Abstract: Objective To evaluate the relationship between preoperative inflammatory status and postoperative delirium(POD) in the elderly patients using latent class analysis based on peripheral inflammatory markers. Methods A total of 240 elderly patients scheduled for unilateral total hip replacement at Affiliated Jinling Hospital, Medical School of Nanjing university were enrolled in this study. Arterial blood samples were collected before anesthesia, and the levels of C-reactive protein(CRP), tumor necrosis factor-alpha(TNF-α), interleukin-1β(IL-1β), IL-8 and IL-6 were detected using enzyme linked immunosorbent assay method. The incidence of POD was assessed using Confusion Assessment Method(CAM) within 3 days after operation, and the patients were divided into POD group and non-POD group. Latent class analysis was used to categorize the patients based on peripheral inflammatory markers. Multivariate logistic regression analysis was used to investigate the independent risk factors for POD, and receiver operating characteristic(ROC) curve was used to analyze the predictive value of hyperinflammatory phenotype for POD in the elderly patients. Results The levels of CRP, TNF-α, IL-1β, and IL-6 showed significant differences between non-POD group and POD group(P<0.05). Latent class analysis using these four inflammatory markers classified the patients into hyperinflammatory or hypoinflammatory phenotypes. Logistic regression analysis showed that hyperinflammatory phenotype(OR=12.15, 95%CI: 5.57-26.5, P<0.001) and age(OR=1.07, 95%CI: 1.01-1.14, P=0.034) were the independent risk factors for POD. ROC curve analysis showed that the AUC of hyperinflammatory phenotype in predicting POD was 0.779(95%CI:0.716-0.847). Conclusions A preoperative high inflammatory state has a good predictive effect on the occurrence of POD in the elderly patients undergoing unilateral total hip arthroplasty.

Key words: peripheral inflammation, postoperative delirium, aged, latent class analysis