Adolescents Living with HIV (ALWHIV) have increased in the last decade, especially in Sub-Saharan Africa, including Uganda. Optimal ART adherence ( 95%) is crucial for improving long-term patient outcomes and limiting HIV transmission; however, studies indicate low ART adherence among ALWHIV. Therefore, it is critical to understand the drivers of poor adherence and urgently devise interventions in this group. We determined the level of adherence and factors predicting poor ART adherence among ALWHIV.
The study used data from 602 ALWHIV aged 10-16 years who participated in a five-year NICHD-funded Suubi+Adherence cluster-randomized controlled (2012-2017) in 39 clinics in South-western Uganda. The Suubi+Adherence study aimed to examine the impact of an innovative family-based economic empowerment intervention package on HIV treatment adherence among HIV-positive adolescents. We performed multilevel logistic regression analysis to explore to what extent socio-demographic, household, and treatment-related characteristics predict six-month ART adherence. Adherence was determined using a wise-pill real-time medication adherence monitoring device as (the number of times device was opened)/total number of expected doses. We used Stata 15.1 and setting significance levels at 0.05. We stratified the regression analysis according to study arms.
Overall, the mean age was 12.4 2.0 years, and 56.4% were female. All baseline socio-demographic and household characteristics were comparable between the intervention and control groups.
After six months, the mean ART Adherence was 92.5% (CI: 90.9% – 94.2%) in the control arm, and 93.3% (CI: 91.7 – 94.9) in the intervention arm, p value=0.503. Only 70.0% (CI: 64.4% – 75.0%) of the participants in the control arm achieved 95% adherence versus 74.2 % (CI: 68.9% – 78.8%) in the intervention arm, p value=0.261.
Factors predicting level of adherence
In the control arm, baseline viral load of over 1000 copies/ml aOR=0.38 (95% CI: 0.17 – 0.69, P value =0.003) and taking ARVs more than once a day aOR=9.76 (95% CI: 3.22 – 29.53, P value <0.001) were significantly associated with achieving 95% ART adherence.
In the intervention group, age aOR=0.68 (95% CI: 0.55 – 0.84, P value <0.001), number of pills taken per day aOR=0.68 (95% CI: 0.52 – 0.88, P value =0.003), viral load above 1000 copies/ml aOR=0.36 (95% CI: 0.18 – 0.74, P value =0.005), and taking ARVs more than once a day aOR=2.56 (95% CI: 1.04 – 6.30, P value =0.041) were associated with good ART adherence
Conclusions and implications
We found that close to three quarters of ALWHIV had good level of ART adherence, which is still below the 90-90-90 target. Our findings demonstrated that high baseline viral loads are an important predictor of poor ART adherence. Interestingly, high pill burden and taking ART more than once a day were associated with better adherence. Therefore, programs to improve ART adherence should consider incorporating age-specific interventions and intensifying adherence counseling among patients with high baseline viral loads.