Analysis of influencing factors for postoperative brain injury in patients with cardiac surgery.

Objective: To identify risk factors in people with heart disease prior to undergoing cardiac surgery and to provide a prediction model.

Methods: 106 patients who underwent cardiac surgery were selected by convenience sampling method. Case-control study design was used to divide them into control and observation groups according to whether there was postoperative brain injury. General demographic data, perioperative surgery, postoperative recovery, past medical history and preoperative conditions, laboratory test indicators were collected. Univariate and multivariate Logistic regression analyses were performed to select influencing factors. A nomogram model for predicting postoperative brain injury was established.

Results: Univariate Logistic regression analysis showed that age, female, stroke, diabetes mellitus, perioperative administration of dexmedetomidine, intraoperative use of cardiopulmonary bypass were related factors for postoperative brain injury(p < 0.05). Multivariate Logistic regression analysis showed that females, history of stroke, and intraoperative use of cardiopulmonary bypass were risk factors, and perioperative administration of dexmedetomidine was protective factor(p < 0.05). Nomogram model was established with an AUC of 0.847 and accuracy of 84.7%.

Conclusion: Preoperative stroke history and females are risk factors for brain injury after coronary artery bypass grafting under cardiopulmonary bypass, and perioperative use of dexmedetomidine is protective factor.