Validation of nonhome discharge risk score after elective infrainguinal bypass surgery

Nonhome discharge (NHD) contributes to poor patient quality of life and health care costs. Prior Vascular Study Group of New England database-based analysis developed a novel risk score for NHD after infrainguinal lower extremity bypass (LEB). Still, it has yet to be validated in an external dataset. This study hypothesized that the LEB NHD risk score would be externally validated using unique single institutional data.

A single institutional quaternary center electronic data warehouse was queried for elective LEB cases from 2012 to 2020. The primary end point was NHD, defined as discharge to a skilled nursing facility or acute rehabilitation center. This analysis excluded inpatient deaths. A previously developed risk score was applied. The risk score's predictive ability for NHD was assessed using a logistic regression, c-statistic, and Hosmer-Lemeshow test. The risk score was then categorized as low risk (score <5), moderate risk (score 5-9), and high risk (score >9) for NHD.

Among 242 included patients, NHD occurred in 22% of cases. The mean age of this cohort was 69 years. The cohort was 38.0% female and 26.4% non-White. The NHD proportion by risk category was 34.0% in high-risk, 26.0% in moderate-risk, and 4.0% in low-risk cases. High-risk cases represented 17% of the population and 27% of all NHD. On logistic regression, higher-risk groups had significantly higher odds of NHD than the low-risk category (moderate risk: odds ratio [OR], 8.8; 95% confidence itnerval [CI], 2.02-38.4; P = .004; high risk: OR, 13.0; 95% CI, 2.7-63.1; P = .001). The risk score successfully predicted NHD with a c-statistic of 0.702 and Hosmer-Lemeshow P value of .748, suggesting the model fit the data.

A novel NHD risk score was validated in an external single-institutional dataset. This risk score could be used to provide better pre-operative counseling and streamline postdischarge planning. Future studies should prospectively validate the NHD risk score.