Background/Aim: H7N9 has lasted for over five years in China, resulted in more than 1500 cases with a fatality rate of 40.5%. This research aimed to identify the key clinical indices and estimated the fatality risk of infected H7N9 patients admitted to hospital. Methods: We performed a retrospective epidemiological investigation to collect laboratory - confirmed H7N9 viral infected cases from 2013 to 2015 in Guangdong province, China. Data with clinical outcomes and biochemical indices were collected after patients and hospitals permission. Multivariable logistic regression model and classification tree model were used to evaluate the risk factors of H7N9 death. The receiver operating characteristic curve (ROC) and survival analysis were used to compare survival and death distribution and difference. Results: The results indicated that body temperature (BT) average (Odds Ratio (OR) = 3.612, 95% Confidence Interval (CI) = 1.914-6.815), white blood cell (WBC) average (OR = 1.212, 95%CI = 1.092-1.346), WBC maximum (OR = 1.095, 95%CI = 1.043-1.149), hemoglobin minimum (OR = 0.981, 95%CI = 0.965-0.996), platelet average (OR = 0.990, 95%CI = 0.986-0.995), lymphocyte average (OR = 0.230, 95%CI = 0.064-0.831) were identified as the independent risk factor for mortality with H7N9 after adjusting the confounders (underlying disease, age and sex). Moreover, WBC average, WBC maximum and BT average were the most reliable indices for predicting mortality with high sensitivity and specificity. There was an 80.0% chance for a death of H7N9 if BT was over 38.1°C, 67.4% chance if WBC average over 9.5 (10＾9/L) and 86.4% chance if WBC maximum over 23.1 (10＾9/L). Conclusions: We screened a series of clinical variables through rigorous statistical methods to provide a reference for the treatment and survival prediction of H7N9 cases.
展开▼
