From Wiki Journal Club
Jump to: navigation, search
Barbar SD, et al. "Timing of renal-replacement therapy in patients with acute kidney injury and sepsis". The New England Journal of Medicine. 2018. 379(15):1431-1442.
PubMedFull textPDF

Clinical Question

Among ICU with septic shock and AKI without urgent need for dialysis, does an initiation of renal replacement therapy (RRT) using an early-initiation strategy reduce all-cause mortality at 90 days when compared to a delayed-initiation strategy?

Bottom Line

Among ICU with septic shock and AKI without urgent need for dialysis, there was no difference in 90-day all cause mortality when comparing early-initiation vs. delayed-initiation RRT strategies. Overall use of RRT was lower in the delayed-initiation strategy group.

Major Points

Sepsis remains one of the most frequent reasons for ICU admissions. The development of AKI in the context of septic shock is well documented with rates ranging between 40-75% [1]. A common dilemma within the ICU in these patients is whether to begin RRT early upon development of severe acute kidney injury or wait until an acute indication for dialysis arises in these patients. Recent clinical evidence is conflicting. The 2016 multicenter AKIKI study compared early- vs. delayed-initiation RRT in general ICU patients, about 67% of whom had septic shock. There was no difference 28 and 90 day mortality. In contrast, the 2016 single-center ELAIN trial (32% with severe sepsis) found a mortality benefit in general ICU patients with severe AKI (HR 0.66; 95% CI 0.45-0.97).[2] Of note, ELAIN primarily enrolled surgical patients.[3] While there is ongoing debate between these two trials and their methodological differences, a focused trial towards severe AKI in sepsis has not been performed.

Published in 2018, the Initiation of Dialysis Early Versus Delayed in the Intensive Care Unit (IDEAL-ICU) trial was a multicenter, randomized, open-label trial designed comparing early- vs. delayed-initiation strategies for RRT among ICU patients with septic shock and AKI. Importantly, the trial excluded severe hyperkalemia (K>6.5 mmol/L), metabolic acidosis (pH <7.15) and volume overload with pulmonary edema, refractory to diuretics. The trial was expected to enroll 864 patients however was stopped early after the second pre-planned interim analysis for futility after randomization of 488 patients. There was no difference in 90-day all-cause mortality between groups (58% early vs. 54% delayed). Among those assigned to the delayed-strategy group, only 62% received RRT. In contrast, 97% of those in the early-strategy group received RRT. The delayed-strategy group carried 4% and 8% incidence of metabolic acidosis and hyperkalemia in the first 48 hours. There was no difference in ICU stay between groups, but the delayed-initiation group had more days free of RRT. Taken together with AKIKI, the IDEAL-ICU trial supports a delayed-initiation RRT strategy among ICU patients with AKI without an emergent indication for the procedure.


As of March 2019, no guidelines have been published that reflect the results of this trial.


  • Multicenter, randomized, open-label trial
  • N=477
    • Early initiation of dialysis (n=239)
    • Delayed initiation of dialysis (n=238)
  • Setting: 29 intensive care units in France
  • Enrollment: 2012-2016 (stopped early for futility)
  • Follow-up period: 90 days
  • Analysis: Intention-to-treat
  • Primary outcome: Death from any cause at 90 days after randomization


Inclusion Criteria

  • Age ≥18 years
  • Admitted to ICU within 48 hours of development of septic shock
  • Acute kidney injury, per ≥1 of the RIFLE criteria:
    • Oliguria, defined as UOP <0.3 mL/kg/hr for ≥24h
    • Anuria for ≥12h
    • 3x increase in serum creatinine or a rapid increase of creatinine to ≥4 mg/dL (≥350 umol/L)

Exclusion Criteria

  • Prior ESRD on RRT
  • Obstructive AKI
  • Need for emergent RRT (e.g., K >6.5mmol/L, metabolic acidosis with pH <7.15, or refractory volume overload with pulmonary edema)
  • Previously received RRT in the ICU for AKI
  • Pregnancy
  • Expected death in the next 24 hours
  • Pre-existent comorbid condition with expected survival <28 days (End-stage cardiac disease, pulmonary disease, liver disease, poorly controlled cancer, or post-anoxic encephalopathy)
  • DNR

Baseline Characteristics

From the early initiation group.

  • Demographics: Age 69 years, Male sex 58%
  • BMI: 28.8
  • Medical comorbidities: CKD 13%, HTN 59%, DM 33%, HF 8%, chronic respiratory failure 8%, chronic liver disease 13%, immunosuppression 28%
  • SOFA score at enrollment: 12.2
  • Mechanically ventilated: 89%
  • Pressors: 100%; dobutamine 21%
  • ECMO: <1%
  • RIFLE criteria for eligibility: Oliguria 35%, anuria 34%, increase in creatinine 63%


  • Randomization to a group:
  • Early initiation of RRT - Initiation of RRT within 12 hours after documentation of severe AKI
  • Delayed initiation of RRT - Monitored until they meet criteria for acute indication of dialysis (K >6.5mmol/L, pH <7.15, or refractory fluid overload with pulmonary edema). If no acute indication for dialysis, then initiation of RRT is implemented at 48h unless renal recovery is made (decline in creatinine and urine output >1000 mL/24h).


Comparisons are early initiation group vs. delayed initiation group.

Primary Outcomes

90-day mortality
58% vs. 54% (p=0.38)

Secondary Outcomes

28-day mortality
45% vs. 42% (p=0.48)
Median time of initiation of RRT after diagnosed with failure-stage AKI
7.6 vs. 51.5 hours (p<0.001)
Patients who received RRT
97% vs. 62% (p<0.001)
17% of the delayed-strategy group received emergent RRT within 48h for trial criteria, including severe hyperkalemia or metabolic acidosis. Among this subgroup, 28/41 (68%) died.
Median days free of certain interventions
RRT: 12 vs. 16 (p=0.006)
Mechanical ventilation: 2 vs. 3 (p=0.19)
Pressors: 16 vs. 17 (p=0.87)
Median length of ICU stay
11 days vs. 10 days (p=0.91)
Cumulative fluid balance at day 7
5570 mL vs. 5878 mL (p=0.75)

Adverse Events=

Development of hyperkalemia
0% vs. 4% (p=0.03)


  • Inconsistency between using KDIGO or RIFLE criteria for inclusion criteria exists between AKI studies in ICU populations
  • High mortality rate (68%) among those in the delayed initiation group who required RRT in the first 48 hours after randomization. Unclear if this subgroup would have benefitted from early RRT.[4]
  • The time window delineated by this trial is arbitrary. Clinical tools to guide timing for initiation of RRT would be useful.[3]
  • Heterogenous population, some patients may have benefitted from early RRT[5]


French Ministry of Health

Further Reading

  1. Alobaidi R et al. Sepsis-associated acute kidney injury. Semin. Nephrol. 2015. 35:2-11.
  2. Zarbock A et al. Effect of Early vs Delayed Initiation of Renal Replacement Therapy on Mortality in Critically Ill Patients With Acute Kidney Injury: The ELAIN Randomized Clinical Trial. JAMA 2016. 315:2190-9.
  3. 3.0 3.1 Zhang Z & No "optimal timing" of renal-replacement therapy in critically ill patients with acute kidney injury. Ann Transl Med 2018. 6:S112.
  4. Xue C et al. Timing of Renal-Replacement Therapy in Acute Kidney Injury and Sepsis. N. Engl. J. Med. 2019. 380:398-9.
  5. Uhel F et al. Initiation of renal replacement therapy in patients with sepsis: more to it than meets the eye. Ann Transl Med 2018. 6:S130.