Abstract / Summary
Hospital readmission within 14 days of discharge carries clinical and economic costs, particularly in nephrology patients whose inflammatory and nutritional status influence early recovery. The red cell distribution width-to-albumin ratio (RAR) is an inexpensive index that captures both inflammation (RDW) and nutritional/acute-phase status (albumin). RAR has been linked to mortality in several disease cohorts, but its value for very-early (14-day) readmission in nephrology inpatients, and the shape of the dose–response, are unclear. We retrospectively reviewed 6,123 adults admitted to the nephrology ward of Changhua Christian Hospital, Taiwan, between January 2017 and May 2020. After excluding eGFR < 15 mL/min/1.73 m 2 , in-hospital death or planned readmission, and missing data, 2,679 patients were analyzed. RAR was divided into quartiles. Inverse probability of treatment weighting (IPTW) balanced baseline covariates. Logistic regression yielded crude, multivariable-adjusted and IPTW-adjusted odds ratios (OR) for 14-day readmission, with the second quartile (3.84 to < 4.63) as the reference. Sensitivity analyses comprised an expanded cohort including patients with eGFR < 15 mL/min/1.73 m², a Fine-Gray model with death as a competing risk, and prespecified subgroup analyses. The 14-day readmission rate increased across RAR quartiles from 6.7% to 21.7%. After IPTW adjustment, patients with RAR ≥ 5.80 had a 2.46-fold higher risk than the reference group (95% CI 1.75–3.46; p < 0.001), and those with RAR 4.63 to < 5.80 had a 1.47-fold higher risk (95% CI 1.02–2.12; p = 0.039). RCS analysis showed a non-linear positive association (p for non-linearity < 0.001), with risk rising sharply above an inflection point near RAR 4.63. Male sex, higher white blood cell count, higher blood urea nitrogen, higher Charlson Comorbidity Index, and faster heart and respiratory rates were independent risk factors, while higher red blood cell count, sodium and diastolic blood pressure were protective. RAR at discharge independently predicts 14-day readmission in nephrology inpatients, with a non-linear dose–response and a clinically usable threshold around 4.63. Adding RAR to routine discharge assessment may help identify patients who need intensified follow-up.