The adherence-outcome relationship is not altered by diary-driven adjustments of microelectronic monitor data

Measuring patient engagement with HIV care in sub-Saharan Africa: a scoping study

INTRODUCTION

Engagement with HIV care is a multi-dimensional, dynamic process, critical to maintaining successful treatment outcomes. However, measures of engagement are not standardized nor comprehensive. This undermines our understanding of the scope of challenges with engagement and whether interventions have an impact, complicating patient and programme-level decision-making. This study identified and characterized measures of engagement to support more consistent and comprehensive evaluation.

METHODS

We conducted a scoping study to systematically categorize measures the health system could use to evaluate engagement with HIV care for those on antiretroviral treatment. Key terms were used to search literature databases (Embase, PsychINFO, Ovid Global-Health, PubMed, Scopus, CINAHL, Cochrane and the World Health Organization Index Medicus), Google Scholar and stakeholder-identified manuscripts, ultimately including English evidence published from sub-Saharan Africa from 2014 to 2021. Measures were extracted, organized, then reviewed with key stakeholders.

RESULTS AND DISCUSSION

We screened 14,885 titles/abstracts, included 118 full-texts and identified 110 measures of engagement, categorized into three engagement dimensions ("retention," "adherence" and "active self-management"), a combination category ("multi-dimensional engagement") and "treatment outcomes" category (e.g. viral load as an end-result reflecting that engagement occurred). Retention reflected status in care, continuity of attendance and visit timing. Adherence was assessed by a variety of measures categorized into primary (prescription not filled) and secondary measures (medication not taken as directed). Active self-management reflected involvement in care and self-management. Three overarching use cases were identified: research to make recommendations, routine monitoring for quality improvement and strategic decision-making and assessment of individual patients.

CONCLUSIONS

Heterogeneity in conceptualizing engagement with HIV care is reflected by the broad range of measures identified and the lack of consensus on "gold-standard" indicators. This review organized metrics into five categories based on the dimensions of engagement; further work could identify a standardized, minimum set of measures useful for comprehensive evaluation of engagement for different use cases. In the interim, measurement of engagement could be advanced through the assessment of multiple categories for a more thorough evaluation, conducting sensitivity analyses with commonly used measures for more comparable outputs and using longitudinal measures to evaluate engagement patterns. This could improve research, programme evaluation and nuanced assessment of individual patient engagement in HIV care.

Measuring patient engagement with HIV care in sub-Saharan Africa: a scoping study

INTRODUCTION

Engagement with HIV care is a multi-dimensional, dynamic process, critical to maintaining successful treatment outcomes. However, measures of engagement are not standardized nor comprehensive. This undermines our understanding of the scope of challenges with engagement and whether interventions have an impact, complicating patient and programme-level decision-making. This study identified and characterized measures of engagement to support more consistent and comprehensive evaluation.

METHODS

We conducted a scoping study to systematically categorize measures the health system could use to evaluate engagement with HIV care for those on antiretroviral treatment. Key terms were used to search literature databases (Embase, PsychINFO, Ovid Global-Health, PubMed, Scopus, CINAHL, Cochrane and the World Health Organization Index Medicus), Google Scholar and stakeholder-identified manuscripts, ultimately including English evidence published from sub-Saharan Africa from 2014 to 2021. Measures were extracted, organized, then reviewed with key stakeholders.

RESULTS AND DISCUSSION

We screened 14,885 titles/abstracts, included 118 full-texts and identified 110 measures of engagement, categorized into three engagement dimensions ("retention," "adherence" and "active self-management"), a combination category ("multi-dimensional engagement") and "treatment outcomes" category (e.g. viral load as an end-result reflecting that engagement occurred). Retention reflected status in care, continuity of attendance and visit timing. Adherence was assessed by a variety of measures categorized into primary (prescription not filled) and secondary measures (medication not taken as directed). Active self-management reflected involvement in care and self-management. Three overarching use cases were identified: research to make recommendations, routine monitoring for quality improvement and strategic decision-making and assessment of individual patients.

CONCLUSIONS

Heterogeneity in conceptualizing engagement with HIV care is reflected by the broad range of measures identified and the lack of consensus on "gold-standard" indicators. This review organized metrics into five categories based on the dimensions of engagement; further work could identify a standardized, minimum set of measures useful for comprehensive evaluation of engagement for different use cases. In the interim, measurement of engagement could be advanced through the assessment of multiple categories for a more thorough evaluation, conducting sensitivity analyses with commonly used measures for more comparable outputs and using longitudinal measures to evaluate engagement patterns. This could improve research, programme evaluation and nuanced assessment of individual patient engagement in HIV care.

Reasons for discordance between antiretroviral adherence measures in adolescents

Adolescents with suboptimal medication taking may deceive caregivers about non-adherence. We conducted a 5-year longitudinal study of adolescents monitored simultaneously with both easily manipulated (e.g., self-report) and hard to manipulate (e.g., microelectronic data) strategies. Adolescents with repeatedly high adherence on the former and low adherence on the latter were invited along with their parental figures ("parents") to participate. We conducted focus groups and semi-structured interviews, separately for adolescents and parents, to elucidate drivers of discordant measures. Forty-seven adolescents and 26 parents participated in focus groups and 4 adolescents were interviewed. Adolescents described hiding pills, discarding pills, and lying. Their motivations included fear of disappointing those who care about them, desire to avoid admonishment by parents and clinic staff, and desire to avoid remedial adherence counseling. Both adolescents and parents considered negative feedback for prior poor adherence to be key motivation to hide current poor adherence from clinic staff. Providing positive feedback for truth-telling, rather than for "evidence" of excellent adherence, might help adolescent patients and their parents to develop stronger treatment alliances with each other and with clinic staff. Such alliances would allow adherence interventions to be better targeted and more fruitful in increasing adherence.

Trends in HIV Treatment Adherence Before and After HIV Status Disclosure to Adolescents in Botswana

PURPOSE

This study aimed to determine if disclosure of HIV status to adolescents impacted their medication adherence and how medication autonomy might explain observed effects.

METHODS

Three hundred adolescents on HIV treatment using electronic medication monitors were followed for 24 months while undergoing routine care. One hundred six of the adolescents were HIV disclosure-naïve and HIV status disclosure in this group was assessed quarterly. Analyses included data from the 75 adolescents who experienced disclosure during the study providing adherence and autonomy data both predisclosure and postdisclosure. Segmented generalized estimating equations were used to examine the trend of adherence and autonomy predisclosure and postdisclosure. Covariates assessed include age at disclosure, sex, and orphan status.

RESULTS

Median age at study entry was 12.2 years (interquartile range 11.6-12.9). Incident disclosure occurred in 75 (71%) of the adolescents at a median age of 13.1 years (interquartile range 12.5-13.9). Adherence decreased by 11% (95% confidence interval [CI] 7-15, p < .001) during the predisclosure period and by 22% (95% CI 9-36, p = .001) during the postdisclosure period. Adolescents' autonomy over their medication-taking increased over time, but disclosure did not impact the rate of increase in measured medication-taking autonomy. On a scale of 1-4 assessing autonomy (1 = receiving directly observed therapy and 4 = taking medicines mostly without supervision), autonomy increased by an average of .03 units/month (95% CI .02-.03, p < .001) predisclosure and by .05 units/month (95% CI -.01 to .11, p = .42) postdisclosure.

CONCLUSIONS

The findings suggest that, among perinatally HIV infected adolescents, HIV status disclosure may adversely impact treatment adherence. Postdisclosure support to HIV infected adolescents should be intensified.

Outcome measures for adherence data from a medication event monitoring system: A literature review

WHAT IS KNOWN

Currently, medication bottles with an electronic cap are frequently used to measure medication adherence. This system is termed medication event monitoring system (MEMS). To our knowledge, the optimal method to summarize data from MEMS has not yet been determined.

OBJECTIVE

Look for best practices on how to quantify adherence data from MEMS.

METHODS

Review of PubMed, Embase and Cochrane databases for the articles on medication adherence with MEMS.

RESULTS

Of 1493 identified articles, 207 were included in this review. The MEMS cap was used for a median of 3 months (IQR: 4; range: 1 week to 24 months) in various health conditions. Many different outcome measures were used. Most studies computed an adherence score, expressed as the percentage of days on which the correct dose of medication was taken. The threshold to mark people as adherent was most frequently, arbitrarily, set at 80% (range: 67%-95%). We found no data to support a specific threshold.

DISCUSSION

Although the commonly used definition of adherence has face validity, we found no validation studies, and not all studies used the same cut-off for adherence. Ideally, a cut-off should be defined and validated in the context of the specific drug and its pharmacokinetic and dynamic characteristics, and perhaps other contextual factors, rather than generically. In addition, there was large heterogeneity in the definition of what "correct intake" of medication is.

WHAT IS NEW AND CONCLUSION

Outcome measures for MEMS data lacked standardization, and no demonstrable effort to validate any definition against a relevant clinical outcome is available. Consensus on the definition of adherence is urgently needed.

Adherence to antiretroviral therapy in a clinical cohort of HIV-infected children in East Africa

OBJECTIVE

To describe antiretroviral therapy (ART) adherence and associated factors for a large HIV-infected pediatric cohort followed by sites of the East Africa International Epidemiologic Databases to Evaluate AIDS (IeDEA) consortium.

METHODS

This study utilized prospectively collected clinical data from HIV-infected children less than 13 years of age who initiated ART within 4 clinical care programs (with 26 clinical sites) in Kenya, Uganda, and Tanzania and were followed for up to 6 years. Programs used one of 3 adherence measures, including 7-day quantitative recall, 7-day categorical recall, and clinician pill assessments. We fit a hierarchical, three-level, logistic-regression model to examine adherence, with observations nested within patient, and patients within the 26 sites providing pediatric HIV data to this analysis.

RESULTS

In East Africa, 3,304 children, 52.0% male, were enrolled in care and were subsequently observed for a median of 92 weeks (inter-quartile range [IQR] 50.3-145.0 weeks). Median age at ART initiation was 5.5 years ([IQR] 3.0-8.5 years). "Good" adherence, as reported by each clinic's measures, was extremely high, remaining on average above 90% throughout all years of follow-up. Longer time on ART was associated with higher adherence (adjusted Odds Ratio-aOR-per log-transformed week on ART: 1.095, 95% Confidence Interval-CI-[1.052-1.150].) Patients enrolled in higher-volume programs exhibited higher rates of clinician-assessed adherence (aOR per log-500 patients: 1.174, 95% CI [1.108-1.245]). Significant site-level variability in reported adherence was observed (0.28), with even higher variability among patients (0.71). In a sub-analysis, being an orphan at the start of ART was strongly associated with lower ART adherence rates (aOR: 0.919, 95% CI [0.864-0.976]).

CONCLUSIONS

Self-reported adherence remained high over a median of 1.8 years in HIV care, but varied according to patient-level and site-level factors. Consistent adherence monitoring with validated measures and attention to vulnerable groups is recommended.

Distinctive barriers to antiretroviral therapy adherence among non-adherent adolescents living with HIV in Botswana

Levels of adherence to HIV treatment are lower among adolescents compared with older and younger individuals receiving similar therapies. We purposely sampled the most and least adherent adolescents from a 300-adolescent longitudinal HIV treatment adherence study in Gaborone, Botswana. Multiple objective and subjective measures of adherence were available and study participants were selected based on sustained patterns of either excellent or poor adherence over a one-year period. Focus group discussions (FGD) and in-depth interviews (IDI) were conducted with the adolescents and a subset of their caregivers with the goal of revealing barriers and facilitators of adherence. Focus groups were segregated by adherence classification of the participants. Following coding of transcripts, matrices were developed based on participants' adherence classifications in order to clarify differences in themes generated by individuals with different adherence characteristics. 47 adolescents and 25 adults were included. The non-adherent adolescents were older than the adherent adolescents (median age 18 years (IQR 16-19) vs. 14 years (IQR 12-15 years)), with median time on treatment near 10 years in both groups. Interference with daily activities, concerns about stigma and discrimination, side effects, denial of HIV status, and food insecurity arose as challenges to adherence among both those who were consistently adherent and those who were poorly-adherent to their medications. Low outcome expectancy, treatment fatigue, mental health and substance use problems, and mismatches between desired and received social support were discussed only among poorly adherent adolescents and their caregivers. Challenges raised only among adolescents and caregivers in the non-adherent groups are hypothesis-generating, identifying areas that may have a greater contribution to poor outcomes than challenges faced by both adherent and non-adherent adolescents. The contribution of these factors to poor outcomes should be explored in future studies.