie) van ijzertekort en het risico
van ijzerbehandeling met name bij HIV geïnfecteerde kinderen die een verminderde afweer
tegen infecties hebben.
In hoofdstuk 3 worden de studieresultaten besproken van een onderzoek naar het optreden
van morbiditeit (o.a. malaria, luchtweginfecties) bij HIV geïnfecteerde anemische kinderen
P a g e | 146
tijdens en na behandeling met ijzer. In het betreffende onderzoek word ook gekeken of de
ijzerbehandeling het voorkomen van anemie verminderde, het hemoglobinegehalte in het
bloed verhoogde en invloed had op het immuunsysteem van het HIV geïnfecteerde kind. De
belangrijkste uitkomst van het onderzoek was dat HIV geïnfecteerde kinderen die ijzer kregen
een sterkere verbetering van hun hemoglobine hadden en dat minder kinderen na de
behandeling nog anemie hadden. Daarbij verbeterde het immuunsysteem van de kinderen die
ijzer kregen. Wel bleken deze kinderen meer kans te hebben op het krijgen van malaria, maar
weer minder kans op luchtweginfecties. Onze studie is de eerste studie die het effect van ijzer
behandeling op HIV geïnfecteerde kinderen met anemie in zoveel details heeft onderzocht. De
conclusie uit onze studie is dat ijzerbehandeling veel voordelen heeft maar wel moet worden
gecombineerd met een preventieve behandeling ter voorkoming van een malaria infectie.
In hoofdstuk 4 wordt het onderzoek besproken naar het effect van ijzerbehandeling op de aan
ijzer gerelateerde waarden in het bloed bij HIV geïnfecteerde anemische kinderen. Op basis
van de laboratorium waarden feritine en soluble transferrin receptor (sTfR) kon
geconcludeerd worden dat kinderen die ijzer hadden gekregen na 3 en 6 maanden, minder
kans hadden nog ijzerdeficient te zijn. Dit is opmerkelijk aangezien er aanwijzingen waren dat
bij kinderen die infecties bij zich dragen, ijzer mogelijk niet uit hun darmen wordt opgenomen.
Deze laboratorium studie bevestigde dat ook bij HIV geïnfecteerde kinderen met anemie, ijzer
supletie zinvol en veilig is, mits gegeven in combinatie met malaria profylaxe.
In hoofdstuk 5, worden de resultaten beschreven van een onderzoek naar de ontwikkeling

van betere richtlijnen voor het geven van een bloedtransfusie aan kinderen met ernstige
anemie in ontwikkelingslanden. In het nieuwe protocol worden kinderen ingedeeld in drie
groepen: ‘gecompliceerde anemie’, ‘ongecomplceerde anemie’ en ‘anemie gecombineerd met
ernstige ondervoeding’. De nieuw ontwikkelde, eenvoudige, richtlijn bleek goed bruikbaar
voor de medische staf van het ziekenhuis. Negenentachtig procent van de kinderen kreeg het
juiste volume en dit werd bij 86% in de juiste snelheid gegeven. Belangrijk was dat bij 29% van
de kinderen die volgens de oude richtlijnen een transfusie hadden gekregen, volgens de
nieuwe richtlijnen een transfusie niet nodig bleek. Deze bevinding is met name van belang
voor de drukke Afrikaanse ziekenhuizen met een overbelaste bloedbank door te veel
aanvragen voor bloed in combinatie met te weinig donoren.
P a g e | 147
In hoofdstuk 6 wordt de studie beschreven die als doel had de WHO bloedtransfusie
richtlijnen te evalueren, door te kijken naar het effect van de transfusie op de klinsche
kenmerken van de kinderen en de veranderingen in hun laboratoriumwaarden. Na evaluatie
van 128 kinderen bleek dat bij 30 kinderen (23.4%) het hemoglobine gehalte in het bloed
onvoldoende gestegen was na transfusie (nog steeds onder de 6.0 gram/dl). Vooral kinderen
met een lage start-hemoglobine waarde, een malaria infectie of een trage afname van een te
snelle ademhaling hadden een verhoogde kans om bij de transfusie-falers te behoren. De
studie laat het belang zien van goede malariadiagnostiek bij kinderen met ernstige anemie
evenals het belang van controle van hun hemoglobine gehalte nadat ze een bloedtransfusie
hebben ontvangen.
In hoofdstuk 7 worden de resultaten gepresenteerd van een studie naar het effect van
intermitterende malaria behandeling (‘intermittent preventive therapy post-discharge’) na een
episode van ernstige anemie op het voorkomen (preventie) van sterfte en een tweede episode
van ernstige malaria of anemie. We vonden dat 14% van de 1414 kinderen binnen 6 maanden,
waren gestorven of opnieuw waren opgenomen in het ziekenhuis. Intermiterende behandeling
met het antimalaria middel artemether-lumefantrine gaf een 31% bescherming tegen deze
complcaties dat vooral zichtbaar was in de eerste 3 maanden (de periode dat de middelen
werden ingenomen). De studie toonde aan dat twee aanvullende antimalaria kuren, na 1 en 2
maanden, een duidelijke bescherming boden tegen deze ernstige complicaties. Aanvullende
studies zijn echter nodig om deze bevindingen te bevestigen en om te onderzoeken of dit
beleid ook in de praktijkte te implementeren is.
In hoofdstuk 8, het laatste hoofstuk van dit proefschrift, worden alle resultaten samengevat
en algemene aanbevelingen gegeven voor de praktijk.
P a g e | 148
Acknowledgements
P a g e | 149
The work reported in this PhD thesis has taken five years to complete, and a lot of people have
contributed to the final product. I will try as much as possible to name everyone that was
involved in the process, but I apologize if I am unable to mention everyone in this section- you
all know who you are, and I sincerely thank you for all your help.
Firstly, I would like to thank the Almighty God, without who’s strength, knowledge and
wisdom I would not have had the courage to embark on this journey- I give Him all the glory. I
would also like to thank my wonderful wife, Olamide, without whose patience, sacrifice, love
and support I would not have been able to complete this PhD. Olamide, you have been
extremely supportive at every step of the way, and encouraged me on each time things didn’t
go as planned. Thank you for always being there. To the rest of my family: father Chief
Olubunmi Esan; sisters Oluseyi, Mosope and ‘Ranti; brother Samuel and brothers-in-law
Adeyemi, Howard and Ope- I thank you all for your encouragement.
This project would not have been possible without the enthusiasm and cooperation of the
parents and children that participated in all the studies. I thank you all very much.
The hard working and dynamic IPTpd and HIV-Iron Study teams gave everything each day for
4 years to make this PhD dream a reality. Without their zeal, efforts, innovation and dedication,
this work would not have been possible. I would like to sincerely thank Ernest Nkhoma,
Zinenani Truwah, Harriet Khofi, Bridgette Mangochi, Olivia Jelenje, Francis Munthali, David
Kachala, Elizabeth Kalanga, Georgina Makuta, Sarah Mwanamanga, Gift Dombola, Mercy Buluzi,
Wezi Neba, Mwayi Mundoli, Misho Chibwe, Thokko Ganiza, Boston Kadzamira, Nedson
Kadzuwa, Isaac Chirambo, Fatsani Gadama, Chifundo Ndamala, Elvis Moyo, Davies Kazembe,
Ben Nkumbira, Andrew Naunje, Mary Kandeya, Patrick Sululu, Dan Mwandumba, Mapopa
Nyirongo, Shadreck Magombo and Edward Banda.
The Malawi Liverpool Wellcome Trust Blantyre was the engine room and base for the IPTpd
and HIV-Iron trials conducted in this thesis and my sincere thanks goes to the current director,
Professor Robert Heyderman, and the past director, Professor Malcolm Molyneux for thier
support throughout the conduct of these studies. Professors Malcolm Molyneux, Neil French
and Victor Mwapasa made up the trial steering commitee for the two trials and deserve special
thanks. I would also like to thank the programme manager Danielle De Clercq; the former
human resources managers, Mrs Grace Mapunda and Mr. Stevie Kauka; the clinical manager
P a g e | 150
Neema Mtunthama; Maureen Mvula and Elia Ng’ambi for their invaluable help with recruiting
and managing study staff and resources; the IT team (former and current IT managers Pelani
Malange, Augustine Mtambalika and Souza Simbota, IT officer Lucia Nuka); the statistics
department (Sarah White, Mavuto Mukaka and Crispin Musicha); the laboratory manager Mike
Moore, his assistant Brigette Denis, Mavis Manyere and the MLW stores team (Richard
Kagomo, Paschal Banda and Felix Nyirenda) for all their help with logistics support for the
trials.
The pediatrics department of Queen Elizabeth Central Hospital, College of Medicine hosted the
transfusion studies and provided invaluable support with safety monitoring for the IPTpd and
HIV-Iron trials. I owe a debt of gratitude to former heads of department Prof. Elizabeth
Molyneux and Prof. Geert Tom Heikens, Drs. Queen Dube, Macpherson Mallewa, Limangheni
Makhambo Peter Moons and Baljit Cheema. In addition, the late Mr. Hastings Kamwendo
(clinical officer) was invaluable to the successful completion of the transfusion studies and
provided key hands-on clinical support for the IPTpd study team in Blantyre, thus deserves
special mention here. May his soul continue to rest in perfect peace.
The Research Support Centre of the College of Medicine, University of Malawi provided
monitoring services, data support, good clinical practice (GCP) and study protocol training for
our study sites and played a key role in the conduct and successful completion of the studies
discussed in this thesis. Special thanks goes to former directors Professors Exnervia Gomo and
Victor Mwapasa, the management team: Drs. Linda Kalinani-Phiri and Gertrude Kalanda, Mr.
Reuben Ndindi; the clinical research associates: John Chabuka, Atusaye Ngwira, Hajj Daitoni;
and the information officer Elizabeth Zambezi. I would also like to thank Sian Roberts of the
Liverpool School of Tropical Medicine, UK for her help with the drafting of serious adverse
event (SAE) reporting protocols at the early stages of the two clinical trials.
I would like to thank the past and present hospital directors, and paediatric department staff of
Thyolo Distric hospital, Queen Elizabeth Central Hospital, Zomba Central Hospital and
Chikhwawa District hospital for their support throughout the conduct of these studies. In
particular I would like to thank Drs. Charlotte Adamczick, Kevin Clarke and Okey Onah of
Zomba Central Hospital for their clinical input and support with the IPTpd and HIV-Iron trials.
I would like to thank my colleagues at the Emma Children’s Hospital, AMC, University of
Amsterdam who played their part in the completion of this thesis. I cannot but make special
P a g e | 151
mention of the former secretary of the global child health group, Caroline Veraart who
provided invaluable logistical support for my trips to the Netherlands. In addition, I would like
to thank Drs. Job Calis and Femkje Jonker for their support and encouragement throughout my
PhD. My sincere thanks to Prof. Rob Kraaijenhagen of the Meander Medisch Centrum,
Amersfoort for his help with the iron function tests reported in this thesis.
I cannot conclude without mentioning the good people of the MLW PhD room, a crazy bunch of
future professors who provided both sound scientific insights and comical relief during the
whole PhD process. I would like to thank Drs. Lloyd Bwanaisa, Chris Moxon, Patrick Nguipdop-
Djomo, Chisomo Msefula, Esther Gondwe, Kondwani Jambo, Darmie Iwajomo and Enoch
Sepako; Herbert Longwe, Victoria Ewing, Dumi Tembo, Benard Kulohoma, Harold Ocholla, Jen
Cornick, Arox Kamn’gona, and Tonney Nyirenda.
Finally, I would like to thank my promotor Professor Bernard Brabin, co-pomotors Drs.
Michael Boele van Hensbroek and Kamija Phiri, and Prof. Feiko ter Kuile for their faith in me
and belief that I had what it took to complete this task, even when at many times I doubted
myself. You are all wonderful mentors and I thank you all very much. Michael Boele has been a
motivational figure for me throughout this PhD, providing much needed encouragement,
scientific advice and support- I have learnt so much from his wealth of wisdom. Michael, I hope
I can be as patient, considerate, detail-oriented and yet amiable as you are in my future
dealings with others. Thank you for helping to make my PhD dream a reality. Kamija, you have
been with me at every step of this five year journey from it’s conception to its conclusion, and
being your first PhD student, I hope I have not let you down thus far. You have looked out for
me as one would do for a younger brother, helping me to settle down to life in Malawi and
taking out time to provide the necessary guidance, advice and training needed to help me
perform and complete the task at hand. You have been a fantastic mentor, I deeply appreciate
everything you have done for me and I am indeed very grateful. Thank you.
P a g e | 152
About the author

P a g e | 153
Michael was born in Ikeja, Lagos Nigeria on July 19, 1976 and is the fourth of five children. He
started his secondary education at King’s College, Lagos and completed it at Suleja Academy,
Niger State in 1992. He studied medicine at the University of Lagos from 1993 to 1999, and
after obtaining his bachelors degree in medicine (MB.BS), he worked as a medical intern at the
Lagos University Teaching Hospital (LUTH) from July 2000 to June 2001. He completed his
National Youth Service posting at State Hospital Sokenu, Ogun State in 2002, and worked as a
medical officer in pediatrics for 2 years till 2004. He then moved to England, UK to attend the
diploma in tropical child health (DTCH) course at the Liverpool School of Tropical Medicine in
September 2004, which he completed in April 2005. He worked as a locum senior house officer
at several hospitals in the northwest of England for a year before returning to the Liverpool
School of Tropical Medicine in September 2006 for the Master of Tropical Paediatrics course,
after obtaining a Leverhulme scholarship. It was during the Masters programme that he had
his first contact with Malawi, where he spent 2 months in Blantyre for his research project in
April 2007. After completing his Masters degree, he worked briefly at the Royal Belfast
Hospital for Sick Children in Northern Ireland before he was awarded a PhD studentship under
the College of Medicine Malawi-Amsterdam-Liverpool (COMMAL) partnership, funded by the
Netherlands-African partnership for capacity development and clinical interventions against
poverty-related diseases (NACCAP). He relocated to Malawi for his PhD in December 2007. He
is married to Olamide since 2008, and hopes to go back to clinical training after completing his
PhD studies.

Thesis

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

Thesis

Uploaded by

Copyright:

Available Formats

UvA-DARE (Digital Academic Repository)

Interventions for anaemia in children living in a resource-poor setting: Malawi

Citation for published version (APA):

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Michael Olusegun Esan

Interventions for anaemia in children living in a resource-

poor setting: Malawi

© Copyright 2012, Michael O. Esan.

Chapter 2 & 7: Copyright © 2012, Elsevier Ltd.

Chapter 5: Copyright © 2010, British Blood Transfusion Society.

Chapter 6: Copyright © 2011, Blackwell Publishing Ltd.

Interventions for anaemia in children living in

a resource-poor setting: Malawi

ter verkrijging van de graad van doctor

Michael Olusegun Esan

geboren te Lagos, Nigeria

Promotor: Prof. dr. B. J. Brabin, University of Amsterdam

Co-promotors: Dr. K. S. Phiri, University of Malawi

Overige leden: Prof. dr. M. Grobush, University of Amsterdam

‘All things are possible to him who believes…’ Mark 9:23

Chapter 2. Iron deficiency in children with HIV-associated anaemia: a systematic 31

Chapter 3. Benefits and risks of iron supplementation in HIV-infected children: 48

Chapter 4. Iron supplementation in HIV-infected children improves iron status: 67

Chapter 5. Development and evaluation of a new paediatric blood transfusion 78

Chapter 6. High transfusion failure rates in Malawian children with severe 98

Chapter 7. Intermittent Preventive Therapy in the post-discharge management 105

Chapter 8. Discussion & Conclusions. 127

WHO World Health Organisation

G6PD Glucose 6 phosphate dehydrogenase

HIV Human immunodeficiency virus

MDG Millenium Development Goals

HAART Highly active anti-retroviral therapy

IPTp Intermittent preventive therapy in pregnancy

IPTi Intermittent preventive therapy in infants

IPTc Intermittent preventive therapy in children

IPTpd Intermittent preventive therapy post-discharge

ACTs Artemisinin combination therapies

sTfR Soulble transferrin receptor

sTfR-F Soluble transferrin receptor-log ferritin index

MCV Mean corpuscular volume

MCHC Mean cell haemoglobin concentration

aMD Adjusted mean difference

aPR Adjusted prevalence ratio

Delta RR Change in respiratory rate

Delta HR Change in heart rate

Delta Hb Change in haemoglobin concentration

Delta CD4% Change in CD4 percentage

PC Packed red blood cells

the WHO millennium development goal number 4 (MDG4) - a reduction in child

Table 1: Haemoglobin levels to diagnose anaemia at sea-level (g/l)

Figure 2: Aetiological factors associated with development of anemia in resource-limited settings

Universal iron supplementation is recommended for the prevention and treatment of

resulting in increased iron accumulation in tissues.47, 48 These result in anaemia of chronic

or recrudescent malaria infections and helminth infections.74-77 In areas with moderate to

Interventions for anaemia in resource-limited settings

Table 2: Interventions to reduce anaemia prevalence in children: a population-based

ii. Iron fortification of foods

2. Malaria prophylaxis (in malaria-endemic regions)

3. Micronutrient malnutrition control

4. Control of hookworm and helminth infections

5. Improved environmental sanitation

7. Improved community-based primary health care

Table 3: Interventions for anaemia: a health-facility based approach*