Objective To explore the prognosis of elderly patients suffered from hospital-acquired pneumonia (HAP) by T cell subsets and clinical pulmonary infection score (CPIS).Methods A cohort of 125 elderly patients admitted in ICU & ED (Emergency Department) from Aug,2012 to Jul,2013 were enrolled for a prospective and observational study.The patients were divided into 3 groups:HAP survival group (n =50,group A),HAP death group (n =40,group B) and non-HAP group (n =35,control group).The criteria of exclusion were patients with auto-immune diseases,immunodeficiency,allergic disorders,malignancies,diabetes,trauma,surgical diseases,or patients with recent use of immunosuppressive agents or cyclooxygenase-inhibitors (Aspirin etc.).In the control group,patients with nosocomial pneumonia and other diseases afecting the CPIS were excluded.APACHE Ⅱ scores of all patients were recorded.Blood T cell subsets (including values of CD3,CD4 +,CD8 +,and CD4 +/CD8 +)were measured on the admission day,the 1st day of HAP onset and the 5th day after onset of HAP in HAP patients whereas these measurements were tested only on the admission day in controls.Meanwhile,the CPISs were recorded on the admission day,the 1st day of HAP onset and the 5th day after onset of HAP in HAP patients.Flow cytometer (FCM) was used to detect T cell subsets.Data of statistical analysis were represented as Mean ± SD.The significant differences in T cell subsets and CPIS between survival group and death group were analyzed by independent t test.The paired samples t test was employed in survival group and death group.Linear correlation analysis was made between CD4 +/CD8 + ratio and CPIS in survival and death groups,respectively.Results There were no significant differences in demographics and clinical features (including age,sex,length of stay,APACHE Ⅱ scores) of patients in survivors and non-survivors (P ＞ 0.05).The values of CDs (CD3,CD4 + and CD4 +/CD8 + ratio) between patients of control group and patients of HAP groups were not significantly different on the admission day (P ＞ 0.05).The values of CDs on the admission day were much lower than those on the 1 st day of HAP onset in both survivors and nonsurvivors (P ＜ 0.05).The values of CDs on the 5th day after onset of HAP were higher than those on the 1 st day of HAP onset in the survival group (P ＜ 0.05),while there were no significant differences in CDs between different intervals after HAP onset in the death group (P ＞ 0.05).There were no significant changes in values of CD8 + in any group (P ＞ 0.05).Both survivors and non-survivors had much higher CPIS values on the 1st day of HAP onset than those on the admission day (P ＜0.01).The survival group had higher CPIS on the 5th day after onset of HAP compared to the 1st day of HAP onset (P ＜0.01),while there was no significant change in the death group.Linear correlation analysis showed negative correlation between CD4 +/CD8 + ratio and CPIS on both the 1 st day of HAP onset (survival group:R =-0.740,P =0.004 ; death group:R =-0.613,P =0.035) and the 5th day after onset of HAP (survival group:R =-0.639,P =0.009; death group:R=-0.686,P=0.021).Conclusions The hospital-acquired pneumonia appears as an immune imbalance disorder.The difference in CDs is a promising objective tool,aiding in prediction of prognosis of HAP in the elderly,the lower the CDs,the higher severity.The CD4 + / CD8 + ratio showed a negative correlation with CPIS.Monitoring of T cell subsets and CPIS may provide clinical value for the treatment of hospital-acquired pneumonia in the elderly.