Abstract
Previous studies showed that the Confusion Assessment Method based delirium severity evaluation tool (CAM-S) had good reliability and validity. However, there is no Chinese version of the CAM-S. Therefore, we set out to perform a prospective investigation in older Chinese patients who had total joint replacement surgery under general anesthesia in Tenth People's Hospital in Shanghai, P.R. China. A total of 576 participants, aged 60 years or older, were screened, 179 participants were enrolled, and 125 of them were included for the final analysis. Pre-operative evaluations were conducted one day before the surgery. Postoperative evaluations were conducted twice daily from postoperative day 1 to day 3. The incidence of postoperative delirium was 24.8%. The Chinese version of CAM-S [including a Short Form (CAM-S Short Form) and a Long Form (CAM-S Long Form)] had an optimal reliability reflected by internal consistency (Cronbach's α= 0.748 and 0.839 for CAM-S Short Form and CAM-S Long Form respectively), split-halves reliability (Pearson correlation coefficient = 0.372 and 0.384 for CAM-S Short Form and CAM-S Long Form respectively), and inter-rater reliability (intra-class correlation coefficients = 0.629 and 0.945 for CAM-S Short Form and CAM-S Long Form respectively). Additionally, the Chinese version of CAM-S also showed a good discriminate validity. The domain scores of CAM-S were inversely correlated with corresponding domain scores of the MMSE. Finally, a receiver operating characteristic (ROC) analysis obtained an optimal cutoff point of 2.5 for CAM-S Short Form and 3.5 for CAM-S Long Form in recognizing delirium diagnosed by CAM. The areas under the ROC were 0.989 (95% CI 0.972 - 1.000, p < 0.001) and 0.964 (95% CI 0.946 - 0.982, p < 0.001), respectively. These data suggest that the Chinese version of CAM-S has good reliability and validity in evaluating postoperative delirium in geriatric Chinese patients and may be a useful tool to assess the severity of delirium.