Predictive Value Of Three Pneumonia-related Scoring Systems And PCT For Elderly CAP Patients

Objective:We retrospectively analyzed the case data of elderly CAP patients hospitalized in the respiratory department of our hospital from March 2012 to March 2015,and explore the prognostic value of commonly used scoring systems（PSI score,CURB-65 score and SMART-COP score）and biomarkers（CRP and PCT）.Methods:A retrospective analysis of the case data of 183 patients with elderly CAP.Taking28-day survival as the end point,it was divided into survival group and death group.Compare the PSI score,CURB-65 score and SMART-COP score,CRP,PCT,WBC,PLT,Pa O2,Pa CO2 levels of the two groups,and draw ROC curve to compare the prognostic value.Statistical analysis uses SPSS 24 software.The comparison of measurement data that meets the normal distribution should use t/t?test;the measurement data that does not conform to the normal distribution should use the Mann-Whitney U test;the comparison of rates should use the?² test.Multivariate analysis was performed using binary classification Logistic regression.The Z-test is used to compare the area under the ROC curve.P<0.05 determined that the difference was statistically significant.Results:1.Case data and clinical characteristics:The age range of patients included was65-93 years old（average age 77.32±6.55;median age 78（72,82））,among them,154 were in the survival group and 29 were in the death group.Age,combined coronary heart disease,and the need for mechanical ventilation support were statistically significant between the survival and death groups.2.The PSI score,CURB-65 score,SMART-COP score and PCTwere statistically significant between the survival and death groups.However,there was no statistical difference between the two groups in CRP,PCT,Pa O₂,Pa CO₂,WBC,and PLT.The PSI score and PCT are independent risk factors for death risk（P<0.05）.3.The receiver operating curve（ROC）point that the PSI score,CURB-65 score and PCT level can predict the 28-day mortality risk,the area under the curve（AUC）is0.774,0.698,0.627.The Z-test point that the PSI score,CURB-65 score and PCT had no statistically significant difference in predicting the 28-day all-cause mortality.4.The PCT was combined with the PSI score,CURB-65 score,and SMART-COP for ROC analysis of the risk of death.The area under the curve was 0.824,0.753,and 0.667（P<0.05）.Z test point that PCT can improve the predictive power of PSI score and SMART-COP score for 28-day all-cause mortality.Conclusion:1.Combined with previous studies,age and combined underlying diseases should be included in the assessment of the severity and mortality risk of elderly CAP patients.2.PSI score,CURB-65 score and PCT level can predict the 28-day all-cause mortality,and the three are equally effective,all of which can provide a reference for clinical diagnosis and treatment.3.PCT can improve the predictive power of PSI score and SMART-COP score for 28-day all-cause mortality,but the improvement of CURB-65 score is not good.The area under the ROC curve drawn by PSI and PCT is 0.824,which is a good model for mortality risk assessment.