Managing medication compliance of tuberculosis patients in Haiti with medication monitors

SETTING: A tuberculosis clinic in Haiti using self-administered medication. OBJECTIVE: To determine if medication monitors could be used along with directly observed therapy in developing countries to help solve the problem of compliance with medication. DESIGN: Patients were randomized into three groups: Group A took medication from medication monitors and were given counseling based on the monitor record; Group B took medication from medication monitors, but the record was not available for counseling; and Group C took medication from simple containers. RESULTS: Good monitor records in the first 11 weeks predicted less default from treatment (P < 0.01), and better compliance (P < 0.01) in the last 9 months. Counseling based on the monitor record appeared to reduce treatment abandonment by about half. CONCLUSION: This study suggests that medication monitoring of self administered treatment would be useful in settings where directly observed therapy cannot be delivered for the entire duration of treatment, especially in rural areas, by 1) identifying poor compliance early in therapy, 2) reducing the frequency of clinic visits for patients with good monitor records who live excessive distances from the clinic, 3) counseling patients about their monitor record to improve treatment completion rates, and 4) lengthening the duration of therapy when poor compliance is found.