Prediction and risk evaluation of delirium after surgery in older patients: development and internal validation of an algorithm from the prospective BioCog cohort study - PubMed
3 hours ago
- #older patients
- #postoperative delirium
- #risk prediction
- Postoperative delirium (POD) affects around 20% of older surgical patients and is linked to poor clinical outcomes and increased mortality.
- The BioCog study aimed to identify major POD risk factors and develop a multivariate algorithm for individual POD risk prediction.
- The study involved patients aged 65+ without preoperative dementia undergoing surgeries lasting at least 60 minutes, screened for POD using DSM 5 criteria.
- A gradient-boosted trees (GBT) algorithm using preoperative data, intervention characteristics, and postoperative lab changes achieved high accuracy (AUC 0.83) and good calibration (Brier score 0.12).
- The findings suggest that combining preoperative and intraoperative factors can predict POD with high accuracy, potentially aiding clinical decision-making.
- The study has ongoing external validation to assess generalizability and real-world performance.
- Several authors reported grants and conflicts of interest, including patents and licensing related to POD prediction tools.