Background: With the implementation of Directly Observed Treatment Supervision (DOTS), TB treatment in South Africa is typically supervised by trained healthcare workers or lay caregivers based at community organizations. However, patients may also choose to pick up treatment themselves and ‘self-administer’ or ask a family member to pick up the treatment from the clinic. We reviewed clinic TB cards to describe the proportion of patients receiving supervised treatment and to examine whether supervised treatment resulted in improved treatment outcomes relative to family or self-administered treatment. Methods: A de-identified retrospective medical record review was performed for a census of adult (≥18 years) TB cases registered 1 April 2011–31 March 2012 at three primary healthcare clinics (PHCs) in Johannesburg, South Africa. Treatment supervisor was classified in the files as facility, community caregiver, employer, family member or as self-administered treatment. Treatment supervision was recorded separately for the initial intensive phase of treatment (lasting 2-3 months) and the subsequent continuation phase (lasting 4-5 months). Outcomes, defined as per guidelines, were captured from TB clinic cards. Treatment success includes both patients who were cured and those who completed treatment. We modelled the risk of treatment success using a log-binomial model. Results: During the study period, 618 cases registered for TB treatment across the three facilities. 48 (9% of 544) patients transferred out and one was missing an outcome; 495 patients were included in the analysis (57% male, median age 35 (IQR: 29-41). 87% were new TB cases, 82% had pulmonary TB, and 72% were TB/HIV co-infected. 222 (45%) cured; 172 (35%) completed treatment; 55 (11%) defaulted; 6 (1%) failed; 40 (8%) died. Supervision type was missing for 99 persons in the intensive phase and 185 in the continuation phase. Of those whose supervision type was reported, family supervision was the most commonly reported supervision type (intensive phase 64% of 396, continuation phase 75% of 310) but there was large variation in family supervision rates across clinics (intensive phase 44-87%, continuation phase 52-98%). For the 64 retreatment cases, 8 (13%) were reported as self-supervised and 15 (23%) had no treatment supervisor listed although all had documentation of receiving anti-TB injections from the clinic; thus, for the analysis all re-treatment cases were assumed to be on facility-based supervision during the intensive phase. In a log-binomial model adjusted for HIV status, patient category, and clinic, there was no difference in treatment success for family supervision compared to community, PHC, or employer supervision (aRR intensive: 1.01, 95% CI: 0.88-1.16; aRR continuation: 1.09, 95% CI: 0.94-1.27).
Conference: SA TB conference 2014, Durban, SA