Objectives: To determine rates and causes of switching
from first- to second-line antiretroviral treatment (ART)
regimens in a large treatment-naive cohort (a South
African community-based ART service) where a targeted
adherence intervention was used to manage initial
Methods: ART-naive adults (n=929) commencing first-line
non-nucleoside-based ART [according to WHO (2002)
guidelines] between September 2002 and August 2005
were studied prospectively. Viral load (VL) and CD4+T-cell
counts were monitored every 4 months. All drug switches
were recorded. Counsellor-driven adherence interventions
were targeted to patients with a VL >1,000 copies/ml at
any visit (virological breakthrough) and the VL measure-
ment was repeated within 8 weeks. Two consecutive VL
measurements >1,000 copies/ml was considered virolog-
ical failure, triggering change to a second-line regimen.
Results: During 760 person-years of observation [median
(IQR) 189 (85–441) days], 823 (89%) patients were retained
on ART, 2% transferred elsewhere, 7% died and 3% were
lost to follow-up. A total of 893 (96%) patients remained
on first-line therapy and 16 (1.7%) switched to second-line
due to hypersensitivity reactions (n=9) or lactic acidosis
(n=7). A Kaplan–Meier estimate for switching to second-
line due to toxicity was 3.0% at 32 months. Virological
breakthrough occurred in 67 (7.2%) patients, but, following
use of a targeted adherence intervention, virological failure
was confirmed in just 20 (2.2%). Kaplan–Meier estimates at
32 months were 20% for virological breakthrough but only
5.6% for confirmed virological failure.
Conclusion: Regimen switches were due to virological failure
or toxicity. Although follow-up time was limited, over 95%
of individuals remained on first-line ART using a combination
of viral monitoring and a targeted adherence intervention.
Conservation of first-line antiretroviral treatment
regimen where therapeutic options are limited
Catherine Orrell1*, Guy Harling1, Stephen D Lawn1,2, Richard Kaplan3, Matthew McNally3,
Linda-Gail Bekker1,4and Robin Wood1
1The Desmond Tutu HIV Centre, Institute for Infectious Disease and Molecular Medicine, Faculty of Health Sciences, University of Cape
Town, South Africa
2Clinical Research Unit, Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, UK
3Hannan Crusaid Treatment Centre, Gugulethu Community Health Centre, Department of Health, Provincial Authority of the Western
Cape, South Africa
4The Department of Medicine, Faculty of Health Sciences, University of Cape Town, South Africa
*Corresponding author: Tel: +27 21 650 6958; Fax: +27 21 650 6963; E-mail: email@example.com
Antiviral Therapy 12:83–88
The major impact of antiretroviral therapy (ART) on
HIV-associated morbidity and mortality in high-income
countries has been well documented [1,2]. Early data
from resource-limited countries have shown similar
clinical and virological successes [3–11]. However, in
many resource-limited settings, ART options are limited
to a single non-nucleoside reverse transcriptase
inhibitor (NNRTI)-based first-line regimen, followed
by a single protease inhibitor (PI)-based second-line
regimen [12,13]. In South Africa, the national ART
programme supplies first- and second-line regimens free
of charge, but thereafter offers no further treatment
options . Only four NRTIs are available as back-
bone therapy within the two available regimens.
Treatment options following development of either
drug toxicity or virological failure are often limited due
to the high costs of alternative medications. Even within
the limited options available, second-line therapy is on
average three times more costly than first line.
For long-term health to be maintained within a frame-
work of very limited ART options, it is important to
maximise the clinical benefits derived from each regimen.
Reports from high-income countries have raised
concerns about rapid exhaustion of treatment options,
leading to high rates of virological failure [14,15]. In the
UK the reported median duration of the first-line ART
regimen was 17 months and 38% of patients experienced
all three classes of antiretrovirals within 2 to 5 years of
commencing therapy [14,15]. There are no similar
studies from resource-limited countries, where such high
rates of therapy switching would result in a large propor-
tion of the ART-treated population quickly exhausting
© 2007 International Medical Press 1359-6535
their limited treatment options. A balance must be main-
tained between switching early with resultant loss of regi-
mens and continuing a failing regimen, which may
encourage development of viral resistance.
The utility of viral load (VL) monitoring in resource-
limited settings has not been well established. However,
our experiences in a community-based public sector
ART service in Cape Town suggest that use of regular
follow-up VL measurements coupled with an intensive
peer counsellor adherence intervention to promptly
manage initial virological breakthrough is associated
with a low rate of confirmed failure, thereby retaining
individuals on first-line therapy and conserving more
expensive second-line therapy for future use. Here we
report the rates of initial virological breakthrough and
confirmed virological failure in this service and the rate
of switching to second-line therapy for reasons of both
virological failure and toxicity.
The Hannan Crusaid Treatment Centre is the ART roll-
out clinic for the Nyanga district, a township near
Cape Town, South Africa where people live in condi-
tions of low socioeconomic status [11,16,17]. An esti-
mated 300,000 people live in this district, which had an
antenatal HIV-1 seroprevalence rate of 28% in 2003.
The site began recruiting in September 2002 and by
September 2005 cared for more than 1,500 adults and
children needing ART. Drugs and laboratory investiga-
tions are supplied free of charge.
The centre accepts referrals from nine local HIV
clinics and the local midwife obstetric unit. The
majority of those referred to the centre are ART-naive.
Patients are referred to the clinic based on eligibility for
ART under the WHO 2002 guidelines, which include
those with a CD4+T-cell count of <200 cells/μl or a
prior AIDS-defining illness as defined by the WHO
classification system .
Employment of local peer counselling staff is enabled
by donor funding. One counsellor is assigned up to 50
patients entering the clinic. Counsellors are responsible
for preparing the patients for ART at three small group
education sessions (‘treatment readiness’ sessions), and
for on-treatment adherence support. This care includes
scheduled visits to patients at home every 2 weeks pre-
treatment and monthly after starting treatment until
viral suppression is achieved. Thereafter the counsellors
make a routine home visit to each patient once every 3
to 4 months and also whenever the need arises. Active
follow-up of patients by counsellors was used when
patients failed to attend appointments.
Patients are medically reviewed every 2 weeks prior
to ART, and then at 4, 8 and 16 weeks of ART and
16-weekly thereafter. HIV RNA load and CD4+T-cell
count assessments are made pre-treatment and every 16
weeks during ART. Viral load assays were done using
the branch DNA hybridization technique [Bayer HIV-1
RNA 3.0 assay (branch DNA); Bayer Healthcare,
Leverkusen, Germany]. Additional safety tests are
performed according to the provincial ART protocol
and vary according to the treatment regimen .
First-line ART is comprised of stavudine (d4T) and
lamivudine (3TC) plus an NNRTI [efavirenz (EFV) or
nevirapine (NVP)]. Pregnant women are started on
zidovudine (AZT), 3TC and NVP and the AZT is
switched to d4T after delivery. The second-line regimen
for those failing first-line treatment comprises
lopinavir/ritonavir (LPV/r), AZT and didanosine.
Switches to a regimen containing LPV/r was consid-
ered to be second-line therapy, whatever the reason for
the switch. For example, patients with symptomatic
hyperlactataemia (lactate >5–10 mmol/l with clinical
symptoms of hyperlactataemia) were switched to an
NRTI-sparing regimen of EFV and LPV/r.
Details of all switches in ART were recorded, but the
primary focus of this study was switches to a second-
Patients with a VL >1,000 copies/ml at any follow-up
visit during ART (defined as initial virological break-
through) were reviewed and received a targeted adher-
ence intervention. This included being issued with a pill
box and a dosing diary. The frequency of counsellor
home visits was increased to weekly and the patient was
required to re-attend the three education sessions. The
VL measurement was repeated after 6 to 8 weeks of this
intensive counselling. Virological failure was defined by
a second VL >1,000 copies/ml and triggered a switch to
second-line therapy. A VL between 50 and 1,000
copies/ml prompted continuation on first-line therapy,
with high frequency home visits continuing. Only once
the patient’s VL had fallen below 50 copies/ml was the
alert status removed and routine visits recommenced.
Patients commencing ART at the clinic between 2
September 2002 and 11 August 2005 were considered
for this analysis. Those <15 years of age and those
who were non-naive to ART were excluded. Data
were accessed from structured clinical and laboratory
records that were maintained on all patients screened
on entry to the ART programme and which were
transferred on a weekly basis to a database.
Outcomes were assigned to all patients, including
dates of leaving the programme.
C Orrell et al.
© 2007 International Medical Press
Patients were right-censored on 11 August 2005 if
they remained in care. Patients who were more than 4
weeks late for a scheduled visit were considered lost to
follow-up, and were considered lost to the programme
28 days after their last visit to the clinic, at which point
they would have run out of drugs. Patients who were
transferred out were considered lost to the programme
(censored) on the date they agreed with the clinic to
leave the programme.
Analysis was carried out using Stata version 9
(StataCorp, College Station, TX, USA). Patient charac-
teristics were described using median values and inter-
quartile ranges (IQR) for continuous variables, and
counts and percentages for categorical data.
Differences of proportion were tested using a χ2test.
Product-limit (Kaplan–Meier) estimates of survival
were calculated using all eligible patients. Where
appropriate, confidence intervals were also calculated
from the product-limit estimator.
The analysis included 929 people who commenced
treatment between 2 September 2002 and 11 August
2005 (Table 1). The median age at entry was 33 years
(IQR 28–38). The analysis cohort was predominantly
female (n=671, 72%) of whom 111 (12%) were preg-
nant at enrolment. The majority of individuals had
advanced symptomatic disease, with 488 (54%) having
WHO stage 3 disease and 255 (28%) having stage 4
disease. The median CD4+T-cell count was 95 cells/μl
(IQR 47–151) and the median HIV RNA level was
4.83 log10copies/ml (IQR 4.48–5.23). Following initia-
tion of ART, 760 person-years of observation accrued,
with a median duration of follow-up of 189 days per
person (IQR 85–441).
Of those who started ART, 89% (n=823) remained in
care at this clinic, 2% (n=18) were transferred to
another ART programme and 6.7% (n=62) of patients
died, with the majority of these deaths occurring during
the first 4 months of treatment, with no deaths after 16
months of ART. Death rates and causes of early and
late mortality have been published elsewhere [16,17].
Losses to follow-up were few (n=26, 2.8%).
Kaplan–Meier estimate of survival at 3 years was 90%
and of overall retention within this service was 84%.
Regimen changes due to toxicity
At the time of reporting, changes to treatment regimens
were infrequent in this cohort. Changes within first-line
therapy that did not necessitate a change to a PI
regimen (n=45, 4.8%) were more common than
regimen switches. These included switches from d4T to
AZT among 25 people (2.7%) and were largely due to
ACTG grade 3 or 4 peripheral neuropathy or subjec-
tively significant lipodystrophy. AZT was switched to
d4T in six (0.6%) patients due to either ACTG grade 3
or 4 anaemia or a protocol-specified change at the end
of pregnancy. EFV was switched to NVP in nine (1.0%)
patients due to EFV-related central nervous system
toxicity (n=1) or intended or confirmed pregnancy
(n=8). NVP was switched to EFV in the remaining five
(0.5%) patients for reasons that included commence-
ment of concurrent anti-tuberculous therapy, NVP
hypersensitivity and end of a pregnancy.
Just 16 patients (1.7%) changed from first to
second-line therapy due to drug toxicity: nine (1.0%)
had a hypersensitivity reaction to EFV necessitating
use of LPV/r. A further seven patients (0.7%)
changed to an NRTI-sparing regimen of LPV/r and
EFV due to development of symptomatic hyperlac-
tataemia (serum lactate >5 mmol/l with clinical
symptoms). Kaplan–Meyer survival analysis esti-
mates that toxicity would result in 3.0% of individ-
uals switching to second-line therapy at 32 months
and would contribute less to regimen switching than
would virological failure (Figure 1).
Virological and immunological outcomes
Viral load and CD4+T-cell counts were measured
every 4 months and these data were available for
92.9% of HIV RNA tests and 93.0% of CD4+T-cell
counts. Each patient had a median of 3.37 (IQR
2.99–3.92) VL measurements per year. The propor-
tions of patients at each timepoint with viral suppres-
sion to <1,000, <400 and <50 copies/ml were
94–100%, 93–98% and 77–88%, respectively (Figure
2A). The median CD4+T-cell count increased steadily
throughout follow-up, rising from 95 cells/μl at base-
line to 404 cells/μl at 32 months.
Antiviral Therapy 12:1
Conservation of first-line ART regimens
Table 1. Baseline characteristics of treated cohort
Female, n (%)
WHO stage 3, n (%)
WHO stage 4 (AIDS), n (%)
CD4+T-cell count, cells/μl
CD4+T-cell count <50 cells/μl, n (%)
HIV RNA level, log10 copies/ml
Starting regimen, n (%)
Unless stated all figures are medians and interquartile ranges. 3TC, lamiduvine;
AZT, zidovudine; d4T, stavudine; EFV, efavirenz; NVP, nevirapine.
Virological failure and change to second-line therapy
Rapid expansion of the programme in the latter part of
the study period resulted in a majority of the cohort
having a short duration of follow-up relative to the
total duration of the study (median 189 days, IQR
85–441). A total of 67 people (7.2%) had ≥1 follow-up
VL measurements of >1,000 copies/ml. Of these, 43
had a second VL test prior to data censorship and 24
people were awaiting this test. At censor date, 20
people (2.2%) were confirmed as virological treatment
failures and had commenced or were shortly to
commence the second-line treatment regimen. The
demographic characteristics of patients with ≥1 VL
measurement >1,000 copies/ml did not significantly
differ from those of the rest of the cohort.
The estimated proportions of individuals who had had
one VL above each of three virological thresholds were
plotted (Figure 2B). At 32 months, the cumulative risk of
having one or more VL measurements >50 copies/ml was
70% with a rate of 38.8 per 100 person-years. In
contrast, the risk of having a VL >400 copies/ml was
just 23% or 9.6 per 100 person-years. The risks for a
single VL of >1,000 (20% or 7.9 per 100 person-years)
was similar to that of >400 copies/ml (Figure 2B).
Further analyses used 1,000 copies/ml as specified by
the ART programme guidelines used in this service.
To determine the likelihood of virological failure
above >1,000 copies/ml, we plotted Kaplan–Meier
proportion estimates (Figure 3). At 32 months, 20%
(CI 14.7–25.1) of patients would be estimated to
have a ≥1 VL measurement >1,000 copies/ml and
receive the targeted adherence intervention. It was
estimated that virological failure would subsequently
be confirmed in 29% of these; comprising just 5.6%
(CI 3.8–11.4) of the whole cohort. This corresponds
to a rate of 2.2 per 100 person-years.
There are few data concerning switching ART to second-
line regimens in a resource-limited setting. We have
demonstrated that in this setting ART can be provided to
a large number of patients attending a public-sector clinic
while retaining more than 95% of patients on a first-line
treatment regimen over 3 years. At the time of reporting,
with a median follow-up time of 189 days (IQR 85–144),
switches to second-line therapy were infrequent (3.9%)
and were attributable to either drug toxicity (1.7%) or
virological failure (2.2%). Using a system of early detec-
tion of virological breakthrough leading to implementa-
tion of a targeted adherence strategy, 23 (53%) of 43
people with initial VL >1,000 copies/ml subsequently
achieved full virological suppression.
Rates of regimen switch due to toxicity were low.
Less than 5% of patients required a single drug switch
C Orrell et al.
© 2007 International Medical Press
Figure 1. Kaplan–Meier survival estimate for time on first-
Censoring occurred at time of treatment switch to a protease-based ART
regimen due to toxicity or confirmed virological failure for 929 naive individ-
uals commencing therapy in a large community-based clinic in South Africa.
048 12 16 20 24 28 32
Duration on treatment, months
929 636 414 311 212 151 117 8045
Toxicity and Virological Failure
Number at risk at
end of period
Proportion of patients surviving
Figure 2. Virological outcomes of the cohort over time
(A) Percentage of patients with a viral load (VL) of <50, <400 or <1,000
copies/ml at each timepoint for the on-treatment cohort (n=929). (B) Cumulative
percentage of patients with a first VL >50, >400 or >1,000 copies/ml.
Measurements are made 4-monthly, hence the step-wise increase in the curves.
048 1216 20242832
Viral load <1000 copies/ml
Viral load <400 copies/ml
Viral load <50 copies/ml
Duration on treatment, months
Percentage of patients
048 121620 24 2832
Months on treatment
Percentage of patients on programme
within the first-line regimen and all changes were
attributed to well-recognised adverse effects of the
drugs used . In particular, widespread use of d4T in
a first-line treatment regimen is frequently associated
with drug switching due to peripheral neuropathy,
lipodystrophy and hyperlactataemia. Only 1.7% of
patients changed to second-line therapy as a result of
toxicity and the estimated cumulative proportion of
patients requiring such a switch was just 3% at 32
months. Lack of drug choices may have served to
significantly limit both the potential and the inclination
of the health-care provider to switch drugs unless
strongly indicated. In addition, the rate of switch due
to drug-related toxicity, and possibly virological
failure, may be expected to rise as the follow-up time of
this cohort increases.
On the date of censorship, 2.2% of this cohort had
switched to second-line therapy due to virological
failure; Kaplan–Meier estimates predicted failure rates
of just 5.6% at 32 months. More than 50% of people
who developed initial virological breakthrough (VL
>1,000 copies/ml) did not progress to confirmed viro-
logical failure. Recent studies in the developed world
have noted that initial or primary virological failure
(one elevated VL) is more likely to be due to treatment
interruption or poor adherence rather than viral resis-
tance [20,21]. Thus, management of initial virological
failure using targeted programmatic interventions to
improve adherence has the potential to have a positive
impact on maintaining low rates of transfer to second-
The threshold of virological failure has not been
clearly defined. The lower the virological failure
threshold, the higher the cumulative incidence of
failure that will be detected. In our cohort, the cumula-
tive incidence of having at least one VL of
>50 copies/ml was 70% in 32 months, although the
risk at any given timepoint was 13% or less. Viral loads
between 50 and 400 copies/ml typically return to viral
suppression without specific intervention and
stochastic variation about the mean is a likely explana-
tion for a majority of these VL increases .
A much smaller cumulative proportion of the cohort
reached a VL of >400 or 1,000 copies/ml (23% or
20%, respectively) over 32 months. These proportions
were not significantly different from each other and
were similar to those of ART-naive patients in the
EuroSIDA cohort . Random variation about the
mean is far less likely at a VL level of 400 or 1,000
copies/ml than at 50 copies/ml and any value >400
copies/ml is more likely to reflect significant viral
breakthrough . Our programme defined people
with a VL >1,000 copies/ml as being potentially at risk
of failure – a figure justified by these observations.
In the absence of a controlled trial it is not possible
to definitively attribute the low rates of virological
failure to the adherence strategy targeted at those with
virological breakthrough. However, we believe that this
intervention plays a very important role and would
consider a randomized trial examining this to be uneth-
ical. Virological breakthrough is likely to represent a
combination of poor adherence and emerging viral
resistance. As reported previously, the demographics
(age, gender and baseline WHO clinical stage) of
patients within this cohort with virological break-
through did not differ from those of the rest of the
cohort . Among individuals with an initial VL
>1,000 copies/ml, 53% subsequently achieved VL
suppression and therefore remained on first-line
therapy. It is possible that our results may be affected
by survival bias due to mortality or losses to follow-up.
However, rates of these losses were low compared with
other African cohorts [7,8,25,26]. Moreover, a
majority of mortality occurred in the first 4 months
prior to any follow-up VL measurements and we have
previously shown that patients lost to follow-up did
not have poor response to ART prior to leaving the
The completeness of the data is an important strength
of this study. In addition to implementation of targeted
adherence support, the peer counsellors played an
important role in patient retention and data integrity.
There is always concern about the generalisability of
data such as those presented here. The clinical and
socioeconomic characteristics of this cohort reflect those
of many other African cohorts being treated on govern-
ment programmes [3,4,7,26]. However, the national
programmes in this and most other countries in sub-
Saharan Africa do not provide the intensive counselling
staff levels that may have contributed to our high levels
Antiviral Therapy 12:1
Conservation of first-line ART regimens
Figure 3. Kaplan–Meier failure estimate for time to first,
then second consecutive HIV RNA level >1,000 copies/ml
048 12 16202428 32
929 641 421 328 229 162 127 86
Patients at risk of starting second-line therapy
First HIV RNA >1000 copies/ml
First and second consecutive
HIV RNA >1000 copies/ml
Duration on treatment, months
Proportion of patients on programme
of success. The cost-effectiveness of such a system on a
wide scale needs to be assessed. Use of VL monitoring in
community-based settings has been debated and a
system for detecting failure through clinical and
immunological means has recently been suggested by
Colebunders and colleagues . However, the
proposed algorithm had very low sensitivity for
detecting virological failures in this cohort [28,29]. At
present, all South African ART sites have access to VL
monitoring; data from other programmes are needed to
further evaluate different patient management strategies.
A public health approach to ART requires optimal
use of limited treatment regimens. Use of VL measure-
ment, coupled with a targeted adherence intervention,
was associated with a low rate of virological failure
and resulted in more than 95% of patients remaining
on first-line therapy in this resource-limited setting.
This study demonstrates that it is possible conserve
future ART options by maintaining the majority of
individuals on first-line regimens.
Catherine Orrell, Linda Gail-Bekker and Robin Wood
receive partial support from the NIH through a
CIPRA-SA grant. Guy Harling receives support from
the City Bridge Foundation. Stephen Lawn is funded by
the Wellcome Trust, London, UK.
1. Egger M, May M, Chene G, et al. Prognosis of HIV-1
infected patients starting highly active antiretroviral
therapy: a collaborative analysis of prospective studies.
Lancet 2002; 360:119–129.
Hogg RS, Yip B, Chan KJ, et al. Rates of disease progres-
sion by baseline CD4 count and viral load change after
initiating triple drug therapy. JAMA 2001; 286:2568–2577.
Laurent C, Gueye NFN, Ndour CT, et al. Long term bene-
fits of highly active antiretroviral therapy in Senegalese
HIV-1 infected adults. J Acquir Immune Defic Syndr 2005;
Coetzee D, Hildebrand K, Boulle A, et al. Outcomes after
two years of providing antiretroviral therapy in
Khayelitsha, South Africa. AIDS 2004; 18:887–895.
Kumarasamy N, Solomon S, Chaguturu SK, et al. The
safety tolerability and effectiveness of generic antiretroviral
drug regimens for HIV-infected patients in South India.
AIDS 2003; 17:2267–2269.
Severe P, Leger P, Charles M, et al. Antiretroviral therapy in
a thousand patients with AIDS in Haiti. N Engl J Med
Wools-Kaloustian K, Kimaiyo S, Diero L, et al. Viability
and effectiveness of large scale HIV treatment initiatives in
sub-Saharan Africa: experience from Western Kenya. AIDS
Tassie JM, Szumilin E, Calmy A, Goemaere E. Highly
active antiretroviral therapy in resource poor settings: the
experience of Médecins Sans Frontières. AIDS 2003,
Djomand G, Roels T, Ellerbrock T, et al. Virologic and
immunologic outcomes and programmatic challenges of an
antiretroviral treatment pilot project in Abidjan, Côte
d’Ivoire. AIDS 2003; 17(Suppl 3):S5–S15.
10. Ivers LC, Kendrick D, Doucette K. Efficacy of antiretro-
viral therapy programs in resource poor settings: a
meta-analysis of the published literature. Clin Infect Dis
11. Bekker L-G, Myer L, Orrell C, Lawn SD, Wood R. Rapid
scale up of community-based HIV treatment service:
programme performances over 3 consecutive years in
Guglethu, South Africa. S Afr Med J 2006; 96:315–320.
12. South African National Antiretroviral Treatment
Guidelines, National Department of Health, South Africa.
First edition, 2004 [cited 2007 Jan 25]. Available from:
13. World Health Organisation. Scaling Up Antiretroviral
Therapy in Resource Limited Settings – Guidelines for a
Public Health Approach. Geneva: WHO; 2002.
14. Mandalia S, Parmar D, Fisher M, et al. Naively changing
HAART. HIV Med 2002; 3:254–262.
15. Sabin C, Hill T, Lampe F, et al. Treatment exhaustion of
highly active antiretroviral therapy (HAART) among indi-
viduals infected with HIV in the United Kingdom: a
multicentre cohort study. BMJ 2005; 330:695–699.
16. Lawn SD, Myer L, Orrell C, Bekker LG, Wood R. Early
mortality among patients accessing a community-based
antiretroviral programme in South Africa: implications for
programme design. AIDS 2005; 19:2141–2148.
17. Lawn SD, Myer L, Harling G, Orrell C, Bekker L-G, Wood
R. Determinants of mortality and nondeath losses from an
antiretroviral treatment service in South Africa: implications
for program evaluation. Clin Infect Dis 2006; 43:770–776.
18. Antiretroviral Treatment Protocol. Provincial
Administration of the Western Cape, South Africa. 2nd
Edition, 2004 [cited 2006 August 9]. Available from:
19. Martinez E. The reassurance of experience. Int J STD
AIDS 2003; 14(Suppl 1):20–28.
20. Masquelier B, Pereira E, Peytavin G, et al. Intermittent
viraemia during first line, protease inhibitor containing
therapy: significance and relationship with drug resistance.
J Clin Virol 2005; 33:75–78.
21. Røge BT, Barfod TS, Kirk O, et al. Resistance profiles and
adherence at primary virological failure in three different
highly active antiretroviral therapy regimens: analysis of
failure rates in a randomised study. HIV Med 2004;
22. Nettles N, Kieffer T, Kwon P, et al. Intermittent HIV-1
viraemia (blips) and drug resistance in patients receiving
HAART. JAMA 2005; 293:817–829.
23. Mocroft A, Ruiz L, Reiss P, et al. Virological rebound after
suppression on highly active antiretroviral therapy. AIDS
24. Orrell C, Badri M, Wood R. Measuring adherence in a
community setting: which measure most valuable? 15th
International AIDS Conference. 11–16 July 2004,
Bangkok, Thailand. Abstract WePEB57.
25. Weidle P, Malamba S, Mwebaze R, et al. Assessment of a
pilot antiretroviral drug therapy programme in Uganda:
patient’s response, survival and drug resistance. Lancet
26. Oosterhout J, Bodasing N, Kumwenda J, et al. Evaluation
of antiretroviral therapy results in a resource setting in
Blantyre, Malawi. Trop Med Int Health 2005; 10:464–470.
27. Colebunders R, Moses KR, Laurence J, et al. A new model
to monitor the virological efficacy of antiretroviral treat-
ment in resource poor countries. Lancet Infect Dis 2006;
28. Lawn SD, Orrell C, Wood R. Evaluating a model for moni-
toring the virological efficacy of antiretroviral treatment in
resource-limited settings. Lancet Infect Dis 2006;
29. Smith-Rohrberg D, Basu S, Subbaraman R, Sharma SK.
Compromising comprehensive AIDS management will lead
to failure. Lancet Infect Dis 2006; 6:384–385.
Accepted for publication 13 September 2006
C Orrell et al.
© 2007 International Medical Press