Back to 2015 Annual Meeting Posters

---

Improving Perioperative Counseling for Emergency Abdominal Surgery: Creation of a Scoring Tool to Predict One-Year Mortality in the Elderly
*Olubode A. Olufajo¹, *Gally Reznor¹, *Stuart R. Lipsitz², *Zara Cooper¹, *Ali Salim¹, *Joseph S. Hanna³, *Erika L. Rangel^1
1Brigham and Women's Hospital, Boston, MA;²Brigham and Women's Hospital, Center For Surgery and Public Health, Boston, MA;³Rutgers-Robert Wood Johnson Medical School, New Brunswick, NJ

Objective: To develop an objective scoring tool to predict long-term mortality in elderly patients undergoing emergency abdominal surgery
Design: Retrospective chart review cross-referenced with the Social Security Death Index. Covariates included demographics, body mass index (BMI), preoperative laboratory values, presence of acute kidney injury (AKI), presence of atrial fibrillation, smoking status, dialysis dependence, alcohol dependence, immunosuppression, Charlson Score, American Society of Anesthesiologists (ASA) score, and surgical procedure. Procedures were classified as major, moderate, or minor severity. A scoring tool was created by using the log-odds ratios of variables in the best-performing multivariate logistic regression model to assign weighted values. The model was cross-validated and the discrimination of both models was evaluated using the c-statistic.
Setting: 793-bed academic teaching hospital
Patients: Patients ≥70 years who had emergency abdominal surgery from 2006-2011
Main Outcome Measure: One-year mortality
Results: 390 patients were included and one-year mortality was 32%. A 10-point scale was developed to predict one-year mortality using six covariates. These variables and their scores were:albumin<3.5 mg/dL (1), Charlson Score≥4 (1), AKI (2), ASA≥4 (2), major operative severity (2), and BMI [≥30 kg/m²(0); 18.5-29.9 kg/m²(1);<18.5 kg/m²(2)]. Associated mortality is shown in the table. The C-statistic was 0.84 in the training dataset and 0.78 in the validation dataset.
Conclusions: We created a practical scoring tool that can be used at the bedside to predict the risk of mortality one year after emergency abdominal surgery in the elderly. Early assessment of long-term outcome is useful for perioperative patient counseling and for prioritizing goals of care.

Emergency Abdominal Surgery Risk Score and Associated Mortality

Risk Score	One-year mortality	Sample size (% total)
0-2	0-3%	94 (24%)
3-4	11-39%	176 (45%)
5-7	60-70%	100 (26%)
8-10	94-100%	20 (5%)

---

Back to 2015 Annual Meeting Posters