A Predictive Model for Postoperative Cognitive Dysfunction in Elderly Patients with Gastric Cancer: a Retrospective Study

Objective: To explore the risk factors of postoperative cognitive dysfunction (POCD) in elderly patients with gastric cancer after radical resection and to establish a risk prediction model.

Methods: A retrospective analysis of the clinicopathological data of 687 elderly patients who underwent radical gastric cancer surgery from January 2014 to January 2020 in the Third Department of Surgery, Fourth Hospital of Hebei Medical University was conducted. The degree of cognitive impairment was divided into POCD positive group (n=141, 20.52%) and POCD negative group (n=546, 79.48%). The general data of the two groups were compared. Multivariate logistic regression was used to analyze the risk factors for POCD in elderly gastric cancer patients after radical surgery. A risk prediction model was established. The receiver operating characteristic (ROC) curve was used to evaluate the effectiveness of the model.

Results: Multivariate logistic regression analysis showed that preoperative ASA classification (OR=4.674, 95% CI: 1.610~12.651, P=0.020), age (OR=3.130, 95% CI: 1.307~8.669, P=0.001), operation time (OR=2.724, 95% CI: 1.232~7.234, P=0.031), preoperative PG-SGA score (OR=4.023, 95% CI: 1.011-10.883, P=0.048), and preoperative hemoglobin (OR=4.158, 95% CI: 2.255~8.227, P=0.001) were independent risk factors for POCD. Intraoperative application of dexmedetomidine (OR=0.172, 95% CI: 0.078~0.314, P=0.002) and maintaining a deeper anesthesia state (OR=0.151, 95% CI: 0.122~0.283, P=0.018) were protective factors. The area under the ROC curve of the POCD risk prediction model for elderly gastric cancer patients after surgery was 0.820 (95% CI: 0.742-0.899) (P<0.01).

Conclusion: The occurrence of postoperative POCD in elderly patients with gastric cancer is closely related to a variety of risk factors. By establishing a risk prediction model for the occurrence of POCD, high-risk patients can be effectively identified during the perioperative period, to intervene earlier.