Kidney replacement therapy in acute kidney injury: Evidence to guide clinical decision-making

Acute kidney injury (AKI) affects more than half of critically ill patients, with a substantial subgroup progressing to severe stages requiring kidney replacement therapy (KRT). We aimed to review the characteristics of the available KRT modalities and to provide an evidence-based, critically appraised framework to support modality selection at the bedside, recognizing that KRT in the critically ill is prescribed and delivered by multidisciplinary teams of both nephrologists and intensivists. A structured, non-systematic search of the peer-reviewed literature was performed through March 2026, prioritizing systematic reviews, meta-analyses, and randomized controlled trials. The available evidence, largely derived from heterogeneous populations and frequently limited by small sample sizes, crossover between modalities, and single-center designs, does not demonstrate a consistent mortality difference among modalities. KRT modalities should therefore be regarded as complementary rather than competing, and selection should be individualized according to the clinical phenotype, therapeutic priorities, hemodynamic status, resource availability, and local expertise. Intermittent hemodialysis (IHD) is well suited to hemodynamically stable patients and to metabolic emergencies such as severe hyperkalemia and exogenous intoxication. To avoid rapid osmotic shifts, continuous therapies are generally preferred in patients at risk of cerebral edema, with peritoneal dialysis (PD) as a selected-patient option; CKRT is generally preferred when hemodynamic instability precludes safe fluid removal by intermittent techniques, although no vasopressor-dose threshold is evidence-based and the thresholds sometimes cited (e.g., norepinephrine >0.4 µg/(kg/min) or two vasopressors) reflect pragmatic expert opinion. Prolonged intermittent kidney replacement therapy offers a hybrid alternative, and PD remains a valuable option in selected patients and in resource-limited settings. Beyond mortality, outcomes such as dialysis dependence, kidney recovery, fluid balance, and length of stay should inform the assessment of adequacy, within a dynamic approach in which the modality is adapted throughout the clinical course.