DOI: 10.1111/j.1468-1293.2011.00928.

x
HIV Medicine (2011), 12, 562–569 r 2011 British HIV Association
ORIGINAL RESEARCH

Adherence to antiretroviral therapy: factors independently

associated with reported difficulty taking antiretroviral
therapy in a national sample of HIV-positive Australians
J Grierson, RL Koelmeyer, A Smith and M Pitts
Australian Research Centre in Sex, Health and Society, La Trobe University, Melbourne, Australia

Objectives
Given the importance of adherence to combination antiretroviral therapy (cART) for the reduced
morbidity and improved mortality of people living with HIV infection (PLWH), we set out to
determine which of a number of previously investigated personal, socioeconomic, treatment-related
and disease-related factors were independently associated with self-reported difficulty taking
antiretroviral therapy (ART) in an Australian sample of PLWH.
Methods
Using data from a national cross-sectional survey of 1106 PLWH, we conducted bivariate and
multivariable analyses to assess the association of over 70 previously investigated factors with self-
reported difficulty taking ART. Factors that maintained an association with reported difficulty taking
ART at the level of a 5 0.05 in the multivariable logistic regression analysis were considered to be
independently associated with reported difficulty taking ART.
Results
A total of 867 (78.4%) survey respondents were taking antiretroviral medication at the time of
completing the HIV Futures 6 survey. Overall, 39.1% of these respondents reported difficulty taking
ART. Factors found to be independently associated with reported difficulty taking ART included
younger age, alcohol and party drug use, poor or fair self-reported health, diagnosis of a mental
health condition, living in a regional centre, taking more than one ART dose per day, experiencing
physical adverse events or health service discrimination, certain types of ART regimen and specific
attitudes towards ART and HIV.
Conclusions
Thirteen previously investigated factors were found to be independently associated with reported
difficulty taking ART, reaffirming the dynamic nature of adherence behaviour and the ongoing
importance of addressing adherence behaviour in the clinical management of PLWH.
Keywords: adherence, antiretroviral therapy, Australia, HIV, nonadherence
Accepted 8 February 2011

gical response and reduced morbidity and mortality in

Introduction
Combination antiretroviral therapy (cART) has revolutio-
nized the course of HIV disease, transforming HIV infection
from a life-threatening infection to a manageable chronic
condition, particularly in developed countries [1–3].
However, a key challenge is the high level of adherence
to cART that is required for viral suppression, immunolo-
Correspondence: Dr Jeffrey Grierson, Australian Research Centre in Sex,
Health and Society, La Trobe University, 215 Franklin Street, Melbourne,
Victoria 3000, Australia. Tel: 1 61 3 9285 5356; e-mail:
j.grierson@latrobe.edu.au
individuals with HIV/AIDS [4–6]. Studies have demon-
strated a requirement for adherence levels of at least 95%
in order to achieve adequate viral suppression for regimens
including unboosted protease inhibitor (PI) therapy [4,7].
While a lower level of adherence has been shown to lead to
viral suppression when using nonnucleoside reverse
transcriptase inhibitor (NNRTI) regimens [7], the develop-
ment of treatment-resistant mutations has been shown to
peak at approximately 70% adherence and there is an
ongoing decline in the rate of developing treatment-
resistant mutations towards 100% adherence, providing a

562

further imperative for achieving near-perfect adherence to
cART [8].
A number of studies have been conducted to elucidate
the factors that are associated with suboptimal adherence
to cART. Such factors can be broadly classified into four
categories: (i) personal factors, (ii) socioeconomic factors,
(iii) treatment-related factors and (iv) disease-related
factors. Of the personal factors studied, lower age, lower
self-efficacy for adherence, psychiatric comorbidity, active
substance use, alcohol consumption, stressful life events
and certain beliefs about treatment and HIV have been
found to be independently associated with nonadherence
to cART [9]. Gender, a history of injecting drug use, risk
factor(s) for HIV infection and marital status have generally
not been associated with nonadherence to cART [9,10].
Socioeconomic factors have generally not been found to be
associated with nonadherence to cART, although a lack of
social support and unstable housing have been associated
with nonadherence [9,11]. Of those treatment-related
factors investigated, a greater number of doses per day
and certain adverse events, typically physical symptoms,
have been associated with nonadherence [4,9,12–16]. The
number and type of prescribed antiretroviral drugs, and the
total number of pills per day have been inconsistently
associated with nonadherence to cART [9,10,14,17]. The
length of time on treatment and the prior number of cART
regimens have generally not been associated with non-
adherence [9,17]. Disease-related factors [CD4 cell count,
duration of HIV infection and diagnosis of an AIDS-
defining illness (ADI)] have generally not been associated
with nonadherence [9,10,18–20].
There is considerable inconsistency in which factors are
independently associated with nonadherence to cART. This
is probably attributable to five factors: (i) the use of
different measures and definitions of adherence [9,21–23];
(ii) variation in the factors assessed in each study [9]; (iii)
differences in the demographics of the study samples
[9,24]; (iv) the cross-sectional nature of most studies [9];
and (v) the dynamic nature of adherence behaviour [9]. A
further limitation of the existing literature is the fact that it
is dominated by studies conducted in the USA, as well as
studies of specific subgroups of HIV-positive individuals
(e.g. injecting drug users, homeless individuals, incarcer-
ated individuals and clinic-based samples of patients) [24].
We previously conducted a national, community-based
survey of HIV-positive people in Australia (the HIV Futures
6 survey), assessing a broad range of factors associated
with the lived experience of being HIV-positive in Australia
[25]. We have information on the demographics, socio-
economic characteristics, health, relationships and atti-
tudes of people living with HIV infection (PLWH) in
Australia. In the current study, we set out to determine
r 2011 British HIV Association HIV Medicine (2011), 12, 562–569
Adherence to ART in HIV-positive Australians 563
which personal, socioeconomic, treatment-related and
disease-related characteristics were independently asso-
ciated with reported difficulty taking antiretroviral therapy
(ART) in those respondents who were taking ART at the
time of completing the HIV Futures 6 survey.
Methods
The HIV futures 6 survey
The HIV Futures 6 survey was an anonymous, self-
complete, cross-sectional survey. The survey contained
189 items organized into eight sections: demographics;
accommodation; health and treatments; services and
communities; sex and relationships; employment; re-
creational drug use; and finances. The survey was largely
based on the HIV Futures 5 survey [26], which was in
turn based on the four previous surveys [27–30]. The
content of the survey was developed in consultation with a
number of organizations and individuals in the HIV/AIDS
sector.
Survey respondents were recruited through community
organizations and clinical settings, as well as through
online and paper-based advertisements in community
organization and gay media within Australia. Previous
survey respondents who indicated that they were interested
in participating in future research projects were also
approached. Any HIV-positive individual residing in
Australia was eligible to complete the survey. Data were
collected from October 2008 to April 2009.
This study
Outcome variable
The HIV Futures 6 survey included two items that asked
respondents about their adherence to ART over the
previous 2 days: ‘How many doses (dose times) of
antiretroviral drugs did you miss yesterday?’ and ‘How
many doses (dose times) of antiretroviral drugs did you
miss the day before yesterday?’, with scores in the range 0–
5 (a score of 5 representing  5 missed doses). The data
from these survey items were highly skewed, with only
1.5% [13] of those respondents currently taking ART
indicating any nonadherence in the previous 2 days. As a
result, we needed to use a proxy variable to assess factors
associated with nonadherence to cART. We considered
using two other survey items: (i) self-reported most recent
viral load (detectable vs. undetectable) and (ii) self-reported
difficulty taking ART (‘Do you experience any difficulties
in taking antiretroviral drugs?’; yes/no responses). The
viral load variable was also fairly skewed, with only 48
respondents currently taking ART (5.5%) reporting a
564 J Grierson et al.
detectable viral load. Hence, we chose to use self-reported
difficulty taking ART as our outcome variable. This
variable was found to be highly associated with both
self-reported adherence (Fisher’s exact test; P 5 0.001) and
respondents’ most recent viral load test result (detectable
vs. undetectable viral load; w2-test; P 5 0.018), and was
therefore deemed to be a suitable proxy variable for
investigating factors associated with poor adherence to
ART.
Potential explanatory variables
We conducted an extensive literature review to identify
variables within our data set that were likely to be
associated with adherence to cART and/or reported
difficulty taking ART. We identified over 70 personal,
socioeconomic, treatment-related and disease-related
characteristics within the HIV Futures 6 data set that were
likely to be associated with treatment adherence and/or
difficulty taking ART. A full list of the potential
explanatory variables included in this analysis is provided
in Figure 1.
Most continuous exposure variables were categorized for
inclusion in our analysis. Categorization was based on the
distribution of the specific variable and/or logical cate-
gories for the variable. The respondent’s most recent CD4
cell count was categorized based on whether the respon-
dent had moderate to severe immune system damage (CD4
count o500 cells/mL) or little immune system damage (CD4
count  500 cells/mL). The ‘timing of HIV diagnosis’
variable was categorized according to the ART period at
the time at which the respondent was diagnosed (1983–
1988, pre-ART period; 1989–1995, early ART/monother-
apy period, and 1996 onwards, post-cART period), as
previously defined by Rawstorne et al. [31]. The ‘period of
commencing ART’ variable was categorized in a similar
manner (prior to 1996, pre-cART era; 1996–2003, early
cART era; 2004–2009, late cART era).
Our data set contained a number of attitude variables
which captured respondents’ views about ART/cART and
the impact HIV infection had on respondents’ health,
physical appearance, health management strategies, rela-
tionships and sex life. These variables were scored on Likert
scales (1 5 strongly disagree, 2 5 disagree, 3 5 agree, and
4 5 strongly agree). To reduce the total number of attitude
variables included in our analysis, we conducted principal
components analysis with oblique rotation to identify
appropriate attitude scales that could be included in our
analysis. Mean scores were computed for each scale when
responses had been given for at least two-thirds of the
variables in the scale. Where a suitable scale could not be
identified, attitude variables were analysed as separate
variables.
Statistical analysis
Bivariate associations between the potential explanatory
variables and our dichotomous outcome variable were
assessed using the w2-test or Fisher’s exact test for
categorical exposure variables and the t test for continuous
exposure variables (mean scale scores for attitude scales).
Variables that showed a significant association at the
level of a 5 0.2 in bivariate analyses were included in
multivariable analyses. The multivariable analysis con-
sisted of a two-step logistic regression modelling procedure
based on backwards stepwise logistic regression using the
likelihood ratio statistic. At step 1, we computed four
separate logistic regression models including factors that
were expected to exhibit a high degree of collinearity,
using a 5 0.1 as the exit criterion. Variables that remained
significant at a 5 0.1 during step 1 modelling were entered
into a single step 2 model where a 5 0.05 was set as the
exit criterion. Unadjusted odds ratios were computed using
univariable backwards stepwise logistic regression based
on the likelihood ratio statistic.
To reduce missing data in the multivariable analyses, we
included a ‘not specified’ category to include nonresponders
who had missing data on a limited number of variables.
Where the ‘not specified’ category was not significantly
different from the reference category, ‘not specified’
responses have not been presented in the Results section.
Reliability analysis (Cronbach’s a) was performed using
STATA/SE 11.0 (StataCorp LP, College Station, TX, USA). All
other analyses were performed using IBM SPSS STATISTICS 18
(formerly known as PASW STATISTICS 18 and SPSS STATISTICS,
IBM Corporation, Armonk, NY, USA).
Results
Current ART
A total of 1106 HIV-positive individuals completed the HIV
Futures 6 survey. This represents approximately 6.6% of the
estimated HIV-positive population within Australia. Of these
respondents, 867 (78.4%) were taking ART at the time of
completing the survey. Most respondents (57.9%) were on
regimens composed of three antiretroviral drugs and took their
antiretroviral medication once (44.6%) or twice (47.1%) a day.
In terms of the actual combinations taken, 22.6% were taking
a regimen composed of two nucleoside reverse transcriptase
inhibitors (NRTIs) and one NNRTI; 15.0% were taking two
NRTIs and one PI; 9.8% were taking both PIs and NNRTIs with
or without an NRTI backbone, and 52.6% were taking another
type of regimen. Of those taking a regimen including a PI,
50.1% were also taking ritonavir. A small proportion of the
sample (7.6%) were taking agents from newer antiretroviral
drug classes (integrase and fusion inhibitors).
r 2011 British HIV Association HIV Medicine (2011), 12, 562–569

Personal Factors
• Current age
• Gender
• Sexual orientation
• Substance use within the last 12 months
(use of alcohol, cigarettes, marijuana, crystal
meth and party drugs)
• Current health and wellbeing
• Change in health and wellbeing
since commencing ART
• Presence/absence of major health
conditions other than HIV/AIDS
• Lifetime diagnosis of hepatitis A, hepatitis B
and hepatitis C
• Lifetime diagnosis of a mental health
condition
• Sexually transmitted infection within the
last 12 months (gonorrhoea, chlamydia,
genital herpes, syphilis or other)
• Doses of medications taken for other
medical conditions each day
• Use of psychiatric medications (for
depression, anxiety and/or psychotic
disorders) in last 6 months
• Disclosure of HIV status to close friends
• Stressful life events (cared for someone with
HIV/AIDS in the last 2 years and death of
someone close as a result of AIDS)
Socioeconomic Factors
• Highest education level
• Employment status
• Income
• Urbanicity
• Living alone
• Living with dependent children
• Level of support from partner, friends,
family, pets, medical professionals, PLWH
groups and religious/spiritual advisers
Fig. 1 Potential explanatory variables investigated in this study. ART, antiretroviral therapy; PLWH, people living with HIV infection.
Reported reasons for difficulty taking ART
Of those respondents on ART at the time of completing the
survey, 820 (94.6%) indicated whether or not they
experienced any difficulty taking ART; 39.1% of respon-
dents expressed difficulty taking ART. Table 1 provides an
overview of the reported reasons for such difficulty.
Remembering to take drugs on time and side effects were
the most common reasons for reported difficulty taking
ART. Commonly stated side effects included gastrointest-
inal disturbances (diarrhoea, nausea and flatulence in
particular), lipodystrophy and fatigue/sleep disturbance. Of
those who specified ‘other reasons’ for difficulty taking
r 2011 British HIV Association HIV Medicine (2011), 12, 562–569
Adherence to ART in HIV-positive Australians 565
Attitudes
• ‘Do not believe in the benefits of ART’
scale (three items; Cronbach α=0.485)
• ‘Safe sex is not essential because of
treatment effects’ scale (two items;
Cronbach α=0.678)
• ‘Positive health attitude/strategies adopted’
scale (three items; Cronbach α=0.745)
• ‘Negative body image/changes’ scale
(three items; Cronbach α=0.657)
• ‘Prefer to be in a seroconcordant
relationship’ scale (two items; Cronbach
α=0.765)
• ‘HIV has had a negative impact on sex and
relationships’ scale (five items; Cronbach
α=0.688)
• ‘HIV treatments will stop me dying of
AIDS’
• ‘I am worried that in the future my
medication will stop working for me’
• ‘Taking tablets gives me an unwanted
reminder that I have HIV’
• ‘As long as I am well I prefer not to think
about HIV/AIDS’
• ‘I feel more confident about unprotected sex
because of the new treatments’
• ‘Withdrawing before ejaculating (cumming)
is a way to reduce the risk of passing on
HIV’
Treatment and Disease Factors
• ART dosing frequency
• Number of antiretroviral drugs
• Type of antiretroviral drugs/regimen
• Number of prior ART regimens
• Length of time on ART/period of
commencing ART
• Experience of health service discrimination
• Experience of adverse events
• Most recent CD4 cell count
• Timing of HIV diagnosis
• Diagnosis of an AIDS-defining illness
ART, such reasons included pill size, travel commitments
and restrictions, and the inconvenience of obtaining
medication.
Bivariate analyses
Of the personal factors evaluated (see Fig. 1), age, party
drug use (use of any of noninjected cocaine, ecstasy,
lysergic acid diethylamide, injected and noninjected
speed, and g-hydroxybutyric acid), alcohol use, cigarette
smoking, self-reported health and wellbeing, a diagnosis of
herpes within the last 12 months, diagnosis of a mental
566 J Grierson et al.

Table 1 Reported reasons for difficulty taking ART (n 5 820)* tion in the last 2 years were associated with reported
difficulty taking ART at a level of a 5 0.05. No additional
Reported by
variables met the criterion for inclusion in the multi-
Reason Number % variable analyses.
Of the disease-related factors assessed (outlined in Fig. 1),
Remembering to take drugs on time 165 20.1
Side effects 155 18.9
diagnosis of an ADI was associated with reported difficulty
Carrying/transporting medication 124 15.1 taking ART at a level of a 5 0.05. The respondent’s most
Taking medication in public 107 13.0 recent CD4 cell count also met the criterion for inclusion in
Organizing meals around medication 95 11.6
Taking a large number of tablets 67 8.2
multivariable analyses.
Difficulty taking ART and medication 38 4.6
for other health conditions Multivariable analyses
Other 37 4.5
Variables that had shown a significant association in
ART, antiretroviral therapy.
*Multiple responses possible. bivariate analyses at the level of a 5 0.2 were included in
multivariable analysis. Initially, we set up logistic regres-
sion models of clusters of variables that were expected to
health condition, use of psychiatric medications and
exhibit a high degree of collinearity (step 1 models). At step
disclosure to close friends showed a significant association
1, we created four models: (i) a substance use model, (ii) an
with reported difficulty taking ART at the level of a 5 0.05.
other personal factors and attitudes model, (iii) a socio-
Crystal meth use, recreational marijuana use, diagnosis of
economic factor model, and (iv) a treatment-related and
hepatitis A and death of someone close as a result of HIV/
disease-related factor model.
AIDS also met the criterion for inclusion in multivariable
Variables that remained significantly associated with
modelling. Interestingly, those who had disclosed their
reported difficulty taking ART at step 1 at the level of
status to close friends were more likely to report difficulty
a 5 0.1 were included in the step 2 logistic regression
taking ART than those who had not disclosed their status to
model. The following variables maintained an independent
close friends.
association with reported difficulty taking ART at the level
Of the attitudes evaluated (outlined in Fig. 1), not
of a 5 0.05 during step 2 modelling: age, urbanicity,
believing in the benefits of ART, concern about the
current health, lifetime diagnosis of a mental health
effectiveness of ART in the future, reporting that tablets
condition, alcohol and party drug use, ART dosing
were an unwanted reminder of HIV infection, negative
frequency, the type of regimen taken by the respondent,
body image/changes, a negative impact of HIV/AIDS on
the experience of physical adverse events in the last 12
sex and relationships and a high degree of confidence that
months or health service discrimination in the last 2 years,
unprotected sex was not a risky behaviour were associated
and the following attitude variables: the ‘Do not believe in
with increased likelihood of reporting difficulty taking ART
the benefits of ART’ scale, ‘I am worried that in the future
at a level of a 5 0.05. A positive health attitude and/or the
my medication will stop working for me’ and ‘Taking
adoption of positive strategies to manage one’s health was
tablets gives me an unwanted reminder that I have HIV’
associated with a reduced likelihood of reporting difficulty
(Table 2). Of these factors, experiencing physical adverse
taking ART. Deeming safe sex to be nonessential because of
events or health service discrimination had the strongest
treatment effects also met the criterion for inclusion in
association with reporting difficulty taking ART, increasing
multivariable analysis.
the odds of reporting difficulty taking ART by approxi-
The level of support from a range of sources (HIV-
mately four- to fivefold. Taking more than one ART dose
positive friends, close friends, parents, family in general, a
per day, reporting poor to fair health and living in a
counsellor and the respondent’s doctor) was the only
regional centre were associated with a two- to threefold
socioeconomic factor associated with reported difficulty
increase in the odds of reported difficulty taking ART.
taking ART at a level of a 5 0.05. Education level,
Being older than 50 years of age, taking an ART regimen
urbanicity and additional support variables (support from
composed of an NNRTI and two NRTIs, and disagreeing
partner/spouse and PLWH groups) also met the criterion for
with negative attitudes about ART were estimated to at
entry into multivariable analyses (see Fig. 1 for the full list
least halve the odds of reporting difficulty taking ART.
of socioeconomic factors investigated).
Of the treatment-related variables assessed (see Fig. 1),
dosing frequency, the type of regimen taken, the length
Discussion
of time on ART, and experiencing physical adverse We found that a number of personal and treatment-related
events in the last 12 months or health service discrimina- factors were independently associated with reported

r 2011 British HIV Association HIV Medicine (2011), 12, 562–569

 Adherence to ART in HIV-positive Australians 567

Table 2 Final logistic regression model of factors independently difficulty taking ART, while social and disease-related
associated with reported difficulty taking antiretroviral therapy factors were not. Of more than 70 personal, socioeconomic,
(ART)§: factors independently associated at a 5 0.05w (n 5 744) treatment-related and disease-related factors investigated
Odds ratioz (95% confidence interval) in our study, we found that 13 distinct variables were
independently associated with reported difficulty taking
Factor Unadjusted Adjusted ART. By chance alone we would have expected three or
Current age (years) four significant associations.
18–34 0.94 (0.49–1.80) 0.80 (0.35–1.84) Specifically, poor or fair self-reported health, diagnosis
35–49 1 1 of a mental health condition, alcohol and party drug use,
50–64 0.52 (0.38–0.71)*** 0.52 (0.34–0.78)**
 65 0.24 (0.12–0.48)*** 0.37 (0.15–0.92)* living in a regional centre, not believing in the benefits of
Urbanicity ART, worrying about ART efficacy, thinking tablets were an
Capital/inner city 1 1 unwanted reminder of HIV, taking more than one ART dose
Outer suburban 1.38 (0.90–2.13) 1.41 (0.81–2.45)
Regional centre 1.64 (1.13–2.38)* 2.74 (1.69–4.43)*** per day, and experiencing health service discrimination or
Rural locality 1.00 (0.59–1.71) 1.16 (0.60–2.23) physical symptoms were each independently associated
Current health with increased odds of reporting difficulty taking ART.
Poor/fair 2.88 (2.09–3.97)*** 2.11 (1.42–3.14)***
Good/excellent 1 1 Being 50 years of age or older and taking an ART regimen
Lifetime diagnosis of a mental health condition composed of an NNRTI and two NRTIs was associated with
Yes 2.32 (1.72–3.13)*** 1.87 (1.29–2.71)** reduced odds of reporting difficulty taking ART.
No 1 1
Alcohol use The findings of our study fit well with the existing
Yes 1.47 (1.03–2.09)* 1.69 (1.08–2.65)* literature about factors that are associated with nonadher-
No 1 1 ence to cART. We found that a number of factors that had
Party drug use
Yes 1.66 (1.20–2.30)** 1.71 (1.11–2.64)* previously been shown to be consistently or inconsistently
No 1 1 associated with cART nonadherence demonstrated an
‘Do not believe in the 1.73 (1.29–2.31)*** 1.55 (1.09–2.20)* independent association with reported difficulty taking
benefits of ART’, mean
score ART – in particular, the association of medication side
‘I am worried that in the future my medication will stop working for me’ effects, dosing frequency, age, alcohol consumption,
Strongly disagree 0.16 (0.07–0.35)*** 0.26 (0.10–0.70)** psychiatric comorbidity, health-related quality of life, and
Disagree 0.50 (0.35–0.73)*** 0.69 (0.44–1.09)
Agree 1 1 knowledge and beliefs about HIV and its treatment [9]. We
Strongly agree 2.37 (1.48–3.79)*** 1.52 (0.85–2.72) also found that factors previously defined as generally
‘Taking tablets gives me an unwanted reminder that I have HIV’ not associated with nonadherence to cART were not
Strongly disagree 0.21 (0.11–0.40)*** 0.35 (0.16–0.77)**
Disagree 0.45 (0.30–0.66)*** 0.58 (0.37–0.91)* associated with reported difficulty taking ART in our
Agree 1 1 study [9].
Strongly agree 1.74 (1.18–2.57)** 1.30 (0.81–2.10) Our findings advance the literature by defining factors
ART dosing frequency (dose/day)
1 1 1 that are independently associated with reported difficulty
1–2 1.46 (1.08–1.98)* 1.97 (1.33–2.93)** taking ART/nonadherence to ART when a broad range
42 1.77 (0.83–3.80) 2.83 (1.07–7.52)* of personal, socioeconomic, treatment-related and
ART regimen
Two NRTIs 1 PI 1 1 disease-related characteristics are considered. Such infor-
Two NRTIs 1 NNRTI 0.52 (0.32–0.85)** 0.39 (0.21–0.72)** mation will assist clinicians to target individuals with
PI 1 NNRTI  NRTI 1.11 (0.62–1.99) 1.09 (0.54–2.21) higher likelihood of experiencing difficulty taking ART.
Other regimen 0.94 (0.62–1.42) 0.72 (0.43–1.21)
Experienced physical adverse events in the last 12 months Many past studies investigating nonadherence to cART
Yes 1 1 have investigated a smaller number of factors than assessed
No 0.11 (0.05–0.29)*** 0.21 (0.08–0.57)** in our study, making it difficult to be certain which factors
Experienced health service discrimination in the last 2 years
Yes 4.75 (2.74–8.21)*** 3.70 (1.92–7.16)*** are truly independently associated with nonadherence to
No 1 1 cART.
§ Our study also provides data on reported difficulty
The final model was highly significant (Po0.001) with a w2-value of
249.099 with 31 degrees of freedom. The final model correctly predicted taking ART in a best-practice context, given that Australia
73.9% of cases. has been recognized as having a best-practice population
w
*0.05 4P  0.01; **0.014P  0.001; ***Po0.001.
z
Reference categories chosen based on statistical grounds. health response to the HIV epidemic [32].
The findings of our study are potentially limited by the
cross-sectional nature of the available data and the use of a
proxy variable to assess factors associated with nonadher-
ence to cART. Given the cross-sectional nature of the data,

r 2011 British HIV Association HIV Medicine (2011), 12, 562–569

568 J Grierson et al.
we are unable to assess causal relationships or determine
which factors are associated with long-term reported
difficulty taking ART. The use of a proxy variable for
adherence behaviour means that we cannot be certain
which independently associated factors are associated with
concerning levels of nonadherence; however, we believe
that our proxy variable is providing relevant information
to the study of factors associated with nonadherence to
cART, given that our proxy variable was found to be
associated with self-reported nonadherence and reporting a
detectable viral load, and that our findings broadly agree
with the existing literature about nonadherence to cART. A
further potential limitation of the current study is its use of
self-report data. However, self-report measures have been
widely used in adherence studies [23] and have been shown
to correlate with more objective measures of adherence
such as those provided by medication event monitoring
caps and pharmacy records [21,22,33].
We expect the results of our study to be highly
generalizable to the broader Australian population of
PLWH and HIV-positive men who have sex with men.
The generalizability of our findings to heterosexual
and injecting drug user populations of PLWH is limited
because of the demographics of the Australian population
of PLWH [34].
Given the multitude of factors found to be independently
associated with reported difficulty taking ART, our study
reaffirms the dynamic nature of adherence behaviour and
highlights how important it is that adherence discussions
and interventions remain an integral component of the
clinical management of HIV infection.
Acknowledgements
We thank the 1106 HIV-positive Australians who completed
the HIV Futures 6 survey and shared their experiences of
living with HIV in Australia. We also thank individuals and
organizations in the HIV/AIDS sector within Australia who
assisted with recruiting participants for our study, as well
as our colleagues at the Australian Research Centre in Sex,
Health and Society (ARCSHS) who processed the data from
the HIV Futures 6 survey and assisted with preparing the
‘HIV futures six: Making positive lives count’ report. We
thank the Commonwealth Department of Health and
Ageing for funding the Living with HIV Program at
ARCSHS. We also thank Gilead Sciences for providing
funding for this analysis.
References
1 Mocroft A, Vella S, Benfield TL et al. Changing patterns of
mortality across Europe in patients infected with HIV-1. Lancet
1998; 352: 1725–1730.
2 Palella FJ, Delaney KM, Moorman AC et al. Declining morbidity
and mortality among patients with advanced human
immunodeficiency virus infection. N Engl J Med 1998; 338:
853–860.
3 Hoy J, Lewin S, Post JJ, Street A. (eds). HIV Management in
Australasia: A Guide for Clinical Care. Darlinghurst, NSW:
Australasian Society for HIV Medicine, 2009.
4 Paterson DL, Swindells S, Mohr J et al. Adherence to protease
inhibitor therapy and outcomes in patients with HIV infection.
Ann Intern Med 2000; 133: 21–30.
5 Bangsberg DR, Perry S, Charlebois E et al. Non-adherence to
highly active antiretroviral therapy predicts progression to
AIDS. AIDS 2001; 15: 1181–1183.
6 Hogg R, Heath K, Bangsberg D et al. Intermittent use of
triple-combination therapy is predictive of mortality at
baseline and after 1 year of follow-up. AIDS 2002; 16:
1051–1058.
7 Bangsberg D. Less than 95% adherence to nonnucleoside
reverse-transcriptase inhibitor therapy can lead to viral
suppression. Clin Infect Dis 2006; 43: 939–941.
8 Raffa JD, Tossonian HK, Grebely J, Petkau AJ, DeVlaming S,
Conway B. Intermediate highly active antiretroviral therapy
adherence thresholds and empirical models for the
development of drug resistance mutations. J Acquir Immune
Defic Syndr 2008; 47: 397–399.
9 Ammassari A, Trotta MP, Murri R et al. Correlates and
predictors of adherence to highly active antiretroviral therapy:
overview of published literature. J Acquir Immune Defic Syndr
2002; 31 (Suppl 3): S123–S127.
10 Spire B, Duran S, Souville M, Leport C, Raffi F, Moatti JP.
Adherence to highly active antiretroviral therapies (HAART) in
HIV-infected patients: from a predictive to a dynamic
approach. Soc Sci Med 2002; 54: 1481–1496.
11 Falagas ME, Zarkadoulia EA, Pliatsika PA, Panos G.
Socioeconomic status (SES) as a determinant of adherence to
treatment in HIV infected patients: a systematic review of the
literature. Retrovirology 2008; 5: 13.
12 Ammassari A, Antinori A, Cozzi-Lepri A et al. Relationship
between HAART adherence and adipose tissue alterations.
J Acquir Immune Defic Syndr 2002; 31 (Suppl 3):
S140–S144.
13 Ammassari A, Murri R, Pezzotti P et al. Self-reported symptoms
and medication side effects influence adherence to highly
active antiretroviral therapy in persons with HIV infection.
J Acquir Immune Defic Syndr 2001; 28: 445–449.
14 Atkinson MJ, Petrozzino JJ. An evidence-based review of
treatment-related determinants of patients’ nonadherence to
HIV medications. AIDS Patient Care STDS 2009; 23: 903–914.
15 Eldred LJ, Wu AW, Chaisson RE, Moore RD. Adherence to
antiretroviral and pneumocystis prophylaxis in HIV disease.
J Acquir Immune Defic Syndr Hum Retrovirol 1998; 18:
117–125.
r 2011 British HIV Association HIV Medicine (2011), 12, 562–569

16 Sherr L, Lampe F, Norwood S et al. Adherence to antiretroviral
treatment in patients with HIV in the UK: a study of
complexity. AIDS Care 2008; 20: 442–448.
17 Kleeberger CA, Phair JP, Strathdee SA, Detels R, Kingsley L,
Jacobson LP. Determinants of heterogeneous adherence to HIV-
antiretroviral therapies in the Multicenter AIDS Cohort Study.
J Acquir Immune Defic Syndr 2001; 26: 82–92.
18 Aloisi MS, Arici C, Balzano R et al. Behavioral correlates of
adherence to antiretroviral therapy. J Acquir Immune Defic
Syndr 2002; 31 (Suppl 3): S145–S148.
19 Braitstein P, Justice A, Bangsberg DR et al. Hepatitis C
coinfection is independently associated with decreased
adherence to antiretroviral therapy in a population-based HIV
cohort. AIDS 2006; 20: 323–331.
20 Horne R, Buick D, Fisher M, Leake H, Cooper V, Weinman J.
Doubts about necessity and concerns about adverse effects:
identifying the types of beliefs that are associated
with non-adherence to HAART. Int J STD AIDS 2004; 15:
38–44.
21 Bangsberg DR, Hecht FM, Charlebois ED et al. Adherence to
protease inhibitors, HIV-1 viral load, and development of
drug resistance in an indigent population. AIDS 2000; 14:
357–366.
22 Lu M, Safren SA, Skolnik PR et al. Optimal recall period and
response task for self-reported HIV medication adherence.
AIDS Behav 2008; 12: 86–94.
23 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: 227–245.
24 Mills EJ, Nachega JB, Bangsberg DR et al. Adherence to
HAART: a systematic review of developed and developing
nation patient-reported barriers and facilitators. PLoS Med
2006; 3: e438.
25 Grierson JG, Power J, Pitts M et al. HIV Futures 6: Making
Positive Lives Count. Melbourne, Australia: The Australian
r 2011 British HIV Association HIV Medicine (2011), 12, 562–569
Adherence to ART in HIV-positive Australians 569
Research Centre in Sex, Health and Society, Latrobe University,
2009.
26 Grierson J, Thorpe R, Pitts M. HIV Futures 5: Life As We Know
It. Melbourne, Australia: The Australian Research Centre in
Sex, Health and Society, Latrobe University, 2006.
27 Grierson J, Thorpe R, Saunders M, Pitts M. HIV Futures 4: State
of the [Positive] Nation. Melbourne, Australia: The Australian
Research Centre in Sex, Health and Society, Latrobe University,
2004.
28 Grierson J, Misson S, McDonald K, Pitts M, O’Brien M. HIV
Futures 3: Positive Australians on Services, Health and Well-
Being. Melbourne, Australia: The Australian Research Centre in
Sex, Health and Society, Latrobe University, 2002.
29 Grierson J, Bartos M, de Visser R, McDonald K. HIV Futures II:
The Health and Well-Being of People with HIV/AIDS in
Australia. Melbourne, Australia: The Australian Research
Centre in Sex, Health and Society, Latrobe University, 2000.
30 Ezzy D, De Visser R, Bartos M, McDonald K, O’Donnell D,
Rosenthal D. HIV Futures Community Report: Health,
Relationships, Community, and Employment. Carlton, Victoria,
Australia: Centre for the Study of Sexually Transmissible
Diseases, La Trobe University, 1998.
31 Rawstorne P, Prestage G, Grierson JG, Song A, Grulich A,
Kippax S. Trends and predictors of HIV-positive community
attachment among PLWHA. AIDS Care. 2005; 17: 589–600.
32 Commonwealth of Australia. National HIV/AIDS Strategy:
Revitalising Australia’s Response 2005–2008. Canberra:
Commonwealth of Australia, 2005.
33 Fairley CK, Permana A, Read TRH. Long-term utility of
measuring adherence by self-report compared with pharmacy
record in a routine clinic setting. HIV Med 2005; 6: 366–369.
34 National Centre in HIV Epidemiology and Clinical Research.
HIV/AIDS, Viral Hepatitis and Sexually Transmissible
Infections in Australia Annual Surveillance Report 2009.
Sydney, NSW: National Centre in HIV Epidemiology and
Clinical Research, The University of New South Wales, 2009.