#92 Quantification of the risk of recurrence in tuberculosis

PURPOSE: The authors intend to quantify the risk of recurrent tuberculosis among patients treated in South Carolina, USA. Recurrent tuberculosis occurs due to a complex of behavioral, therapeutic and clinical factors. METHODS: This registry-based, case-control study was conducted using data provided by the Department of Health and Environmental Control of South Carolina on tuberculosis patients seen between 1970 and 1999. 430 patients who developed subsequent tuberculosis more than one year after discharge from previous supervision met the criteria for recurrent tuberculosis and were enrolled as cases. Tuberculosis patients with no reported recurrence, matched on year of initial diagnosis, were enrolled as systematically sampled controls. Cases and controls were evaluated with respect to compliance with treatment, treatment regimen, duration of treatment, positive initial culture, adverse effects of medication, whether and when the course of treatment was finished, presence of risk factors such as family history of TB or alcohol consumption, and demographic information. RESULTS: Poor compliance with treatment was the strongest risk factor associated with relapse (OR = 3.24, 95%CI 1.41–7.44). Patients treated with low potency regimen, those with adverse effects to medication, and patients with a history of alcohol abuse were more likely to have recurrent tuberculosis. Although there was a disproportionate representation of African-Americans in the study population, no association was observed between race and recurrence. CONCLUSIONS: Our findings suggest that despite the availability of potent drugs, recurrent TB is still a public health problem. Evaluation of patients for compliance through directly observed treatment improves case holding and reduces the likelihood of acquired drug resistance.