|
Back to 2016 Annual Meeting
Identifying Predictors for Frequent Utilizers of Health Resources Following Bariatric Surgery
Andrew Bates*1, Jie Yang2, Mark A. Talamini1, Aurora Pryor1, Dana A. Telem1
1Surgery, Stony Brook University Hospital, Oyster Bay, NY; 2Stony Brook University Hospital, Stony Brook, NY
Background
Patients with a high utilization of emergency and inpatient services following bariatric surgery represent a disproportionate cost on health systems. Recent data has demonstrated the effectiveness of early-intervention programs in easing transitions of care and reducing readmissions. The identification of this subgroup of patients preoperatively and perioperatively would allow for targeted interventions to reduce emergency department visits and readmissions.
Methods
Using the New York Statewide Planning and Research Cooperative System (SPARCS) database, a longitudinal and comprehensive reporting system, 38,000 patients were identified as undergoing bariatric surgery (gastric bypass, sleeve gastrectomy, laparoscopic adjustable gastric band) from 2010 to 2013. 30-day readmissions, ED revisits, ED admissions, and patient demographic and surgical information were collected. Frequent utilization was defined as 2 or more ED visits or hospital admissions within 30 days postoperatively. Generalized linear mixed models and multivariate analyses were employed.
Results
Patients with younger age (p<0.0001), Medicare/Medicaid insurance status (p=0.0002), and those who underwent roux-en-Y gastric bypass (p=0.039) were significantly more likely to have multiple ED visits within 30 days postoperatively. Race (p=0.0003), insurance type (p=0.003), surgery type (p<0.0001), and postoperative complications were associated with higher hospital readmission rates. Postoperative complications associated with readmissions and ED visits included pulmonary embolism (p=0.013), abscess (p=0.0002), pneumonia (p=0.0008), respiratory failure (p=0.003), surgical error (p=0.0008), and hemorrhage (p=0.0003). Preoperative conditions associated with readmission included diabetes (p=0.006), psychoses (p=0.021), fluid/electrolyte disorders (p=0.002), cirrhosis (p<0.0001), and chronic blood loss anemia (p=0.024).
Conclusion
Patients who utilize a disproportionate amount of emergency and hospital resources postoperatively are more likely to be of black ethnicity, have undergone RYGB, and have suffered postoperative complications. Additionally, numerous preoperative conditions are predictive of high utilization. This data may serve as a basis for predictive models to identify high-risk patients preoperatively.
Back to 2016 Annual Meeting
|
