Simple questionnaire to monitor HIV treatment adherence in children

Adherence to antiretroviral treatment (ART) is worryingly poor among children living with HIV in three low- and middle-income countries (LMICs). But a simple, cost-effective questionnaire could help identify children with adherence below 90%, and those experiencing treatment interruptions and viral non-suppression.

In a recent study published in the Journal of the International AIDS Society (JIAS), researchers developed a 10-question questionnaire which could effectively predict 48-hour treatment interruptions and adherence below 90% in children.

Most of the tools currently available to monitor adherence are either costly or burdensome for the health facility, which include pill counts or electronic dose monitoring that rely on either staff time or expensive equipment. As a result many clinics use self-reports to monitor adherence, which can be unreliable as care-givers and children may be reluctant to admit non‐adherence to health providers.  

Given the poor levels of adherence in this population and the subsequent impact on morbidity and mortality, there is a need to identify more effective tools to monitor and understand adherence in children.

In the present study, 319 children aged 0 to 16, on first- or second-line ART in Kenya (n=110), South Africa (n=109) and Thailand (n=100) were enrolled with their caregivers for six months of adherence monitoring.  All participants were given a Medication Event Monitoring Systems® (MEMS®) bottle, which records the time and date of a pill bottle opening. This has been shown to offer a highly reliable adherence measure, because of its association with viral outcomes.

Participants were also given the adherence questionnaire during their routine clinic visits at baseline, at month three and at month six. The questionnaire was developed in Kenya and was informed by a literature review, cognitive interviews, and focus group discussions with children, caregivers, and providers. An initial pool of 48 questions was reduced to 10 questions found to have the largest impact on adherence.

The questionnaire was administered to the child, the caregiver, or both parties depending on who reported primary responsibility for the child’s medication-taking. The survey questions were not included in the journal article but are available upon request from the authors.

The average age of the children included in the study was 11 and most (54%) were female. The median age of children from Thailand was older (14 years), compared to 10 years in Kenya and South Africa. Viral suppression was highest in Thailand at 97%, followed by South Africa (81%) and Kenya (69%). 57%  of children in Thailand had MEMS® adherence prevalence of over 90%, compared to 58% in South Africa and 40% in Kenya.

The results of the questionnaire were validated against data downloaded and reviewed from the MEMS® bottle, verifying its ability to predict the percentage of participants adhering above or below 90% according to the MEMS® measures.

The questionnaire was significantly associated with 48-hour treatment interruption and effectively predicted whether MEMS® adherence was above or below 90% in all settings. It was the most sensitive at predicting adherence outcomes in Kenya, where it was also associated with viral suppression. It had a lower sensitivity in South Africa and Thailand, but was still associated with MEMS® adherence.

The results of the electronic dose monitoring showed that just under half (48%) of the children took 90% of their prescribed doses, and about 40% had experienced at least one treatment interruption of more than 48 hours every three months, and 82% were virally suppressed.

Similarly, child reports using the questionnaire showed that 52% were adherent at baseline, 57% at month three and 60% at month six.

Caregiver‐reports using the questionnaire showed that 50% of children were adherent at baseline, 53% at month three and 60% at month six. This high level of consistency across the three different approaches to adherence measuring demonstrates the potential of the questionnaire in assessing adherence levels in these populations.

Although the authors noted that the questionnaire may need further refinement to specific contexts for future roll-out in clinical settings, they commented that “this study provides evidence for a 10-item questionnaire that may be further revised and improved in different contexts for routine adherence screening in a clinic setting.” These findings are important as regular and accurate adherence evaluation is critical to the support of children living with HIV, most of whom live in LMICs.