Tuberculosis patient care decentralised to district clinics with community-based directly observed treatment in a rural district of South Africa.

SETTING: A rural district of the Northern Province, South Africa.
OBJECTIVES: To measure the effect of decentralisation of a tuberculosis service in a rural area on treatment outcomes.
DESIGN: An intervention study that measured treatment outcomes of patients attending district clinics for tuberculosis treatment and compared these with outcomes of patients attending the district hospital.
RESULTS: Over the 4-year period 1992-1995, 928 patients were admitted to the tuberculosis unit of the district hospital. In the initial pre-intervention phase, the best estimate of completed treatment for all 503 cases was 61%, and for 206 new smear-positive patients it was 67%. The intervention process established a tuberculosis control programme with directly observed treatment for all patients, and training and supervision of clinic staff. Ninety per cent of all patients had community-based supporters. For most patients, cure was not proven, but assuming success from completion of and proven adherence to treatment, successful outcomes for new smear-positive cases rose to 82% for decentralised (clinic) treatment, and 88% for those patients who lived in the district but attended the hospital for treatment.
CONCLUSIONS: These findings suggest that district clinics can achieve the same good results as the hospital. It is recommended that tuberculosis control needs a dedicated co-ordinator at district level to manage the necessary infrastructural and staff resources.