Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Oct 1.
Published in final edited form as: AIDS Care. 2021 Jan 23;33(10):1340–1349. doi: 10.1080/09540121.2021.1874275

Poor quality of life and incomplete self-reported adherence predict second-line ART virological failure in resource-limited settings

Thiago S TORRES 1,2, Linda J HARRISON 2, Alberto M LA ROSA 3, Lu ZHENG 2, Sandra W CARDOSO 1, Godwin ULAYA 4, Nazim AKOOJEE 5, Dileep KADAM 6, Ann C COLLIER 7, Michael D HUGHES 2; AIDS Clinical Trials Group (ACTG) A5273 Study Team
PMCID: PMC8298588  NIHMSID: NIHMS1664507  PMID: 33487029

Abstract

We evaluated health-related quality of life (QoL) and self-reported incomplete adherence as predictors of early second-line antiretroviral (ART) virological failure (VF) (confirmed HIV-1 RNA [VL]>400 c/mL after 24 weeks of ART). ACTG A5273 was a randomized clinical trial of second-line ART comparing lopinavir/ritonavir (LPV/r) + raltegravir with LPV/r + nucleos(t)ide reverse transcriptase inhibitors in participants failing a first-line regimen at 15 sites in 9 resource-limited settings. Participants completed the ACTG SF-21 which has 8 QoL domains with a standard score ranging from 0 (worst) to 100 (best). We used exact logistic regression to assess the association of QoL at baseline and week 4 with early VF adjusted for self-reported adherence. Of 500 individuals (51% women, median age 39 years) in this analysis, 79% and 75% self-reported complete adherence (no missing doses in the past month) at weeks 4 and 24, respectively. Early VF was experienced by 7% and more common among those who self-reported incomplete adherence. Participants with low week 4 QoL scores had higher rates of early VF than participants with high scores. After adjusting for self-reported adherence at week 4, viral load and CD4 at baseline, cognitive functioning, pain and mental health domains were significantly associated with subsequent early VF. In this post-hoc analysis, poorer QoL adds to self-reported incomplete adherence after 4 weeks of second-line ART in predicting VF at week 24. Evaluation is needed to assess whether individuals with poorer QoL might be targeted for greater support to reduce risk of VF.

Keywords: HIV, quality of life, virological failure, treatment adherence, second-line therapy, antiretroviral therapy

Introduction

Antiretroviral therapy (ART) is a key component in plans to end the HIV epidemic. Effective ART decreases the risk of AIDS-related and non-AIDS related events (The INSIGHT START Study Group, 2015) as well as HIV transmission, which is typically referred as treatment as prevention (TasP) (Smith et al., 2011). Efficacy of TasP has been shown among serodiscordant heterosexual and couples of men who have sex with men (Bavinton et al., 2018; Cohen et al., 2011; Rodger et al., 2019). In this context, the message “undetectable equals untransmittable” (U=U) has been created to raise HIV prevention awareness and to reduce stigma (Rendina & Parsons, 2018; Torres et al., 2020), emphasizing the importance of suppressed viral load values. With availability of first line ART increasing and prevailing challenges of non-adherence and drug resistance, the World Health Organization has projected an increasing number of people on second-line ART (WHO, 2015). Consequently, it is essential to understand and identify the predictors of virologic success and failure, particularly on second-line ART.

Adherence to ART is closely linked to viral suppression (Arnsten et al., 2001). Although the monitoring of adherence to ART is highly recommended by health organizations, there is no gold standard for assessing adherence (WHO, 2016). An individual’s self-report is the easiest and most frequently used method to assess adherence in resource-limited settings (RLS) (Costa et al., 2018; Ortego et al., 2011).

Health-related quality of life (QoL) reflects the impact of disease and treatment on a person’s well-being and ability to carry out daily activities, taking into account biological and psychological effects of disease, including physical, social, cognitive and psychological functioning, as well as subjective sense of health, comfort and well-being. QoL measurements are important to assess a person’s perception of his/her own health (Wilson, 1995). Both poorer QoL and incomplete self-reported adherence have been found to be independent predictors of virologic failure among participants in a large clinical trial of initial ART undertaken in the U.S. (Schackman et al., 2007). In addition, in a large clinical trial of people starting first-line ART in RLS, worse QoL was found to predict incomplete self-reported adherence to ART (Safren et al., 2014). Furthermore worse QoL during treatment was found to predict treatment failure even after adjustment for self-reported level of adherence (Safren et al., 2014). Evaluating QoL early in treatment may therefore be an important marker of increased risk of virologic failure separate from that identified by a person’s self report of incomplete adherence. In turn this might help in identifying people in need of interventions to help improve QoL and potentially reduce the risk of subsequent virologic failure.

We previously reported results about QoL among individuals from RLS who were failing first-line ART before starting second-line ART in the AIDS Clinical Trials Group (ACTG) A5273 randomized trial (Torres, Harrison, La Rosa, Lavenberg, et al., 2018), and subsequent improvement of QoL scores after one year of second-line ART with no difference between randomized treatments (lopinavir/ritonavir [LPV/r] + raltegravir compared to LPV/r + nucleos(t)ide reverse transcriptase [NRTI]-containing inhibitors) (Torres, Harrison, La Rosa, Cardoso, et al., 2018). Mean QoL scores were poorer among participants with higher plasma HIV-1 RNA viral load (VL) and lower CD4 cells count before second-line ART initiation (Torres, Harrison, La Rosa, Lavenberg, et al., 2018), but these discrepancies disappeared after one year of treatment (Torres, Harrison, La Rosa, Cardoso, et al., 2018). There was also no difference between randomized treatments in incidence of virologic failure (La Rosa et al., 2016). Our objective in this post-hoc analysis is to build upon the findings described above for people receiving first-line ART (Safren et al., 2014; Schackman et al., 2007) to explore whether poorer QoL predicts incomplete self-reported level of adherence to ART, and whether poorer QoL predicts virologic failure independent of self-reported level of adherence among people starting second-line ART in RLS. We focused on predicting early VF defined as HIV-1 RNA >400 c/mL at 24 weeks of second-line ART with subsequent confirmation, as the majority of virologic failures occurred at 24 weeks, and because predictors of individuals experiencing VF after 24 weeks might differ from those individuals who have achieved virologic suppression at 24 weeks.

Materials and Methods

A5273 study

ACTG A5273 was a randomized clinical trial, which enrolled and followed participants between January 2012 and October 2014, and showed non-inferior virologic efficacy of LPV/r + raltegravir compared to LPV/r + NRTIs as second-line ART in participants failing non-nucleoside reverse transcriptase inhibitor (NNRTI) based ART (clinicaltrials.gov NCT01352715). Details of study design and main results were described elsewhere (La Rosa et al., 2016).

Eligible participants were HIV-infected men and women (≥18 years) who had first-line VF confirmed by two consecutive VL ≥1000 copies/mL at least one week apart after at least 24 weeks on an NNRTI-containing first-line regimen. Participants were enrolled at 15 sites in 9 RLS: Brazil (1 site), India (3 sites), Kenya (1 site), Malawi (2 sites), Peru (2 sites), South Africa (3 sites), Tanzania (1 site), Thailand (1 site) and Zimbabwe (1 site).

We included in the present analysis participants followed for at least 24 weeks (at which time early VF was determined) who had an assessment for QoL at week 4.

Early Second-line Virological Failure

Early second-line VF was defined as having VL>400 copies/mL at week 24 with confirmation at the next HIV-1 RNA measurement. Note that week 24 was the time of the first HIV-1 RNA measurement at which VF was evaluated in the A5273 study.

Self-reported adherence measure

We used the question “When was the last time you missed any of your ART medications?” for our assessment of self-reported adherence at 4 and 24 weeks after starting second-line ART. There were 6 potential responses: never skip medications, more than 3 months ago, 1–3 months ago, 2–4 weeks ago, 1–2 weeks ago, and within the past week. We dichotomized self-reported adherence as incomplete (self-report of missed doses within the past 4 weeks) and complete (self-report of never skip medications, or missed doses 1–3 months ago or more than 3 months ago) (R. A. Parker et al., 2017).

Quality of life measures

Participants were interviewed at week 0 (baseline, just prior to starting second-line ART), and weeks 4 and 24 using a modified version of the SF-21 measure (ACTG SF-21). The ACTG SF-21 questions cover eight domains: General Health Perceptions (GHP), Physical Functioning (PF), Role Functioning (RF), Social Functioning (SF), Cognitive Functioning (CF), Pain (P), Mental Health (MH), and Energy/Fatigue (E/F). A standardized score ranging from 0 (worst QoL) to 100 (best QoL) was calculated for each domain using standard methods previously described (Safren et al., 2012; Torres, Harrison, La Rosa, Cardoso, et al., 2018; Torres, Harrison, La Rosa, Lavenberg, et al., 2018). High scores for Pain and Energy/Fatigue mean less pain and less fatigue, respectively. The ACTG SF-21 tool was administered in a face-to-face interview by study staff in the participant’s local language.

Statistical Analysis

For univariable analysis, we used exact logistic regression to evaluate the association of early second-line VF with self-reported adherence at week 4, age, VL and CD4 at baseline. Continuous variables were categorized as: age into 18–34, 35–49 or ≥50 years; VL into <10,000, 10,000–99,999 or ≥100,000 copies/mL and CD4 count into <50, 50–199 or ≥200 cells/mm3.

We calculated the mean change of QoL from baseline to week 4, and from week 4 to week 24, and tested whether the differences were significantly different from zero using paired t-test. Then, we calculated mean QoL scores at baseline according to self-reported adherence (complete or incomplete) at week 4 and mean QoL scores at week 4 according to self-reported adherence at week 24. We compared the self-reported adherence groups using the two-sample t-test.

We created three categories for QoL scores in each domain measured at baseline and week 4: high (score 100), medium (75-<100) and low (<75). Subsequently, we calculated the proportion of individuals with early VF and incomplete self-reported adherence at week 4 for each QoL category measured at baseline and week 4. We used exact logistic regression to assess whether QoL at baseline or week 4 as a quantitative variable was significantly associated with early VF taking into account self-reported adherence (complete or incomplete) at week 4. Finally, we evaluated the association of early VF with QoL adjusted also for HIV disease status as measured by baseline VL and CD4 count as we had previously found an association of QoL scores for most domains (Torres, Harrison, La Rosa, Lavenberg, et al., 2018). Data is not publicly available or accessible, and was analyzed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Of the 512 participants who participated in the A5273 study, 500 individuals were included in this analysis. Twelve were excluded because of death (n=4), loss to follow-up (n=4) or withdrawal of informed consent (n=2) before 24 weeks, or because of missed week 4 (n=1) or week 24 study evaluations (n=1). Median age was 39 years (interquartile range [IQR]: 34–44) and 51% of participants were women. Participants were predominantly from India (31%), Malawi (21%) and South Africa (20%). Median baseline CD4 count was 138 cells/mm3 (IQR: 55–271) and VL 33,037 copies/ml (IQR: 7,963–137,925). Median duration of first-line ART was 4.1 years (IQR: 2.2–6.2).

At week 4, 79% (397/500) self-reported complete adherence during the past month. Incomplete self-reported adherence during the past month was reported by 103 (21%) participants: 23 (5%) within the past week, 49 (10%) 1–2 weeks ago and 31(6%) 2–4 weeks ago. At week 24, 81% (401/497) had self-reported complete adherence during the past month. Incomplete self-reported adherence was reported by 19% (96/497) participants: 8 (2%) within the past week, 23 (5%) 1–2 weeks ago and 65 (13%) 2–4 weeks ago (percentages for the breakdown add to 20% rather than 19% because of rounding).

Thirty-seven (7.4%) of the 500 participants experienced early VF at 24 weeks after starting second-line ART (Table 1). Early VF was more common among participants who self-reported incomplete adherence at week 4 (14 [13.6%] of 103) versus complete adherence (23 [5.8%] of 397) (OR: 2.56; 95%CI: 1.16–5.42). Associations of early VF with sex, age, baseline VL and baseline CD4 count were not statistically significant (Table 1).

Table 1.

Association of early virologic failure (VF) on second-line antiretroviral therapy (ART) with self-reported adherence at week 4, demographic and baseline clinical factors.

Characteristic Total n Early VF n(%) OR(95%CI)a p-value
Total 500 37 (7.4)
Self-reported adherence at week 4
Complete 397 23 (5.8) Ref. 0.019
Incomplete 103 14 (13.6) 2.55 (1.16–5.42)
Sex
Female 254 16 (6.3) Ref. 0.43
Male 246 21 (8.5) 1.39 (0.67–2.92)
Baseline Age (years)
18–34 141 11 (7.8) 2.15 (0.45–20.6) 0.46b
35–49 306 24 (7.8) 2.17 (0.51–19.5)
50+ 53 2 (3.8) Ref.
Baseline Viral Load (HIV-1 RNA copies/mL)
<10,000 142 5 (3.5) Ref. 0.24b
10,000–99,999 202 17 (8.4) 2.51 (0.86–8.92)
≥100,000 154 15 (9.7) 2.95 (0.98–10.6)
Baseline CD4 count (cells/mm³)
<50 115 9 (7.8) Ref. 0.15b
50–199 193 12 (6.2) 0.78 (0.29–2.17)
≥200 187 16 (8.6) 1.10 (0.44–2.94)
Open in a new tab
a.

Odds ratios (OR) and 95% confidence interval (95%CI) for early virological failure (VF) at week 24 from exact logistic regression

b.

p-values from logistic regression considering the quantitative variable.

Table 2 summarizes mean QoL scores at baseline, week 4 and week 24, as well as mean changes in QoL scores from baseline to week 4 and from week 4 to week 24. Mean QoL scores improved for all domains during the first four weeks of second-line ART, though the mean change for the pain score was not statistically significant. Five of the 8 QoL domains improved significantly between week 4 and week 24. In general, these improvements were smaller in magnitude than those observed between baseline and week 4, though mean changes for the role functioning and pain scores were larger. In those with subsequent early VF, QoL did not improve from baseline to week 4 and from week 4 to week 24 (p>0.1; data not shown).

Table 2.

Mean QoL scores at baseline, week 4 and week 24 and mean change at weeks 4 and 24.

Mean QoL score (95% CI) Change in QoL score from baseline to week 4 Change in QoL score from week 4 to week 24
Baseline Week 4 Week 24 Mean (95% CI) p-valuea Mean (95% CI) p-valuea
General Heath Perceptions (GHP) 67 (65, 69) 72 (71, 74) 74 (73, 75) 5 (3, 7) <.001 2 (0, 3) 0.030
Physical Functioning (PF) 92 (90, 93) 94 (93, 95) 97 (96, 97) 2 (1, 4) 0.003 2 (1, 4) 0.001
Role Functioning (RF) 81 (78, 83) 84 (82, 86) 89 (88, 91) 3 (1, 6) 0.008 6 (3, 8) <.001
Social Functioning (SF) 91 (90, 93) 93 (92, 94) 95 (94, 96) 2 (0, 3) 0.013 2 (0, 3) 0.004
Cognitive Functioning (CF) 91 (90, 93) 95 (94, 96) 95 (94, 96) 3 (2, 5) <.001 0 (−1, 2) 0.60
Pain (P) 83 (81, 85) 85 (83, 86) 88 (86, 90) 2 (0, 3) 0.13 3 (1, 5) 0.001
Mental Health (MH) 85 (84, 86) 89 (88, 90) 90 (89, 91) 4 (2, 5) <.001 1 (0, 2) 0.10
Energy / Fatigue (E/F) 80 (78, 82) 83 (82, 85) 85 (83, 86) 3 (1, 5) 0.001 1 (−1, 3) 0.22
Open in a new tab
a

p-value from Student’s paired sample t-test. QoL: quality of life; CI: confidence interval. Possible scores for all domains ranged from 0 (worst) and 100 (best).

Lower QoL scores in role functioning and pain at baseline were associated with subsequent self-reported incomplete adherence at the week 4 and 24 evaluations (Table 3). QoL scores at week 4 were significantly lower in several domains (physical functioning, role functioning, social functioning, pain and mental health) in participants reporting incomplete compared to those reporting complete adherence at the week 4 and 24 evaluations.

Table 3.

Mean QoL scores at baseline and week 4 according to self-reported adherence at weeks 4 and 24.

Mean QoL score at baseline (95% CI)
Participants self-reporting incomplete adherence at week 4 (n=103) Participants self-reporting complete adherence at week 4 (n=397) p-value Participants self-reporting incomplete adherence at week 24 (n=96) Participants self-reporting complete adherence at week 24 (n=401) p-value
General Heath Perceptions (GHP) 68 (64, 71) 67 (65, 69) 0.71 68 (65, 71) 67 (65, 69) 0.64
Physical Functioning (PF) 90 (87, 94) 92 (90, 94) 0.41 91 (88, 94) 92 (90, 94) 0.68
Role Functioning (RF) 67 (61, 74) 84 (81, 87) <.001 64 (57, 72) 85 (82, 87) <.001
Social Functioning (SF) 90 (87, 92) 92 (90, 93) 0.22 89 (87, 93) 92 (90, 93) 0.15
Cognitive Functioning (CF) 93 (91, 95) 91 (89, 92) 0.14 93 (91, 95) 91 (89, 92) 0.13
Pain (P) 76 (71, 81) 85 (83, 87) <.001 77 (73, 82) 84 (83, 87) 0.003
Mental Health (MH) 85 (82, 87) 85 (84, 87) 0.56 86 (84, 88) 85 (83, 87) 0.22
Energy / Fatigue (E/F) 80 (77, 83) 80 (78, 82) 0.92 81 (78, 85) 80 (78, 82) 0.73
Mean QoL score at week 4 (95% CI)
General Heath Perceptions (GHP) 70 (67, 73) 73 (71, 74) 0.15 71 (68, 74) 73 (71, 74) 0.35
Physical Functioning (PF) 89 (85, 93) 95 (94, 97) 0.003 89 (85, 93) 96 (94, 97) 0.002
Role Functioning (RF) 73 (68, 78) 87 (84, 89) <.001 69 (63, 75) 87 (85, 90) <.001
Social Functioning (SF) 88 (86, 91) 94 (93, 96) 0.001 89 (85, 92) 94 (93, 95) 0.001
Cognitive Functioning (CF) 93 (91, 95) 95 (94, 96) 0.19 92 (90, 95) 95 (94, 96) 0.08
Pain (P) 77 (72, 81) 87 (85, 88) 0.001 78 (74, 83) 86 (84, 88) 0.002
Mental Health (MH) 87 (85, 89) 89 (88, 90) 0.06 86 (83, 88) 89 (88, 91) 0.005
Energy / Fatigue (E/F) 80 (77, 84) 84 (82, 86) 0.09 81 (78, 85) 84 (82, 86) 0.25
Open in a new tab

p-values from Student’s two sample t-test.

Table 4 provides a descriptive summary of the percentage of participants experiencing early VF by category of QoL score for each domain (high: score of 100; medium: 75-<100; low: <75). Also shown are results from a trend test using the quantitative scores for the association between QoL score and proportion experiencing early VF. Only for the pain domain was early VF associated with lower QoL scores. This association was attenuated when adjusted for self-reported adherence at week 4 (Table 4).

Table 4.

Associations of QoL domains at baseline with early virologic failure (VF) on second-line antiretroviral therapy (unadjusted and adjusted)

QoL Domain QoL Score Categorya N Incomplete self-reported adherence at week 4 N (%) Early VF N (%) OR for early VF (95%CI) Trend p-valueb OR for early VF (95%CI) adjusted for self-reported adherence at week 4 Trend p-valueb
General Health Perception (GHP) High 31 5 (16.1) 2 (6.4) Ref 0.87 Ref 0.84
Medium 212 44 (20.7) 18 (8.5) 1.34 (0.30–12.5) 1.28 (0.28–12.0)
Low 257 54 (21.0) 17 (6.6) 1.03 (0.23–4.68) 0.97 (0.21–9.16)
Physical Functioning (PF) High 344 61 (17.7) 22 (6.4) Ref 0.19 Ref 0.22
Medium 114 37 (32.5) 11 (9.6) 1.56 (0.66–3.50) 1.34 (0.56–3.06)
Low 41 5 (12.2) 4 (9.8) 1.58 (0.38–5.02) 1.70 (0.40–5.44)
Role Functioning (RF) High 327 48 (14.7) 19 (5.8) Ref 0.47 Ref 0.91
Medium 27 5 (18.5) 5 (18.5) 3.66 (0.98–11.5) 3.57 (0.94–11.3)
Low 146 50 (34.2) 13 (8.9) 1.58 (0.70–3.49) 1.29 (0.55–2.93)
Social Functioning (SF) High 326 49 (15.0) 20 (6.1) Ref 0.45 Ref 0.52
Medium 125 47 (37.6) 13 (10.4) 1.77 (0.78–3.89) 1.43 (0.61–3.24)
Low 49 7 (14.3) 4 (8.2) 1.36 (0.32–4.32) 1.37 (0.32–4.39)
Cognitive Functioning (CF) High 275 53 (19.3) 20 (7.3) Ref 0.67 Ref 0.53
Medium 167 45 (26.9) 13 (7.8) 1.08 (0.50–2.35) 0.99 (0.43–2.17)
Low 58 5 (8.6) 4 (6.9) 0.94 (0.23–2.98) 1.08 (0.25–3.45)
Pain (P) High 249 38 (15.3) 16 (6.4) Ref 0.033 Ref 0.08
Medium 111 20 (18.0) 4 (3.6) 0.54 (0.13–1.74) 0.53 (0.12–1.70)
Low 140 45 (32.1) 17 (12.1) 2.01 (0.92–4.41) 1.73 (0.77–3.86)
Mental Health (MH) High 124 10 (8.1) 5 (4.0) Ref 0.12 Ref 0.11
Medium 263 74 (28.1) 20 (7.6) 1.96 (0.69–6.83) 1.56 (0.53–5.60)
Low 113 19 (16.8) 12 (10.6) 2.82 (0.89–10.6) 2.56 (0.80–9.66)
Energy/Fatigue (E/F) High 129 10 (7.7) 6 (4.6) Ref 0.29 Ref 0.28
Medium 234 73 (31.2) 19 (8.1) 1.81 (0.67–5.68) 1.39 (0.49–4.53)
Low 137 20 (14.6) 12 (8.8) 1.96 (0.66–6.58) 1.82 (0.60–6.13)
Open in a new tab
a

QoL score categories: high = 100; medium = 75-<100; low = <75;

b

p-values from logistic regression model including QoL score as a quantitative score

In contrast, at week 4, participants with low QoL scores (<75) had higher rates of early VF than participants with high scores (=100) in all domains and were statistically significant for the general health perceptions, cognitive functioning, pain, mental health and energy/fatigue domains (Table 5). Associations were slightly attenuated with adjustment for self-reported adherence at week 4, and with further adjustment for baseline VL and CD4 count, but remained significant for the cognitive functioning, pain and mental health domains. In the analyses adjusted for self-reported adherence at week 4, as well as baseline VL and CD4 count, across QoL domains, the odds of virologic failure were higher for participants with low scores compared with participants with scores of 100.

Table 5.

Associations of QoL domains at week 4 with early virologic failure (VF) on second-line antiretroviral therapy (unadjusted and adjusted)

QoL Domain QoL Score Categorya N Incomplete self-reported adherence at week 4 N (%) Early VF N (%) OR for early VF (95%CI) Trend p-valueb OR for early VF (95%CI) adjusted for self-reported adherence at week 4 Trend p-value aOR(95%CI)c Trend p-valueb
General Health Perception (GHP) High 35 5 (14.3) 1 (2.9) Ref 0.048 Ref 0.063 Ref. 0.063
Medium 256 52 (20.3) 17 (6.6) 2.41 (0.36–104) 2.27 (0.33–98.0) 2.30 (0.33–99.7)
Low 209 46 (22.0) 19 (9.1) 3.39 (0.50–145) 3.13 (0.46–135) 3.16 (0.46–136)
Physical Functioning (PF) High 382 57 (14.9) 22 (5.8) Ref 0.08 Ref 0.18 Ref. 0.24
Medium 95 39 (41.0) 12 (12.6) 2.36 (1.02–5.22) 1.91 (0.79–4.38) 1.87 (0.78–4.30)
Low 23 7 (30.4) 3 (13.0) 2.45 (0.43–9.24) 2.15 (0.38–8.27) 1.91 (0.33–7.44)
Role Functioning (RF) High 347 49 (14.1) 20 (5.8) Ref 0.48 Ref 0.14 Ref. 0.17
Medium 20 5 (25.0) 1 (5.0) 0.86 (0.02–6.00) 0.78 (0.02–5.51) 0.84 (0.02–5.98)
Low 133 49 (36.8) 16 (12.0) 2.23 (1.04–4.71) 1.84 (0.83–4.00) 1.77 (0.79–3.87)
Social Functioning (SF) High 342 51 (14.9) 21 (6.1) Ref 0.13 Ref 0.28 Ref. 0.41
Medium 131 41 (31.3) 12 (9.2) 1.54 (0.67–3.40) 1.32 (0.56–2.96) 1.28 (0.54–2.89)
Low 27 11 (40.7) 4 (14.8) 2.65 (0.61–8.82) 2.10 (0.47–7.23) 1.78 (0.39–6.31)
Cognitive Functioning (CF) High 332 55 (16.6) 19 (5.7) Ref 0.006 Ref 0.009 Ref 0.014
Medium 137 41 (29.9) 12 (8.8) 1.58 (0.68–3.55) 1.38 (0.58–3.14) 1.35 (0.57–3.08)
Low 31 7 (22.6) 6 (19.3) 3.93 (1.18–11.5) 3.75 (1.11–11.1) 3.38 (0.99–10.1)
Pain (P) High 253 37 (14.6) 14 (5.5) Ref 0.001 Ref 0.006 Ref. 0.008
Medium 112 22 (19.6) 6 (5.4) 0.97 (0.30–2.77) 0.92 (0.28–2.65) 0.92 (0.28–2.65)
Low 135 44 (32.6) 17 (12.6) 2.45 (1.10–5.58) 2.10 (0.91–4.86) 2.00 (0.87–4.64)
Mental Health (MH) High 159 20 (12.6) 9 (5.7) Ref 0.015 Ref 0.023 Ref. 0.029
Medium 268 66 (24.6) 19 (7.1) 1.27 (0.53–3.28) 1.11 (0.46–2.90) 1.10 (0.45–2.88)
Low 73 17 (23.3) 9 (12.3) 2.33 (0.78–6.98) 2.09 (0.69–6.31) 2.07 (0.68–6.25)
Energy/Fatigue (E/F) High 161 20 (12.4) 6 (3.7) Ref 0.048 Ref 0.07 Ref 0.10
Medium 232 58 (25.0) 17 (7.3) 2.04 (0.75–6.46) 1.80 (0.65–5.75) 1.79 (0.64–5.72)
Low 107 25 (23.4) 14 (13.1) 3.87 (1.34–12.7) 3.49 (1.20–11.6) 3.30 (1.28–11.0)
Open in a new tab
a

QoL score categories: high = 100; medium = 75-<100; low = <75;

b

p-values considering quantitative QoL domains scores;

c

adjusted for self-reported adherence at week 4, VL and CD4 count at baseline

Discussion

In RLS, QoL measured 4 weeks after second-line ART initiation adds to self-reported incomplete adherence in predicting early VF. The associations between low QoL scores and VF persisted with adjustment for self-reported adherence at week 4 and baseline HIV disease status as measured by baseline VL and CD4. These findings indicate that QoL and adherence assessments after 4 weeks of ART initiation could be implemented as real-time measurements to identify individuals at higher risk of subsequent VF in RLS. These individuals could benefit from interventions to improve QoL, such as social self-value empowerment (Bhatta & Liabsuetrakul, 2017) and yoga (Hari Chandra et al., 2019), or to optimize adherence, such as text-message reminders (Garofalo et al., 2016).

Different from QoL measured at week 4, most QoL domains measured prior to second-line ART initiation (baseline) could not predict early VF (except pain domain). This is consistent with a study conducted in Italy among multiexperienced individuals receiving raltegravir based-regimens, in which baseline characteristics, including QoL measures, did not predict VF after 48 weeks of ART initiation (Bucciardini et al., 2012). Our findings could be explained by significant QoL score improvements between baseline and week 4 for most domains. Thus, four weeks of second-line ART resulted in improvement of QoL among most individuals but not among those subsequently experiencing early VF.

Associations of low QoL at week 4 and early VF were more pronounced for cognitive functioning, pain and mental health domains. These domains aim to measure: (i) for cognitive functioning, the degree of difficulty participants have experienced with respect to their cognitive abilities; (ii) for pain, the intensity of physical pain (e.g. headache, muscle pain, stomach ache) and degree of interface with daily activities; (iii) for mental health, anxiety, depression, and overall psychological well-being (Torres, Harrison, La Rosa, Lavenberg, et al., 2018).

Cognitive dysfunction has been previously demonstrated to be associated with a higher risk of VF in resource rich-settings (Hinkin et al, 2002; Shahani et al., 2018), as it may impact ability to adhere to ART (Gorman et al., 2009). Pain symptoms are more prevalent among people living with HIV (PLWH) than the general population (Canan et al., 2019) due to multiple factors, including HIV infection, certain antiretroviral drugs, co-existing psychiatric illness, and substance abuse (Krashin et al. 2012). Point prevalence of clinically significant pain among PLWH ranges from 54% (95% CI 51.14, 56.09) to 83% (95% CI 76, 88), considering a three-month recall period (R. Parker et al., 2014). Chronic pain has a negative impact on HIV status, including decreased adherence to ART (Surratt et al., 2015) and higher rates of unsuppressed HIV viral loads (Denis et al., 2019). However, chronic pain is undertreated in many settings, including in PLWH due to complexity of ART regimens, potential drug-drug interactions with pain medications, risk of side effects (Isenberg, MA et al., 2017), psychiatric and substance use comorbidities (Miaskowski et al., 2011) and patient beliefs (e.g. no connection between pain and the disease or fear of what the pain may mean) (Breitbart et al., 1998). Mental health disorders are also reported more often by PLWH than the general population in RLS, including depression and posttraumatic stress disorders (Remien et al., 2019). Elevated depressive symptoms are correlated with a decrease of adherence to ART (Uebelacker et al., 2015) and virological failure (Anastos et al., 2005; Barfod et al., 2005).

In this study, a single question about self-reported adherence in the first 4 weeks of second-line ART was useful in predicting early VF in RLS. In the SECOND-LINE trial, which also evaluated a raltegravir-based regimen for second-line ART in RLS, poor adherence was a major determinant of VF (Boyd et al., 2015). Enhanced adherence counseling could be considered as a potential strategy to prevent VF; this intervention was effective in a Swaziland study to promote viral load suppression in PWLH with high VL (Etoori et al., 2018). Two-way mobile phone intervention might also be used to improve adherence, as it was shown to have a modest impact on virological failure rates in a multi-center study among individuals failing second-line ART in RLS (Gross et al., 2019).

We have also observed that low QoL scores measured at week 4 could predict incomplete self-reported adherence 24 weeks after starting second-line ART for most domains. Better QoL with respect to general health perceptions at baseline was associated with higher rates of complete self-reported adherence through 72-weeks of first-line ART in a multicentre trial conducted in RLS (Safren et al., 2014). Adherence over time was associated with greater QoL in prospective cohort studies (Geocze et al., 2010). This indicates that QoL measures could be useful to identify individuals at risk of incomplete adherence in RLS.

This study has limitations. Data were not collected on factors such as employment and educational status, sexual behavior and social stigma that might be associated with QoL. The population studied was from a clinical trial and so may differ from those in the routine clinical practice. We did not have sufficient number of early VFs to evaluate whether associations between QoL and VF might have varied among countries/cultures. Adherence data was self-reported by participants and may be subject to recall bias. First-line ART virological failure was identified at study entry and it is possible that participants may have had unidentified virologic failure for an extended period of time, particularly in locations where viral load testing was not part of routine care. It is also possible, therefore, that some HIV-infected people experiencing unidentified virologic failure might die or be lost from care and so not be represented in our study versus a study that was identifying virologic failure proximal to its time of occurrence. Caution should therefore be taken before generalizing our findings to other clinical and cultural settings.

In this post-hoc analysis, poorer QoL adds to self-reported incomplete adherence after 4 weeks of second-line ART in predicting VF at week 24. Evaluation is needed to assess whether patients with poorer QoL might be targeted for greater support to improve QoL and adherence, as well as to reduce risk of VF.

Acknowledgments

Presented in part at the Conference of Retroviruses and Opportunistic Infections (CROI) 2019, Seattle, 06 March 2019. We thank the study participants and the other members of the ACTG A5273 Study Group.

Funding

ACTG A5273 study was supported National Institute of Allergy and Infectious Diseases (award number U01AI068636), National Institute of Mental Health (NIMH), and the National Institute of Dental and Craniofacial Research (NIDCR). The study was also funded via the following grants: AI069432, AI069438, AI069481, 2UM1AI069423 (including a subcontract to the Pittsburgh Virology Specialty Laboratory), UM1 AI069471, and UM1 AI068634 (ACTG Statistical and Data Management Center) and the Investigator-Initiated Studies Program of Merck Sharp & Dohme. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases, the National Institutes of Health, or Merck Sharp & Dohme. Abbott Laboratories, Gilead Sciences, ViiV Healthcare/GSK, and Merck Sharp & Dohme provided the study drugs.

Footnotes

Disclosure Statement

The authors declare no competing interest.

References

  1. Anastos K, Schneider MF, Gange SJ, Minkoff H, Greenblatt RM, Feldman J, Levine A, Delapenha R, & Cohen M (2005). The association of race, sociodemographic, and behavioral characteristics with response to highly active antiretroviral therapy in women. Journal of Acquired Immune Deficiency Syndromes, 39(5), 8, 537–544 [PubMed] [Google Scholar]
  2. Arnsten JH, Demas PA, Farzadegan H, Grant RW, Gourevitch MN, Chang C, Buono D, Eckholdt H, Howard AA, & Schoenbaum EE (2001). Antiretroviral Therapy Adherence and Viral Suppression in HIV‐Infected Drug Users: Comparison of Self‐Report and Electronic Monitoring. Clinical Infectious Diseases, 33(8), 1417–1423. 10.1086/323201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barfod TS, Gerstoft J, Rodkjaer L, Pedersen C, Nielsen H, Møller A, Kristensen LH, Sørensen HT, & Obel N (2005). Patients’ Answers to Simple Questions About Treatment Satisfaction and Adherence and Depression Are Associated with Failure of HAART: A Cross-Sectional Survey. AIDS Patient Care and STDs, 19(5), 317–325. 10.1089/apc.2005.19.317 [DOI] [PubMed] [Google Scholar]
  4. Bavinton BR, Pinto AN, Phanuphak N, Grinsztejn B, Prestage GP, Zablotska-Manos IB, Jin F, Fairley CK, Moore R, Roth N, Bloch M, Pell C, McNulty AM, Baker D, Hoy J, Tee BK, Templeton DJ, Cooper DA, Emery S, … Pinto AN (2018). Viral suppression and HIV transmission in serodiscordant male couples: An international, prospective, observational, cohort study. The Lancet HIV, 5(8), e438–e447. 10.1016/S2352-3018(18)30132-2 [DOI] [PubMed] [Google Scholar]
  5. Bhatta DN, & Liabsuetrakul T (2017). Efficacy of a Social Self-Value Empowerment Intervention to Improve Quality of Life of HIV Infected People Receiving Antiretroviral Treatment in Nepal: A Randomized Controlled Trial. AIDS and Behavior, 21(6), 1620–1631. 10.1007/s10461-016-1546-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boyd MA, Moore CL, Molina J-M, Wood R, Madero JS, Wolff M, Ruxrungtham K, Losso M, Renjifo B, Teppler H, Kelleher AD, Amin J, Emery S, & Cooper DA (2015). Baseline HIV-1 resistance, virological outcomes, and emergent resistance in the SECOND-LINE trial: An exploratory analysis. The Lancet HIV, 2(2), e42–e51. 10.1016/S2352-3018(14)00061-7 [DOI] [PubMed] [Google Scholar]
  7. Breitbart W, Passik S, McDonald MV, Rosenfeld B, Smith M, Kaim M, & Funesti-Esch J (1998). Patient-related barriers to pain management in ambulatory AIDS patients: Pain, 76(1), 9–16. 10.1016/S0304-3959(98)00018-9 [DOI] [PubMed] [Google Scholar]
  8. Bucciardini R, D’ettorre G, Baroncelli S, Ceccarelli G, Parruti G, Weimer LE, Fragola V, Galluzzo CM, Pirillo MF, Lucattini S, Bellagamba R, Francisci D, Ladisa N, Antoni AD, Guaraldi G, Manconi PE, Vullo V, Preziosi R, Cirioni O, … Floridia M (2012). Virological failure at one year in triple-class experienced patients switching to raltegravir-based regimens is not predicted by baseline factors. International Journal of STD & AIDS, 23(7), 459–463. 10.1258/ijsa.2012.011391 [DOI] [PubMed] [Google Scholar]
  9. Canan C, Alexander GC, Moore R, Murimi I, Chander G, & Lau B (2019). Medicaid trends in prescription opioid and non-opioid use by HIV status. Drug and Alcohol Dependence, 197, 141–148. 10.1016/j.drugalcdep.2018.11.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, Hakim JG, Kumwenda J, Grinsztejn B, Pilotto JHS, Godbole SV, Mehendale S, Chariyalertsak S, Santos BR, Mayer KH, Hoffman IF, Eshleman SH, Piwowar-Manning E, Wang L, … Fleming TR (2011). Prevention of HIV-1 Infection with Early Antiretroviral Therapy. New England Journal of Medicine, 365(6), 493–505. 10.1056/NEJMoa1105243 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Costa J. de M., Torres TS, Coelho LE, & Luz PM (2018). Adherence to antiretroviral therapy for HIV/AIDS in Latin America and the Caribbean: Systematic review and meta-analysis. Journal of the International AIDS Society, 21(1), e25066. 10.1002/jia2.25066 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Denis CM, Morales KH, Wu Q, Metzger DS, & Cheatle MD (2019). Association Between Diagnoses of Chronic Noncancer Pain, Substance Use Disorder, and HIV-Related Outcomes in People Living With HIV: JAIDS Journal of Acquired Immune Deficiency Syndromes, 82, S142–S147. 10.1097/QAI.0000000000002179 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Etoori D, Ciglenecki I, Ndlangamandla M, Edwards CG, Jobanputra K, Pasipamire M, Maphalala G, Yang C, Zabsonre I, Kabore SM, Goiri J, Teck R, & Kerschberger B (2018). Successes and challenges in optimizing the viral load cascade to improve antiretroviral therapy adherence and rationalize second-line switches in Swaziland. Journal of the International AIDS Society, 21(10), e25194. 10.1002/jia2.25194 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Garofalo R, Kuhns LM, Hotton A, Johnson A, Muldoon A, & Rice D (2016). A Randomized Controlled Trial of Personalized Text Message Reminders to Promote Medication Adherence Among HIV-Positive Adolescents and Young Adults. AIDS and Behavior, 20(5), 1049–1059. 10.1007/s10461-015-1192-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Geocze L, Mucci S, De Marco MA, Nogueira-Martins LA, & Citero V. de A. (2010). Qualidade de vida e adesão ao tratamento anti-retroviral de pacientes portadores de HIV. Revista de Saúde Pública, 44(4), 743–749. 10.1590/S0034-89102010000400019 [DOI] [PubMed] [Google Scholar]
  16. Gorman AA, Foley JM, Ettenhofer ML, Hinkin CH, & van Gorp WG (2009). Functional Consequences of HIV-Associated Neuropsychological Impairment. Neuropsychology Review, 19(2), 186–203. 10.1007/s11065-009-9095-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gross R, Ritz J, Hughes MD, Salata R, Mugyenyi P, Hogg E, Wieclaw L, Godfrey C, Wallis CL, Mellors JW, Mudhune VO, Badal-Faesen S, Grinsztejn B, & Collier AC (2019). Two-way mobile phone intervention compared with standard-of-care adherence support after second-line antiretroviral therapy failure: A multinational, randomised controlled trial. The Lancet Digital Health, 1(1), e26–e34. 10.1016/S2589-7500(19)30006-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hari Chandra B, Ramesh M, & Nagendra H (2019). Effect of yoga on immune parameters, cognitive functions, and quality of life among HIV-positive children/adolescents: A pilot study. International Journal of Yoga, 12(2), 132. 10.4103/ijoy.IJOY_51_18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hinkin CH, Castellon SA, Durvasula RS, Hardy DJ, Lam MN, Mason KI, Thrasher D, Goetz MB, & Stefaniak M (2002). Medication Adherence among HIV + Adults, 59, 12. 10.1212/01.WNL.0000038347.48137.67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Isenberg MA, Maragh-Bass, MPH, PhD SR, Ridgeway, MSPH AC, Beach, MD, MPH, M. C. K, & Knowlton, MPH, ScD, A. R. (2017). A qualitative exploration of chronic pain and opioid treatment among HIV patients with drug use disorders. Journal of Opioid Management, 13(1), 5. 10.5055/jom.2017.0363 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Krashin DL, Merrill JO, & Trescot AM (2012). Opioids in the management of HIV-related pain. Pain Physician, 15(3 Suppl):ES157–68 [PubMed] [Google Scholar]
  22. La Rosa AM, Harrison LJ, Taiwo B, Wallis CL, Zheng L, Kim P, Kumarasamy N, Hosseinipour MC, Jarocki B, Mellors JW, & Collier AC (2016). Raltegravir in second-line antiretroviral therapy in resource-limited settings (SELECT): A randomised, phase 3, non-inferiority study. The Lancet HIV, 3(6), e247–e258. 10.1016/S2352-3018(16)30011-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Miaskowski C, Penko JM, Guzman D, Mattson JE, Bangsberg DR, & Kushel MB (2011). Occurrence and Characteristics of Chronic Pain in a Community-Based Cohort of Indigent Adults Living With HIV Infection. The Journal of Pain, 12(9), 1004–1016. 10.1016/j.jpain.2011.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ortego C, Huedo-Medina TB, Llorca J, Sevilla L, Santos P, Rodríguez E, Warren MR, & Vejo J (2011). Adherence to Highly Active Antiretroviral Therapy (HAART): A Meta-Analysis. AIDS and Behavior, 15(7), 1381–1396. 10.1007/s10461-011-9942-x [DOI] [PubMed] [Google Scholar]
  25. Parker RA, Rabideau DJ, Sax PE, Tierney C, Daar ES, Collier AC, Losina E, & Freedberg KA (2017). Impact of Medication Adherence on Virologic Failure in A5202: A Randomized, Partially Blinded, Phase 3B Study. Clinical Infectious Diseases, 64(11), 1612–1614. 10.1093/cid/cix176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Parker R, Stein DJ, & Jelsma J (2014). Pain in people living with HIV/AIDS: A systematic review. Journal of the International AIDS Society, 17(1), 18719. 10.7448/IAS.17.1.18719 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Remien RH, Stirratt MJ, Nguyen N, Robbins RN, Pala AN, & Mellins CA (2019). Mental health and HIV/AIDS: The need for an integrated response. AIDS, 33(9), 1411–1420. 10.1097/QAD.0000000000002227 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Rendina HJ, & Parsons JT (2018). Factors associated with perceived accuracy of the Undetectable = Untransmittable slogan among men who have sex with men: Implications for messaging scale-up and implementation. Journal of the International AIDS Society, 21(1), e25055. 10.1002/jia2.25055 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rodger AJ, Cambiano V, Bruun T, Vernazza P, Collins S, Degen O, Corbelli GM, Estrada V, Geretti AM, Beloukas A, Raben D, Coll P, Antinori A, Nwokolo N, Rieger A, Prins JM, Blaxhult A, Weber R, Van Eeden A, … Janeiro N (2019). Risk of HIV transmission through condomless sex in serodifferent gay couples with the HIV-positive partner taking suppressive antiretroviral therapy (PARTNER): Final results of a multicentre, prospective, observational study. The Lancet, 393(10189), 2428–2438. 10.1016/S0140-6736(19)30418-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Safren SA, Biello KB, Smeaton L, Mimiaga MJ, Walawander A, Lama JR, Rana A, Nyirenda M, Kayoyo VM, Samaneka W, Joglekar A, Celentano D, Martinez A, Remmert JE, Nair A, Lalloo UG, Kumarasamy N, Hakim J, Campbell TB, & for the PEARLS (ACTG A5175) Study Team. (2014). Psychosocial Predictors of Non-Adherence and Treatment Failure in a Large Scale Multi-National Trial of Antiretroviral Therapy for HIV: Data from the ACTG A5175/PEARLS Trial. PLoS ONE, 9(8), e104178. 10.1371/journal.pone.0104178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Safren SA, Hendriksen ES, Smeaton L, Celentano DD, Hosseinipour MC, Barnett R, Guanira J, Flanigan T, Kumarasamy N, Klingman K, & Campbell T (2012). Quality of Life Among Individuals with HIV Starting Antiretroviral Therapy in Diverse Resource-Limited Areas of the World. AIDS and Behavior, 16(2), 266–277. 10.1007/s10461-011-9947-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Schackman BR, Ribaudo HJ, Krambrink A, Hughes V, Kuritzkes DR, & Gulick RM (2007). Racial Differences in Virologic Failure Associated With Adherence and Quality of Life on Efavirenz-Containing Regimens for Initial HIV Therapy: Results of ACTG A5095. JAIDS Journal of Acquired Immune Deficiency Syndromes, 46(5), 547–554. 10.1097/QAI.0b013e31815ac499 [DOI] [PubMed] [Google Scholar]
  33. Shahani L, Salazar L, Woods SP, & Hasbun R (2018). Baseline Neurocognitive Functioning Predicts Viral Load Suppression at 1-Year Follow-Up Among Newly Diagnosed HIV Infected Patients. AIDS and Behavior, 22(10), 3209–3213. 10.1007/s10461-018-2025-5 [DOI] [PubMed] [Google Scholar]
  34. Smith K, Powers KA, Kashuba AD, & Cohen MS (2011). HIV-1 treatment as prevention: The good, the bad, and the challenges: Current Opinion in HIV and AIDS, 1. 10.1097/COH.0b013e32834788e7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Surratt HL, Kurtz SP, Levi-Minzi MA, Cicero TJ, Tsuyuki K, & O’Grady CL (2015). Pain Treatment and Antiretroviral Medication Adherence Among Vulnerable HIV-Positive Patients. AIDS Patient Care and STDs, 29(4), 186–192. 10.1089/apc.2014.0104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. The INSIGHT START Study Group. (2015). Initiation of Antiretroviral Therapy in Early Asymptomatic HIV Infection. New England Journal of Medicine, 373(9), 795–807. 10.1056/NEJMoa1506816 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Torres TS, Cox J, Marins LM, Bezerra DR, Veloso VG, Grinsztejn B, & Luz PM (2020). A call to improve understanding of undetectable equals untransmittable (U = U) in Brazil: A web-based survey. Journal of the International AIDS Society, 23(11),e25630. 10.1002/jia2.25630 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Torres TS, Harrison LJ, La Rosa AM, Cardoso SW, Zheng L, Ngongondo M, Some F, Lalloo UG, Mwelase T, Collier AC, & Hughes MD (2018). Quality of life improvement in resource-limited settings after one year of second-line antiretroviral therapy use among adult men and women. AIDS(London, England), 32(5):583–593. 10.1097/QAD.0000000000001738 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Torres TS, Harrison LJ, La Rosa AM, Lavenberg JA, Zheng L, Safren SA, Ngongondo M, Poongulali S, Matoga M, Samaneka W, Collier AC, & Hughes MD (2018). Quality of life among HIV-infected individuals failing first-line antiretroviral therapy in resource-limited settings. AIDS Care, 30(8), 954–962. 10.1080/09540121.2018.1427207 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Uebelacker LA, Weisberg RB, Herman DS, Bailey GL, Pinkston-Camp MM, & Stein MD (2015). Chronic Pain in HIV-Infected Patients: Relationship to Depression, Substance Use, and Mental Health and Pain Treatment. Pain Medicine, 16(10), 1870–1881. 10.1111/pme.12799 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. WHO. (2015). Global forecasts of antiretroviral demand for 2014–2018, projection modelling for new antiretroviral formulations for 2015–2024 and update on hepatitis B & C. https://www.who.int/hiv/amds/ARV-forecasting-meeting-2015/en/
  42. WHO. (2016). Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection: Recommendations for a public health approach. http://www.ncbi.nlm.nih.gov/books/NBK374294/ [PubMed]
  43. Wilson IB (1995). Linking Clinical Variables With Health-Related Quality of Life: A Conceptual Model of Patient Outcomes. JAMA, 273(1), 59. 10.1001/jama.1995.03520250075037 [DOI] [PubMed] [Google Scholar]

RESOURCES