Content uploaded by Yuri Quintana
Author content
All content in this area was uploaded by Yuri Quintana on Apr 05, 2020
Content may be subject to copyright.
A Systematic Review on Promoting Adherence
to Antiretroviral Therapy in HIV-infected
Patients Using Mobile Phone Technology
Yuri Quintana1,2 Eduardo A. Gonzalez Martorell2Darren Fahy1Charles Safran1,2
1Division of Clinical Informatics, Department of Medicine, Beth Israel
Deaconess Medical Center, Boston, Massachusetts, United States
2Harvard Medical School, Boston, Massachusetts, United States
Appl Clin Inform 2018;9:450–466.
Address for correspondence Yuri Quintana, PhD, Division of Clinical
Informatics, Department of Medicine, Beth Israel Deaconess Medical
Center, 1330 Beacon Street, Suite 400, Brookline, MA 02215,
United States (e-mail: yquintan@bidmc.harvard.edu).
Keywords
►HIV
►acquired
immunodeficiency
syndrome
►medication
adherence
►antiretroviral therapy
►highly active
►mHealth
Abstract Objective Adherence to antiretroviral therapy (ART) is paramount to successful long-
term suppression of human immunodeficiency virus (HIV). For poorly adherent
patients with HIV, barriers to remaining adherent may be overcome by the imple-
mentation of targeted interventions delivered via mobile devices. This systematic
review is focused specifically on mobile phone technologies to deliver adherence
interventions in HIV/acquired immunodeficiency syndrome (AIDS) populations.
Methods This review (PROSPERO #CRD42017065131) systematically extracted data
from published literature from five databases on mobile phone interventions to
improve adherence to ART for HIV. The reported studies had been conducted between
2007 and 2017. Risk of bias was assessed using the Cochrane method ranking each
criterion as low, high, or unclear risk of bias.
Results Of the 835 articles returned, we identified 26 randomized controlled trials
(RCTs), retrospective and prospective cohort trials, or mixed method studies with a
comparison group that fit criteria for inclusion. No standard measure of adherence was
consistent throughout the examined studies, and assessments by self-report, pill
counting, and medication event monitoring system (MEMS) were utilized. The studies
reported mixed results, with 17 reporting significant improvements to adherence, 3
reporting improvements without supplying p-values, and 6 reporting no significant
change or a reduction in adherence.
Conclusion The mixed nature of the results exemplifies the need for more compre-
hensive approaches and larger scale trials to confirm results observed in limited cohort
sizes. To better retain satisfactory adherence within the HIV population, and especially
in low-resource settings, we recommend that future interventions incorporate multiple
strategies: mobile-based reminders, social support structures, and personalized
content.
received
January 8, 2018
accepted after revision
May 2, 2018
Copyright © 2018 Schattauer DOI https://doi.org/
10.1055/s-0038-1660516.
ISSN 1869-0327.
Review Article450
Background and Significance
A key concern for optimizing treatment of human immuno-
deficiency virus (HIV) is the adherence to antiretroviral
therapy (ART). This paper presents a systematic review of
mobile phone technologies to improve HIV medication
adherence. There were an estimated 36.7 million people
living with HIV (PLWHIV) worldwide in 2015.1The impact
of HIV/acquired immunodeficiency virus (AIDS) has been
particularly hard on countries with limited resources. ART
can help people with HIV infection to live longer, healthier
lives, but adherence to treatment can be challenging, espe-
cially in low-resource settings.
Non-adherence to HIV/AIDS treatment is a significant
public health issue.2Patients who remain adherent to ART
are significantly less likely to transmit an infection to sexual
partners;3however, the population affected by HIV/AIDS
demonstrates a lack of knowledge about the disease, treat-
ment approaches and the importance of strict adherence.4
Various HIV programs have noted the importance of improv-
ing adherence to attain better outcomes.5–8
Mobile phone technologies have the potential to promote
adherence in these patients. Wireless telecommunications
networks have spread rapidly worldwide, and sending text
messages on wireless mobile telephones has become an
extremely popular means of communication.5Mobile phone
text messaging, also called short messaging service (SMS),
has been explored as an approach to enhancing patient
adherence to ART regimens.
Recent systematic reviews have focused on mobile health
(mHealth) and adherence not specifictoHIV;
9,10 mHealth in
low-resource settings, but not specific to HIV adherence;11–13
HIV adherence in low-resource settings, but not with
mHealth;14 treatment for HIV but not specifically adher-
ence;15,16 and mHealth and disease management.17,18 Two
earlier review studies examined ART adherence, but were not
focused on mobiletechnologies.19,20 Other systematic reviews
have focused on adolescent and young adult populations using
mHealth interventions for adherence, which have seen HIV
diagnoses growing at a disproportionately high rate.21–23 A
review5in 2012 analyzed two randomized control trials (RCT)
using mobile phones for HIV adherence, but there have been
many studies since then. This paper presents a comprehensive
review focused on mobile text messaging to deliver adherence
interventions in HIV/AIDS populations.
Objective
Our research focus was to determine whether using mobile
phone text-based reminder systems are efficacious in enhan-
cing adherence to ART in patients with HIV infection.
Methods
Search Strategy
We conducted a systematic review following Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) guidelines for reporting systematic reviews and
the Cochrane review standards for classification of quality
assessment and data extraction.24 Systematic computerized
literature searches were performed in PubMed/MEDLINE
(NCBI), Embase (Elsevier), Cumulative Index to Nursing
and Allied Health Literature-CINAHL (EBSCO), Cochrane
Central Register of Controlled Clinical Trials (EBSCO), and
Web of Science (Thomson Reuters). The search was designed
to identify studies that evaluated the effectiveness of mobile
phone technologies for enhancing adherence to ART in
PLWHIV. Controlled vocabulary terms were included when
available. No date or language limits were applied, but
meeting abstracts were excluded when possible. We exam-
ined the bibliographies of relevant studies for additional
material. Neither study authors were contacted nor were
gray literature sources examined. The literature search was
completed on December 12, 2017. The search terms are l isted
in ►Appendix A.
Study Selection and Eligibility Criteria
Three of this study’s authors (Y.Q., D.F., and E.A.G.M.) inde-
pendently screened and reviewed the titles and abstracts of
each article identified by the search. The authors were
unblinded because blinding has little effect on systematic
review results.25 The authors used inclusion and exclusion
criteria to assess the full eligibility of the screened abstracts.
Study date was not a criterion for inclusion or exclusion. Only
peer-reviewed journal articles written in English were
included in the review. Selection of English-only articles was
decided because of our language limitations. Additional inclu-
sion criteria were as follows: any experimental or observa-
tional intervention evaluations with a control group or single
cohort group with a before and after comparison, where the
intervention was delivered digitally, not by voice alone, via
mobile phone, or smart phone in any setting. The types of
evaluations included were as follows: retrospective cohort
study, prospective cohort study, randomized controlled trial
(RCT), cohort study, mixed retrospective/prospective cohort
study, or before-and-after study. The exclusion criteria were
applied in the order shown in ►Fig. 1. Any study selection
disputes were settled by a consensus of the authors.
Extraction and Analysis of the Data
Two blinded authors (D.F. and E.A.G.M.) independently con-
ducted the quality assessment and data extraction. The
Covidence (http://www.covidence.org) screening and data
extraction tool for authors were used. The quality assess-
ment was conducted based on the recommendations of the
Cochrane Risk of Bias Comparison: sequence generation,
allocation concealment, blinding of participants, blinding
of outcome assessors, incomplete outcome data, selective
outcome, and other sources of bias. The following data were
extracted: study identification, methods, population, inter-
vention, and outcome variables. p-Values <0.05 were con-
sidered significant. Disagreements between the two blinded
authors (D.F. and E.A.G.M.) were adjudicated by a third
author (Y.Q.), who was not involved in the data extraction.
Four authors (C.S., Y.Q., D.F., and E.A.G.M.) reviewed the
results and participated in the analysis and write-up.
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 451
Results
After examining the search results for inclusion and exclusion
criteria, described in ►Fig. 1, 26 articles for the qualitative
synthesis remained: one short report,26 and 25 peer-reviewed
journal articles.8,27–50 Of these 26 studies, 23 studies used SMS
alone for reminders, and 3 used SMS plus counseling as noted
in ►Table 1. Six reports were from the United States;35–37,42–44
three each from Kenya8,31,40 and South Africa;28,39,48 two each
from Uganda,38,45 China,34,47and Cameroon;30,41and one each
from Italy,26 Brazil,29 Nigeria,32 New Zealand,27 India,33 Argen-
tina,46 Canada,49 and Malaysia.50 The types ofstudies included
19 RCTs,8,27,29–32,34–36,38,39,41–45,47,50 threeprospective cohort
studies,26,37,46two ambidirectionalstudies (retrospective/pro-
spective cohort study),28,49 one cluster RCT,40 one quasi-
experimental cohort study,33 and one retrospective cohort.48
The studies had been performed between 2007 and 2015 and
published between2010 and 2016 (►Table 1). For t he RCTs, the
number of participants ranged from 11 to 401 in the control
groups and from 14 to 314 in the intervention groups. The
intervention periodplus follow-up rangedfrom 1 to 24 months,
with a median of 10 months (quartiles 5–12). The primary
outcomes addressed in the studies included one ART usage
among study participants,40 one adherence measured through
prescription refills,48 two viral load adherences,37,46 six adher-
encesover time,31,33,34,38,39,43 and 16 self-reported adherences
(►Table 2).8,26–30,32,35,36,41,42,44,45,47,49,50 Adherence inter-
ventions were measured as self-report, ART usage, adherence
over timeusing a medication eventmonitoring system (MEMS)
or an electronicadherence monitoring device, such as Wisepill,
pharmacy refills, pill count, and clinical outcomes (viral load,
HIV-ribonucleic acid [RNA] suppression, and cluster of differ-
entiation 4 [CD4] count). Additionally, a summary of the out-
come data, adherence assessment, metrics, relative effect (risk
and odds ratios [OR]), and a measure of precision (95% con-
fidence interval [CI]) for each study are shown in ►Table 2.
Detailson the types and levels of bias inherent in the individual
studies are summarized in ►Fig. 2.
Fig. 1 PRISMA flow char t of study selection.24 AIDS, acquired immunodeficiency virus; HIV, human immunodeficiency virus; PRISMA, Preferred
Reporting Items for Systematic Reviews and Meta-Analyses.
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.452
Table 1 Summary of studies
Authors Study year Countr y Type of study Reminder
method
Target population n(Control
group)
n(Intervention
group)
Inclusion/
Exclusion
criteria
reported
Adherence
measure
reported
Abdulrahman et al50 2014 Malaysia RCT SMS Adult HIV po sitive patients who had
completed fou r weeks of vitamin
training and were newly in itiating ART
121 121 Yes/ Yes Yes
Ammassari et al26 NR Italy PC SMS Patients repor ting any degree of sub-
optimal adh erence
0 145 Yes/ Yes Yes
da Costa et al29 2009–2010 Brazil RCT SMS Patients with viral load below 400
copies/mL for at least 3 months and
patients with CD4þcell counts great er
than 200/mm
3
15 14 Yes/ Yes Yes
Evans et al 28 2011–2014 South Africa RC/PC SMS Patients receiving a second-line ART
regimen containing lopinavir/ritonavir
or atazanavir/ritonavir and
experien ced a single elevat ed viral load
(400 copies/mL) on second-line ART
intervention
401 49 and 314 Yes/No Yes
Garofalo et al35 2010–2014 USA RCT SMS HIVþpatients on ART for 1month
with adherence problem s
54 52 Yes/ Yes Yes
Georgette et al48 2012–201 4 South Afric a RC SMS Patient s with pre-progr am prescription
coverage <100% and patients who
initiat ed ART within 2 years of th e start
of the SMS program
2,255 2,255 Yes/No Yes
Haberer et al 38 2013–2014 Uganda RC T SMS This study involved two types of
participants: individual s taking ART
(‘study par ticipant s’)andtheir‘social
suppor ters’
21 21 and 20 Yes/Yes Yes
Hardy et al36 2008 USA RCT SMS HIV-infected men and wo men
receiving HIV primary care on a stable
regimen of antiretroviral ther apy for at
least 3 months and repor ting less than
85% adherence to ART over the prior
7days
Not repor ted 23: 12 and 11 Yes/No Yes
Kalichman et al42 2011–2015 USA RCT SMS Patients receiv ing ART and
self-repo rted less than 95% adherence
in the past month as per a validated
visual analog adheren ce scale
149 and 151 150 and 150 Yes/ No Yes
Kassaye et al40 2012–2013 Kenya Cluster RCT SMS Pregnant women were eligible to
enroll in the study if t hey were less than
32 weeks of gestational age, were not
currently receiving ART, were planning
to remain in the area for the dur ation
of the study period, and agreed to
follow-up of their in fants until 6 weeks
following delivery. Male pa rtners of
women were also permi tted to enroll
in the study.
270 280 Yes/No No: impr ove PMTCT
(Continued)
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 453
Table 1 (Continued)
Authors Study year Countr y Type of study Reminder
method
Target population n(Control
group)
n(Intervention
group)
Inclusion/
Exclusion
criteria
reported
Adherence
measure
reported
King et al49 2013–2014 Canada RC/PC SMS Patie nts attendin g the clinic for at leas t
1 year prior to study entry, with an
indicati on for cART (at time of study
developmen t, CD4 <500 cells/mm
3
),
detectable VL (200 copies/mL), with
high risk for disengagement in
treatment
80 80 Yes/ Yes Yes
Lester et al82007–2008 Kenya RCT SMS Patients who were in itiating ART for
the first time
265 273 Yes/No Yes
Lewis et al37 2010 USA PC SMS HIV-positive men having sex with men 0 52 Yes/No Yes
Linnemayr et al45 2014–2015 Uganda RCT SMS Patients on ART or cotrimoxazole
prophylaxis against com mon oppor tu-
nistic infe ctions, 15–22 years of age
112 110 Yes/ Yes Yes
Mbuagbaw et al 41 2010 Came roon RCT SMS Patients who ha d been on ART for
at least 1 mo nth
99 101 Yes/Yes Yes
Moore et al44 NR USA RCT SMS Documented HIV in fection, diagnosis
of bipolar disorder I or II via the
composite internatio nal diagno stic
interview, and currently taking at least
one ARV and on e PSY medication to
treat HIV and BD, respectively.
Medication non-adherence was not a
study entry requir ement
28 30 Yes/ Yes Yes
Nsagha et al30 2011 Cameroon RCT SMS People living with HIV and AIDS and
who had been on ARVs for at least
1month
45 45 Yes/ Yes Yes
Orrell et al39 2012–2014 South Afri ca RCT SMS Partic ipants wh o were ART-naive
adults and adolescents (15 years old)
commencing treatment at the Hannan
Crusaid Treatment Centre
115 115 Yes/No Yes
Perera et al27 2012–2013 New Zealand RCT Mobile App Indiv iduals who had been on ART for at
least 6 months
11 17 Yes/No Yes
Rodrigues et al33 2010–2011 In dia Quasi-experimental
cohort study
SMS HIV-infe cted adults who followed u p at
the clinic as outpatients and who were
on ART for at least a month prior to
enrollment
0 150 Yes/ Yes Yes
Ruan et al47 2013–2014 Ch ina RCT SMS HIV-positive patie nts on ART for no
more than 3 months
50 50 Yes/No Yes
Sabin et al34 2012–2013 China RCT SMS Patients who were receiving or
initiating ART and wer e deemed at ri sk
for poor adh erence by clinicians or
themselves
21 and 35 23 and 40 Yes/No Yes
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.454
The Cochrane method was used to evaluate bias and is
summarized in ►Fig. 2. Most of the studies we reviewed fall
under the high-risk category for performance bias. For proto-
cols that call for large numbers of subjects, it may not be
economically feasible to supply each study participant with a
mobile phone to ensure proper blinding. The alternative—
either using a subject’s phone or providing phones only to the
interventional cohort—introduces the risk of performance
bias. Nine of the 19 RCTs were deemed to be at low risk for
detection bias, 4 at high risk, and 6 at unclear risk. All the
studies were deemed to be at low risk for attrition bias, with
two exceptions. Two studies—one by Evans et al28 and another
by Lewis et al37—had a high risk of reporting bias because of
missing data, but in the Evans et alstudy, the authors noted the
missing data and addressed the issue by including a sensitivity
analysis. A major factor in determining the risk of bias was the
inadequate reasoning for missing data with low attrition rates.
Similarly, reporting bias was lowacross all studies, with all but
one RCT deemed as low risk. One prospective cohort trial was
evaluated as high risk and one as unclear. We determined a
high risk of bias in one study due to non-reporting of HIV
laboratory data,43 which were not consistently gathered.
Four studies contained other potential sources of bias.
One study noted that 37% of screened participants were
excluded because of lack of mobile phone access.38 We
determined that this met the criteria for a high risk of bias
owing to the study design and problems with the recruit-
ment of participants with mobile phone access. Another
study was evaluated as having a high risk of bias due to
the lack of information l isted regarding sources of funding or
conflicts of interest.43 Two studies48,49 declared authors
with a material interest in the development of adherence
promoting tools or procedures.
Small sample sizes and missed recruitment goals were
other concerning issues in some studies. Recruitment ranged
from 28 to 715 subjects, and only 11 (55%) studies had more
than 100 enrolled. Two studies had low recruitment num-
bers—28 and 29—limiting the generalizability of the
reported results.27,29 Additionally, although 200 patients
were recruited, the sample size was insufficient to ade-
quately statistically power the study.41 The study by Lewis
et al had limited enrollment criteria because of a second,
concurrently running study at the same site.37
Nineteen studies reported statistically significant adherence
outcomes (►Table 2).Onestudyfoundthattextmessages
significantly increased the proportion of those achieving adher-
ence (76.9% versus 55.8%, p¼0.02).32 Another reported that
even though adherence increased significantly with weekly
reminders, daily SMS reminders did not have a significantly
different effect on study subjects versus the control group,
which the authors attributed to the possibility of habituation.31
One study showed that increased adherence, from 85% to 91%,
had been maintained 6 months after the intervention was
discontinued (p¼0.016).33 One study using real-time remin-
ders significantly improved adherence.34 Another study con-
cluded that mobile phone SMS interventions might be effective
tools for improving patient outcomes in resource-limited set-
tings.8One study found a higher proportion of adherent subjects
Table 1 (Continued)
Authors Study year Countr y Type of study Reminder
method
Target population n(Control
group)
n(Intervention
group)
Inclusion/
Exclusion
criteria
reported
Adherence
measure
reported
Stankievic h et al46 2014 Argentina PC SMS Patients who were receiving ART and
who had suboptimal adhe rence
(incomplete viral suppression
determined by VL 1000 copies/mL)
22 22 Yes/ Yes Yes
Studies with SMS reminder and additional counseling
Ingersoll et al43 2012 US A RCT SMS Patien ts who had an active prescripti on
for ART and re ported less t han 95% ART
adherence in the past 2 weeks
30 33 Yes/no Yes
Maduka and Tobin32 20 11 Nigeria RCT SMS HIV-positive patie nts who had been
HAART experienced fo r at least
3 months an d had a histor y of
non-ad herence (adheren ce below 95%)
to HAART at the time of the study
52 52 Yes/yes Yes
Pop-Elecheset al.31 2007–2008 Kenya RCT SMS Patients who had initiated ART less
than 3 months prior to enrollment
139 70 and 72 and 73
and 74
Yes/yes Yes
Abbreviations: AIDS, acquired immunodeficiency syndrome; ART, antiretroviral ther apy; ARV, antiretrovir al; BD, bipolar disorder; cART, combination antiretroviral therapy; HAART, highly active antiretroviral
therapy ; HIV, human immuno deficiency virus ; NR, not reporte d; PC, prospecti ve cohort; PMTC T, prevention of moth er-to-child tran smission; PSY, psychia tric; RC, retrospec tive cohort; RC T, randomized contro lled
trial; SMS , short me ssaging service; VL, viral load.
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 455
through the use of pill counting and MEMS methods.28 An
additional study estimated that the average effect over the
6-month intervention period was significant for 90% adher-
ence (OR ¼2.12, 95% CI 1.01–4.45, p<0.05) and was main-
tained at 12 months.35 One studyshowed significant adherence
when it was measured by MEMS (mean difference standard
deviation [SD]: 33.4 9.1, p¼0.002) and composite adher-
ence score (27.1 9.2, p¼0.009).36 In contrast, significance
was not attained while assessing adherence measured by pill
count (13.7 9.1, p¼0.153) or self-report (20.2 10.3,
p¼0.069). One study showed that SMS improved adherence
and that the key constraints affecting adherence to antiretro-
viral medication can be addressed by using SMS.30 Another
study showed that greater usage of the extra components of an
augmented phone application was associated with a greater
perceived understanding of HIV infection and increased per-
ceived necessity for ART.27 In addition to significant adherence
improvement, some authors reported improvement in clinical
outcomes. One study observed a significant reduction in viral
load,37 and in another study, the proportion of subjects with
undetectable viral load increased.26 One study reported that
scheduled SMS reminders significantly improved adherence in
the context of real-time monitoring (p¼0.02), and HIV RNA
suppression was seen, although it was not statistically signifi-
cant (p¼0.14).38
The remaining seven studies described non-significant
outcomes (►Table 2). Six yielded no significant difference in
adherence over the course of their studies.39–42,45,46 One
study observed that in patients with an elevated viral load on
a second-line ART, electronic adherence monitoring was
associated with a modest, but not significant, improvement
in viral suppression.28 Those authors concluded that adher-
ence strategies increase the durability of second-line ART,
decrease the need for costly third-line regimens, and prevent
unnecessary genotyping tests.
Discussion
Our review shows that using mobile phone text-based
reminder systems are efficacious in enhancing adherence
to ART in patients with HIV infection in a wide variety of
global settings. Of the 23 studies that use d text messaging, 21
had positive outcomes, and of the 3 studies that used text-
messaging with counseling, all 3 had positive outcomes.
While a majority of studies had positive outcomes, the
sample sizes will need to be larger to evaluate effect size
and variations in delivery methods. All the studies contrib-
uted to a better understanding of how to deliver these
interventions via mobile devices.
Several studies showed that wireless devices could pro-
vide real-time b ehavior data28,34,38 that allow messages to be
tailored to the patient and promote faster behavior correc-
tion. Four studies monitored the patients’ART adherence
electronically in real time by providing them with a medica-
tion dispenser device.28,34,38,39 Three of those studies would
trigger an SMS reminder if the monitoring device were not
activated within 30 minutes to 2 hours of a scheduled
dose.34,38,39 One study achieved high outcomes (optimal
adherence 95%) and showed this technology to be feasible
and acceptable in China, with reliable monitoring of adher-
ence over time.34 Combined, these findings suggest that
wireless technologies can be an aid to adherence programs.
Three studies combined the strategies of adherence coun-
seling and text-message reminders. The first evaluated the
differences between the lengths of ‘short’and ‘long’mes-
sages and the difference in daily and weekly frequencies.31
The second used a similar strategy, sending up to four text
messages to those in the intervention cohort regarding
whether a medication was taken, the subject’s mood, and
if any substance use had taken place.43 The third, in Nigeria,
evaluated subjects receiving adherence counseling monthly
for 4 months in addition to receiving adherence text-message
reminders.32 Text messages were crafted to address beha-
vioral barriers and adherence support and to serve as a
reminder to continue the ART. All three reported positive
outcomes as increases in adherence in the intervention
cohorts. The optimal frequency of reminders is debatable,
as two of the studies sent messages daily,31,43 while the third
called for reminders twice a week on weekdays only.32 The
results were mixed, with one study finding no difference
between the control group and those who received daily
messages,31 while the other study found clinically relevant
Fig. 2 Types and levels of bias inherent in the individual studies.
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.456
Table 2 Summary of outcomes, outcome variables, and statistical significance
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
Kassaye et al40 201 6 Antiretr oviral usage
among study
partic ipants
N,% Anymisseddosepast
week
Risk ratio (95 %) for both
outcomes (ART usage
and missed d ose)
Antiretrov iral usage
among study
particip ants
Slightly hig her % in
control grou p than in
interven tion group;
no p-value reported
X––
Any missed dose
during past week
Slightly lo wer % in
control grou p than in
interven tion group;
no p-value reported
X––
Lewis et al37 201 3 VL Median, hig h, and low,
Improved, rema ined
adherent, n on-
adherent
CD4 count, Me dication
adherence - VL (HIV -1
RNA copies pe r mL);
medication adherence -
CD4 count (a bsolute
count per mm
3
)
–VL Significant im provement
in VL from bas eline to
follow-u p, p<0.012
XX
CD4 count Signi ficant increase in CD4
counts, p<0. 037
XX
Medicatio n adher-
ence- VL (HI V-1 RNA
copies/mL)
Partic ipants who
improved thei r medica-
tion adhere nce during th e
study or rema ined adher-
ent had signi ficant
improvemen ts in VLs,
p¼0.013
XX
Medicatio n adherence
–CD4 count (abso lute
count per mm
3
)
Partic ipants who
improved thei r medica-
tion adhere nce during th e
study or rema ined adher-
ent had a gene ral trend of
improvemen ts in CD4
counts, p¼0. 051
XX
Pop-Elech es et al31 2011 Adherence ITT
over time (ME MS)
%––Adherence ITT over
time (MEMS)
Weekly SMS remin ders
increased th e percentage
of partic ipants ach ieving
90% adhere nce to ART by
13% to 16% compar ed
with those wi th no remin-
der; no p-value reported
X––
Rodrigues et al33 2 012 Adher ence over time
(IVR)
N, % ––Adhe rence over
time (IVR)
Significant im provement
in propor tion of par tici-
pants adher ent over time,
p¼0.016
XX
Sabin et al34 201 5 Adherence ove r time N,%,mean,SD CD4þcell cou nt,
undetecta ble VL over
time <50 copie s/mL
Risk Ratio (9 5%)
Adherence on ly
Adherence ove r time Adherence was signifi-
cantly grea ter in the
interven tion group than in
the control s at the end of
the study as well as du ring
the entire in tervention
period, p<0.001
XX
Undetecta ble VL over
time <50 copies /mL
Adherence in both the
interven tion and contr ol
groups incr eased, but
propor tions were simi lar
between th em at month
9, p¼0.218
XX
(Continued)
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 457
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
CD4þcell coun t Mean change in CD4-ce ll
count betwe en baseline
and month 9 tr ended
higher, but wa s not
significantly different in
interven tion subjec ts ver-
sus control s, p¼0.297
XX
Orrell et al-39 2 015 Adher ence over time
(EAMD)
OR, CI, N, mean TI only TIs, retention in care, HIV-
RNA >40 copies/mL
Odds ratio (9 5%) for
adherence, o dds ratio
for HIV-RNA , risk ratio
(95%) for TIs
Adherence ove r time
(EAMD)
Median adhe rence by
EAMD was s lightly greater
in the inter vention grou p
than in the co ntrols, but
not significan t, p¼0.642
XX
HIV-RNA >40 copies/
mL
No differe nce in the odds
of virologi cal failure in the
interven tion arm,
p¼0.393
XX
TIs, retenti on in care The inte rvention si gnifi-
cantly red uced the
frequenc y of TIs over
72 h, p¼0.393
XX
Ingersoll et al 43 2015 Adhe rence over
time (phar macy refill)
Mean, %, SD Alcohol and dr ug using
days, Propor tion of
missed visits
–Adherence ove r time
(pharmac y refill)
The inter vention
improved adhe rence,
p¼0.02
XX
Proporti on of missed
visits
There was a tren d toward
improved vis it attendance
in the inter vention gro up,
but it was not signi ficant,
p¼0.12
XX
Alcohol and dr ug
using days
There was impr ovement,
but the inte rvention d id
not reduce subs tance-
using days compa red with
the control, p¼0. 14
XX
Haberer et al38 201 6 Adheren ce over time
(EAMD)
Median, %, IQ R,
mean %, SD
HIV-RNA suppress ion Risk ratio adherence only Adheren ce over time
(EAMD)
Percentage adherence was
11.1% higher (p<0.02)
and >48-h and >96-h
lapses were less frequent
(p<0.02, p<0.001,
respectively) in the sched-
uled SMS arm compared
with the control
XX
HIV-RNA suppress ion No statistically significant
differences in HIV RNA
suppression were seen
between study arms,
p¼0.14
XX
Maduka et al32 2013 Self-reported
adherence ove r time
N, % CD4 þcell count Risk ratio (9 5%)
adherence on ly
Self-repor ted
adherence ove r time
Text message reminders
significantly improved drug
adherence, p¼0.022
XX
CD4þcell count Median C D4þcell count of
the inter vention group
increased, p¼0.007
XX
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.458
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
Lester et al82010 Self-reported
adherence ove r time
N, % Rate of attr ition, HIV- 1
VL RNA Suppr ession
<400 copies/m L
Risk ratio (9 5%) Self-re ported
adherence ove r time
Greater ad herence to ART
was repor ted for patient s
receiving th e SMS inter-
vention, p¼0.006
XX
HIV-1 VL RNA
suppressio n
<400 copies/mL
Suppresse d VLs were
reporte d in more patie nts
in the SMS gro up,
p¼0.04
XX
Ammassari et al 26 2011 Self-reported
adherence ove r time
Mean, %, SD Undetectabl e HIV RNA
VL <50 copies/mL
–Self-repor ted
adherence ove r time
Significant im provement
in the propor tion of ART
doses taken ove r the pre-
ceding month at a ll study
time point s, p<0.001
XX
Undetecta ble HIV RNA
VL <50 copies/m L
Significant im provement
in the propor tion of sub-
jects with undete ctable
HIV RNA VL , p<0.001
XX
da Costa et al29 2012 Self-reported
adherence ove r time
N, % MEMS adhere nce over
time, pill cou nting
adherence ove r time
–Self-repor ted
adherence ove r time
Not significa nt, p¼0.243 X X
Pill countin g
adherence ove r time
Not significa nt,
p¼0.6038
XX
MEMS adhere nce
over time
Significant, p¼0.1946 X X
Garofalo et al 35 2016 Self- rep or ted
adherence ove r time
N, mean, SD Undet ectable VL 75
copies/mL ove r time
Odds ratio (9 5%) Self-reporte d
adherence ove r time
The average ef fect esti -
mate over th e 6-month
interven tion was
significant for 90%
adherence, p<0.05, and
maintaine d at 12 months
XX
Undetecta ble VL 75
copies/mL ove r time
Improved in in terventi on
group, p<0.05
XX
Hardy et al36 2011 Self-reported
adherence ove r time
Mean, range MEM S adherence over
time, CAS adhere nce
over time, pill cou nt
adherence ove r time
Odds ratio (9 5%) for
MEMS only
Self-repor ted
adherence ove r time
All the sampl es resulted in
asignificant diffe rence
between th e mean adher-
ence in the t wo inter ven-
tion groups a t both week 3
and week 6 (p-value
ranges 0.00 4–0.043 at
week 3 and 0.00 4–0.026
at week 6)
XX
Pill count adhe rence
over time
All the sampl es resulted in
asignificant diffe rence
between th e mean adher-
ence in the t wo inter ven-
tion groups a t both week 3
and week 6 (p-value
ranges 0.00 4–0.043 at
week 3 and 0.00 4–0.026
at week 6)
XX
(Continued)
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 459
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
MEMS adhere nce over
time, CAS ad herence
over time
All the sampl es resulted in
asignificant diffe rence
between th e mean adher-
ence in the two inter ven-
tion groups a t both week 3
and week 6 (p-value
ranges 0.00 4–0.043 at
week 3 and 0.00 4–0.026
at week 6)
XX
Kalichman et al42 2016 Self-reported
adherence ove r time
N, mean, SD Medic ation adher ence
self-efficacy, HIV RNA VL
suppress ion <100
copies/mL
Odds ratio (9 5%) for HIV
RNA viral
suppress ion only
Self-repor ted
adherence ove r time
Self-repo rted adhere nce
improved 90% dem on-
strating a cl inical mean-
ingful imp rovement,
p<0.01 & p<0.05
XX
HIV RNA VL
suppressio n <100
copies/mL
Modest but si gnificant
effect on the HIV RNA VL
suppressi on, p<0.05
XX
Medicatio n
adherence
self-effica cy
Significant ef fect,
p<0.05
XX
Mbuagbaw et al 41 2012 Self-reported
adherence ove r time
N, %, mean, SD Pharmacy re fill rate, self-
reporte d number of
doses misse d
Risk ratio (9 5%) for self-
reporte d
adherence an d doses
missed only
Self-repor ted
adherence ove r time
No significant effect on
adherence
XX
Self-repor ted
number of dos es
missed
p>0.9 X X
Pharmac y refill rate No significant ef fect X X
Moore et al44 2015 Self-reported
adherence ove r time
%, SD ––Self-reported
adherence ove r time
Adherence fo r 80% of the
partici pants in the st udy
was >90% and ART d ose
timing signi ficantly
improved with th e inter-
vention comp ared with
the control, no p-va lue
reporte d
X––
Nsagha et al30 2016 Self-reported
adherence ove r time
N,% ––Self-repor ted
adherence ove r time
Self-repo rted adhere nce
was higher in th e inter-
vention grou p, p¼0.05
XX
Perera et al27 2014 Self-reported
adherence ove r time
N, mean, SD, 95% CI Pr escribed dos es taken,
pharmacy
dispensing , HIV VL
(log
10
copies/mL)
–Self-repor ted
adherence ove r time
Self-repo rted adhere nce
significan tly improved,
p¼0.03
XX
HIV VL (log
10
copies/
mL)
Decreased, p¼0.02 X X
Prescribe d doses
taken, pharma cy
dispensin g
Non-adher ence
decreased, p¼0. 18, not
significant
XX
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.460
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
Evans et al 28 2016 Self-reported
adherence ove r time
N, % VL suppress ed
<400 copies/m L
–Self-repor ted
adherence ove r time
Only 44.9% of participan ts
in the inter vention resu p-
pressed th eir VL, modest
improvemen t in the pri-
mary outco me. It could
not be demonst rated that
EAMD could si gnificantly
improve adher ence
XX
VL suppress ed
<400 copies/mL
Modest, but not sign ifi-
cant improve ment in viral
suppressi on
XX
Abdulrahm an et al50 2017 Self-reported
adherence ove r time
N, mean, SD, 95% CI Ad herence, CD4 cou nt
(cells/mL), VL (l og
10
),
Weight (kg)
–Adherence Mean ad herence to ART
increased fr om baseline
values in the in tervent ion
group as compar ed with
the control gr oup after
6 months foll ow-up
(p¼0.035). The pro por-
tion of respon dents who
had good adhe rence
(>95%) was signi ficantly
higher in th e interventi on
group (n¼107, 92.2%) as
compared with con trol
group (n¼59, 54 .6%)
(p¼0.001) af ter
6 months foll ow-up.
XX
CD4 count A sign ificantly high er rise
in CD4 count (p¼0. 017)
was observe d in the inter-
vention grou p after
6 months foll ow-up. CD4
count increase d by 146.01
cells/μL in the inter ven-
tion group wh ereas an
increase of 93. 62 cells/μL
was obser ved in the con-
trol group.
XX
VL It was found that 99 .1% of
the inter vention group
who achieve d optimal/
good adhere nce, >95%,
had viral sup pression
compared with 89.3% in
the control gr oup,
p¼0.028
XX
Georgette et al48 2017 ART prescripti on
coverage
N,%,95%CI,IQR Programeffect:
exposed, un exposed,
and unknown
Adjusted od ds ratio
(95%)
Self-repor ted
adherence ove r time
Self-repor ted adher ence
significan tly improved,
p¼0.03
XX
King et al49 201 7 VL, CD4 coun t, and
self-repor ted adherence
N, mean, median , % Geometric mean VL
(copies/mL), mean (95%
CI); CD4 (cells/mm
3
),
median (IQR); cART regi-
men, N(%); attendance
(%), mean (95% CI)
Odds ratio (95%) HIV VL (log
10
copies/mL) Decreased , p¼0.02 X X
(Continued)
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 461
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
Prescribe d doses
taken, pharma cy
dispensin g
Non-adher ence
decreased, p¼0. 18, not
significant
XX
Self-repor ted adher-
ence over time
Only 44.9% of participan ts
in the inter vention resu p-
pressed th eir VL, modest
improvemen t in the pri-
mary outco me. It could
not be demonst rated that
EAMD could si gnificantly
improve adher ence
XX
Linnemayr et al 45 2017 Self- reported ad herence
over time
N, mean, 95% CI MEMS ad herence over
time for cont rol, SMS
only, and SMS þ
response: I TT and Com-
plete case
–VL suppress ed <400
copies/mL
Modest but n ot significant
improvemen t in viral
suppressi on
XX
Adherence Mean ad herence to ART
increased fr om baseline
values in the in tervent ion
group as compar ed with
the control gr oup after
6 months foll ow-up
(p¼0.035). The pro por-
tion of respon dents who
had good adhe rence
(>95%) was signi ficantly
higher in th e interventi on
group (n¼107, 92.2%) as
compared with con trol
group (n¼59, 54 .6%)
(p¼0.001) af ter
6 months foll ow-up.
XX
CD4 count A sign ificantly high er rise
in CD4 count (p¼0. 017)
was observe d in the inter-
vention grou p after
6 months foll ow up. CD4
count increase d by 146.01
cells/μL in the inter ven-
tion group wh ereas an
increase of 93. 62 cells/μL
was obser ved in the con-
trol group.
XX
Ruan et al47 2017 HI V-related an d HIV
mediation knowledge
and self-rep orted
adherence
N,%,mean,SD,Z-score VAS, CPCR A
Antiretrov iral
medication
self-repor t, CD4 count
–VAS Th e interven tion group
had a significa ntly higher
VAS mean score
(Z¼2.735, p¼0.006)
and lower su boptimal
adherence ra te
(Z¼2.208, p¼0.027) at
the end of th e study.
XX
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.462
differences.43 This difference could be explained by the
bidirectional nature of the inter vention, which could account
for increased subject par ticipation and engagement. Subjects
receiving daily texts may become accustomed to the content
of the messages, thus diminishing the effect through habi-
tuation. In studies with combined strategies, it may be
impossible to distinguish the isolated effect of the SMS
intervention alone, and for future work, examining the
efficacy of mobile reminders, the research methodology of
combined interventions should be tailored to differentiate
the effects via multiple cohorts.
Studies that included both adherence and clinical outcomes
could provide a more comprehensive view of their outcomes
than those that relied on self-report alone. Onestudy that had
both self-report and clinical outcomesfound that self-reported
ART adherence was >95% of prescribed doses in the past
30 days at both 6- and 12-month follow-up visits, and plasma
HIV-1 viral RNA load suppression was <400 copies/mL at
12 months.8In all, 11 studies collected self-reports in addition
to clinical metrics.8,26–28,32,34,35,37–39,42
Several studies used self-adherence reporting to measure
adherence without clinical measures. Six of the studies
reviewed had ‘self-reported adherence’as the primary out-
come measure, and lacked any clinical measure.29–32,44
While patient-reported non-adherence has been found to
be an accurate predictor of virologic outcomes,51–53 it is not a
replacement for clinically derived measures such as HIV R NA
viral load. Self-reported adherence is subject to inherent
cognitive bias, such as response bias and recall bias. Subjects
may not recall whether they have taken their medication,
especially in cases of cognitive dysfunction or decline54,55
and may incorrectly report as adherent. Another study
discusses the challenges and discrepancies due to measure-
ment methods.56
Illiteracy is a barrier to ART adherence.42 Of the papers
reviewed, only one had inclusion criteria that clearly
addressed illiteracy and owning or sharing a mobile phone.8
Six studies stated illiteracy and/or inability to read as specific
exclusion criteria, while the remaining authors did not
mention these criteria.8,26,29,30,41,43 Five studies required
the participants to be able to speak, read, and wr ite English at
the fourth-grade level.28,35,38,43,44
In most resource-limited settings, access to the second-line
treatment is scarce and has a significantly higher cost com-
pared with the first-line therapy, and accessto third-line ART is
non-existent.19 Of the reviewed studies, only one assessed the
effectiveness of an electronic adherence monitoring device
among patients failing the second-line ART, as measured by
viral load suppression (<400 copies/mL).28 The results of this
study are encouraging since patients who re-suppress viral
load at the first follow-up are more likely to remain virally
suppressed.
According to the International Telecommunications
Union,57 in 2016, per 100 people mobile cellular subscriptions,
for developing countries had a mobile cellular subscription rate
of 60 and for middle-income countries had a rate of 101. Sub-
Saharan Africa, for example, had a rate of 74. Middle-income
countrieswere further subdividedinto upper middle andlower
Table 2 (Continued)
Author Year Study focus Metric Outcome variab le Relative effe ct Outcome Notes Posit ive
outcome
Non-
positive
outcome
p<0.05 p0.05
CPCRA ant iretroviral
medication se lf-repor t
The percent age of
people with sub optimal
adherence in the cont rol
group (27%) was sign ifi-
cantly high er than that in
the inter vention group
(10.7%) in th e post-test
(p¼0.027).
XX
Stankievic h et al46 2017 ART ad herence, VL N, % VL (copies/ mL) –VL After the str ategy imple -
mentation, 20/22 VL
results we re availabl e.
13/20 (65%) were
undetecta ble, 14/20
(70%) had VL <1000
copies/mL. 6/20 (30%) VLs
had no change s;
no p-value reported.
X––
Abbreviations: AIDS, acquired immunodeficiency syndrome; AOR, adjusted odds ratio; ART, antiretroviral therapy; cART, combination antiretroviral therapy; CAS, composite adherence score; CI, confidence
interval; CPCRA, Community Programs for Clinical Research on AIDS; EAMD, electronic adherence monitoring device; HIV, human immunodeficiency virus; IQR, interquartile range; ITT, intention to treat; IVR,
interactive voice response; MEMS, medication event monitoring system; OR, odds ratio; SD, standard deviation; SMS, short message service; TI , treatment interruption; VAS, visual analog scale; VL, viral load.–>
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 463
middle, with rates of 108 and 95, respectively.57 This data
suggest that cell phone health interventions could be scaled to
large numbers of people even in low- and middle-income
countries.
One of the limitations of this review is that our search
included only articles written in English. However, we feel that
the papers in this review represent a global perspective on
these approaches. Additionally, completing a meta-analysis of
the reviewed studies is not possible due to the wide range of
research study designs, and included papers were not all RCTs.
Given that many of the published workson this subject are not
RCTs speaks to the continual evolution of the field and is a
reflection that the body of evidence is in development. Future
studies could include more reported data and sensitivity
analysis.
Privacy concerns are an opportunity for an investigation
to understand differences in adherence according to age, sex,
socioeconomic level, and geographical differences. Although
most of the studies we reviewed developed methods for
ensuring the privacy of the text reminders8,31–39,41–44 that
ranged from omitting references to HIV or ART to common
message greetings, there was no examination of whether
privacy concerns affected the viewing of the reminder and
maintenance of adherence.
Based on this review, it appears that studies that combine
multiple delivery methods (SMS and counseling) and measure
both adherence and clinical outcomes show particular promise
in improving care. Future mHealth ARTstudies shouldconsider
people who access mobile phones via a partner or family
member, include both self-report and clinically measurable
outcomes, evaluate the training for both patients and health
professionals, and pay more careful attention to randomization
and blinding of participants to reduce possible effects of bias.
The mental health of patients could be part of future study
designs. Interventions should be individualized in real time to
promote faster behavior correction by adjusting the content,
frequency, and length of messages. Studies should consider the
use of devices that can track medication pill dispensing. A cost
analysis would be useful for policymakers. The definition of
adherence needs to be more clearly defined and reported to
allow comparison of approaches. Future investigations could
examine the role of personalized messages and cultural differ-
ences in approaches.
Conclusion
This review shows that text message in low-cost cell phones
can be effective for promoting adherence, and this is particu-
larly useful to low- and middle-income countries. While the
majority of studies reviewed had positive outcomes in many
diverse global settings, there needs to belarger scale studies to
evaluate effect sizes.The analysis of potential bias showed that
the main potential source of bias was in the lack of blinding of
participants and personnel. While our search criteria did not
exclude apps, only one study used a mobile app as a supple-
mentary tool.27 All the studies that had text messaging com-
bined with additional support mechanisms had positive
outcomes.There is a need for more comprehensiveapproaches
and larger scale trials to confirm results observed in limited
cohort sizes and evaluate the impact of complimentary ser-
vices, such as counseling. The feasibility of such interventions
is increasingly possible even within low-resource settings due
to the expanding prevalence of mobile phoneavailability.57 To
better retain satisfactory adherence within the HIV popula-
tion, and especially in low-resource settings, we recommend
that future interventions incorporate multiple strategies:
mobile-based reminders, social support structures, and per-
sonalized content. Further workneeds to be done to determine
the optimal frequency of reminders, content length, and
tailoring to local customs and user preferences.
Clinical Relevance Statement
Phone text-based reminder systems are efficacious in enhan-
cing HIV ART adherence. Mobi le reminder systems have been
successfully implemented in low-resource settings. Content
and length should be tailored to local customs and user
preferences.
Multiple Choice Question
Which of the following should be considered while tailoring
text messaging-based adherence reminders for patients liv-
ing with HIV/AIDS?
a. Privacy
b. Length of message
c. Local customs and preferences
d. All of the above
Correct Answer: The correct answer is option d, all of the
above.
Protection of Human and Animal Subjects
Human or animal subjects were not included in thisproject.
Funding
None.
Conflict of Interest
None.
Acknowledgments
The searches were conducted by Paul Pain, Reference and
Education Librarian, Countway Library of Medicine, Har-
vard Medical School. We appreciate comments by Patricia
Stephens, PhD; Brian Haynes, MD; and Roger Davis, ScD,
on earlier versions of this paper.
References
1WHO. 2017. Number of people (all ages) living with HIV Estimates
by WHO region. Global Health Observatory data repository. World
Health Organization. Available at: http://apps.who.int/gho/data/
view.main.22100WHO?lang¼en. Accessed January 3, 2018
2Dryden-Peterson S, Bennett K, Hughes MD, et al. An augmented
SMS intervention to improve access to antenatal CD4 testing and
ART initiation in HIV-infected pregnant women: a cluster rando-
mized trial. PLoS One 2015;10(02):e0117181
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.464
3Anglemyer A, Rutherford GW, Horvath T, Baggaley RC, Egger M,
Siegfried N. Antiretroviral therapy for prevention of HIV trans-
mission in HIV-discordant couples. Cochrane Database Syst Rev
2013;4(04):CD009153
4Dyrehave C, Rasmussen DN, Hønge BL, et al; Bissau HIV Cohort
Study Group. Nonad herence is associated wit h lack of HIV-related
knowledge: a cross-sectional study among HIV-infected indivi-
duals in Guinea-Bissau. J Int Assoc Provid AIDS Care 2016;15(04):
350–358
5Horvath T, Azman H, Kennedy GE, Rutherford GW. Mobile phone
text messaging for promoting adhe rence to antiretroviral therapy
in patients with HIV infect ion. Cochrane Database Syst Rev 2012;3
(03):CD009756
6Joint United Nations Programme on HIV/AIDS (UNAIDS). Global
AIDS Update 2016. May 31, 2016. Available at: http://www.
unaids.org/sites/default/files/media_asset/global-AIDS-update-
2016_en.pdf. Accessed January 3, 2018
7U.S. Central Intelligence Agency (CIA). 2016. The World Factbook:
Country comparison: hiv/aids–adult prevalence ratE. Available at:
https://www.cia.gov/library/publications/the-world-factbook/ran-
korder/2155rank.html. Accessed January 3, 2018
8Lester RT, Ritvo P, Mills EJ, et al. Effects of a mobile phone short
message service on antiretroviral treatment adherence in
Kenya (WelTel Kenya1): a randomised trial. Lancet 2010;376
(9755):1838–1845
9Anglada-Martinez H, Riu-Viladoms G, Martin-Conde M, Rovira-
Illamola M, Sotoca-Momblona JM, Codina-Jane C. Does mHealth
increase adherence to medicatio n? Results of a systematic review.
Int J Clin Pract 2015;69(01):9–32
10 Checchi KD, Huybrechts KF, Avorn J, Kesselheim AS. Electronic
medication packaging devices and medication adherence: a sys-
tematic review. JAMA 2014;312(12):1237–1247
11 Aranda-Jan CB, Mohutsiwa-Dibe N, Loukanova S. Systematic
review on what works, what does not work and why of imple-
mentation of mobile health (mHealth) projects in Africa. BMC
Public Health 2014;14:188
12 Déglise C, Suggs LS, Odermatt P. SMS for disease control in
developing countries: a systematic review of mobile health
applications. J Telemed Telecare 2012;18(05):273–281
13 James DC, Harville C II, Sears C, Efunbumi O, Bondoc I. Participa-
tion of African Americans in e-Health and m-Health studies: a
systematic review. Telemed J E Health 2017;23(05):351–364
14 Bärnighausen T, Chaiyachati K, Chimbindi N, Peoples A, Haberer J,
Newell M-L. Interventions to increase antiretroviral adherence in
sub-Saharan Africa: a systematic review of evaluation studies.
Lancet Infect Dis 2011;11(12):942–951
15 Catalani C, Philbrick W, Fraser H, Mechael P, Israelski DM. mHealth
for HIV treatment & prevention: a systematic review of the
literature. Open AIDS J 2013;7:17–41
16 Devi BR, Syed-Abdul S, Kumar A, et al. mHealth: An updated
systematic review with a focus on HI V/AIDS and tuberculosis long
term management using mobile phones. Comput Methods Pro-
grams Biomed 2015;122(02):257–265
17 Hall AK, Cole-Lewis H, Bernhardt JM. Mobile text messaging for
health: a systematic review of reviews. Annu Rev Public Health
2015;36:393–415
18 Burns K, Keating P, Free C. A systematic review of randomised
control trials of sexual health interventions delivered by mobile
technologies. BMC Public Health 2016;16(01):778
19 Jeffery R A, Navarro T, Wilczynski NL, et al. Adherence measure-
ment and patient recruitment methods are poor in intervention
trials to improve patient adherence. J Clin Epidemiol 2014;67
(10):1076–1082
20 Chaiyachati KH, Ogbuoji O, Price M, Suthar AB, Negussie EK,
Bärnighausen T. Inter ventions to improve adherence to antire-
troviral therapy: a rapid systematic review. AIDS 2014;28
(Suppl 2):S187–S204
21 Belzer ME, MacDonell KK, Clark LF, et al; the Adolescent Medicine
Trials Network for HIV/AIDS Interventions. Acceptability and
feasibility of a cell phone support intervention for youth living
with HIV with nonadherence to antiretroviral therapy. AIDS
Patient Care STDS 2014;6(29):338–345
22 Hightow-Weidman LB, Muessig KE, Bauermeister J, Zhang C,
LeGrand S. Youth, technology, and HIV: recent advances and
future directions. Curr HIV/AIDS Rep 2015;12(04):500–515
23 Navarra AD, Gwadz MV, Whittemore R, et al. Health technology-
enabled interventions for adherence support and retention in
care among US HIV-infected adolescents and young adults: an
integrative review. AIDS Behav 2017;21(11):3154–3171
24 Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group.
Preferred reporting items for systematic reviews and meta-ana-
lyses: the PRISMA statement. PLoS Med 2009;6(07):e1000097
25 Berlin JA; University of Pennsylvania Meta-analysis Blinding
Study Group. Does blinding of readers affect the results of
meta-analyses? Lancet 1997;350(9072):185–186
26 Ammassari A, Trotta MP, Shalev N, et al. Timed short messaging
service improves adherence and virological outcomes in HIV-1-
infected patients with suboptimal adherence to antiretroviral
therapy. J Acquir Immune Defic Syndr 2011;58(04):e113–e115
27 Perera AI, Thomas MG, Moore JO, Faasse K, Petrie KJ. Effect of a
smartphone application incorporating personalized health-
related imagery on adherence to antiretroviral therapy: a rando-
mized clinical trial. AIDS Patient Care STDS 2014;28(11):579–586
28 Evans D, Berhanu R, Moyo F, Nguweneza A, Long L, Fox MP. Can
short-term use of electronic patient adherence monitoring
devices improve adherence in patients failing second-line anti-
retroviral therapy? Evidence from a pilot study in Johannesburg,
South Africa. AIDS Behav 2016;20(11):2717–2728
29 da Costa TM, Barbosa BJ, Gomes e Costa DA, et al. Results of a
randomized controlled trial to assess the effects of a mobile SMS-
based intervention on treatment adherence in HIV/AIDS-infected
Brazilian women and impressions and satisfact ion with respect to
incoming messages. Int J Med Inform 2012;81(04):257–269
30 Nsagha DS, Lange I, Fon PN, Nguedia Assob JC, Tanue EA.
A randomized controlled trial on the usefulness of mobile text
phone messages to improve the quality of care of HIV and AIDS
patients in Cameroon. Open AIDS J 2016;10:93–103
31 Pop-Eleches C, Thirumurthy H, Habyarimana JP, et al. Mobile
phone technologies improve adherence to antiretroviral treat-
ment in a resource-limited setting: a randomized controlled trial
of text message reminders. AIDS 2011;25(06):825–834
32 Maduka O, Tobin-West CI. Adherence counseling and reminder text
messages improve uptake of antiretroviral therapy in a tertiary
hospital in Nigeria. Niger J Clin Pract 2013;16(03):302–308
33 Rodrigues R, Shet A, Antony J, et al. Supporting adherence to
antiretroviral therapy with mobile phone reminders: resul ts from
a cohort in South India. PLoS One 2012;7(08):e40723
34 Sabin LL, Bachman DeSilva M, Gill CJ, et al. Improving adherence to
antiretroviral therapy with triggered real-time text message
reminders: the China adherence through technology study.
J Acquir Immune Defic Syndr 2015;69(05):551–559
35 Garofalo R, Kuhns LM, Hotton A, Johnson A, Muldoon A, Rice D. A
randomized controlled trial of personalized text message remin-
ders to promote medication adherence among HIV-positive ado-
lescents and young adults. AIDS Behav 2016;20(05):1049–1059
36 Hardy H, Kumar V, Doros G, et al. Randomized controlled trial of a
personalized cellular phone reminder system to enhance adher-
ence to antiretroviral therapy. AIDS Patient Care STDS 2011;25
(03):153–161
37 Lewis MA, Uhrig JD, Bann CM, et al. Tailored text messaging
intervention for HIV adherence: a proof-of-concept study. Health
Psychol 2013;32(03):248–253
38 Haberer JE, Musiimenta A, Atukunda EC, et al. Short message
service (SMS) reminders and real-time adherence monitoring
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al. 465
improve antiretroviral therapy adherence in rural Uganda. AIDS
2016;30(08):1295–1300
39 Orrell C, Cohen K, Mauff K, Bangsberg DR, Maartens G, Wood R. A
randomized controlled trial of real-time electronic adherence
monitoring with text message dosing reminders in people start-
ing first-line antiretroviral therapy. AIDS 2015;70(05):495–502
40 Kassaye SG, Ong’ech J, Sirengo M, et al. Cluster-randomized
controlled study of SMS text messages for prevention of
mother-to-child transmission of HIV in rural Kenya. Aids Res
Treat 2016;2016:1289328
41 Mbuagbaw L, Thabane L, Ongolo-Zogo P, et al. The Cameroon
Mobile Phone SMS (CAMPS) trial: a randomized trial of text
messaging versus usual care for adherence to antiretroviral
therapy. PLoS One 2012;7(12):e46909
42 KalichmanSC, Kalichman MO, Cherry C, EatonLA, Cruess D, Schinazi
RF. Randomized factorial trial of phone-delivered support counsel-
ing and daily text message remindersfor HIV treatment adherence.
J Acquir Immune Defic Syndr 2016;73(01):47–54
43 Ingersoll KS, Dillingham RA, Hettema JE, et al. Pilot RCT of
bidirectional text messaging for ART adherence among nonurban
substance users with HIV. Health Psychol 2015;34S:1305–1315
44 Moore DJ, Poquette A, Casaletto KB, et al; HIV Neurobehavioral
Research Program (HNRP) Group. Individualized texting for
adherence building (iTAB): improving antiretroviral dose timing
among HIV-infected persons with co-occurring bipolar disorder.
AIDS Behav 2015;19(03):459–471
45 Linnemayr S, Huang H, Luoto J, et al. T ext messaging for improving
antiretroviral adherence: no effect after 1 year in a randomized
controlled trial among adolescents and young adults. Am J Public
Health 2017;107(12):1944–1950
46 Stankievich E, Malanca A, Foradori I, Ivalo S, Losso M. Utility of
mobile communication devices as a tool to improve adherence to
antiretroviral treatment in HIV-infected children and young
adults in Argentina. Pediatr Infect Dis J 2018;37(04):345–348
47 Ruan Y, Xiao X, ChenJ, Li X, Williams AB, Wang H. Acceptability and
efficacy of interactive short message serviceinterventionin improv-
ing HIV medication adherence in Chinese antiretroviral treatment-
naïve individuals. Patient Prefer Adherence 2017;11:221–228
48 Georgette N, Siedner MJ, Petty CR, Zanoni BC, Carpenter S,
Haberer JE. Impact of a clinical program using weekly Short
Message Service (SMS) on antiretroviral therapy adherence sup-
port in South Africa: a retrospective cohort study. BMC Med
Inform Decis Mak 2017;17(01):18
49 King E, Kinvig K, Steif J, et al. Mobile text messaging to improve
medication adherence and viral load in a vulnerable Canadian
population living with human immunodeficiency virus: a
repeated measures study. J Med Internet Res 2017;19(06):e190
50 Abdulrahman SA, Rampal L, Ibrahim F, Radhakrishnan AP, Kadir Shahar
H, Othman N. Mobile phone reminders and peer counseling improve
adherence and treatment outcomes of patients on ART in Malaysia: a
randomized clinical trial. PLoS One 2017;12(05):e0177698
51 Haubrich RH, Little SJ, Currier JS, et al; California Collaborative
Treatment Group. The value of patient-reported adherence to
antiretroviral therapy in predicting virologic and immunologic
response. AIDS 1999;13(09):1099–1107
52 Nieuwkerk PT, Oort FJ. Self-reported adherence to antiretroviral
therapy for HIV-1 infection and virologic treatment response: a
meta-analysis. J Acquir Immune Defic Syndr 2005;38(04):445–448
53 Simoni JM, Kurth AE, Pearson CR, Pantalone DW, Merrill JO, Frick
PA. Self-report measures of antiretroviral therapy adherence: a
review with recommendations for HIV research and clinical
management. AIDS Behav 2006;10(03):227–245
54 Hinkin CH, Hardy DJ, Mason KI, et al. Medication adherence in
HIV-infected adults: effect of patient age, cognitive status, and
substance abuse. AIDS 2004;18(Suppl 1):S19–S25
55 Bonnet F, Amieva H, Marquant F, et al; S CO3 Aquitaine Cohort.
Cognitive disorders in HIV-infected patients: are they HIV-
related? AIDS 2013;27(03):391–400
56 Ajose O, Mookerjee S, Mills EJ, Boulle A, Ford N. Treatment
outcomes of patients on second-line antiretroviral therapy in
resource-limited settings : a systematic review and meta-analysis.
AIDS 2012;26(08):929–938
57 Telecommunication Union World Telecommunication/ ICT Indi-
cators database; World Development Indicators database (IT.CEL.
SETS.P2, IT.MLT.MAIN.P2). Available at: https://data.worldbank.
org/indicator/IT.CEL.SETS.P2. Accessed January 3, 2018
Appendix A: PubMed Search Strategy
PubMed
2017–12–11
371 Records
(“HIV Infections”[mesh] OR “Antiretroviral Therapy,
Highly Active”[mesh] OR “Anti-HIV Agents”[mesh] OR
human immunodeficiency virus[tiab] OR immunodeficiency
syndrome[tiab] OR HIV[tiab] OR aids[tiab] OR antiretroviral
[tiab] OR antiretroviral[tiab] OR haart[tiab])
AND
(“Cell Phone”[mesh] OR “Smartphone”[mesh] OR “Mobile
Applications”[mesh] OR cellular phone
[tiab] OR cellular
telephone
[tiab] OR cellular device
[tiab] OR cell pho-
ne
[tiab] OR cellphone
[tiab] OR mobile telephone
[tiab]
OR mobile phone
[tiab] OR smartphone
[tiab] OR smart
phone
[tiab] OR wireless phone
[tiab] OR wireless telepho-
ne
[tiab] OR mobile app
[tiab] OR text messag
[tiab] OR text
reminder
[tiab] OR short message service
[tiab] OR short
messaging service
[tiab] OR short message system[tiab] OR
short messaging system[tiab] OR sms[tiab] OR texting[tiab])
AND
(“Patient Compliance”[mesh] OR compliance[tiab] OR
adherence[tiab] OR noncompliance[tiab] OR nonadherence
[tiab] OR attendance[tiab] OR retention[tiab] OR return to
clinic[tiab])
Applied Clinical Informatics Vol. 9 No. 2/2018
Promoting Adherence to Antiretroviral Therapy Quintana et al.466