For patients with dialysis-requiring acute kidney injury, persistent need for dialysis is associated with increased morbidity and mortality, high health care use, and poor quality of life.

To test the effect of a conservative dialysis strategy on recovery of kidney function.

This multicenter, unblinded, randomized superiority trial enrolled participants with dialysis-requiring acute kidney injury who had a baseline estimated glomerular filtration rate greater than 15 mL/min/1.73 m2, had already initiated kidney replacement therapy, and were hemodynamically stable, with planned intermittent hemodialysis between January 23, 2020, and March 10, 2025, at 4 clinical sites in the United States. Of 909 patients assessed for eligibility, 221 were randomized and 220 received the allocated intervention. The date of last follow-up was June 10, 2025.

With the conservative dialysis strategy, participants received dialysis only when specific metabolic or clinical indications were met. Participants in the conventional dialysis group received dialysis 3 times per week until urine output or creatinine clearance criteria were met.

The primary study end point was unadjusted kidney function recovery at hospital discharge, defined as being alive and not receiving dialysis, with at least 14 consecutive days without dialysis (including after discharge). Two prespecified key secondary end points were the number of dialysis sessions per week and the number of dialysis-free days to day 28.

The mean age of participants was 56 years (SD, 16 years). A total of 148 of 220 participants (67%) were male, 28 (13%) were Black, and 133 (60%) were White. Mean baseline estimated glomerular filtration rate was 64.8 mL/min/1.73 m2 (SD, 30.9 mL/min/1.73 m2). Participants had started kidney replacement therapy a median of 9 days (IQR, 5-18.5 days) before randomization. In the conservative dialysis intervention group, 70 of 109 participants (64%) achieved kidney function recovery at hospital discharge vs 55 of 109 participants (50%) in the conventional dialysis control group (difference, 13.8% [95% CI, 0.8%-26.8%; P = .04]; unadjusted odds ratio, 1.76 [95% CI, 1.02-3.03; P = .04]; after prespecified adjustment: odds ratio, 1.56 [95% CI, 0.86-2.84; P = .15]). Participants in the conservative dialysis group received fewer dialysis sessions per week (median, 1.8 [IQR, 0-2.6] vs 3.1 [IQR, 2.6-3.5]; difference, -1.4 [95% CI, -1.8 to -1.0]) and recovered earlier (21 [IQR, 0-28] vs 5 [IQR, 0-21] consecutive dialysis-free days to day 28; difference, 16 days [95% CI, 5-27]). Dialysis-associated hypotension occurred less frequently in the conservative dialysis group (69 vs 97 events).

A conservative dialysis strategy in dialysis-requiring acute kidney injury resulted in a shorter time to and higher rates of recovery of kidney function in the unadjusted analysis. Given uncertainty regarding the estimated effect size, this approach should be tested in a larger study population.

ClinicalTrials.gov Identifier: NCT04218370.