Investigators examined 284,230 adult bariatric surgery hospitalizations from the National Inpatient Sample spanning 2016–2022. They stratified admissions by renal status: no CKD; chronic kidney disease stages 1–2, 3, 4, 5; and dialysis-dependent ESRD. The primary outcome was in-hospital mortality. Crude mortality increased with CKD severity: 0.3% for hospitalizations without CKD, 2.0% for stage 3, 5.9% for stage 4, and 4.9% for dialysis-dependent ESRD. In survey-weighted multivariable regression, stage 4 disease (adjusted odds ratio [aOR] 3.85, 95% CI 2.58–5.75) and ESRD (aOR 7.34, 95% CI 5.33–10.12) were linked to greater adjusted odds of in-hospital death. The study team observed that stages 1–3 were not significantly associated with mortality after adjustment.
The pattern is clear and clinically relevant: advanced renal impairment—particularly stage 4 and dialysis-dependent ESRD—carries substantially higher perioperative in-hospital risk in this national cohort. That said, the data come from administrative hospitalizations and the authors caution about overinterpretation of procedure-level differences because residual confounding may remain.
Secondary outcomes included acute kidney injury, major bleeding, sepsis, acute respiratory failure, new dialysis initiation, length of stay, and hospital charges. Data indicated that advanced CKD correlated with higher rates of several nonfatal complications. The cohort with advanced CKD experienced greater bleeding, more sepsis, and more acute respiratory failure compared with those without CKD. Resource-use measures—length of stay and hospital charges—were also higher in advanced stages.
Stage 4 CKD showed a particularly strong relationship with new dialysis initiation in the perioperative period: the primary model produced an aOR of 16.47 (95% CI 9.19–29.51) for new dialysis. The authors reported these associations after applying survey-weighted multivariable regression to account for the national sample design.
The investigators compared crude event rates between procedure types across CKD strata. Sleeve gastrectomy had lower crude event rates than Roux-en-Y gastric bypass across stages of CKD. The procedure-by-stage interaction for a composite adverse event reached modest statistical significance (P = 0.048). The study team urged caution in interpreting differences between procedures because the observational design allows for residual confounding that may partially explain the observed pattern.
These procedure-level observations suggest possible differences in short-term safety profiles across renal strata, but the authors did not provide randomized comparisons or protocolized patient selection criteria in the abstract; therefore, the report does not establish causation.
Primary models were survey-weighted multivariable regressions that evaluated associations between CKD stage and outcomes. A separate fully specified mortality model included CKD stage plus age group, sex, diabetes, frailty, procedure type, and hospital volume. The fully specified model reaffirmed that stage 4 CKD and dialysis-dependent ESRD remained independently associated with increased adjusted odds of in-hospital mortality while earlier stages did not retain a significant association after adjustment.
Authors presented adjusted effect sizes for stage 4 and ESRD specifically; they did not report adjusted estimates for every secondary outcome within the abstract, and detailed covariate coefficients beyond those named were not supplied in the source abstract. The analysis therefore focuses attention on stage-stratified risk even after controlling for common clinical and procedural covariates.
The authors recommended CKD-stage-specific risk assessment, multidisciplinary perioperative management, and individualized procedural planning based on their findings. They also explicitly stated that observed differences between procedures should be interpreted cautiously because of the potential for residual confounding. The abstract did not provide granular methodological details such as coding algorithms, specific frailty definitions, or the full set of covariates and model diagnostics; those details would be found in the full manuscript.
Unanswered clinical questions that remain in this dataset include how patient selection, perioperative optimization protocols, and center-level practices modify these risks; the authors did not report randomized comparisons or prospective perioperative pathways. Those questions will require either more detailed registry work or controlled prospective studies to address.
Limitations and next steps are concrete in the authors' framing: use CKD stage in preoperative counseling, apply multidisciplinary perioperative pathways for patients with advanced CKD, and interpret comparative procedure safety cautiously pending further study.