You are on page 1of 8

Journal of Clinical Pharmacy and Therapeutics (2008) 33, 393400


Prevalence of possible drugdrug interactions between

antiretroviral agents in different age groups in a section of
the private health care sector setting in South Africa
N. L. Katende-Kyenda* M Pharm , M. S. Lubbe PhD , J. H. P. Serfontein PhD
and I. Truter PhD
*Department of Pharmacology, Walter Sisulu University, Mthatha, Pharmacy Practice, North-West
University, Potchefstroom and Department of Pharmacy, Nelson Mandela Metropolitan University, Port
Elizabeth, South Africa

senting 473% of the total number of prescriptions

claimed during the study period (N = 993 804).
Background: The chronic nature of human HIV patients received an average of 236 061
immunodeficiency virus (HIV) infection requires ARVs per prescription. Only 495% of the pre-
lifelong highly active antiretroviral (ARV) scriptions had one ARV medicine item, 5604%
therapy (HAART) to continuously suppress HIV- two, 3710% three, 175% four and <1% had more
1 viral replication, thus reducing morbidity than four.Of 960 DDIs identified, 188% were for
and mortality. HAART is restricted by complex patients 6 years, 427% for patients >6 years and
dosing, drugdrug interactions (DDIs) and 12 years, 063% for patients >12 and 19 years,
toxicities. 3240% for patients <19 years and 40 years,
Objective: To determine the prevalence of possible 6021% for patients <40 years and 60 years and
DDIs between ARV drugs in different age groups 063% for patients >60 years with patients
in a section of the private primary health care <40 years and 60 years having the highest num-
sector in South Africa. ber of DDIs and patients older than 60 years the
Methods: A quantitative, retrospective drug utili- lowest. The majority of DDIs between the ARVs
zation review was performed on 47 085 ARV presented in significance levels 2 and 4. The most
prescriptions claimed through a national medi- important interactions were between: indinavir
cine claims database during 2006. Possible DDIs (IDV) and ritonavir (n = 199); efavirenz (EFV) and
identified were classified according to a clinical lopinavir ritonavir (n = 65) and EFV and IDV
significance rating as described by Tatro [Drug (n = 60) all interacting at level 2.
Interaction Facts 2005. St Louis, MO: Facts and Conclusion: The importance of using drug utili-
Comparisons (2005)]. zation study as an identification tool to provide
Results: The total number of patients who received insight into the prescribing and utilization pat-
prescriptions that were claimed through the med- terns of ARV drugs, to provide optimal therapy
icine claims database was 275 424, of whom 2511% for patients infected with HIV is emphasized.
were males, 2828% were females and the gender of
4661% patients was unknown. Of the total num- Keywords: age-groups, antiretroviral agents, drug
ber of patients, 327% were HIV patients of which drug interactions, human immunodeficiency
an average of 523 386 ARV prescriptions virus patients, private health care
(n = 47 085) per patient were claimed for repre-

Received 6 November 2007, Accepted 9 April 2008

Correspondence: N. L. Katende-Kyenda, Department of Phar- INTRODUCTION
macology, Faculty of Health Sciences, Walter Sisulu University,
Private Bag X 1, Mthatha, 5117, Eastern Cape, South Africa.
The prognosis of human immunodeficiency virus
Tel.: +27 0475 022873; fax: +27 4753 269249; e-mail: kyendano (HIV) can be improved by highly active antiretro- viral (ARV) therapy (HAART). Although HAART

 2008 Blackwell Publishing Ltd 393

394 N. L. Katende-Kyenda et al.

has been proved to be effective in suppressing HIV patients clinical outcome, an understanding of the
replication, decreasing morbidity and mortality fundamental mechanism of HIV DDIs may allow
associated with the virus, and improving quality of for early detection or avoidance of troublesome
life in adults as well as infected children, the regimens and prudent management if unwanted
treatment of HIV together with its associated con- trends do develop.
ditions remains highly complex. The combination The prevalence of DDIs between ARVs in
of these drugs can present with potential drug patients prescriptions in the different age groups
drug interactions (DDIs), an important cause of has not been studied in the private health care
adverse drug reactions (ADRs) (1). sector of South Africa. Therefore, the aim of this
As the success of HAART ushered in a new era study is to determine the prevalence of DDIs
in the treatment of HIV-infection, it concurrently between ARVs prescribed to patients of different
introduced a new level of complexity in its phar- age groups in a section of the private health care
macological management. The long-term use of sector in South Africa. The results of this study will
multiple medications to treat HIV infection, HIV- provide information on which age group presents
related comorbid conditions and underlying med- with the most DDIs between ARVs and which ARV
ical disorders and the associated toxicities and drugs presented with the most clinically significant
side-effects of these medications has introduced DDIs in the different age groups.
the potential for numerous DDIs (2).
Highly active ARV therapy consists of combina-
tions such as two nucleoside reverse transcriptase
inhibitors (NRTIs) plus one non-nucleoside reverse METHODS
transcriptase inhibitor (NNRTI), three NRTIs or two
Study design
NRTIs and a protease inhibitor (PI). The ARVs
available in South Africa are divided into three Permission to conduct the study was granted from
therapeutic classes: NRTIs that include zidovudine Interpharm Datasystems and approved by the
(AZT), lamivudine (3TC), abacavir, stavudine Research and Ethical Committees of the North-
(d4T), didanosine (ddl) and zalcitabine (ddC); West University, Potchefstroom campus and Wal-
NNRTIs: nevirapine (NVP) and efavirenz (EFV); ter Sisulu University, Mthatha campus. This was a
and PIs: nelfinavir, indinavir (IDV), ritonavir (RTV), quantitative, retrospective drug utilization study
saquinavir (SQV) (hard gel capsule), SQV (soft gel performed on 47 087 ARV prescriptions claimed
capsule), amprenavir and lopinavir (LPV) (3). through a medicine claims database during the
Two ARV treatment (ART) regimens are rec- period 1 January to 31 December 2006. During
ommended for adult use in South Africa: 1a) d4T 2006, this medical scheme administrator adminis-
plus 3TC plus EFV; Protocol 1b) d4T plus 3TC plus tered 36 medical schemes data.
NVP; 2) AZT with ddl and LPV RTV (4). The The focus of this study was on the prevalence of
recommended ART regimen for paediatrics is as possible DDIs between ARVs prescribed to patients
follows: First-line therapy: 6 months3 years: of different age groups. During the retrospective
d4T 3TC LPV RTV; >3-year old and >10 kg: drug utilization study, prescriptions with more
d4T 3TC EFV. Then the second-line therapy: than one ARV drug were identified with the
6 months3 years: AZT DDI 3TC; >3-year old and Statistical Analysis System, SAS 9.1 (10). These
>10 kg: AZT DDI LPV RTV (4). prescription were then evaluated by the researcher
Much concern surrounds DDIs in patients to determine if the combination of ARVs could
receiving multi-drug therapy (5). Such interactions cause a possible DDI according to Tatro (11). The
are an important cause of ADRs and may lead to an possible DDIs identified were classified according
increased risk of hospitalization and higher health to a clinical significant rating expressed as a num-
care costs (6). Studies conducted in various coun- ber, assigned to each DDI based on the severity and
tries report that in general rates of potential DDIs documentation of the interaction, and the formula
range from approximately 1% to 66% (79). expressed as 1 (major); 2 (moderate); 3 (minor); 4
As DDIs may result in toxicity, treatment failure, (major moderate) and 5 (minor any) as described
or loss of effectiveness and can significantly affect a by Tatro (11).

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
Prevalence of DDIs between ARV agents in different age groups in the private health care sector in SA 395

Three degrees of severity were identified, Age band 5: older than 4060 years; and
namely major, moderate and minor. Age band 6: 61 years and older.

Major effects. They are potentially life threatening

Study protocol
capable of causing permanent damage. Additional
treatment, hospitalization or extension of hospital The data analysed consisted of ARV drug names
stay may be necessary. that were classified according to the pharmaco-
logical groups as described in the Monthly Index of
Moderate effects. They may cause deterioration of a Medical Specialities (12). During the retrospective
patients clinical status. Additional treatment, hos- drug utilization study, prescriptions with more
pitalization or extension of hospital stay may be than one ARV drug were identified with the
necessary. Statistical Analysis System, SAS 9.1 (10). These
prescription were then evaluated by the researcher
Minor effects. They are usually mild, having both- to determine if the combination of ARVs could
ersome or unnoticeable consequences, without cause a possible DDI according to Tatro (11) and
significantly affecting the therapeutic outcome. other literature.
Additional treatment is usually not required (11).
The following documentation levels can be dis-
Statistical analysis
tinguished namely established, probable, sus-
pected, possible and unlikely. According to Tatro The data were obtained directly from Interpharm
(11), the scale represents an evaluation of the Datasystems and analysed using the Statistical
quality and clinical relevance of the primary liter- Analysis System, SAS 9.1 (10). There was no direct
ature supporting the occurrence of an interaction. manipulation of the data by the researcher.
Drug interactions assigned documentation levels of Research was conducted from the viewpoint that
established, probable, or suspected are considered all data obtained from the medicine claims data-
to be well substantiated and have significance base were correct and accurate. Data for the anal-
ratings of 1, 2 or 3. These interactions have a ysis were obtained from one medicine claims
probability of occurring, whereas interactions of database, thus limiting external validity, implying
significance ratings 4 or 5 are not substantiated that results can only be generalized to the specific
having documentation levels of possible or unli- database used, as well as to the specific study
kely. population.
As a limitation of the study it was not possible to Each prescription record contained a unique
identify newly treated patients from the sample, number to identify each patient, medical practice,
because no demographic and clinical information pharmacy or medical scheme. These numbers were
was available on the database. All prescriptions randomly allocated by the medical scheme
that were claimed during the study period were administrator providing the data to ensure confi-
included and those prescriptions with more than dentiality. Thus, no specific patient, medical prac-
one ARV were evaluated according to Tatro (11). tice, pharmacy or medical scheme could be
identified. Thus, confidentiality of information was
maintained throughout the study.
Study population
Of the study population, 8999 were HIV patients,
who received 47 085 ARV prescriptions. The
prevalence of DDIs was analysed according to the A total 275 424 patients visited the clinic during the
age groups of the patients. The following age year 2006, of whom 2511% were males, 2828%
groups were used: were females and the gender of 4661% patients
Age band 1: birth to 6 years; were unknown. The age distribution of patients is
Age band 2: older than 612 years; given in Table 1.
Age band 3: older than 1219 years; Of the total number of patients, 8999 (327%)
Age band 4: older than 1940 years; were HIV patients, by whom an average of

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
396 N. L. Katende-Kyenda et al.

Table 1. Distribution of patients

Total database HIV patients per age group (n = 275 424)
(N = 275 424) (N = 8999)

Age band (years) (n) (%a) (n) (%a)

Birth to 6 years 29 749 1080 171 190

Older than 612 23 451 851 230 256
Older than 1219 21 457 779 76 084
Older than 1940 93 273 3387 3033 3370
Older than 4060 80 718 2931 5347 5942
Older than 60 years 16 431 597 142 158
Unknown 10 345 376
Total 275 424 100 8999 100

Percentage was calculated according to the total number of patients per age group.

511 389 ARV prescriptions (n = 47 085) per in Table 1 which demonstrates that these two age
patient were claimed during 2006. The average groups presented the highest number of HIV
number of ARV prescriptions per year according to patients resulting in having the highest number of
age groups is shown in Table 2. ARV prescriptions ARV prescriptions. This could be explained by the
represented 472% of the total number of pre- fact that people in these age groups represent the
scriptions claimed during the study period working and economically active class that could
(n = 996 787). HIV patients received an average of afford a medical scheme and use the private health
236 061 ARVs per prescription. care services. Age group 3 (older than 1219 years)
The results in Table 2 demonstrate that age group had the lowest number of ARV items (096%) and
5 (older than 4060 years) accounted for the highest 095% ARV prescriptions. Only 495% (n = 2332) of
number of ARV prescriptions, 5561% of the total the prescriptions had only one ARV medicine item,
number of ARV prescriptions, followed by age 5604% (26 387) two, 3709% (n = 17 468) three,
group 4 (older than 2040 years), with 3758% ARV 174% (n = 822) four, and <1% (n = 76) had more
prescriptions. These results are confirmed by results than five ARV medicine items per prescription.

Table 2. Average number of ARV

Total number prescription per patient per year
of according to age group
Average number of ARV prescrip-
HIV patients ARV prescriptions tions
(N = 8999) per patient per yeara (N = 47 085)

Age band (n) (n) (%b)

Birth to 6 years 171 544 428 958 203

Older than 612 230 569 419 1332 283
Older than 1219 76 506 436 445 095
Older than 1940 3033 494 381 17 694 3758
Older than 4060 5347 522 391 26 185 5561
Older than 60 years 142 431 360 470 100
Total 8999 511 389 47 085 100

Average number was calculated according to the number of ARV items claimed in each
age group during 1 year.
Percentage was calculated according to the total number of ARV prescriptions claimed in
each age group during 1 year.

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
Prevalence of DDIs between ARV agents in different age groups in the private health care sector in SA 397

Table 3. Total number of possible DDIs identified highest number of both patients ARV prescriptions
according to the total number of ARV prescriptions per (Tables 1 and 2).
age band The results of this study showed that of the
possible DDIs between the ARVs in the different
Total number of
age bands. The most important interactions were
ARV prescriptions Number of
between: IDV and RTV (2073%; n = 199); EFV and
Age band n (%)a DDIs n (%)b
LPV RTV (677%; n = 65); and EFV and IDV
Birth to 6 years 958 (203) 18 (188) (625%; n = 60) all interacting at clinical signifi-
Older than 612 1332 (283) 41 (427) cance 2 (moderate) as shown in Table 4.
Older than 1219 446 (095) 6 (063) The age band with the highest number of inter-
Older than 1940 17 694 (3758) 311 (3240) acting ARVs was 5 (older than 4060 years), with
Older than 4060 26 185 (5561) 578 (6021) the most DDIs between IDV and RTV (n = 199).
Older than 60 years 470 (100) 6 (063) This age band presented with the highest number
Total 47 085 (100) 960 (100) of both DDIs and ARV prescriptions (Table 3). Age
group 6 (older than 60 years) presented with the
Percentage was calculated according to the total number of
smallest number of DDIs and interacting ARVs.
ARV prescriptions in each age group.
Percentage was calculated according to the total number of The ARVs with the most DDIs were between IDV
DDIs identified in each age group. and RTV, EFV and IDV and EFV and LPV RTV.

A total of 960 DDIs were identified according to

age bands of the patients. A comparison of the DDI
rates against the prescription rates for each band A quantitative, retrospective drug utilization review
are shown in Table 3. Patients <40 and 60 years was performed on ARV prescriptions claimed
presented with highest number of possible DDIs through a medicine claims database during 2006 to
(6021%) and prescription rates (5561%), followed determine the prevalence of DDIs between ARV
by patients 19 and 40 years with 3240% possible drugs in different age groups in a section of the
DDIs and 3758% ARV prescription rates, com- private health care sector in South Africa.
pared with patients older than 60 years, with the As demonstrated in Table 1, age group 5 (older
lowest possible number of DDIs (063%) and ARV than 4060 years) presented with the highest
prescription rates of 100%. As already stated number of HIV patients accounting for 5942%
patients in 4 and 5 age bands presented with the (n = 53 479), followed by age group 4 (older than

Table 4. Number of possible

interacting ARVs at moderate Age band Interacting ARVs Prevalence
in the different age bands
Birth to 6 years Nelfinavir and nevirapine 5
Older than 612 years Saquinavir and efavirenz 1
Saquinavir and lopinavir ritonavir 1
Efavirenz and lopinavir ritonavir 22
Nelfinavir and nevirapine 6
Older than 1219 years Indinavir and ritonavir 2
Older than 1940 years Efavirenz and lopinavir ritonavir 19
Efavirenz and indinavir 60
Efavirenz and saquinavir 2
Indinavir and ritonavir 7
Saquinavir and ritonavir 5
Older than 4060 years Efavirenz and lopinavir ritonavir 14
Efavirenz and nelfinavir 2
Indinavir and ritonavir 188
Older than 60 years Indinavir and ritonavir 2

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
398 N. L. Katende-Kyenda et al.

1940 years) with 3370% (n = 3033). This could be be performed on ARV dosages and their combi-
explained by the fact that these two age groups are nations to verify whether they adhere to the rec-
the most sexually active groups of society, and ommended treatment guidelines.
probably engaging in unprotected sex, which puts A total of 960 possible DDIs were identified
them at higher risk of HIV infection, and the most between the ARVs with the most important DDIs
economically advantaged so could afford to visit a between RTV and IDV (1615%; n = 155); EFV and
private health care centre. Age group 6 (61 years IDV (625%; n = 60); EFV and LPV RTV (354%;
and older) presented with the lowest number of n = 34); and SQV and LPV RTV (229%; n = 22), all
patients, being the less sexually active, and possi- interacting at level 2. RTV, IDV and LPV RTV all
bly using condoms or abstaining from sex, with the belong to the group of PIs and EFV to the group of
result that their chances of contracting the disease NNRTIs. All currently available PIs are metabo-
are more limited (13). According to South Africa lized by the cytochrome P450 (CYP) enzyme sys-
HIV and AIDS Statistics Summary for 2006, women tem, and are all inhibitors of CYP3A4, ranging from
are more likely to be infected than men (13). weak inhibition for SQV to very potent inhibition
The gender distribution of the study population for RTV (16). Thus in this study these were iden-
revealed female predominance. According to Sta- tified as the most important possible DDIs, as they
tistics South Africa Census 2001 (14), there were are predicted to have numerous drug interactions.
more females than males in South Africa in 2001. Single-PI regimens significantly reduced the
As a limitation of the study, no demographic or morbidity and mortality associated with HIV fol-
clinical information of the patients was given in the lowing their introduction (17). More recently,
database. boosted PI regimens combined RTV with a second
The results of this study showed that ARVs were PI to achieve higher sustained levels of the second
prescribed to approximately 3% of all patients one than seen when it was given as part of a single-
whose prescriptions were claimed through this PI-regimen (18).
medical scheme administrator in the 1-year period. In this study, it was established that RTV com-
ARV prescriptions accounted for 427% of the total bined with IDV may interact at clinical level 2. Both
number of prescriptions claimed through the are PIs and according to Malaty (16), PIs also
database for the 1-year period. interact with each other, and these interactions are
Human immunodeficiency virus patients being explored for their potential therapeutic ben-
received an average of 511 389 ARVs per pre- efits. This is in line with data from Saah et al. (19)
scription. Only 495% of the prescriptions had one which suggested that higher IDV doses (800 mg)
ARV medicine item, 5604% two, 3709% three, and or RTV doses (>100 mg) might provide better
174% four and <1% had more than five. Combi- efficacy, but might also contribute to greater
nation therapy is recommended because combina- toxicity and therefore may need dose adjustment to
tions of ARVs create multiple obstacles to HIV IDV RTV 667 100 mg regimen.
replication to keep the number of offspring low The issue of DDIs should be a major clinical
and reduce the possibility of a superior mutation concern for all clinicians, hence the need for mul-
(15). No individual ARV drug has been demon- tiple reminders and warnings whenever more than
strated to suppress HIV infection for long. The two medicines are administered. Results of this
results of this study indicated that most patients study revealed that combining EFV with LPV RTV
appeared to being prescribed therapy that is may interact at clinical significance 2. It was
consistent with the recommendations, therefore observed in a study that the NNRTIs NVP and EFV
complying with the standard care of using combi- lower plasma levels of PIs in adults and children.
nations of ARVs, though it is worrying that the Therefore, it was recommended that coadminis-
mean number of ARV per prescription is worry- tration of LPV RTV with NVP and EFV necessi-
ingly low at much less than 3. Another limitation in tates a 30% increase in the dose of LPV RTV in
this study was that no dosages of the different ARV adults (20).
drugs were supplied. The focus was on the prin- In another study by Aarnoutse et al. (21) on
ciples rather than on specific interactions. It is healthy volunteers, the effect of EFV (600 mg once
therefore recommended that further investigations daily) on the pharmacokinetics of RTV IDV

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
Prevalence of DDIs between ARV agents in different age groups in the private health care sector in SA 399

(100 mg 800 mg twice daily) was examined the age groups with both the highest number of
resulting in decreased plasma concentrations of patients and ARV prescriptions. This knowledge
both IDV and RTV, on addition of EFV to will impact positively on the healthy professionals
RTV IDV. The study therefore recommended that caring for HIV AIDS patients in this age band
a dose of 200 mg of RTV with 800 mg of IDV can be group, in that strategies will have to sought out and
used whenever EFV is used concomitantly, in interventions performed in this age band group,
treatment-experienced patients with more resistant thus improving on HIV medical practice in South
virus. The above observations support the results Africa.
obtained in this study that EFV, a NNRTI, and IDV, Combination ART is a potent and effective
a PI, may interact at clinical significance 2 therapy for HIV infection. Unfortunately, ARV
accounting for (625%) (n = 60). drugs frequently interact among themselves, as
Drugs commonly taken by HIV patients have a well as with other anti-infectives. These interac-
strong potential to interact with the PIs. In partic- tions determine positive or negative consequences
ular, the NNRTIs are also metabolized by CYP and resulting in recommendations to avoid some com-
have been shown to interact with PIs (16). There- binations or to adjust the dosage of coadministered
fore, pharmacists and physicians must always be drugs. The effective clinical use of ARVs requires
vigilant for drug interactions, both those that are detail knowledge of interaction mechanisms by all
already documented and those that are predictable healthcare professionals dealing with HIV AIDS,
from pharmacokinetic profiles, in patients receiv- and the consequent recommendations in their
ing PIs. management when coadministering interacting
More supporting evidence to the results of this ARVs.
study is by Moreno et al. (22) to verify whether RTV,
a PI, could be used to prevent adverse interactions
between NNRTIs and PIs, extended this strategy to
concomitant EFV and IDV and their results revealed The financial assistance of the South African Med-
that in the absence of RTV, EFV decreased the IDV ical Research Council (MRC) and the South African
area under the concentration curve (AUC) by 31% National Research Foundation (NRF) towards the
and the Cmax by 16%. The study therefore recom- research is hereby acknowledged. Opinions
mended that the IDV dose be increased from 800 to expressed in this study, and conclusions arrived at,
1000 mg every 8 h. are those of the authors. Thank you to the manag-
Although HIV drug interactions are usually ers of the medicines claims database that provided
thought of as detrimental, resulting in a loss of the data and to Mrs. M. Terblanche for assisting in
therapeutic effect or toxicity, some drug interac- proofreading the manuscript.
tions such as RTV boosted PI-based ARTs are
beneficial and are commonly used in clinical
practice (23). Therefore, pharmacists need to
understand drug interaction mechanisms, remem- 1. Highleyman L (2005) Drug interactions and anti-HIV
ber key drug interactions, and vigilantly monitor therapy. The body: the complete HIV AIDS
resource. Available at:
patients for potential complications. It is also rec-
content/treat/art2579.html (accessed 2 August 2007).
ommended that analysis on the management of the
2. Agins BD, Mojica-Rogers C (2003) Drug interactions
most important clinically significant DDIs between associated with HAART. AIDS Read, 13, 433450.
ARVs be performed. Available at:
461892-4 (accessed 29 July 2007).
CONCLUSION 3. Wood R (2006) Antiretroviral medicines (ARVs).
Expert annexure. Available at:
The aim of the study was to determine the preva- Document/DrugcomapaniesCC/Tau-V-GSK
lence of DDIs between ARV agents in different age woodaffidavit.doc (accessed 18 July 2007).
bands. It was established that age band 4 (older 4. National Department of Health South Africa (2004)
than 1940 years) and 5 (older than 4060 years) National antiretroviral treatment guidelines. Depart-
had the highest prevalence of possible DDIs, being ment of Health: Pretoria, South Africa. Available at:

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400
400 N. L. Katende-Kyenda et al. 16. Malaty LI, Kuper JJ (1999) Drug interactions of HIV

artguideline04-f.html (accessed 10 May 2008). protease inhibitors. Drug Safety, 20, 147169.
5. Crucio-Souza JM, Thomoson JC (2006) Prevalence of 17. Palella FJ Jr, Delaney KM, Moorman AC et al. (1998)
potential drug-drug interactions and its associated Declining morbidity and mortality among patients
factors in a Brazilian teaching hospital. Journal of with advanced human immunodeficiency virus
Pharmacy Pharmaceutical Sciences, 9, 427433. infection. New England Journal of Medicine, 338, 853
6. Juurlink DN, Mamdani M, Kopp A et al. (2003) 860.
Drug-drug interactions among elderly patients 18. Rhame FS, Rawlins SL, Petruschke RA et al. (2004)
hospitalized for drug-toxicity. Journal of the American Pharmacokinetics of indinavir and ritonavir admin-
Medical Association, 289, 16521658. istered at 667 and 100 milligrams, respectively, every
7. Geppert U, Beindl W, Hawranek T et al. (2003) Drug 12 hours compared with indinavir admininistered at
interactions in clinical practice. A pilot project 800 mg every 8 hours in human immunodeficiency
for quality assurance in prescribing. Hautarzt, 54, virus-infected patients. Antimicrobial Agents Chemo-
5357. therapy, 48, 42004208.
8. Bobb A, Gleason K, Husch M et al. (2004) The epi- 19. Saah AJ, Winchell GA, Nessly ML et al. (2001)
demiology of prescribing errors: the potential impact Pharmacokinetic profile and tolerability of indinavir-
of computerised prescriber order entry. Archives of ritonavir combinations in healthy volunteers. Anti-
Internal Medicine, 164, 785792. microbial Agents Chemotherapy, 45, 27102715.
9. Klarin I, Wimo A, Fastbom J (2005) The association of 20. Groot R, de Burger DM, Hartwig NG et al. (2005)
inappropriate drug use with hospitalization and Increased dose of lopinavir ritonavir compensates
mortality. Population-based study of the very old. for efavirenz-induced drug-drug interaction in
Drugs and Aging, 22, 6982. HIV-1-infected children. Journal of Acquired Immune
10. SAS Institute (2005) Statistical analysis system (SAS Deficiency Syndrome, 39, 6368.
9.1) software version 9.1. Cary, NC: SAS Institute. 21. Aarnoutse RE, Burger DM, Hugen PWH et al. (2000)
11. Tatro DS (2005) Drug interaction facts 2005. St. Louis, A pharmacokinetic (PK) study to investigate the
MO: Facts and Comparisons, 11699. influence of efavirenz (EFV) on a BID indinavir
12. Synman JR (2007) Antimicrobials. In: Snyman JR, ed. (IDV) ritonavir(RTV) regimen (800 100 mg) in
Mims medical specialities, Vol. 44. Pinegowrie: Joh- healthy volunteers. Program and abstracts of the
nic Mims, 251306. 40th Interscience Conference on Antimicrobial
13. South African HIV & AIDS Statistics Summary Agents and Chemotherapy, September 1720,
(2006) The South African Department of Health Toronto, ON: Abstract 423.
Study, 2006. Available at: 22. Moreno A, Casado JL, Marti-Belda P et al. (2001)
safricastats.htm (accessed 21 August 2007). Concomitant use of non-nucleoside reverse trans-
14. Statistics South Africa (2003) Census, 2001. Pretoria, criptase inhibitors (NNRTIs) does not decrease the
South Africa: Statistics South Africa. Available at: inhibitory quotient of dual ritonavir-indinavir-based therapy. Program and abstracts of the 41st Inter-
CIB2003.pdf (accessed 27 March 2007). science Conference on Antimicrobial Agents and
15. United States Department of Health and Human Chemotherapy, December 1619, Chicago, IL: Ab-
Sciences (2004) A guide to primary care for people stract 11728.
with HIV AIDS, 2004 Edition. Available at: http:// 23. Krikorian SA (2005) Drug-drug interactions and HIV (accessed 6 therapy: what should pharmacist know? Journal of
December 2007). Pharmacy Practice, 18, 278294.

 2008 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 33, 393400