ABSTRACT
BACKGROUND: Studies on early postoperative readmissions after bariatric surgery (BS) have examined readmissions as a single entity, regardless of urgency. Strategies to lower nonurgent readmissions would reduce unnecessary hospital utilization. OBJECTIVES: To identify predictors of urgent readmissions (UR) versus nonurgent readmissions (NUR) at 30 days post-BS. SETTING: Single academic institution. METHODS: Patients undergoing primary BS over 2 years (n = 589) were retrospectively reviewed. Baseline demographic, medical, and hospitalization data were compared between readmitted patients, stratified by urgency, and nonreadmitted patients. Multivariate regression models of UR and NUR were created using variables with a P value ≤ .2 on univariate analyses. A P value ≤ .05 was considered statistically significant. RESULTS: There were 39 documented instances of 30-day readmissions, of which 44% (n = 17) were NUR; NUR patients were more likely to be female (100% versus 78.2% male; P = .03) and trended toward being younger, experiencing ≥2 perioperative complications, and having a longer index hospital length of stay (LOS). Patients with URs had a higher baseline BMI (52.5 ± 11.4 kg/m2 versus 48.7 ± 8.3 kg/m2, respectively; P = .04), were more likely to have sleep apnea (77.3% versus 56.1%, respectively; P = .05), had a longer LOS (3 versus 2 d, respectively; P = .007), and were more likely to have ≥2 postoperative complications (46% versus 17.0%, respectively; P = .003) compared with those with an NUR. Independent predictors of NUR included public insurance (odds ratio [OR] = 3.7; 95% confidence interval [CI], 1.17-11.67; P = .03), younger age (OR = 1.05; 95% CI, 1-1.01; P = .04), and female sex, while URs were independently predicted by LOS (OR = 1.3; 95% CI, 1.04-1.5; P = .02). CONCLUSIONS: Public insurance appears to be associated with NURs, while LOS predicts URs after BS. This suggests an important dichotomy within readmissions based on urgency, which has important implications for targeted quality initiatives.