Technology-based self-care methods of improving antiretroviral adherence: a systematic review.

Abstract

As HIV infection has shifted to a chronic condition, self-care practices have emerged as an important topic for HIV-positive individuals in maintaining an optimal level of health. Self-care refers to activities that patients undertake to maintain and improve health, such as strategies to achieve and maintain high levels of antiretroviral adherence.
Technology-based methods are increasingly used to enhance antiretroviral adherence; therefore, we systematically reviewed the literature to examine technology-based self-care methods that HIV-positive individuals utilize to improve adherence. Seven electronic databases were searched from 1/1/1980 through 12/31/2010. We included quantitative and qualitative studies. Among quantitative studies, the primary outcomes included ARV adherence, viral load, and CD4+ cell count and secondary outcomes consisted of quality of life, adverse effects, and feasibility/acceptability data. For qualitative/descriptive studies, interview themes, reports of use, and perceptions of use were summarized. Thirty-six publications were included (24 quantitative and 12 qualitative/descriptive). Studies with exclusive utilization of medication reminder devices demonstrated less evidence of enhancing adherence in comparison to multi-component methods.
This systematic review offers support for self-care technology-based approaches that may result in improved antiretroviral adherence. There was a clear pattern of results that favored individually-tailored, multi-function technologies, which allowed for periodic communication with health care providers rather than sole reliance on electronic reminder devices.

Full text

Page 1

Technology-Based Self-Care Methods of Improving
Antiretroviral Adherence: A Systematic Review
Parya Saberi*, Mallory O. Johnson
Department of Medicine, University of CaliforniaSan Francisco, San Francisco, California, United States of America
Abstract
Background: As HIV infection has shifted to a chronic condition, self-care practices have emerged as an important topic for
HIV-positive individuals in maintaining an optimal level of health. Self-care refers to activities that patients undertake to
maintain and improve health, such as strategies to achieve and maintain high levels of antiretroviral adherence.
Methodology/Principal Findings: Technology-based methods are increasingly used to enhance antiretroviral adherence;
therefore, we systematically reviewed the literature to examine technology-based self-care methods that HIV-positive
individuals utilize to improve adherence. Seven electronic databases were searched from 1/1/1980 through 12/31/2010. We
included quantitative and qualitative studies. Among quantitative studies, the primary outcomes included ARV adherence,
viral load, and CD4+ cell count and secondary outcomes consisted of quality of life, adverse effects, and feasibility/
acceptability data. For qualitative/descriptive studies, interview themes, reports of use, and perceptions of use were
summarized. Thirty-six publications were included (24 quantitative and 12 qualitative/descriptive). Studies with exclusive
utilization of medication reminder devices demonstrated less evidence of enhancing adherence in comparison to multi-
component methods.
Conclusions/Significance: This systematic review offers support for self-care technology-based approaches that may result
in improved antiretroviral adherence. There was a clear pattern of results that favored individually-tailored, multi-function
technologies, which allowed for periodic communication with health care providers rather than sole reliance on electronic
reminder devices.
Citation: Saberi P, Johnson MO (2011) Technology-Based Self-Care Methods of Improving Antiretroviral Adherence: A Systematic Review. PLoS ONE 6(11):
e27533. doi:10.1371/journal.pone.0027533
Editor: Susanne Hempel, RAND Corporation, United States of America
Received June 15, 2011; Accepted October 19, 2011; Published November 3 , 2011
Copyright: ? 2011 Saberi, Johnson. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Institute of Mental Health (F32MH086323 and K24MH087220) The content is solely the responsibility of the
authors and does not necessarily represent the official views of the National Institutes of Health. The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: parya.saberi@ucsf.edu
Introduction
As HIV infection has evolved from an acute to a chronic illness,
much of the medical treatment of HIV-positive patients has shifted
from critical care to outpatient settings. Consequently, self-care
practices of individuals living with HIV have emerged as a
significanttopic for disease
[1,2,3,4,5,6]. Optimal adherence to antiretroviral (ARV) therapy
is among the most important aspects of these practices and an
emergent strategy to improve ARV adherence is the use of
technology-based methods. The strength of technology lies in its
ability to transcend borders, cultures, and languages; therefore,
understanding self-care technology-based strategies used by HIV-
positive individuals to improve adherence is critical for providers
and researchers who seek to support patients in enhancing
adherence while simultaneously utilizing existing resources and
limiting cost.
Individual self-care has been defined in numerous ways
[7,8,9,10,11,12]. A broad definition of self-care refers to ‘‘those
activities individuals undertake in promoting their own health,
preventing their own disease, limiting their own illness, and
restoring their own health [7,8,9].’’ These activities are generally
treatment andmanagement
informed by technical knowledge of health care professionals and
lay experience, but are undertaken without professional support.
Self-care has also been defined as the ‘‘naturalistic decision making
process involving the choice of behaviors that maintain physiologic
stability (maintenance) and the response to symptoms when they
occur (management)’’ [11]. Therefore, self-care maintenance
includes health-promoting habits, adhering to treatment regimens,
and monitoring and managing symptoms. More explicitly, HIV-
specific self-care behaviors include ARV adherence and engage-
ment in care [13].
High ARV adherence is associated with enhanced CD4+ cell
count, reductions in HIV viral load, and decreased morbidity and
mortality [14,15,16]. Conversely, non-adherence may result in
virologic rebound, ARV drug resistance, transmission of drug-
resistant virus, and progression to AIDS [17,18,19,20,21]. Despite
the necessity of high adherence, in the U.S. and Europe the
percentage of prescribed doses taken has been estimated to range
from 60–70% [22,23,24,25,26,27]. ‘‘Forgetfulness’’ is commonly
cited as the top reason for missing doses [28]; therefore, many
researchers have investigated the role of electronic reminder
devices, such as alarms and pagers, to improve adherence. The
U.S. Department of Health and Human Services [29], the British
PLoS ONE | www.plosone.org1 November 2011 | Volume 6 | Issue 11 | e27533
0

Page 2

HIV Association [30] and the World Health Organization [31]
have acknowledged the supportive role of technology-based
methods to improve adherence. This recognition underscores
the need for stronger evidence of the effectiveness of these
technologies and the identification of cost-containing strategies for
improving adherence.
We conducted a systematic review of studies that explored the
use and impact of technology-based methods by HIV-positive
individuals for improving ARV adherence. The purpose of this
review was to extend prior reviews examining the impact of
electronic reminder devices [32] and the efficacy of interventions
[33] on adherence. Specifically, we focused on the use of self-care
technology-based adherence strategies.
Methods
Objective
The primary objective of this systematic review was to evaluate
the impact of self-care technology-based methods on ARV
adherence. We report the efficacy (adherence, HIV viral load,
and CD4+ cell count) and other secondary outcomes of using self-
care technology-based methods.
Data Sources
Initially, we searched PubMed, EMBASE, Cochrane Central,
Web of Science, and PsycINFO from 1/1/1980 through 12/31/
2010. Additionally, we screened the references of all pertinent
articles to identify additional relevant publications.
Search Strategy
The search strategy was in the style of Cochrane Highly
Sensitive Search Strategy [34] for identifying reports of random-
ized, non-randomized, observational, and qualitative studies in
PubMed, as well as the appropriate MeSH terms, and a wide
range of relevant search terms in all databases. The detailed search
strategy used for PubMed can be found in Table S1 This strategy
was modified as appropriate for use in other databases. We
included all quantitative and qualitative studies (including
descriptive studies) published in the English language.
Inclusion/Exclusion Criteria
We included research regarding the impact of technology-based
methods used by HIV-positive individuals on our primary
outcomes (ARV adherence, HIV viral load, and CD4+ cell
count), secondary outcomes (quality of life, adverse effects, and
feasibility/acceptability data), and outcomes of qualitative/de-
scriptive studies (interview themes, reports of use, and perceptions
of use). Among quantitative studies with our primary outcomes,
findings were contrasted across groups receiving and not receiving
the intervention or in before-after comparisons. Only studies
published in English but regardless of geographical location were
included in the review.
Technology-based methods were defined as devices such as
electronic reminder devices (including alarms, electronic pillboxes,
and pagers), mobile telephones (for automated functions such as
automated text messages and automated alarms), personal digital
assistants (PDAs), computer software, and Internet and mobile
applications. These included tools that may have been initially set
up or implemented by a researcher/clinician, but that the
participant/patient could use independent of the researcher/
clinician for adherence self-care. This decision was made to set
apart self-care techniques that an individual could utilize
independent of their health-care providers from methods that
required constant interaction/supervision of a health professional.
Therefore, reviewed studies did not include adherence monitoring
devices (e.g., medication event monitoring systems or MEMs) or
any method that clinicians used to monitor patients’ adherence to
give feedback. We did not include studies that examined
technologies that facilitated the interactions between patients/
participants and clinicians/researchers (such as email, text
messaging, or telephone) because we did not view these methods
as strictly promoting self-care. Multifactorial interventions con-
taining at least one self-care technology-based method were
included.
Review Method and Data Abstraction
Using the EndNote software package, relevant studies were
located in the above-mentioned data sources and duplicates and
irrelevant articles were removed by one author. One author and
the research assistant read the remaining citations and identified
eligible studies based on pre-specified inclusion/exclusion criteria.
All uncertainties and disagreements were arbitrated by the second
author. Using a data abstraction form, one author and the
research assistant summarized pertinent information from includ-
ed articles.
Outcome Variables
Primary outcomes included ARV adherence (based on self-
report, pill-counts, pharmacy refill records, MEMS), HIV viral
load, and CD4+ cell count. Secondary outcomes consisted of
quality of life, adverse effects, and feasibility/acceptability data.
For qualitative and descriptive studies, interview themes, reports of
use, and perceptions of use were summarized.
Results
From 1,207 gross results, 36 publications met our eligibility
criteria and were included (Figure 1). Among these publications,
24 were quantitative, from which 16 reported on our primary
outcomes (adherence, viral load, CD4+ cell count) [35,36,37,
38,39,40,41,42,43,44,45,46,47,48,49,50] and 8 stated information
regarding our secondary outcomes (quality of life and feasibility/
acceptability) [51,52,53,54,55,56,57,58]. Table 1 and Table S2
summarize these studies. An additional 12 qualitative and
descriptive studies were identified that are summarized in Table
S3 [59,60,61,62,63,64,65,66,67,68,69,70].
Quantitative Research
Publications with Primary Outcome.
were mainly published between 2001 through 2010 (with the
exception of one published in 1992 [48]) and were primarily
conducted in the U.S. (75%). Baseline sample size ranged from
23–928 (median=98); from studies where mean age is presented,
mean age ranged from 36–43 years; percentage of male
participants ranged from 0–98% (median=80%); and within the
U.S. studies, the percentage of participants who were Black ranged
from 20–100% (median=47%). The most common method of
adherence assessment was self-report (63%), followed by a
combination of self-report and another method (such as MEMS
caps and pill counts) (31%), and solely MEMS caps (6%).
Studies not examining a research intervention- In five of the 16 studies,
the relationship between technology-based methods and adher-
ence was reported [37,38,39,40,48]. These studies presented
conflicting results on the positive [37,39,48] or neutral [38,40]
effect of these strategies on adherence. In the only quantitative
study examining Internet use among regular Internet users [39],
those who did not use the Internet to seek health information were
more non-adherent than those who used it for this purpose.
These 16 studies
Technology and ARV Adherence Self-Care
PLoS ONE | www.plosone.org2 November 2011 | Volume 6 | Issue 11 | e27533

Page 3

Stand-alone technology-based interventions-Two studies reported the
effect of stand-alone technology-based interventions [35,46],
consisting of electronic reminder devices, such as a pager or a
programmable medication reminder device providing verbal
reminders. In participants receiving individualized adherence
counseling sessions, the use of the Disease Management Assistance
System (DMAS) device, an electronic device that produces a timed
voice message to prompt subjects to take ARVs, resulted in a mean
adherence of 80% in the intervention arm versus 65% in the
control arm, which was not statistically significant [35]. Post-hoc
analyses noted an effect in memory impaired individuals.
Surprisingly, the use of DMAS was associated with some
deterioration in quality of life (see ‘‘Publications with Secondary
Outcomes’’) [58]. Safren and colleagues reported a statistically
significant increase in adherence with the use of pagers; however,
this improvement was not clinically significant, as adherence
remained poor (#70%) at points of outcome assessment in both
arms of the study [46].
Multi-component interventions including a technology-based method-We
identified nine publications that examined the effects of multi-
component interventions, including a self-care technology-based
method [36,41,42,43,44,45,47,49,50]. These interventions also
included individualizedcounseling
47,49,50], group sessions [42,45], or a combination of one-on-
one and group sessions [44]. The technology-based adherence
strategies consisted of mobile telephone automated text messages,
alarms, beepers, and wrist watches with alarms. In these studies,
four reported enhanced ARV adherence [36,41,42,49], two
revealedatrendforstatistically
[44,45], two showed mixed results (improved adherence with
counseling support but not with electronic reminder devices)
[43,50], and one did not result in changes in adherence [47].
Increased CD4+ cell count was observed in one publication [42];
appointments [36,41,43,
significantimprovements
however, the remaining studies either did not detect any changes
[36,41,43,47,49,50] or did not report this value [44,45]. In one
study, an increase in the number of individuals with undetectable
viral load was reported [42], five did not detect any changes in
viral load [36,41,47,49,50], two did not report this outcome
[44,45], and one showed a statistically significant increased rate of
virologic failure with the use of reminder devices [43]. The median
length of follow-up in these studies was 20 weeks (range=4–52
weeks).
In a 2-by-2 factorial design study, including a medication
manager or medication alarm, the use of individualized,
structured, long-term adherence support strategies from trained
medication managers was associated with higher reports of perfect
adherence and 13% lower rates of virologic failure in comparison
to no medication manager [43]. However, use of a medication
alarm did not produce a significant difference in adherence but
resulted in 25% higher rates of virologic failure in comparison to
not using medication alarms. Similarly, participants were ran-
domized in another 2-by-2 factorial design to a peer-support
intervention or a pager messaging strategy [50]. The use of pager
did not result in increased odds of reporting 100% adherence;
however at six months, there was a trend for decreased adherence.
Publications with Secondary Outcomes
These eight studies were published between 2000 and 2010 and
75% were conducted in the U.S. Sample sizes ranged from 10–
300 (median=30); mean age ranged from 31–43 years; and
percentage of male participants ranged from 0–88% (medi-
an=56%).
Feasibility and acceptability- From the seven publications that
evaluated feasibility and acceptability of technology-based self-care
methods [51,52,53,54,55,56,57], all concluded that these methods
were feasible and acceptable. Among these studies, four examined
Figure 1. Selection process for study inclusion.
doi:10.1371/journal.pone.0027533.g001
Technology and ARV Adherence Self-Care
PLoS ONE | www.plosone.org3 November 2011 | Volume 6 | Issue 11 | e27533

Page 4

Table 1. Summary of quantitative studies with primary outcomes.
Source
Country, City/State
Start-End Year Sample size
Age (years)
% Male% MSM% BL% WH
Inclusion /
Exclusion
Criteria
Study Objectives
If examined
interventions,
description of
intervention
Length of
follow-up
Outcomes
Method of
AdherenceAssessment:
Interval
Adherence
Outcomes
Viral load
(copies/mL)
CD4+ + cell count
(cells/mm3)
OtherOutcomes
Andrade
et al,
2005 [35]
Wu et al,
2006 [58]
United States,
Maryland
1999–2000
N=58
Mean age=38
59%
N/R
88%
N/R
Inclusion: $18 years,
able to self-medicate,
care at Johns Hopkins
Moore HIV Clinic,
treatment-naı ¨ve &
initiating ARV or
treatment-experienced
& switching ARV (had
#3 prior ARV regimens)
Exclusion: inability to
self-medicate, severe
dementia,
institutionalization
Assess if use of
Disease Management
Assistance System
(DMAS: programmable
medication reminder
device providing
verbal reminders at
ARV dosing times)
device improves ARV
adherence, viral load, and CD4+.
Control: Monthly
individualized 30-minute
adherence counseling
session+standardized
adherence feedback
transcript (education on
barriers of adherence &
hazards of non-adherence)
Intervention: Same as
controls+DMAS
24 weeks
- EDM & self-
report: 4 days
- Overall mean
adherence by
electronic drug- exposure
monitoring
caps: 80%
in DMAS
vs 65%
in control (NS)
- Undetectable viral
load: 34% in DMAS vs
38% in controls (p=0.49)
- 1log10 reduction: 72%
in DMAS vs 41% in
controls (p=0.02)
- Mean reduction:
22.1log10 DMAS vs
20.98log10 controls
(p=0.02)
Mean CD4+: 301+/
2172 DMAS vs 250+/
2172 controls
(p=0.28)
Post-hoc
analysis:
- Mean
adherence in
memory
impaired
(n=31): 77%
DMAS vs 57%
controls
(p=0.001)
- Mean
adherence in
memory intact:
83% DMAS vs77% controls
(p=0.25)
QOL: controls
had improved
& DMAS had
reduced QOL
score.
Technology and ARV Adherence Self-Care
PLoS ONE | www.plosone.org4November 2011 | Volume 6 | Issue 11 | e27533

Page 5

Source
Country, City/State
Start-End YearSample size
Age (years)
% Male% MSM% BL% WH
Inclusion /
Exclusion
Criteria
Study Objectives
If examined
interventions,
description of
intervention
Length of
follow-up
Outcomes
Method of
AdherenceAssessment:
Interval
Adherence
Outcomes
Viral load
(copies/mL)
CD4+ + cell count
(cells/mm3)
OtherOutcomes
Fairley
et al,
2003 [36]
Australia,
Melbourne
2001–2002
N=43
Mean age=38
98%91%
N/R N/R
Inclusion: $18 years,
not planning to
interrupt or change
ARVs in next 3 month,
had missed at least 1
dose of treatment by
self-report in last
month
Determine if a
comprehensive
adherence package
improved self-
reported ARV
adherence pre- and
post-intervention.
Use of adherence package:
an educational program (on
HIV, HIV treatment,
importance of adherence),
medication planner, & choice
of adherence aids (pillbox,
text messaging at scheduled
doses, or medication alarm).
5 months
- Self-report:
4, 7, 28 days
- Missed doses
decreased in last
4 days (0.76 to 0.38,
p=0.03) & last 7 days
(1.5 to 0.74, p=0.005),
but not last 28 days
(2.5 to 2.5, p=0.63)
- Morisky score:
pre- intervention=2.9,post- intervention=3.3
(p=0.006)
Undetectable viral load:
73% pre-intervention
vs 74% post-intervention
(p=1.0)
Mean CD4+: 513
pre-intervention
vs 551 post-
intervention (p=0.8)
Report of use:
17 began timed
pillbox, 13 usedplain pillbox, 11
SMS text, 6
declined aids
Golin et al,
2002 [37]
United States, North
Carolina
1998–1999
N=117
Mean age=38
80%
N/R
26%16%
Inclusion: English/
Spanish speaking,
newly initiating
PIs or NNRTIs
Examine relationship
between adherence & patient factors, regimen
factors (adherence aids
such as medication lists,
timers, & pillboxes),
clinical interaction, &
social factors.
None
48 weeks
- MEMS, pill count,
self-report: 4 weeks
- Adherence in those
using no adherence
aids=67.5% vsadherence=76%among top quartile of adherence aid users
(p=0.01)
Iroha et al,
2010 [38]
Nigeria, Lagos
2008
N=212
43%: 60–119
months
48.1%
N/R N/RN/R
Inclusion: Caregivers
of children whohad been on ARVs
for at least 30 days
& who consented
to participate
Determine level of ARV
adherence among
pediatric patients &barriers & facilitators of
adherence according to
caregivers.
None
None
- Self-report: 3 days
- Use of reminders
not associated with
adherence
- 30% of those adherent
used reminders vs 31%of those non-adherent
Table 1. Cont.
Technology and ARV Adherence Self-Care
PLoS ONE | www.plosone.org5November 2011 | Volume 6 | Issue 11 | e27533