Mental Health Factors in Cardiothoracic Surgery Risk Prediction Models: A Systematic Review

BACKGROUND: Mental health disorders are associated with substantial morbidity and death in cardiovascular disease, yet their role in perioperative risk stratification remains unclear in patients undergoing cardiothoracic surgery. We evaluated how mental health disorders are currently incorporated into cardiothoracic surgery risk prediction models. METHODS: We conducted a Preferred Reporting Items for Systemic Reviews and Meta-Analysis-concordant systematic review of PubMed and Embase on September 11, 2025 (International Prospective Register of Systematic Reviews: CRD420251274958). We included studies that developed or validated risk models predicting postoperative outcomes in adults undergoing cardiothoracic surgery that either included a mental health disorder as a covariate or validated a model in patients with such disorders. Model characteristics were extracted using the Checklist for Critical Appraisal and Data Extraction for Systemic Reviews of Prediction Modeling Studies, and risk of bias and applicability were assessed using the Prediction Model Risk of Bias Assessment Tool+Artificial Intelligence. Findings were synthesized qualitatively. RESULTS: Of 17 986 articles screened, 9 studies met inclusion criteria and described 17 unique models. No study validated a model specifically in patients with mental health disorders. Depression was the most frequently included mental health covariate (8 models). When incorporated, mental health disorders had associations comparable to established clinical risk factors. The most frequently modeled outcomes were hospital readmission and postoperative depression. Three studies had a high risk of bias. Covariate selection and modeled outcomes varied substantially. CONCLUSIONS: Mental health disorders are rarely incorporated into cardiothoracic surgery risk models despite potentially important prognostic associations. Future studies should evaluate their incremental predictive value, validate models in patients with mental health disorders, and examine intersections with sociodemographic risk factors.