Abstract
BACKGROUND: Dialysis-requiring acute kidney injury (AKI-D) represents the most severe form of AKI and is associated with substantial morbidity. As survival after AKI-D improves, recovery following hospitalization has emerged as a critical but understudied phase of care. Whether recovery trajectories differ across sociodemographic groups and hospital contexts remains poorly understood. Dialysis-requiring AKI affects approximately 2-3% of hospitalized adults nationally and carries high post-acute morbidity, underscoring the importance of understanding recovery trajectories beyond survival. METHODS: We conducted a retrospective, nationally representative study using the 2022 Healthcare Cost and Utilization Project National Inpatient Sample. Adult hospitalizations complicated by AKI-D among patients without pre-existing end-stage kidney disease who survived to discharge were included. The primary outcome was dialysis dependence or non-recovery at discharge, operationalized using discharge disposition as a pragmatic surrogate for post-acute recovery following receipt of acute dialysis. Survey-weighted logistic regression models adjusted for demographics, illness severity, and hospital characteristics were used to estimate adjusted odds ratios. Marginal standardization was applied to derive adjusted probabilities and absolute risk differences. Effect modification by hospital context was examined. RESULTS: Among survivors of dialysis-requiring AKI, approximately 40% were discharged to non-home settings, indicating a substantial burden of incomplete recovery at hospital discharge. After adjustment, adults aged ≥85 years had more than two-fold higher odds of non-recovery compared with those aged 50-64 years (adjusted OR 2.19), while self-pay patients had substantially lower odds compared with Medicare beneficiaries (adjusted OR 0.45). Patients with Medicaid or no-charge encounters-and several racial and ethnic minority groups-also exhibited lower adjusted probabilities of non-recovery. Hospital characteristics modified these associations, with payer-related differences in non-recovery varying by teaching status, bed size, and geographic region. Certain hospital settings exhibited both higher overall non-recovery burden and larger disparities. Sensitivity analyses using alternative outcome definitions and excluding patients with chronic kidney disease yielded consistent findings. CONCLUSION: Incomplete recovery at hospital discharge is common among survivors of dialysis-requiring AKI and is shaped by both patient-level vulnerability and hospital context. Institutional environments appear to modify recovery disparities, highlighting hospitals as potential leverage points for improving equitable post-AKI outcomes. Efforts to enhance recovery after AKI-D should extend beyond the acute hospitalization to address post-discharge transitions and system-level factors.