Optimal strategies for monitoring response to
antiretroviral therapy in HIV-infected adults,
adolescents, children and pregnant women:
a systematic review
Joseph D. Tuckera,b,c,M, Cedric H. Biena,M, Philippa J. Easterbrookd,M,
Meg C. Dohertyd, Martina Penazzatod, Marco Vitoriad and
Rosanna W. Peelingc
Objective: The objective of this review was to examine different monitoring strategies
(clinical, immunologic (CD4þ T cell count measurement) and virologic (viral load
measurement)) to inform revision of the 2013 WHO guidelines for antiretroviral therapy
(ART) in low and middle-income countries.
Design: A systematic review.
Methods: We searched 10 databases, reference lists of included research studies and
contacted experts in an attempt to identify all relevant studies regardless of language or
publication status. We included both randomized controlled trials (RCTs) and observational studies. We selected studies that examined routine clinical monitoring (CM),
immunologic monitoring (IM) or virologic monitoring (VM). CM involved clinical
evaluation and basic laboratory blood testing without CD4þ T cell count or viral load.
Two authors independently assessed study eligibility, extracted data and graded
methodological quality.
Results: A total of six studies were identified, including five RCTs and one observational
study. Two RCTs among adults found an increased risk of AIDS-defining illness and
mortality in CM compared to CM þ IM. Two studies compared CM þ IM to
CM þ IM þ VM, with one finding a mortality advantage in the CM þ IM þ VM group.
Duration of viremia and time to switching to a second-line regimen were longer in
CM þ IM compared to CM þ IM þ VM. Only one trial was conducted in children, and
showed no difference in mortality comparing CM and CMþIM. No studies specifically
studied pregnant women.
Conclusion: CM þ IM was shown to be beneficial in terms of a combined mortality and
morbidity endpoint compared to CM alone. VM was associated with shorter duration of
viremia and higher rates of switching, but an impact on mortality was not consistently
shown. Pooled outcome estimates were possible with comparison of only CM to
CM þ IM. Further HIV research on different VL monitoring strategies is required. These
data support the recommendation in the 2013 WHO ART guidelines for the use of VM to
confirm and diagnose ART failure, and for the use of IM þ CM when VM is not available.
ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
AIDS 2014, 28 (Suppl 2):S151–S160
Keywords: antiretroviral therapy, HIV, immunologic, monitoring, systematic
review, virologic, WHO
a
University of North Carolina Project-China, Guangzhou, China, bInstitute of Global Health and Infectious Diseases, University of
North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA, cLondon School of Hygiene and Tropical Medicine, London, UK,
and dHIV Department, World Health Organization, Geneva, Switzerland.
Correspondence to Joseph D. Tucker, MD, MA, PhD, UNC Project-China, Number 2 Lujing Road, Guangzhou 510095, China.
E-mail: [email protected]
Joseph D. Tucker, Cedric H. Bien and Philippa J. Easterbrook contributed equally to this manuscript.
DOI:110.1097/QAD.0000000000000230
ISSN 0269-9370 Q 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
S151
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
S152
AIDS
2014, Vol 28 (Suppl 2)
Introduction
In the past two years, there has been a 60% increase in the
number of individuals taking antiretroviral therapy (ART)
and there are now 9.7 million individuals on ART
worldwide [1]. Over two-thirds of these individuals are
in low and middle-income countries. This has been
achieved through scaling up systems for delivering HIV
ART at the population level based on a public health
approach of simplified and standardized treatment and
monitoring in accordance with WHO guidelines [2,3].
Yet, this progress has also prompted the need to reevaluate optimal strategies for routine monitoring
of ART response and detection of treatment failure.
Most HIV monitoring guidelines in low-income settings
define treatment failure and ART switching criteria
according to clinical monitoring with or without
immunologic monitoring (CD4þ T cell count measurement), consistent with the 2006 [4,5] and 2010 [6,7]
WHO guidelines. In contrast, routine virologic monitoring (viral load measurement every 3–6 months) has been
the standard of care for detection of treatment failure and
the basis for switching ART regimens in many middle
and high-income nations for over a decade [8].
There are concerns with the use of clinical and
immunologic monitoring alone. Multiple studies,
including a systematic review undertaken to inform
the development of the 2013 WHO guidelines have
highlighted the poor positive predictive value of the 2010
WHO clinical and immunologic criteria to identify
virologic failure [9,10]. This contributes to delayed
switching to second-line regimens as well as misdiagnosis
of treatment failure leading to unnecessary switching.
Virologic monitoring also offers the potential to assess
early ART response, thereby providing an opportunity to
improve adherence [11]. Furthermore, technological
developments in point-of-care immunologic and virologic monitoring allow decentralized monitoring in
many locations where monitoring was not feasible in the
past [12].
In 2012, the WHO started the process of updating
guidelines on the use of ART for treatment and
prevention of HIV infection in adults, pregnant women,
adolescents and children [13]. This systematic review was
undertaken to better understand optimal monitoring
strategies for HIV-infected individuals. Our review
focused on evaluating the medical and public health
benefit of different HIV monitoring strategies without
considering cost or availability of monitoring tools.
We examined studies that compared different strategies
for routine monitoring of ART response, including
clinical monitoring (CM), immunologic monitoring
(IM), and virologic monitoring (VM) among all available
populations, including HIV-infected adults, pregnant
women, adolescents and children. The objective of
this review is to update evidence on ART monitoring
strategies to inform revision of the 2013 WHO guidelines
for ART in low and middle-income countries.
Materials and methods
We examined studies that compared CM, IM and VM
(or combinations of these strategies) among adults,
adolescents, children and pregnant women (Appendix
1, http://links.lww.com/QAD/A500). CM was defined
as clinical evaluation (medical history and physical
examination by a trained healthcare professional) that
may include basic laboratory testing, but not CD4þ T cell
count or viral load measurements, usually at 3–6 month
intervals. IM was defined as measurement of absolute
CD4þ T cell count values for adults or CD4þ T cell
count percentage measurement for children, usually every
6 months. VM referred to routine measurement of plasma
viral load at 6–12 month intervals, but not the use of
confirmatory viral loads in those with evidence of clinical and/or immunologic failure. Outcomes examined
included morbidity and mortality, drug resistance, switch
rates to second-line ART and adherence. We did not
examine strategies for monitoring ART drug toxicities
because these are largely ART-specific and have been
described elsewhere [14–16]. Studies that examined
only different frequencies of monitoring or thresholds
for switching to second-line ART, optimal frequency
of monitoring or cost-effectiveness were excluded.
The review was reported following PRISMA and
AMSTAR guidelines and registered in PROSPERO
(CRD42013004358).
Search strategy and identification of studies
No date, age, geographic or language restrictions were
used in this search, and we included all citations identified
through 20 April 2013. We searched the following
electronic journal and trial databases: Pubmed, EMBASE,
Cochrane Central, Sciverse Scopus, NLM Gateway
(for HIV/AIDS conference abstracts before 2005),
Conference on Retroviruses and Opportunistic Infections, International AIDS Conference and International
AIDS Society Conference on HIV Pathogenesis, Treatment, and Prevention from 2005 to 2012, Web of Science
Conference Proceedings since 1990, Clinicaltrials.gov,
PsycINFO and Current Controlled Trials. We contacted
researchers and relevant organizations and screened
reference lists of all included studies to identify additional
eligible studies. Search terms are included in Appendix 1,
http://links.lww.com/QAD/A500.
Study selection and data abstraction
Two reviewers (J.T. and C.B.) independently screened
all citations and undertook independent data abstraction
using a standardized form (Fig. 1). Discrepancies between
abstracted data accounted for less than 3% of the total data
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
HIV monitoring systematic review Tucker et al.
and were resolved by a third independent individual.
Authors of all five RCTs were contacted for more
information about items not reported in the manuscript
and three replied. Full research articles, brief reports,
abstracts and descriptions of clinical trials evaluating
HIV monitoring were included. We selected studies that
compared combined monitoring strategies of CM, IM or
VM (Appendix 2, http://links.lww.com/QAD/A500).
Quality assessment
Two forms of quality assessment were undertaken to
identify the strengths and weaknesses of the research
studies. First, in accordance with the WHO Guidelines
Review Committee guidance and the Cochrane group,
the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) approach was used
to summarize the quality of the evidence [17,18].
GRADE classifies the quality of evidence into four
categories – high quality, moderate quality, low quality
and very low quality – on the basis of consideration of
study attributes. Evidence from RCTs starts as high but
can be downgraded based on study limitations, including
inconsistency of results, indirectness of evidence, imprecision, risk of bias including publication bias. Evidence
S153
from observational studies starts at low but can be
upgraded if the effect size is very large, or if all possible
confounders would decrease the magnitude of effect of
the intervention [18]. Second, recognizing the limitations
of the GRADE approach, we also examined the quality
of RCTs and observational studies according to
more extensive checklists that have been designed to
evaluate quality of reporting from these types of studies.
The CONSORT checklist [19] evaluated RCT manuscripts based on 25 items. The STROBE checklist [20]
evaluated observational studies based on 22 items within
similar categories. Potential conflicts of interest were
also examined based on reported funding sources.
Data analysis
For each comparison of monitoring strategies, we
reported estimated relative risks (RR) and the 95%
confidence interval (CI) for all key outcomes for each
of the published studies. We also examined the
appropriateness of pooling research findings. Two studies
compared the same monitoring strategies and reported
comparable mortality outcomes, and these outcomes
were therefore pooled using a fixed effects model
to generate a summary effect size [21], in addition to
6125 citations identified
8 studies
identified through
references
2343 from PubMed
1033 from EMBASE
160 from Cochrane Central
832 from Sciverse Scopus
318 from Gateway NLM
799 from Web of Science
201 from ClinicalTrials.gov
302 from Clinicaltrials.com
137 from PsycINFO
1257 duplicates removed
4868 citations excluded
84 abstracts screened
44 abstracts excluded:
Interventions did not compare
monitoring strategies (n = 12)
Outcomes were non-empirical (n = 16)
Studies compared monitoring
algorithms (n = 7)
24 full texts examined
6 articles included in qualitative synthesis
18 full texts excluded:
Studies assessed CD4 monitoring
and VL confirmation of treatment
failure (n = 14)
Studies without clear comparison
groups (n = 4)
--
6 studies included in quantitative synthesis
Fig. 1. Flow chart of study selection process.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
S154
AIDS
2014, Vol 28 (Suppl 2)
reporting the outcomes separately for each study. Other
study results were not pooled because of differences in
monitoring strategies compared or outcomes, or insufficient data. Publication bias was not formally assessed
because there were fewer than 10 included studies.
mortality in the CM group with a pooled hazard ratio
estimate of 1.36 (95% CI 1.14–1.64). One adult RCT
examining switch rates to a second-line regimen found a
higher switch rate to second-line ART of 22% in the
CM þ IM group at 5 years than 19% in the CM group
(P ¼ 0.03) [44].
Results
The only RCT among children showed no difference
between CM compared to CM þ IM in terms of a
combined endpoint of WHO stage four events or
mortality (hazard ratio 1.13, 95% CI 0.73–1.73, P ¼
0.59) [43]. However, in year one, WHO stage four events
or mortality were lower in the CM group (P ¼ 0.20).
During years 2–5, WHO stage 4 events or mortality were
higher in the CM group (P ¼ 0.002).
Our search strategy identified 4138 citations (Fig. 1).
After removing duplicates and titles not relevant to ART
response monitoring, a total of 68 abstracts were further
screened. Twenty-four full-text published studies were
screened and 18 studies were excluded [9,22–38].
Six studies were included in the final review [39–44].
Description of included studies
Table 1 summarizes the characteristics of the included
studies. Five studies were RCTs [39,41–44] and one
study was an observational study [40]. Five studies were
undertaken in Africa [40–44] and one in Asia (Table 1)
[39]. Three studies were undertaken in multiple countries
[40,43,44] and three in a single country [39,41,42]. The
observational study included 99 643 individuals [40]
and the five RCTs ranged in size from 459 [41] to
3321 participants [44]. Five studies included only adults
[39–42,44] and one examined children and adolescents
[43]. Three studies specifically excluded pregnant women
[41,43,44] (Table 3). The median CD4þ T cell count
prior to initiation of monitoring was less than 300 cells/ml
in all studies. Median follow-up time ranged from 2
to 4.9 years. Two studies only included ART-naive
individuals [40,41] (Table 1). Studies compared the
following treatment response monitoring strategies: CM;
CM þ IM; CM þ VM; CM þ IM þ VM. All studies
examined mortality and four studies included WHO
stage 3 and 4 events [41–44].
Comparison of clinical monitoring and clinical
monitoring together with immunologic
monitoring
Three RCTs compared CM with CM þ IM (Table 2)
[42–44]. Both RCTs in adults found a significant
advantage of CM þ IM compared to CM when
examining a combined morbidity and mortality endpoint
(6.0 events per 100 person-years in the CM þ IM group
compared with 7.6 events per 100 person-years in the
CM group in one RCT, [42]; 5.2 events per 100 personyears in the CM þ IM group compared with 6.9 events
per 100 person-years in the CM group in another RCT
[44]). Among the adult RCTs, both found a mortality
benefit in the CM þ IM group. One study found
significantly increased mortality in the CM group (hazard
ratio 1.35, 95% CI 1.10–1.65, P < 0.001) [44] and one
found a nonsignificant trend in the same direction (hazard
ratio 1.43, 95% CI 0.92–2.21, P ¼ 0.10) (Table 3) [42].
Pooling the mortality outcome for these two studies
comparing CM and CM þ IM groups showed increased
Comparison of clinical monitoring and clinical
monitoring along with immunologic monitoring
and virologic monitoring
One RCT compared CM þ IM þ VM to CM [41]
(Table 2). This study found no difference between CM
and CM þ IM þ VM in terms of combined mortality/
morbidity endpoint (26.6 events per 100 person-years
in the CM group compared with 19.9 events per 100
person-years in the CM þ IM þ VM group, P ¼ 0.25) or
mortality alone (11.5 deaths per 100 person-years in
the CM group compared with 8.8 deaths per 100 personyears in the CM þ IM þ VM group; P ¼ 0.11) [41].
The hazard ratio comparing CM to CM þ IM þ VM
was 1.31 (95% CI 0.83–2.06). The CM þ IM þ VM arm
had significantly higher switch rates to second-line
ART (6 compared with 0%), but the frequency of drug
resistance mutations was similar at 10% [41].
Comparison of clinical monitoring along with
immunologic monitoring and clinical monitoring
along with immunologic monitoring and
virologic monitoring
Two studies compared CM þ IM to CM þ IM þ VM
[40,42], including one RCT [42] and one observational
study [40] (Table 2). The RCT showed no difference
between the two groups in terms of mortality, although
the observational study [40] found a 3-year mortality rate
of 6.3% (95% CI 6.0–6.5) in the CM þ IM group and
4.3% (95% CI 3.9–4.8) in the CM þ IM þ VM group
(P < 0.001). Switch rates to second-line ART regimens at
the end of year 3 were also higher in the CM þ IM þ VM
group at 9.8% (95% CI 9.1–10.5) [40] compared with
2.1% (95% CI 2.0–2.3) in the CM þ IM group. None of
these studies directly measured differences in adherence
or onward HIV transmission.
Comparison of clinical monitoring along with
immunologic monitoring and clinical monitoring
along with virologic monitoring
One adult RCT compared CM þ IM with CM þ VM
[39], but this study was only available in abstract form
(Table 2). This study found no difference in either
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Thailand
716
99 643
3321
459
1094
1206
Jourdain
et al.a [39]
Keiser
et al. [40]
DART [44]
Laurent
et al. [41]
Mermin
et al. [42]
ARROW [43]
12.5%, 12.0%
129
Adults with CD4þ T cell count <250 cells/ml
or WHO stage 3 or 4 disease, 67%
women, 18 years and older, clients with
CD4þ T cell count less than 250 cells/ml or
severe HIV diseasec, AST less than five
times the upper limit of normal, creatinine
clearance greater than 0.42 ml/s,
Karnofsky score greater than 40%, no prior
use of nevirapine for PMTCT
Aged 3 months to 17 years eligible for ART,
51% female, age 3 months to 17 years old
who met WHO 2006 criteria for ART
initiation, without acute infection or
taking contraindicated drugs or adherence
issues or laboratory abnormalities that
contraindicate ART or pregnant or
breastfeeding
179, 182
86
Adults with CD4þ T cell count <200 cells/ml,
64% women, able to and likely to attend
follow-up, likelihood of good compliance,
no acute infection or tuberculosis therapy
ongoing, no chemotherapy for
malignancy, not pregnant or
breastfeeding, laboratory values allowed
initiation of ART
Adults, 70% women, 18 years old or older
with confirmed HIV-1 group M infection,
WHO clinical stage 3 or 4 or 2 with total
lymphocyte count of less than
1200 cells/ml, likely to attend follow-up
appointments, without active tuberculosis
or malignancy or psychiatric disorders or
hepatocellular insufficiency or previous
ART or use of steroids or other
experimental drugs, not pregnant
132, 93
144
Adults, 65% women, ART-naive patients
aged 16 years and older who started ART
with an NNRTI-based regimen and who
had at least 1 day of follow-up
Adults, 62% women, CD4þ T cell count
50–250 cells/ml, not hepatitis B or C
coinfected
Population
Median CD4þ T cell
count (cells/ml in adults
and atl % in children)b
5.9 (2.2–9.2)
39 (32–44)
36 (30–44)
36 (18–73)
34 (30–41)
36 (31–41)
Median age
in years (IQR)
RCT
RCT
RCT
RCT
Observational
RCT
Study design
CM
CM
CM
CM
CM þ IM
CM þ IM
Control
arm
CM þ IM
CM þ IM or
CM þ IM þ VM
CM þ IM þ VM
CM þ IM
CM þ IM þ VM
CM þ VM
Intervention
arm(s)
4.0
3
2
4.9
3
NR
Follow-up
(median, years)
CM, clinical monitoring; FU, follow-up time; IM, immunologic monitoring; NR, not reported; VM, virologic monitoring.
a
Only abstract available.
b
Three studies reported overall median CD4þ T cell count and three studies reported median CD4þ T cell count by study arm. All CD4þ T cell counts were at baseline prior to assignment except for
Keiser et al. [40].
Uganda and
Zimbabwe
Uganda
Cameroon
Uganda and
Zimbabwe
South Africa,
Malawi,
Zambia
Location
First author
Sample
size
Table 1. Characteristics of antiretroviral therapy response monitoring studies.
HIV monitoring systematic review Tucker et al.
S155
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
S156
AIDS
2014, Vol 28 (Suppl 2)
Table 2. Main study outcomes for antiretroviral response monitoring studies (for entire study duration unless otherwise noted).
Study
Study Arm
New WHO stage
3–4 events
or deatha
Jourdain
et al. [39]e
CM þ IM
NR
CM þ VM
Keiser
et al. [40]
CM þ IM
NR
CM þ IM þ
VM
Mortality (%)b
Switch rates to
second-line ART
Drug
resistancec
Viremia
>400 copies/mld
3.4% (95% CI 2.0–6.0)
at 3 years
4.3% (95% CI 2.6–7.0)
at 3 yrs
7.2% (95% CI 4.9–10.5)
at 3 years
5.1% (95% CI 3.2–8.0%)
at 3 years
1 case
15.8 m
(IQR 8.5–20.4)
7.2 m
(IQR 5.8–8.0)
6.3% (95% CI 6.0–6.5)
2.1% (95% CI 2.0–2.3)
at 3 years
9.8% (95% CI 9.1–10.5)
at 3 years
NR
NR
4.3% (95% CI 3.9–4.8)
No cases
DART [44]
CM
CM þ IM
6.9 (95% CI 6.3–7.6)
5.2 (95% CI 4.7–5.8)
2.9 (95% CI 2.6–3.4)
2.2 (95% CI 1.9–2.5)
19% at 5 years
22% at 5 years
NR
NR
Laurent
et al. [41]
CM
26.6
11.5
0
10%
NR
CM þ IM þ
VM
19.9
8.8
6%
10%
CM
7.6
4.9
11% at 3 years
NR
CM þ IM
CM þ IM þ
VM
6.0
4.8
4.0
3.7
15% at 3 years
44% at 3 years
CM
2.0
1.1
5%
CM þ IM
1.7
1.3
6%
Mermin
et al. [42]
ARROW
[43]
5.4%
7.5%
4.6%
NRf
23%
22%
CM, clinical monitoring; IM, immunologic monitoring; m, months; NR, not reported; VM, virologic monitoring.
a
WHO stage 3–4 events or death reported as events per 100 person-years.
b
DART, Laurent et al. [42], Mermin et al. [42] and ARROW reported mortality as deaths per 100 person-years; Jourdain et al. [39] and Keiser et al.
[40] reported mortality as percentages.
c
Drug resistance defined as multiple thymidine analogue mutations. (D67N/M41L/L210W/T215Y) in Jourdain study and major mutations
conferring resistance in the Laurent study.
d
Viremia greater than 500 copies/ml reported in Mermin et al. [42].
e
Only abstract available.
f
Resistance analysis was included, but not described in the initial manuscript.
mortality (3.4% CM þ IM vs. 4.3% CM þ VM, P ¼ 0.57)
or a composite endpoint of mortality, new AIDS-defining
event, or CD4þ T cell count of 50 cells/ml or less at
3 years (7.4% CM þ IM vs. 8.0% CM þ VM). There was
a trend towards higher switch rates in the CM þ IM
group (7.2% vs. 5.1%, P ¼ 0.10). The median duration
of viremia of more than 400 copies/ml was twice as
long at 15.8 months in the CM þ IM group compared
with 7.2 months in the CM þ VM group (P ¼ 0.002).
Quality assessment
All four RCTs with full texts available [41–44] were rated
as of good quality on the basis of the CONSORT criteria,
fulfilling 94–97% of 34 key indicators (Appendix 3,
http://links.lww.com/QAD/A500). Minor limitations
were that three out of four studies did not explicitly state
why the study was stopped and none of the articles
had publicly available links to the full trial protocol. The
RCT abstract [39] achieved 12/17 (71%) of CONSORT
abstract indicators (Appendix 4, http://links.lww.com/
QAD/A500) and the observational study [40] achieved
24/30 (80%) of STROBE indicators (Appendix 5, http://
links.lww.com/QAD/A500). On the basis of GRADE
quality of evidence assessment, the observational study
[40] was ranked as low and each of the RCT studies
[39,41–44] as moderate (Appendices 6–8, http://links.
lww.com/QAD/A500). All of the studies described
the research funders and none of the research studies
noted support from manufacturers of CD4þ T cell count
or viral load measurement technology.
Discussion
Understanding optimal strategies for HIV monitoring
is critical as health systems expand routine delivery of
ART worldwide. Our review extends a previous
Cochrane systematic review [45], which only examined
data from two RCTs in abstract form. Our review
benefited from each of these two RCTs being available
as published manuscripts [42,44] alongside recently
available data from four other studies. Our review found
that ART monitoring strategies that included CM along
with IM had better outcomes than strategies that only
provided CM, drawing on RCT data from both adults
and children. The addition of routine VM resulted in
some additional benefit. Although limitations in the data
precluded a formal meta-analysis, this review provides an
evidence base informing recommendations about routine
monitoring of HIV-infected individuals on ART.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
NR
0% CM; 6% CM þ IM þ VM
11% at 3 years CM; 44% at
3 years CM þ IM þ VM
1.30 (0.94–1.80)
1.83 (1.25–2.69)
CM, clinical monitoring; IM, immunologic monitoring; NR, not reported; VM, virologic monitoring.
a
DART, Laurent et al. [41], Mermin et al. [42] and ARROW reported mortality as deaths per 100 person-years; Keiser et al. [40] reported mortality as percentage died.
b
Reference group in each of these comparisons is the group with less monitoring.
1.57 (1.00–2.46)
26.6 CM; 19.9
CM þ IM þ VM
7.6 CM; 4.8
CM þ IM þ VM
1.31 (0.83–2.06)
11.5 CM; 8.8 CM þ
IM þ VM
4.9 CM; 3.7 CM þ
IM þ VM
Laurent
et al. [41]
Mermin
et al. [42]
CM vs. CM þ
IM þ VM
Mermin
et al. [42]
NR
NR
15% at 3 years CM þ IM;
44% at 3 years CM þ
IM þ VM
1.23 (0.82–1.84)
6.0 CM þ IM 4.8
CM þ IM þ VM
NR
NR
1.36 (1.14–1.64)
1.72 (1.52–2.00)
CM þ IM vs.
CM þ IM þ VM
Keiser
et al. [40]
6.3% (95% CI 6.0–6.5)
CM þ IM; 4.3%
(95% CI 3.9–4.8)
4.0 CM þ IM; 3.7 CM þ
IM þ VM
1.10 (0.69–1.75)
NR
11% at 3 years CM; 15% at
3 years CM þ IM
1.49 (1.03–2.13)
1.43 (0.92–2.21)
Mermin
et al. [42]
Pooled
DART [44]
CM vs. CM þ IM
2.9 (95% CI 2.6–3.4) CM;
2.2 (95% CI 1.9–2.5)
CM þ IM
4.9 CM at 3 years;
4.0 CM þ IM
1.35 (1.10–1.65)
6.9 (95% CI 6.3–7.6)
CM; 5.2 (95% CI
4.7–5.8) CM þ IM
7.6 CM; 6.0 CM þ IM
NR
NR
19% at 5 years CM; 22% at
5 years CM þ IM
1.31 (1.14–1.51)
1.06 (0.91–1.23)
Switch rates to
second-line ART
Morbidity or mortality
hazard Ratio and 95% CI
New WHO stage 3–4
events or deatha
Mortality Hazard
Ratio and 95% CIb
Mortality (%)a
Study
Study
Table 3. Adult mortality and switch rate comparison for antiretroviral therapy response arranged by Comparison Group.
Switch rate hazard
ratio and 95% CI
HIV monitoring systematic review Tucker et al.
S157
CM þ IM was shown to be beneficial in terms of
both mortality and a combined mortality/morbidity
endpoint compared with CM based on data from
the three RCTs included in this systematic review,
including one in children [42–44]. This finding supports
the existing 2010 WHO recommendations for routine
CM þ IM [6,7] and recommendations from the 2013
WHO guidelines pertaining to settings without ready
access to viral load monitoring technology [2]. The
reduced mortality in the CM þ IM group was likely
due to earlier switching to second-line regimens
before progression to advanced immunodeficiency and
associated risk of life-threatening opportunistic infections. One RCT included in this review, the DART trial,
estimated that 130 HIV-infected individuals would need
to be monitored using CM þ IM for 1 year to prevent
one death [44], although the number needed to monitor
would decrease with increasing duration of monitoring.
This review also supports the use of routine VM in
addition to CM þ IM, although current data have
not consistently shown a clear impact on mortality.
One RCT demonstrated a longer duration of viremia
in those without VM [39]. Other studies not included
in this review have suggested that earlier switching,
especially in cohorts with protease inhibitor sparing ART
regimens, has clinical benefit [46–48]. However, only the
observational study of our three eligible studies found a
mortality benefit at 3 years at ART sites that used VM
[40]. The two RCTs with comparable follow-up time did
not find a mortality difference between these two groups
[41,42]. One of these RCTs [41] specifically excluded
individuals with potential or proven poor follow-up, who
would be at an increased risk for poor ART adherence
and virologic rebound. In addition, the duration of
follow-up was less than 3 years in most studies. A recent
systematic review of 16 studies also used as part of the
evidence base for the 2013 guidelines showed that
the 2010 WHO clinical and immunologic criteria have
low positive predictive value for identifying virologic
failure [10]. Two additional potential advantages of VM
were not addressed by the available literature: improving
adherence [11] and decreasing onward HIV transmission.
This review contributed to the evidence base for the
revised 2013 WHO guidelines on ART monitoring.
On the basis of the collective findings of this systematic review and another review [10], the 2013 WHO
guidelines made a strong recommendation for use of
VM as the preferred monitoring approach to both
diagnose and confirm ART failure (evidence ranked as
of low quality based on GRADE) [2]. The WHO also
recommended that when VM is not available, CM þ IM
should be used to diagnose treatment failure (strong
recommendation, moderate quality evidence) [2].
Our review identified a number of ART monitoring
implementation issues that are likely to be important for
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
S158
AIDS
2014, Vol 28 (Suppl 2)
scaling up ART monitoring worldwide. Expanding
ART monitoring access, ensuring quality control systems
and strengthening health systems are all linked to the
overall success or failure of HIV monitoring programmes.
First, availability of CM and VM in low and middleincome countries is uneven [49], despite evidence that
decentralized community-based HIV testing is associated
with better access and outcomes [50]. Expanding HIV
monitoring services at clinical and nonclinical settings
may increase the total number of HIV-infected individuals who receive appropriate ART monitoring [51].
Novel financing mechanisms such as UNITAID alongside advances in point-of-care technology can accelerate
this process [12]. Second, comprehensive quality control
systems are also necessary for the implementation of HIV
monitoring technologies [52]. Several quality assessment
projects have been undertaken [53–55], but none have
been implemented on a large scale. Ensuring quality
control will be especially important as HIV monitoring
becomes decentralized and implemented at lower
level healthcare facilities with limited capacity. Finally,
monitoring technology alone does not guarantee
appropriate clinical action resulting from diagnosis of
virologic failure [49]. Local staff training and responsive
health systems are necessary to reap the benefits of
enhanced monitoring [56,57].
Our study has several limitations. First, most of the studies
had follow-up of less than 3 years, and longer follow up
will be necessary to examine the impact on mortality
and development of drug resistance. Second, a pooled
estimate of mortality was only done for one comparison
of CM compared with CM þ IM. For all the other
monitoring strategy comparison studies, there were
substantive differences between studies that precluded
pooling of outcome measures. One key trial [39] was only
available in abstract form during our data analysis.
In addition, HIV drug resistance data were only reported
in one published manuscript [41], although this is
important to understand the limitations of routine monitoring without viral load measurement. Third, costeffectiveness considerations are important with respect to
implementation and scale up in low and middle-income
countries, but were not the focus of this review. However,
cost-effectiveness analyses of routine HIV treatment
monitoring strategies have not been conclusive [58,59].
Fourth, most of the studies were conducted in HIVinfected adults. There were no studies specifically among
pregnant women, and only one RCT among children.
Finally, although we assessed each study for bias, we
were not able to formally evaluate for publication bias
because of the limited number of studies. But given
the complexity and cost of organizing HIV monitoring
research studies, we doubt that unpublished studies exist
that would compromise the review conclusions.
Our review highlights a number of HIV monitoring
research priorities moving forward. In light of the WHO
recommendations for earlier initiation of ART, the
need for scale-up of viral load monitoring relative
to CD4þ T cell count monitoring is likely to increase,
which needs to be informed by timely implementation research. Evaluation of monitoring strategies in the
context of ART initiation at higher CD4þ T cell counts,
especially among pregnant women initiating lifelong
ART (option Bþ), will be particularly important. Further
data on using viral load monitoring as a tool to support
adherence among subsets of key populations at risk for
poor follow-up are also needed.
Acknowledgements
Preliminary findings from this research were presented
at the WHO’s HIV Therapy Guidelines Committee
meeting in December 2012 and subsequently at the 7th
IAS Conference on HIV Pathogenesis, Treatment and
Prevention in Kuala Lumpur, Malaysia, in July 2013. This
research was supported by the WHO (APW 200609817),
an NIH Fogarty Career Development Award (US NIH
1K01TW008200-01A1), the Doris Duke Medical
Foundation and the UNC Center for AIDS Research
(NIAID 5P30AI050410-13). We would like to thank
Larry Han for data extraction and Lei Zhang and Eric
Chow for assistance with pooling results. Thanks to
UNC Project-China for administrative support.
JDT, CHB, and PE were the lead authors. JDT and RWP
conceptualized the study. JDT and CHB scrutinized
identified studies for eligibility, extracted data and assessed
the methodological quality of included studies. MD,
MP, MA, and RWP provided input on the study design.
All authors critically reviewed the manuscript before
submission.
Funding for this study was provided by World Health
Organization, US National Institutes of Health.
Conflicts of interest
None.
References
1. UNAIDS. Report on the global AIDS epidemic 2013. Geneva,
Switzerland: UNAIDS; 2013.
2. World Health Organization. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public
health approach [2010 revision]. Geneva: World Health Organization; 2013, http://whqlibdoc.who.int/publications/2010/
9789241599764_eng.pdf. [Accessed 26 January 2014]
3. Gilks CF, Crowley S, Ekpini R, et al. The WHO public-health
approach to antiretroviral treatment against HIV in resource
limited settings. Lancet 2006; 368:505–510.
4. World Health Organization. Antiretroviral therapy for HIV
infection in adults and adolescents: recommendations for a
public health approach. [2006 revision]. Geneva: World Health
Organization; 2006, http://www.who.int/hiv/pub/guidelines/
artadultguidelines.pdf. [Accessed 26 January 2014]
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
HIV monitoring systematic review Tucker et al.
5. World Health Organization. Antiretroviral therapy for HIV
infection in infants and children towards universal access:
recommendations for a public health approach. [2006 revision].
Geneva: World Health Organization; 2006, http://www.
who.int/hiv/pub/guidelines/paediatric020907.pdf. [Accessed
17 May 2013]
6. World Health Organization. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public
health approach [2010 revision]. Geneva: World Health Organization; 2010, http://whqlibdoc.who.int/publications/2010/
9789241599764_eng.pdf. [Accessed 26 January 2014]
7. World Health Organization. Antiretroviral therapy for HIV
infection in infants and children: towards universal access.
Recommendations for a public health approach [2010 revision].
Geneva: World Health Organization; 2010, http://whqlibdoc.
who.int/publications/2010/9789241599801_eng.pdf. [Accessed
26 January 2014]
8. Walker AS, Gibb DM. Monitoring of highly active antiretroviral
therapy in HIV infection. Curr Opin Infect Dis 2011; 24:27–33.
9. Rawizza HE, Chaplin B, Meloni ST, Eisen G, Rao T, Sankale JL,
et al. Immunologic criteria are poor predictors of virologic
outcome: implications for HIV treatment monitoring in
resource-limited settings. Clin Infect Dis 2011; 53:1283–
1290.
10. Rutherford GW, Anglemyer A, Easterbook PJ, Horvath T, Vitoria
M, Penazzato M, Doherty MC. Predicting treatment failure in
adults and children on antiretroviral therapy: a systematic
review of the performance characteristics of the 2010 WHO
immunologic and clinical criteria for virologic failure. AIDS
2014; 28 (Supp 2):S161–S169.
11. Wilson D, Keiluhu AK, Kogrum S, Reid T, Seriratana N, Ford N,
et al. HIV-1 viral load monitoring: an opportunity to reinforce
treatment adherence in a resource-limited setting in Thailand.
Trans Royal Soc Trop Med Hygiene 2009; 103:601–606.
12. UNITAID. HIV/AIDS diagnostics technology landscape.
Geneva: UNITAID; 2012.
13. Easterbrook PJ, Irvine CJ, Vitoria M, Shaffer N, Muhe LM,
Negussie EK, et al. Developing the 2013 WHO consolidated
antiretroviral guidelines. AIDS 2014; 28 (Supp 2):S93–S100.
14. Ford N, Calmy A, Andrieux-Meyer I, Hargreaves S, Mills EJ,
Shubber Z. Adverse events associated with nevirapine use in
pregnancy: a systematic review and meta-analysis. AIDS 2013;
27:1135–1143.
15. Shubber Z, Calmy A, Andrieux-Meyer I, Vitoria M, RenaudThery F, Shaffer N, et al. Adverse events associated
with nevirapine and efavirenz-based first-line antiretroviral
therapy: a systematic review and meta-analysis. AIDS 2013;
27:1403–1412.
16. Thompson MA, Aberg JA, Hoy JF, Telenti A, Benson C, Cahn P,
et al. Antiretroviral treatment of adult HIV infection: 2012
recommendations of the International Antiviral Society-USA
panel. JAMA 2012; 308:387–402.
17. Guyatt GH, Oxman AD, Kunz R, Vist GE, Falck-Ytter Y,
Schunemann HJ, et al. What is ‘‘quality of evidence’’
and why is it important to clinicians? BMJ 2008; 336:995–998.
18. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, AlonsoCoello P, et al. GRADE: an emerging consensus on rating
quality of evidence and strength of recommendations. BMJ
2008; 336:924–926.
19. Schulz KF, Altman DG, Moher D, Group C. CONSORT 2010
statement: updated guidelines for reporting parallel group
randomised trials. PLoS Med 2010; 7:e1000251.
20. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC,
Vandenbroucke JP, et al. The Strengthening the Reporting of
Observational Studies in Epidemiology (STROBE) statement:
guidelines for reporting observational studies. Lancet 2007;
370:1453–1457.
21. Sutton AJ. Methods for meta-analysis in medical research.
Chichester, NY: John Wiley; 2000.
22. Barry C. Relationship of viral resistance development to CD4
monitoring alone or with viral load monitoring. ClinicalTrials.
gov; 2013. http://clinicaltrials.gov/show/NCT00434070. [Assessed on 15 January 2014]
23. Abouyannis M, Menten J, Kiragga A, Lynen L, Robertson G,
Castelnuovo B, et al. Development and validation of
systems for rational use of viral load testing in adults receiving
first-line ART in sub-Saharan Africa. AIDS 2011; 25:1627–
1635.
S159
24. Albrecht M, Lisa Mukherjee A, Tierney C, Morse GD, Dykes C,
Klingman KL, et al. A randomized clinical trial evaluating
therapeutic drug monitoring (TDM) for protease inhibitorbased regimens in antiretroviral-experienced HIV-infected
individuals: week 48 results of the A5146 study. HIV Clin Trials
2011; 12:201–214.
25. Babiker A, Castro nee Green H, Compagnucci A, Fiscus S,
Giaquinto C, Gibb DM, et al. First-line antiretroviral therapy
with a protease inhibitor versus nonnucleoside reverse transcriptase inhibitor and switch at higher versus low viral load in
HIV-infected children: an open-label, randomised phase 2/3
trial. Lancet Infect Dis 2011; 11:273–283.
26. Badri M, Wood R. Usefulness of total lymphocyte count
in monitoring highly active antiretroviral therapy in resourcelimited settings. AIDS 2003; 17:541–545.
27. Collazos J, Knobel H, Casado JL. CD4 count and viral load timecourses in patients treated with highly active antiretroviral
therapy and association with the CDC staging system. HIV Med
2006; 7:504–513.
28. Davies M-A, Boulle A, Eley B, Moultrie H, Technau K, Rabie H,
et al. Accuracy of immunological criteria for identifying
virological failure in children on antiretroviral therapy –
The IeDEA Southern Africa Collaboration Précision. Trop
Med Int Health 2011; 16:1367–1371.
29. Diaz T, Loth G, Whitworth J, Sutherland D. Surveillance
methods to monitor the impact of HIV therapy programmes
in resource-constrained countries. AIDS 2005; 19:S31–S37.
30. Duong T, Jourdain G, Ngo-Giang-Huong N, Le Coeur S,
Kantipong P, Buranabanjasatean S, et al. Laboratory and
clinical predictors of disease progression following initiation
of combination therapy in HIV-infected adults in Thailand.
PLoS One 2012; 7:e43375.
31. FeregrinoGoyos M, GomezCaro HH, MirelesVieyra MP,
AlvaradoDiez R, HigueraRamirez F, Hidalgo H. Viral load
count as a better guide for antiretroviral therapy of HIV
patients, the Mexican experiences. AIDS 1996; 10:69–169.
32. Kanapathipillai R, McGuire M, Mogha R, Szumilin E, Heinzelmann A, Pujades-Rodriguez M. Benefit of viral load testing for
confirmation of immunological failure in HIV patients treated
in rural Malawi. Trop Med Int Health 2011; 16:1495–1500.
33. Keiser O, MacPhail P, Boulle A, Wood R, Schechter M, Dabis F,
et al. Accuracy of WHO CD4 cell count criteria for virological
failure of antiretroviral therapy. Trop Med Int Health 2009;
14:1220–1225.
34. Lundgren JD, Babiker A, El-Sadr W, Emery S, Grund B,
Neaton JD, et al. Inferior clinical outcome of the CD4(R) cell
count guided antiretroviral treatment interruption strategy in
the SMART study: role of CD4(R) cell counts and HIV RNA
levels during follow-up. J Infect Dis 2008; 197:1145–1155.
35. Lynen L, An S, Koole O, Thai S, Ros S, De Munter P, et al.
An algorithm to optimize viral load testing in HIV-positive
patients with suspected first-line antiretroviral therapy failure
in Cambodia. J Acquir Immune Defic Syndr 2009; 52:40–48.
36. Moore DM, Awor A, Downing R, Kaplan J, Montaner JS,
Hancock J, et al. CD4R T-cell count monitoring does not
accurately identify HIV-infected adults with virologic failure
receiving antiretroviral therapy. J Acquir Immune Defic Syndr
2008; 49:477–484.
37. Kekitiinwa RA, Sekadde M, Kitaaka SB, Kalyesubula I, Pitter C,
Thomas D, et al. Viral load monitoring is critical to assess early
paediatric response to highly active antiretroviral therapy in a
resource limited setting: the Ugandan experience. International
Conference on AIDS (16th: 2006: Toronto, Canada); 13 August
2006; Toronto, ON.
38. Sigaloff KC, Hamers RL, Wallis CL, Kityo C, Siwale M, Ive P,
et al. Unnecessary antiretroviral treatment switches and
accumulation of HIV resistance mutations; two arguments
for viral load monitoring in Africa. J Acquir Immune Defic
Syndr 2011; 58:23–31.
39. Jourdain G, Ngo-Giang-Huong N, Le Coeur S, Traisaithit P,
Barbier S, Techapornroong M, et al. PHPT-3: a randomized clinical trial comparing CD4 vs viral load ART monitoring/switching
strategies in Thailand. 18th Conference on Retroviruses and
Opportunistic Infections; 27 February 2011; Boston, MA.
40. Keiser O, Chi BH, Gsponer T, Boulle A, Orrell C, Phiri S, et al.
Outcomes of antiretroviral treatment in programmes with
and without routine viral load monitoring in southern Africa.
AIDS 2011; 25:1761–1769.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
S160
AIDS
2014, Vol 28 (Suppl 2)
41. Laurent C, Kouanfack C, Laborde-Balen G, Aghokeng AF,
Mbougua JB, Boyer S, et al. Monitoring of HIV viral loads,
CD4 cell counts, and clinical assessments versus CM alone for
antiretroviral therapy in rural district hospitals in Cameroon
(Stratall ANRS 12110/ESTHER): a randomised noninferiority
trial. Lancet Infect Dis 2011; 11:825–833.
42. Mermin J, Ekwaru JP, Were W, Degerman R, Bunnell R,
Kaharuza F, et al. Utility of routine viral load, CD4 cell count,
and CM among adults with HIV receiving antiretroviral
therapy in Uganda: randomised trial. BMJ 2011; 343:d6792.
43. ARROW Trial Team. Kekitiinawa A, Cook A, Nathoo K,
Mugyenyi P, Nahirya-Ntege P, Bakeera-Kitaka S, et al. Routine
versus clinically driven laboratory monitoring and first-line
antiretroviral therapy strategies in African children with HIV
(ARROW): a 5-year open-label randomised factorial trial.
Lancet 2013; 381:1391–1403.
44. Mugyenyi P, Walker AS, Hakim J, Munderi P, Gibb DM, Kityo C,
et al. Routine versus clinically driven laboratory monitoring of
HIV antiretroviral therapy in Africa (DART): a randomised
noninferiority trial. Lancet 2010; 375:123–131.
45. Chang LW, Harris J, Humphreys E. Optimal monitoring
strategies for guiding when to switch first-line antiretroviral
therapy regimens for treatment failure in adults and adolescents living with HIV in low-resource settings. Cochrane
Database Syst Rev (2010):CD008494.
46. Lifson AR, Krantz EM, Grambsch PL, Macalino GE, CrumCianflone NF, Ganesan A, et al. Clinical, demographic and
laboratory parameters at HAART initiation associated with
decreased post-HAART survival in a U.S. military prospective
HIV cohort. AIDS Res Ther 2012; 9:4.
47. Petersen ML, van der Laan MJ, Napravnik S, Eron JJ, Moore RD,
Deeks SG. Long-term consequences of the delay between
virologic failure of highly active antiretroviral therapy and
regimen modification. AIDS 2008; 22:2097–2106.
48. Gsponer T, Petersen M, Egger M, Phiri S, Maathuis MH,
Boulle A, et al. The causal effect of switching to secondline ART in programmes without access to routine viral load
monitoring. AIDS 2012; 26:57–65.
49. Belec L, Bonn JP. Challenges in implementing HIV laboratory
monitoring in resource-constrained settings: how to do more
with less. Future Microbiol 2011; 6:1251–1260.
50. Suthar AB, Ford N, Bachanas PJ, Wong VJ, Rajan JS, Saltzman
AK, et al. Towards universal voluntary HIV testing and counselling: a systematic review and meta-analysis of communitybased approaches. PLoS Med 2013; 10:e1001496.
51. Koenig SP, Leandre F, Farmer PE. Scaling-up HIV treatment
programmes in resource-limited settings: the rural Haiti
experience. AIDS 2004; 18 (Suppl 3):S21–S25.
52. Peeling RW, Ronald A. Diagnostic challenges of sexually
transmitted infections in resource-limited settings. Future
Microbiol 2009; 4:1271–1282.
53. Louis FJ, Anselme R, Ndongmo C, Buteau J, Boncy J, Dahourou
G, et al. Evaluation of an external quality assessment program
for HIV testing in Haiti, 2006–2011. Am J Clin Pathol 2013;
140:867–871.
54. Damond F, Benard A, Balotta C, Boni J, Cotten M,
Duque V, et al. An international collaboration to standardize
HIV-2 viral load assays: results from the 2009 ACHI(E)V(2E)
quality control study. J Clin Microbiol 2011; 49:3491–
3497.
55. Damond F, Benard A, Ruelle J, Alabi A, Kupfer B,
Gomes P, et al. Quality control assessment of human immunodeficiency virus type 2 (HIV-2) viral load quantification
assays: results from an international collaboration on
HIV-2 infection in 2006. J Clin Microbiol 2008; 46:2088–2091.
56. Kawonga M, Blaauw D, Fonn S. Aligning vertical interventions
to health systems: a case study of the HIV monitoring and
evaluation system in South Africa. Health Res Policy Syst 2012;
10:2.
57. Kawonga M, Fonn S, Blaauw D. Administrative integration
of vertical HIV monitoring and evaluation into health systems:
a case study from South Africa. Glob Health Action 2013; 6:
19252.
58. Hamers RL, Sawyer AW, Tuohy M, Stevens WS, Rinke de Wit
TF, Hill AM, et al. Cost-effectiveness of laboratory monitoring
for management of HIV treatment in sub-Saharan Africa: a
model-based analysis. AIDS 2012; 26:1663–1672.
59. Medina Lara A, Kigozi J, Amurwon J, Muchabaiwa L, Nyanzi
Wakaholi B, Mujica Mota RE, et al. Cost effectiveness analysis
of clinically driven versus routine laboratory monitoring of
antiretroviral therapy in Uganda and Zimbabwe. PLoS One
2012; 7:e33672.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.