LIBERATE-D – Conservative Dialysis Strategy

1 May 2026 Samuel Brockbank Intensive Care Medicine, Summaries Leave a comment

LIBERATE-D – A Conservative Dialysis Strategy and Kidney Function Recovery in Dialysis-Requiring Acute Kidney Injury: The Liberation From Acute Dialysis (LIBERATE-D) Randomized Clinical Trial

Liu KD, Siew ED, Tuot DS, Vijayan A, Umemoto GM, Birkelo BC, et al. JAMA. 2026;335(4):326–335. doi:10.1001/jama.2025.21530

Clinical Question

In clinically stable adults with AKI-D receiving IHD and expected to require outpatient dialysis after discharge, does an indication-triggered “conservative” dialysis strategy (vs routine thrice-weekly dialysis) increase kidney function recovery by hospital discharge without unacceptable harms?

Background

Acute kidney injury (AKI) is a common condition amongst intensive care (ICU) patients. AKI-D refers to patients who have AKI requiring dialysis. Kidney replacement therapy (KRT) acts as a mechanism to ‘replace’ the kidneys functions using dialysis or filtration machines
Long term dialysis has associated high mortality and morbidity. Kidney recovery to a point of not requiring long term dialysis would undoubtedly improve patients quality of life, but also carry a significant cost benefit to the healthcare system
The transition from continuous KRT to an intermittent dialysis regime has typically defaulted to a thrice weekly regime as per chronic dialysis patients, but this lacks scientific backing
The most relevant studies have looked into the optimal timing of initiating KRT in ICU patients, these have have indicated that early initiation of dialysis does not improve long-term outcomes. Additionally, studies looking into ‘intensive’ regimes when transitioning to intermittent dialysis have shown no benefit
There are no previous studies looking at more conservative regimes and long term renal recovery

Design

Randomised, multicentre, parallel-group, investigator-initiated clinical trial
- Allocation stratified by site and baseline eGFR category (≥45 vs <45 mL/min/1.73 m²)
- Open-label, unblinded
220 patients would give 80% power to detect a 20% difference in absolute recovery rate from 30% to 50%
Intention to treat analysis

Setting

Conducted at 4 US centres (inpatient wards/ICU step-down settings after stabilisation)
Enrolment between January 2020, and March 2025, last follow up June 2025

Population

909 patients assessed, 220 patients randomised and included in study, 109 patients per treatment/control group, 1 withdrawal, 1 subsequently excluded due to renal transplant.
Haemodynamically stable patients, e.g. not currently requiring mechanical ventilation or significant vasopressor support.
Inclusion: Age ≥18 years; AKI requiring dialysis at least in part due to acute tubular necrosis; currently receiving IHD; clinician assessment that outpatient dialysis/ultrafiltration would likely be required after discharge; baseline eGFR ≥15 mL/min/1.73 m² (known or estimated per protocol rules).
Exclusion: Pre-existing long-term maintenance dialysis; kidney transplant during index hospitalisation; nephrectomy as AKI cause; “non-traditional” dialysis indications (eg intoxication); severe hypoxaemia requiring significant oxygen support (>5l/min); mechanical ventilation; continuous inotrope/vasopressor support; inability to consent/surrogate unavailable; pregnancy; incarceration; anticipated discharge/transfer within 48 hours; clinician non-equipoise regarding conservative strategy
Comparing baseline characteristics of intervention vs control group
- Age: 55 vs 59
- Male: 64% vs 71%
- Race – white: 53% vs 68%
- Primary service at admission
  - Medical: 65% vs 65%
  - Surgical: 34% vs 33%
- ICU admission before randomisation: 81% vs 84%
- Days from 1st kidney replacement therapy to randomisation: 11 vs 8 days
- Baseline eGFR: 65 vs 64.5mL/min/1.73 m2

Intervention

Dialysis only initiated once one or more criteria met:
- Serum urea nitrogen >112 mg/dL
- Hyperkalaemia: potassium >6.0 mEq/L, or >5.5 mEq/L despite medical therapy
- Acidaemia: pH <7.15, or pH <7.25 with bicarbonate <14 mEq/L, despite medical therapy
- Hypoxaemia due to presumed volume overload: oxygen >5 L/min or FiO2 >50% in a patient breathing via tracheostomy (diuretics attempted first when appropriate)
- Clinician judgement (allowed within protocol; reasons recorded)

Control

Scheduled thrice weekly dialysis regime
Extra sessions as per clinical indications
Trial of halting dialysis based on urine output and creatinine values
- Timed urine creatinine clearance >20 mL/min, or
- Urine output >1 L/day without diuretics (or >2 L/day with diuretics) and spontaneous serum creatinine decrease on 2 measurements >12 hours apart without dialysis

Outcome

Primary outcome: kidney function recovery at hospital discharge (defined as being alive and free from dialysis for at least 14 consecutive days, including after discharge). Comparing intervention vs control group
- significantly increased recovery
  - 64.2% vs 50.5%, p=0.04
  - Unadjusted odds ratio 1.76 [95% CI, 1.02-3.03; p = 0.04]
  - Prespecified adjustment: odds ratio, 1.56 [95% CI, 0.86-2.84; p = 0.15]
Key Secondary outcomes:
- Number of dialysis sessions per week – significantly reduced in intervention group
  - 1.8 vs 3.1, p<0.001
- Dialysis-free days up to day 28 – significantly increased in intervention group
  - 21 vs 5 days, p<0.001
Other Secondary outcomes: comparing intervention vs control group
- Day 90 all-cause mortality – no significant difference
  - 14.7% vs 18.5% (95% CI −13.7% to 6.1%; p=0.45)
- Hospital length of stay – no significant difference
- Time to kidney function recovery – significantly reduced in intervention group
  - 2 vs 8.5 days, p<0.001
- Severe or clinically relevant events such as dialysis-associated hypotension, arrhythmias, cardiopulmonary arrest, urgent dialysis needs, and significant electrolyte or acid-base disturbance – no significant difference

Authors’ Conclusions

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

Strengths

Randomised, multicentre design
Clinically relevant outcome measures
Clear separation in the dialysis frequencies

Weaknesses

Unblinded, open-label study
Multiple patients screened but not recruited due to clinician non-equipoise. This led to an overall small sample size which was not powered to assess important safety and adverse effects
Patients only recruited once off vasopressor or ventilator support reducing generalisability to an ICU population
Majority of conservative group decision to dialyse was clinician led, reducing protocolability and standardisation
Primary outcome point was liable to adjustment, with the adjusted analysis showing no significant difference between the groups

The Bottom Line

Adopting a ‘conservative’ approach to dialysis in stable AKI-D patients reduces the amount of dialysis received in both number of sessions and dialysis-free days, with a potential improvement in renal recovery without an increase in significant adverse events
This study adds weight to the ‘less is more’ strategy with KRT in patients with AKI-D, however, there needs to be larger scale research to ensure the accuracy of the results given the liability to the adjustment
This study was looking at the optimal dialysis regime in patients in critical care once they are stable (off vasopressor and ventilatory support) and as such should not be extrapolated to direct KRT in these critically unwell patients

External Links

Metadata

Summary author: Samuel Brockbank
Summary date: 27th March 2026
Peer-review editor: David Slessor

Picture by: Generated through ChatGPT