The outcome of tuberculosis treatment at a rural hospital in southern Ethiopia.

We evaluated the treatment results of 239 patients with sputum positive pulmonary tuberculosis (TB) and describe potential factors associated with reduced programme performance in aTB treatment programme in Ethiopia.The TB registry was incomplete and 64 (26.8%) patients were not recorded in the control programme. Of the 239 patients, 34.3% received short-course chemotherapy (SCC) as a first treatment, 5.9% received SCC having previously been treated with standard long-course chemotherapy (LCC), and 54.4% were initially put on LCC. After excluding the 75 patients (31.4%) who were transferred to other health institutions outside the control area, 100 (61.0%; 95% CI 53.0-68.4) were cured (22.6%) or completed the treatment (38.4%) falling short of the target of 85%. Five months or later during treatment 1.8% remained smear-positive, 7.3% died and 29.9% interrupted their treatment. Sputum tests were done in 78% of the eligible patients at 2 months, in 20% at 5 months and in 60.2% at the expected time of treatment completion. By July 1998, 5.4% of the patients initially on LCC had relapsed and were retreated with SCC. None of those initially cured with SCC needed to be retreated. Compared with patients in theTB registry, non-registered patients had lower treatment completed and cure rates (42.3% versus 65.2%; P = 0.047), more patients were transferred out of the TB programme (59.4% versus 21.1%; P<0.001) and the defaulter rate was higher (57.7% versus 24.6%; P=0.002). Weaknesses in the programme performance include organizational issues such as the under use of theTB registry, deficient follow-up procedures, the common usage of LCC and unsatisfactory rates of defaulting.