ObjectiveTo compare the value of the acute physiology and chronic health evaluationⅡ (APACHEⅡ) scores, the pneumonia severity index (PSI) scores), the CURB-65 scores, and serum procalcitonin (PCT) concentration in prediction of prognosis for inpatients with community-acquired pneumonia (CAP) and discuss the influence factors.MethodsRetrospective analysis was conducted based on the APACHEⅡ scores, the CURB-65 scores, the PSI scores and PCT concentration of hospitalized CAP patients admitted in the Department of Respiratory Medicine of First Hospital of Shanxi Medical University between January 2015 and December 2016, and within 24 hours of their admission. The end point of this study was the clinical outcome of hospitalization (recovery, improvement, exacerbation or death). Receiver operating characteristic (ROC) curve analysis and binary logistic regression models were used to assess the ability of prognostic evaluation and determine the boundary value, to screen risk factors that influence deterioration and death in CAP patients.ResultsTwo hundred and thirty-five CAP patients were enrolled with 146 males and 89 females at an average age of (60.4±18.1) years old. All patients were divided into 2 groups: improving recovery group had 205 cases, and deteriorating group had 30 cases. The rank of areas under the ROC curve for predicting the deterioration and death risk of CAP, from big to small were APACHEⅡ(0.889), PSI (0.850), CURB-65 (0.789), and PCT (0.720). APACHEⅡ score over 11 points and PSI score over 91 points were optimal cut-off values for the prognostic assessment. Moreover, the logistic regression analysis revealed that APACHEⅡ score and PCT were independent risk factors of deterioration and death in CAP patients.ConclusionsThe better predictability of clinic outcome of CAP is APACHEⅡ score, PSI score, CURB-65 score, and PCT respectively in order, while the APACHEⅡ score and PCT concentration were independent risk factors for exacerbation and mortality in CAP patients. The predictive ability of a single PCT measurement is limited. The combination of APACHEⅡ score and PCT may increase specificity, but reduce sensitivity.