People with HIV in Tanzania were more likely to report problems with non-adherence to antiretroviral treatment after their healthcare providers underwent training in patient-centred communication skills and how to discuss adherence with patients, findings from a cohort study published in HIV Medicine show.
Good adherence to treatment is critical for long-term virological suppression, CD4 cell recovery and prevention of treatment failure. Checking for adherence difficulties is an important element of ongoing HIV care. But, the way that providers ask about adherence and the ways in which they respond to information about non-adherence is likely to have a critical effect on the willingness of patients to report adherence.
Some treatment providers and researchers in sub-Saharan Africa think that the negative responses of healthcare workers to disclosures of non-adherence are an important reason for the loss of patients from care. For example, a qualitative study of people with HIV in five African countries found that patients perceived a need to adopt a co-operative or compliant persona if they wished to be treated well by healthcare workers and to submit to the authority of healthcare workers. Disclosure of non-adherence was likely to invite a hostile or disapproving response, or perhaps even referral to a disciplinary committee.
To understand if improving the ability of healthcare providers to solicit information about adherence had an effect on patient outcomes, researchers from the Kilombero and Ulanga Antiretroviral Cohort in Tanzania designed a study to test the effect of a two-day patient-centred communication skills workshop, and an instrument for checking adherence, on patient outcomes.
The intervention consisted of a two-day workshop at which 13 healthcare providers who had direct contact with patients reviewed how to identify and discuss non-adherence and how to improve adherence. They shared examples of difficulties in discussing adherence with patients, then looked at solutions and carried out role-plays to practice communication skills. Participants were also given a checklist of adherence questions based on those developed by the European AIDS Clinical Society (EACS).
The study recruited a prospective cohort of 299 patients who were receiving antiretroviral treatment at the Chronic Diseases Clinic of Ikafara, part of the Kilombero and Ulanga Antiretroviral Cohort in rural Tanzania. Approximately two-thirds (65%) had very advanced HIV disease (CD4 < 200 cells/mm3) at the time of treatment initiation and one third (34%) had been diagnosed with an AIDS-defining illness in the past, predominantly tuberculosis. Participants had been taking antiretroviral treatment for a median of three-and-a-half years.
Two-thirds were taking efavirenz-based treatment and a further 24% were taking nevirapine. The nucleoside reverse transcriptase inhibitor (NRTI) backbones used most commonly were tenofovir and emtricitabine (45%) and zidovudine and lamivudine (54%).
Participants were followed for one year, during which 81% attended all three scheduled clinic visits and underwent adherence assessment. At baseline 3.3% reported missing at least one dose in the previous four weeks; at visit 2 (median 63 days later) 10.7% reported non-adherence (p < 0.001) and at visit 3, 5.7% reported non-adherence, a non-significant difference with baseline. The most common reason for missing a dose was running out of pills (64%) or forgetting (10%).
To check whether non-adherence was under-reported the investigators measured plasma drug concentrations of efavirenz, nevirapine, atazanavir or lopinavir. 6.5% had sub-therapeutic drug concentrations at baseline. The proportion of patients with sub-therapeutic concentrations did not decline during the study, although there was a trend towards a decline in the efavirenz-treated group.
The study found no difference in odds of virological failure over time, but the risk of immunological failure declined (OR 0.75, p = 0.002). Virological failure was measured using the WHO definition (viral load above 1000 copies/ml); at baseline, 7.7% had detectable viral load and at visits 2 and 3, 9.1% and 9.2% respectively had a detectable viral load. At a lower cut-off rate of 500 copies/ml, the virological failure rate was also 9%. The authors say that the high adherence rate and high baseline rate of virological suppression, coupled with a modest sample size, made it unlikely that they could detect a difference.
Three patients were lost to follow-up during the study, six died (five for unknown reasons) and six developed a new AIDS-defining illness. The study does not report the relationship between adherence at baseline and loss to follow-up, clinical illness or death.
The researchers conclude that training in patient-centred communication skills can improve the reporting of non-adherence in people with HIV in a sub-Saharan African setting.