Objective To investigate the clinical characteristics and predicting factors for death in critically ill patients with severe community-acquired pneumonia (CAP). Method A total of 143 hospitalized patients with severe CAP between January 2009 and December 2012 were included and their clinical data were retrospectively analyzed. According to the clinical outcome, patients were divided into survival group and death group, and their clinical features and laboratory test results were compared, and multivariate regression analysis was conducted to search for predicting factors for death. Results In this study, a total of 118 patients survived and 25 patients died, and the mortality rate was 17.5%. The number of underlying diseases in the two groups were different, and death group had more patients with 3 kinds of diseases than the survival group[76.0% (19/25) vs. 22.8% (13/57), P<0.05]. The intubation rate in the death group was significantly higher than that in the survival group[84.0% (21/25) vs. 33.1% (39/118), P<0.05], and the arterial blood pH value (7.15±0.52 vs. 7.42±0.17, P<0.05), HCO3- concentration[(18.07±6.25) vs. (25.07±5.44) mmol/L, P<0.05], PaO₂[(58.92±35.18) vs. (85.92±32.19) mm Hg (1 mm Hg=0.133 kPa), P<0.05] and PaO₂/FiO₂[(118.23±98.02) vs. (260.17±151.22) mm Hg, P<0.05)] in the death group were significantly lower than those in the survival group. And multivariate regression analysis indicated that the number of underlying diseases[OR=0.202, 95%CI (0.198, 0.421), P=0.003], PaO₂[OR=1.203, 95%CI (1.193, 1.294), P=0.011] and PaO₂/FiO₂[OR=0.956, 95%CI (0.927, 0.971), P=0.008] were independent predictors of death in the patients with severe pneumonia. Conclusions Patients who died of severe pneumonia often had severe illnesses before admission, and the number of underlying diseases and PaO₂ have highly predictive value for death.

Citation： XIEZheng-liang, GUOLu, LIUYue-jian. Analysis of Clinical Features of Severe Community-acquired Pneumonia and Predictive Factors of Death. West China Medical Journal, 2015, 30(9): 1618-1621. doi: 10.7507/1002-0179.20150464 Copy