• Mark R Katlic, JoAnn Coleman, Kamran Khan, Susan E Wozniak, and Joseph H Abraham.
• Department of Surgery, Sinai Center for Geriatric Surgery, Sinai Hospital, Baltimore, MD.
• Ann. Surg. 2019 Jan 1; 269 (1): 177-183.

ObjectiveTo develop and validate a simple geriatric screening tool that performs as well as more complex assessments BACKGROUND:: Many tools that predict treatment risk in older adults are impractical for routine clinical use.MethodsWe prospectively conducted comprehensive preoperative evaluations on 1025 patients age ≥75 years who presented to Sinai Hospital of Baltimore for major elective surgery, then retrospectively reviewed patients' medical records for occurrence of postoperative outcomes. Using logistic regression modeling and receiver operating characteristic curve analysis we selected the best combination of simple tests, labeling this the Sinai Abbreviated Geriatric Evaluation (SAGE). The performance of the SAGE was then compared with 3 standard tools in its power to predict postoperative outcomes.ResultsThe SAGE is a statistically significant predictor of postoperative outcomes. Each unit decrease in SAGE score was significantly associated with a 51% (95% CI 1.30-1.77) increase in odds of a complication, a 2-fold increase in odds of postoperative delirium (95% CI 1.65-2.66), a 27% increase in odds of length of hospital stay >2 days (95% CI 1.10-1.47), a 54% increase in odds of a hospital readmission within 30 days (95% CI 1.25-2.88), and a 38% increase in odds of an unanticipated discharge to higher-level care (95% CI 1.18-1.61). We estimated the receiver operating characteristic curve area under the curve (AUC) for the SAGE of 0.69, 0.77, 0.73, 0.66, and 0.78 for the above outcomes, respectively. The SAGE performed as well in predicting postoperative outcomes as Fried's frailty phenotype, Charlson Comorbidity Index, and American Society of Anesthesiologists Physical Status Class (ASA).ConclusionThe SAGE performs as well as other geriatric evaluations that require equipment or memorization.

24 keywords...

hide…