Abstract
Severe flu associated respiratory infections remain actual.
Aim: To evaluate the prognostic model of outcome of severe viral-bacterial community acquired pneumonia (VBCAP) during flu H1N1 pandemia based on hospital admission data.
Materials: 84 patients (43 survivors and 41 non-survivors) admitted with VBCAP in hospitals of Bashkortostan region between November and December 2009 were included in the retrospective study. The endpoint was defined as hospital mortality.
Methods and results: Initially 16 demography, history, functional and laboratory variables were obtained from hospital admission records. In the univariate ROC analysis, 5 variables were correlated with poor outcome of VBCAP.
Cutoffs were used to transfer continuous variables to dichotomous ones. Multivariate stepwise logistic regression analysis identified SpO2% and white blood cell count as independent predictive factors at p<0.05.
Logistic regression equation:
logit P = 3,49דSpO2<90%”+ 2,24דWBC<6,9×109/l” – 1,76, where P- poor outcome probability. Correct classification rate was 82,8%.
Conclusion: Acute respiratory failure and diminished immune response are the most powerful independent prognostic factors of poor outcome of viral-bacterial pneumonia during flu H1N1 pandemia.
- © 2011 ERS