LAUSR

INTRODUCTION Elderly patients, > 65 years of age, undergo an increasing number of surgical operations performed annually in the United States and they present with unique health care needs. Preventing postoperative readmission remains an important challenge to improving surgical care. This study examined whether geriatric-specific variables were independently associated with postoperative readmissions of elderly patients. METHODS The ACS Geriatric Surgery Research File (GSRF) was joined with the ACS NSQIP PUF files for 2014-2016. This data set included 13 GSRF variables and 26 ACS NSQIP variables. Associations between clinically relevant variables and readmission were tested with multivariable logistic regression. RESULTS The data represented 6,039 general surgery patients age 65 years and older. 58% of patients had colorectal operations, 19% pancreatic or hepatobiliary, 15% hernia, 4% thyroid or esophageal, and 3% had appendix operations. 24% of patients experienced a NSQIP defined 30-day postoperative complication, and 3% died within 30 days following surgery. 11% of patients had unplanned 30-day readmissions. Standard NSQIP variables including 30-day composite morbidity (OR 5.11, 95% CI 4.24-6.16, p <0.001), reoperation (OR 2.8, CI 2.07- 3.79, p <0.001), and steroid use (1.42, CI 1.03-1.96, p=0.03) were associated with readmission. In addition, GSRF variables including incompetent on admission (OR 1.63, CI 1.11-2.38, p =0.01), fall risk at discharge (OR 1.42, CI 1.11-1.82, p=0.005), use of mobility aid (OR 1.26 CI 1.02-1.56, p= 0.03), and discharged home with skilled care (OR 1.22, CI 1.0-1.49, p=0.04) were associated with readmission. CONCLUSIONS Four GSRF and 3 current standard ACS NSQIP variables were important in the evaluation of postoperative readmission of elderly patients. Geriatric-specific variables contributed to the explanation of the relationship between clinical variables and readmissions in elderly surgical patients.