Summed score of initial chest radiograph and APACHE II preferably evaluates illness severity and accurately predicts ICU mortality in severe avian influenza H 7 N 9

Objective: It is not rational to predict outcomes by chest imaging scores, regardless of illness severity, or to evaluate illness severity without considering imaging characteristics. Therefore, this study utilized a summed score of initial imaging and APACHE II/SOFA for evaluation of illness severity on ICU admission and prediction of ICU mortality in avian influenza (AI) H7N9. Methods: This retrospective study was conducted to record scores of radiographs, CT scans, APACHE II, SOFA, incidence of ARDS, multiple system organ failure (MSOF), and mortality. The aim of this study was to determine predictive accuracy and cut-off values of initial radiographic or CT scores (IRS/ICS), APACHE II, SOFA, (IRS/ICS + APACHE II), and (IRS/ICS + SOFA) and analyze correlation of each score to mortality in 22 patients with AI H7N9 at an ICU in Wuhan, China. Results: ARDS, MSOF, and mortality, respectively, occurred in 22 (100%), 17 (77.3%) and 10 (45.5%) patients. Cut-off values of (IRS + APACHE II), (IRS + SOFA), (ICS + APACHE II), and (ICS + SOFA) were 37.0, 28.0, 41.5 and 31.5, respectively, with sensitivity of 100%, 77.8%, 62.5%, and 62.5%, and specificity of 90.9%, 81.8%, 100% and 87.5% (P < 0.001, P = 0.020, 0.021, and 0.074). Mortality was positively correlated to (IRS + APACHE II), (IRS + SOFA), and (ICS + APACHE II) (Ρ = 0.839, 0.534 and 0.597; all P < 0.05). Conclusion: The summed score of initial radiograph and APACHE II appears to be a preferable indicator for evaluation of illness severity on admission and for prediction of ICU mortality in AI H7N9.