PURPOSE: The purpose of this study was to determine the clinical and laboratory characteristics, diagnostic methods, and prognostic variables in adults treated for miliary tuberculosis in the rifampicin era. PATIENTS AND METHODS: Computerized records of our community-based university teaching hospital over a 10-year period (1978 to 1987) were analyzed. A total of 109 patients were identified, including 12 who did not have miliary nodules on the chest radiograph (all of whom were shown to have hematogenous dissemination). Predisposing conditions were present in 46 patients. RESULTS: Clinical features were similar to those of previously reported series. Hematologic abnormalities were common: leukopenia (less than 4 x 10(9)/L) was present in 16 of 107 patients (15%), thrombocytopenia (less than 150 x 10(9)/L) in 24 of 104 (23%), and lymphopenia (less than 1.5 x 10(9)/L) in 82 of 94 (87%). Pancytopenia was found in six patients, three of whom recovered. Disseminated intravascular coagulation occurred in four patients, all of whom died. Adenosine deaminase levels were elevated in only seven of 11 serosal exudates and in seven of 12 samples of abnormal cerebrospinal fluid. Fiberoptic bronchoscopy was diagnostic in 44 of 51 patients (86%), bone marrow examination in 19 of 22 (86%), and liver biopsy in all 10 patients. Twenty-six patients (24%) died of miliary tuberculosis a median of 6 days after starting treatment. Survivors were followed up for a median of 51 weeks. Stepwise logistic regression identified aged (greater than 60 years), lymphopenia, thrombocytopenia, hypoalbuminemia, elevated transaminase levels, and treatment delay as independent predictors of mortality. CONCLUSIONS: Miliary tuberculosis commonly causes hematologic derangements, some of which are helpful prognostically. Fiberoptic bronchoscopy compares favorably to liver and bone marrow biopsy in sputum smear-negative cases. Mortality remains high and treatment should be begun as soon as the diagnosis is suspected.
PURPOSE: The purpose of this study was to determine the clinical and laboratory characteristics, diagnostic methods, and prognostic variables in adults treated for miliary tuberculosis in the rifampicin era. PATIENTS AND METHODS: Computerized records of our community-based university teaching hospital over a 10-year period (1978 to 1987) were analyzed. A total of 109 patients were identified, including 12 who did not have miliary nodules on the chest radiograph (all of whom were shown to have hematogenous dissemination). Predisposing conditions were present in 46 patients. RESULTS: Clinical features were similar to those of previously reported series. Hematologic abnormalities were common: leukopenia (less than 4 x 10(9)/L) was present in 16 of 107 patients (15%), thrombocytopenia (less than 150 x 10(9)/L) in 24 of 104 (23%), and lymphopenia (less than 1.5 x 10(9)/L) in 82 of 94 (87%). Pancytopenia was found in six patients, three of whom recovered. Disseminated intravascular coagulation occurred in four patients, all of whom died. Adenosine deaminase levels were elevated in only seven of 11 serosal exudates and in seven of 12 samples of abnormal cerebrospinal fluid. Fiberoptic bronchoscopy was diagnostic in 44 of 51 patients (86%), bone marrow examination in 19 of 22 (86%), and liver biopsy in all 10 patients. Twenty-six patients (24%) died of miliary tuberculosis a median of 6 days after starting treatment. Survivors were followed up for a median of 51 weeks. Stepwise logistic regression identified aged (greater than 60 years), lymphopenia, thrombocytopenia, hypoalbuminemia, elevated transaminase levels, and treatment delay as independent predictors of mortality. CONCLUSIONS:Miliary tuberculosis commonly causes hematologic derangements, some of which are helpful prognostically. Fiberoptic bronchoscopy compares favorably to liver and bone marrow biopsy in sputum smear-negative cases. Mortality remains high and treatment should be begun as soon as the diagnosis is suspected.
Authors: Ekaterina V Kourbatova; Michael K Leonard; Javier Romero; Colleen Kraft; Carlos del Rio; Henry M Blumberg Journal: Eur J Epidemiol Date: 2006-10-27 Impact factor: 8.082