Deconstructing 30-day readmission after bariatric surgery into severity-graded phenotypes: a national analysis

BACKGROUND: Thirty-day readmission is a widely used bariatric quality metric, but as a yes-or-no event it counts a return for hydration the same as one with an anastomotic leak. OBJECTIVES: To classify 30-day readmissions into registry-derived severity grades and to determine whether conversion operations are associated with higher-acuity readmissions. SETTING: National Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program Participant Use File, 2020 to 2024. METHODS: We analyzed 949,507 bariatric operations, examining sleeve gastrectomy, Roux-en-Y gastric bypass, duodenal switch (DS), and single-anastomosis duodeno-ileal bypass (SADI) separately. Readmissions were graded hierarchically as Grade C (major: reoperation, leak, sepsis, thromboembolism, or death), Grade B (reintervention, transfusion, or organ-space infection), or Grade A (isolated). Multinomial logistic regression estimated adjusted relative risk ratios (RRRs) for conversion versus primary operations, with inverse probability of treatment weighting as a sensitivity analysis. RESULTS: Overall, 30,978 operations (3.3%) were readmitted, twice as often after conversion as after primary surgery (6.1% vs 2.9%). Most readmissions were minor (61.0% Grade A, 12.2% Grade B, 26.8% Grade C), but acuity climbed steeply with grade: unplanned intensive care use rose from 2.5% (Grade A) to 15.2% (Grade C), and nearly all 30-day deaths occurred in Grade C (1.9%, versus 0% in Grades A and B). Conversion shifted readmissions toward higher acuity, with adjusted RRRs rising stepwise from 1.29 (Grade A) to 1.64 (Grade B) to 1.66 (Grade C; all P < .001). The shift was most striking for reverse conversions (a prior gastric bypass taken down to a sleeve, DS, or SADI), where the adjusted Grade C risk was roughly sevenfold that of primary surgery (RRR 7.05, 95% confidence interval 5.45-9.12), one in two readmissions met major-complication criteria, and leak rates approached 5%. DS and SADI carried a comparable readmission burden. CONCLUSIONS: Conversion operations were associated with more frequent readmission and a greater share of severe readmissions; reverse conversions warrant pathway-specific consent discussions. After external validation, a severity-aware metric could describe what a readmission means more faithfully.