Is Timing Really Everything With RRT?

Tejas P. Desai, MD


August 08, 2016

Early Dialysis in Acute Kidney Injury

For years we have been told that early interventions result in better patient outcomes. Early cardiac catheterization results in more preserved myocardial function. Early antibiotic therapy can lead to better results in patients with pneumonia. Patients with signs of an ischemic stroke can recover more neurologic function with timely administration of thrombolytic agents. Perhaps surprising to many providers, early dialysis initiation in patients with acute kidney injury (AKI) doesn't follow a similar trend. The multicenter AKIKI trial[1] empirically tests the idea that "timing is everything" in medicine.

Early initiation of renal replacement therapy (RRT) in the intensive care unit (ICU) for patients with AKI has been a matter of debate between intensivists and nephrologists. Intensivists depend on RRT for improved fluid management, which can lead to better ventilation and oxygenation as well as fewer nosocomial complications (eg, ulcers, infections, metabolic derangements). Some nephrologists contend that several critical outcomes are unaltered by early initiation of RRT, and that in some cases, RRT can worsen the chances of kidney recovery, as demonstrated by respondents to a recent Twitter survey (Figure).

Figure. Twitter poll. May 26, 2016.

AKIKI: Artificial Kidney Initiation in Kidney Injury

The AKIKI study helps resolve these frequently opposing perspectives. The study was a multicenter randomized trial of 620 critically ill patients who developed oligo-anuric AKI, meeting the definition of stage 3 AKI by the Kidney Disease: Improving Global Outcomes (KDIGO) group (Table 2).

Table. Stages of Kidney Injury[2]

Stage Serum Creatinine Level Urine Output
1 1.5-1.9 times baseline, or
≥ 0.3 mg/dL (> 26.5 µmol/L) increase
< 0.5 mL/kg/hr for 6-12 hours
2 2.0-2.9 times baseline < 0.5 mL/kg/hr for ≥ 12 hours
3 3.0 times baseline, or
≥ 353.6 μmol/L (≥ 252.6 µmol/L) increase, or
Initiation of RRT, or
(in patients < 18 years old),
eGFR < 35 mL/min/1.73 m2
< 0.3 mL/kg/hr for ≥ 24 hours, or
Anuria for ≥ 12 hours

eGFR = estimated glomerular filtration rate

Across 31 medical centers, patients with AKI as a result of an ischemic event (and therefore presumed to be suffering from acute tubular necrosis) were randomly assigned to receive early or late RRT, provided that they required mechanical ventilation, vasoactive agent support, or both. Any patient with critical laboratory values (eg, blood urea nitrogen [BUN] level > 112 mg/dL, serum potassium level > 6 mmol/L, and/or pH ≤ 7.15), or requiring a fraction of inspired oxygen level ≥ 0.5 (50%), or diagnosed with acute pulmonary edema was excluded from the study. Patients in the early-initiation group received either continuous or intermittent RRT within 6 hours of being identified as having stage 3 AKI. Patients assigned to the delayed group were not given any form of RRT until they developed a critical laboratory value or were anuric for 72 hours after diagnosis of stage 3 AKI. Local providers could choose the method of RRT.

The primary outcome of AKIKI was the 60-day mortality rate. Several secondary outcomes were also measured, including length of stay, days of mechanical ventilation, onset of diuresis (an indicator of kidney recovery), and number of catheter-related infections.

AKIKI Findings

Randomization was approximately equal, with 312 patients (a plurality with sepsis) receiving early RRT and 308 in the delayed group. Nearly half (151 patients) in the delayed group ultimately required RRT, typically because of a persistently elevated BUN (>112 mg/dL) level or anuria lasting longer than 72 hours. The number of deaths was statistically equal in both groups (303 total deaths, 150 in the early group and 153 in the delayed group). The overall estimate for 60-day mortality was 49.1% (48.5% in the early initiation group and 49.7% in the delayed group, a nonsignificant difference). Perhaps even more interesting was the post-hoc analysis comparing patients who did not receive any form of RRT versus both trial groups. In that group, the patients were less ill at baseline and had a 60-day mortality of only 37.1%.

The only favorable secondary outcomes were rates of renal recovery and catheter-related infections, both supporting the delayed-initiation group. Length of ICU and overall hospital stays were statistically equivalent in both groups.

Where Do We Go From Here?

In short, we go back to the drawing board. Unlike other interventions in which timing has a favorable effect on outcomes, starting RRT earlier resulted in no difference in 60-day mortality. Indeed, if you delay RRT and your patient is fortunate enough to be alive after 60 days, he or she has a better chance of kidney recovery and less risk for catheter-related infection. It seems that the baseline illness of a patient (stage 3 AKI, ventilator-dependent, pressor-dependent) has a more deleterious effect on mortality than can be overcome by RRT, regardless of when it is initiated.

Clinicians reading the AKIKI trial (and this report) might consider the following questions:

  • Is it appropriate to initiate RRT using KDIGO stage in the absence of clinical findings (eg, acidosis, fluid overload, etc.)? Should AKI stage alone determine when RRT is initiated?

  • Should the results of AKIKI be interpreted narrowly or broadly, given the variety of conditions in the patient population? A similar study of surgical ICU patients, concurrently published in JAMA,[3] found a mortality difference in those receiving early RRT.

  • Does the fact that both continuous and intermittent modalities were used in this trial affect the results? In the JAMA study[3] of surgical ICU patients, all patients received continuous RRT.

  • Should another measure of kidney function, such as neutrophil gelatinase-associated lipocalin, be used to detect "early" kidney dysfunction? It is possible that by waiting for the development of stage 3 AKI, the investigators delayed RRT for all patients.

These questions are certainly not all-encompassing, but they give us pause while we try to understand the AKIKI results and reconcile the fact that, in treating specific kidney diseases with RRT, early intervention might not be as therapeutic as in other scenarios.

As always, we welcome your comments below.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: