Nosocomial pneumonia in patients having bone marrow transplant attributable mortality and risk factors

Publisher Website
Google Scholar
Abstract
The authors performed a matched historic cohort study to determine the attributable mortality and risk factors for nosocomial pneumonia in bone marrow transplant (BMT) recipients. All patients with nosocomial pneumonia at a university tertiary care center were identified by a prospective surveillance system between 1980 and 1988. Control patients were selected from the population of BMT patients. The crude mortality for 55 patients with nosocomial pneumonia was 74.5% (95% confidence interval [CI95], 63% to 86%). The excess or attributable mortality was 61.8% (CI95,43.7% to 80%). Aspergillus species represented the most frequent etiologic agent in this series, causing 20 of the 55 (36%) episodes. The attributable mortality of Aspergillus species pneumonia alone was 85% (CI95, 58.6% to 100%). For death in the hospital, the risk ratio for all 55 case patients relative to control patients was 9.5 (CI95, 4.1 to 22.1). To evaluate several risk factors simultaneously, a multiple logistic regression analysis using a conditional likelihood method was performed. A mathematical model with three variables best predicted nosocomial pneumonia in our patients: the occurrence of other nosocomial infections before the diagnosis of pneumonia, allogeneic BMT, and the use of methotrexate. The presence of other nosocomial infections before the diagnosis of pneumonia remained a significant independent risk factor, with an odds ratio of 13.27 (CI95, 2.51 to 70.2) after adjustment for the use of methotrexate and allogeneic BMT. Most importantly, effective methods for preventing nosocomial pneumonias in BMT recipients will have an enormous effect on crude mortality. Cancer 1992: 69:2653-2662.