RESUMEN
OBJECTIVE: To verify compliance with cotrimoxazole prophylaxis in human immunodeficiency virus (HIV) infected tuberculosis (TB) patients during the continuation phase of anti-tuberculosis treatment, and to assess the sensitivity, specificity and positive predictive values of verbal verification and pill counts as methods of checking compliance. DESIGN: Cross-sectional study. METHODS: Cotrimoxazole compliance was assessed in a cohort of TB patients who were attending four TB follow-up centres during the continuation phase of anti-TB treatment between months 4 and 6. Verbal verification of drug intake, physical verification of pill count balance, and urine trimethoprim detection by gas chromatography and mass spectrometry were used for assessing compliance. RESULTS: Using urine trimethoprim detection as the gold standard for compliance, trimethoprim was detected in 82 (94%) of 87 patients in the cohort. Verbal verification of cotrimoxazole intake and objective pill count balances showed high sensitivity and positive predictive values compared with the gold standard of urine trimethoprim detection. CONCLUSIONS: In a rural district in Malawi, compliance with cotrimoxazole as an adjunct to anti-tuberculosis treatment in HIV-infected TB patients was good, and can be assessed simply and practically by verbal verification and pill counts.
