Predictors on In-hospital Mortality Following In-hospital Diagnosis of Tuberculosis

Study objectives: To determine the factors associated with mortality after an in-hospital diagnosis of tuberculosis in a region with low levels of HIV coinfection. Methods: From January 2003 to December 2004, all subjects who were > 15 years of age and had received a diagnosis of tuberculosis were registered. The clinical, radiological and laboratory aspects of the patients who died (n=27) were compared with those of an age and gender matched control population(n=54). Logistic regression analyses were carried out, which included age, gender, hospital admission source, initial site of admission, dyspnea, general weakness and initial laboratory data. Results: The mean age of the patients was 60 +/- 16 years and male patients outnumbered female patients. Univariate analysis identified hemoglobin, blood urea nitrogen, albumin, cholesterol, aspartate aminotransferase (AST), C-reactive protein and the risk factors for tuberculosis to be significantly associated with mortality. Among the characteristics of disease presentation and treatment, emergency department admission, intensive care unit, disease severity, general weakness and dyspnea at the time of admission were associated with mortality. Multiple regression analysis revealed the initial management in the intensive care unit and lower serum albumin to be independently associated with mortality. Conclusion: The markers of disease chronicity and severity appear to be associated with in-hospital mortality. Identifying potentially reversible factors such as malnutrition and respiratory failure suggests specific intervention that might lead to an improvement in the patients' outcomes.