Evidence of effective adherence interventions is still limited

This article is more than 10 years old. Click here for more recent articles on this topic

Perhaps tellingly, the Fifth International Conference on HIV Treatment Adherence conference in Miami this week was not bursting with evaluations of adherence interventions which conclusively demonstrated that a given intervention was effective in a specific population.

Indeed, when Mahnaz Charania of the Centers for Disease Prevention and Control (CDC) outlined the methodology of the CDC’s efficacy review of adherence interventions, she indicated that in the draft results, there were only eight interventions considered to have “promising evidence” of efficacy. They found none with “best evidence” of efficacy. (The CDC only considered studies conducted in the United States).

Charania gave only limited details, but did indicate that two of the eight involved specific ways of delivering antiretroviral therapy (via a methadone clinic or mobile van), while the other six were educational or behavioural interventions. Common features of the latter were that they were all delivered by a nurse or primary healthcare provider, all had a cognitive-behavioural component (for example addressing barriers), several featured a nominated support partner and half included problem solving.


retention in care

A patient’s regular and ongoing engagement with medical care at a health care facility. 

control group

A group of participants in a trial who receive standard treatment, or no treatment at all, rather than the experimental treatment which is being tested. Also known as a control arm.

However during questions, Allen Gifford pointed out that the CDC’s methodology involves rigorous checks on the internal validity of an intervention’s evaluation (in other words, does this study really prove that this intervention worked with these people?). Once the CDC have finalised their list of interventions, each one will then be field-tested to see if results can be replicated with other people in other settings.

Gifford suggested that it was only at this stage that it will be possible to see if the interventions have external validity too. However, given that the factors which affect adherence are often culturally specific, and that there will always be some variation in the way in which behavioural interventions are delivered, a number of interventions will then no longer appear to be effective. He suggested that an improved methodology would include consideration of external validity at an earlier stage.

One limitation of many adherence interventions is that they target individuals. However levels of adherence are not just influenced by factors under a patient’s control (such as developing a routine, use of recreational drugs or motivation). Adherence is also influenced by doctors (not explaining the regimen, not considering the person’s lifestyle, or choosing an overly complex regimen). Furthermore, it can be influenced by healthcare systems (introducing costs, delays or other difficulties in accessing treatment).

Some conference presentations did point to interventions which work beyond the individual level.

For example, Mary Catherine Beach presented findings from a small randomised controlled trial that intervened with both doctors and patients in the United States. Doctors received brief training on communication skills with a focus on motivational interviewing in support of adherence. Their HIV-positive patients were coached to discuss adherence issues. Doctors and patients in the control arm did not receive any intervention.

The impact was assessed by recording and analysing interactions the subjects then had. In those who had received the interventions, there was more discussion of adherence, and doctors were more likely to ask the patient’s opinion, talk about emotions and brainstorm solutions to non-adherence. However doctors still did most of the talking, suggesting that further work is needed on the skills required to engage patients.

Also, John Chalker of Management Sciences for Health outlined the development of standardised and validated measures of adherence that can be derived from clinic records in resource-limited settings. These measures record adherence at a clinic level, rather than in terms of individuals. If used systematically across a region or a country, they could help programme planners identify poorly performing facilities which need intervention. Moreover, they could facilitate the evaluation of interventions which aim to improve the delivery of healthcare, and through it adherence and retention in care.

The measures are based on data that can be collected by clinics in a routine way, including patients’ self-report of adherence in recent days, the frequency of pharmacy refills and attendance at appointments. Management Sciences for Health produce an electronic record-keeping tool which helps with clinic management and also produces these indicators automatically.

Trials of interventions to improve clinic performance are under way in Uganda, Tanzania, Rwanda and Kenya. The interventions aim to reduce clinic congestion and waiting time, and improve appointment recording, rates of appointment attendance, record keeping and the follow-up of missing patients.


Charania M. CDC review and dissemination of evidence-based HIV treatment adherence interventions. Fifth International Conference on HIV Treatment Adherence, Miami, 2010.

Beach MC et al. Impact of a patient and provider intervention to improve the quality of communication about medication adherence among HIV patients. Fifth International Conference on HIV Treatment Adherence, abstract 61339, Miami, 2010.

Chalker J. The role of ARV adherence data in monitoring health system performance in resource limited settings. Fifth International Conference on HIV Treatment Adherence, Miami, 2010.

Related topics