Facility-level intervention to improve attendance and adherence among patients on anti-retroviral treatment in Kenya—a quasi-experimental study using time series analysis

BACKGROUND: Achieving high rates of adherence to antiretroviral therapy (ART) in resource-poor settings comprises serious, but different, challenges in both the first months of treatment and during the life-long maintenance phase. We measured the impact of a health system-oriented, facility-based intervention to improve clinic attendance and patient adherence. METHODS: This was a quasi-experimental, longitudinal, controlled intervention study using interrupted time series analysis. The intervention consisted of (1) using a clinic appointment diary to track patient attendance and monitor monthly performance; (2) changing the mode of asking for self-reported adherence; (3) training staff on adherence concepts, intervention methods, and use of monitoring data; (4) conducting visits to support facility teams with the implementation. We conducted the study in 12 rural district hospitals (6 intervention, 6 control) in Kenya and randomly selected 1894 adult patients over 18 years of age in two cohorts: experienced patients on treatment for at least one year, and newly treated patients initiating ART during the study. Outcome measures were: attending the clinic on or before the date of a scheduled appointment, attending within 3 days of a scheduled appointment, reporting perfect adherence, and experiencing a gap in medication supply of more than 14 days. RESULTS: Among experienced patients, the percentage attending the clinic on or before a scheduled appointment increased in both level (average total increase immediately after intervention) (+5.7%; 95% CI = 2.1, 9.3) and trend (increase per month) (+1.0% per month; 95% CI = 0.6, 1.5) following the intervention, as did the level and trend of those keeping appointments within three days (+4.2%; 95% CI = 1.6, 6.7; and +0.8% per month; 95% CI = 0.6, 1.1, respectively). The relative difference between the intervention and control groups based on the monthly difference in visit rates increased significantly in both level (+6.5; 95% CI = 1.4, 11.6) and trend (1.0% per month; 95% CI = 0.2, 1.8) following the intervention for experienced patients attending the clinic within 3 days of their scheduled appointments. The decrease in the percentage of experienced patients with a medication gap greater than 14 days approached statistical significance (-11.3%; 95% CI = -22.7, 0.1), and the change seemed to persist over 11 months after the intervention. All facility staff used appointment-keeping data to calculate adherence and discussed outcomes regularly. CONCLUSION: The appointment-tracking system and monthly performance monitoring was strengthened, and patient attendance was improved. Scale-up to national level may be considered.