Tuberculosis deaths: are we measuring accurately?

Death among tuberculosis patients is one of the major reasons for non-attainment of 85% treatment success target set by World Health Organization. In this short paper, we evaluated whether the overall mortality rate in pulmonary tuberculosis is being affected by other comorbid conditions. All new smear positive pulmonary tuberculosis patients (N =336), who started their treatment at the chest clinic of the Penang General Hospital, between March 2010 and February 2011, were followed-up until December 2011. Tuberculosis treatment outcomes were reported according to six treatment outcome categories recommended by World Health Organization. The outcome category 'died' was defined as 'a patient who died due to tuberculosis or other cause during tuberculosis treatment'. Our findings showed that out of 336 smear positive pulmonary tuberculosis patients, 59 (17.6%) died during treatment (mortality rate = 1.003 cases per 1000 person-days of follow-up). Among the deceased patients, the mean age was 55.8 years (SD =16.17) and 49 were male. According to the mortality review forms, 29 deaths were tuberculosis-related, while the remaining 30 patients died due to reasons other than tuberculosis. Cerebrovascular accident (n =7), septicaemia shock (n =4) and acute coronary syndrome (n =4) were the most common non-tuberculosis related reasons for mortality in the patients. If the 30 patients, for whom tuberculosis was incidental to death, are excluded from the final cohort, the proportion of patients in the 'died' outcome category could be reduced to 9.5%. The treatment outcome criterion (i.e., died) set by World Health Organization has limitations. Therefore, it requires improvement for more objective evaluation of the performance of the National Tuberculosis Program.
Study site: Chest clinic, Hospital Pulau Pinang, Malaysia