Existing evidence has suggested that in patients with severe acute kidney injury (AKI), the initiation of kidney replacement therapy (KRT) should be delayed until indications such as severe hyperkalemia or refractory fluid overload develop. However, there is no clarity on the duration for which the KRT can be delayed safely. 

Various modalities of KRT are intermittent hemodialysis (IHD), continuous kidney replacement therapies (CRTs), and hybrid therapies, also referred to as prolonged intermittent kidney replacement therapies (PIKRTs)like sustained low-efficiency dialysis (SLED) and extended-duration dialysis (EDD). 

The advantages of KRT in patients with AKI lead to the prevention of uremia and immediate death from the complications of kidney failure. It is hypothesized that differences in the timing of starting the therapy, modalities, and/or dosing may bring about clinical outcomes, especially survival. 

The currently accepted urgent indications for KRT in patients with AKI usually include refractory fluid overload, severe hyperkalemia or quickly rising potassium levels, signs of uremia like pericarditis, encephalopathy, severe metabolic acidosis (pH<7.1), and specific alcohol and drug intoxications.

The Artificial Kidney Initiation for AKI 2 (AKIKI 2) trial included 280 critically ill patients with severe AKI, no urgent indications for KRT, and either oliguria for 72 hours or a blood urea nitrogen (BUN) between 112 and 140 mg/dL who were randomly assigned to initiate KRT immediately or to delay KRT till an urgent indication developed or BUN was more than 140 mg/dL. The results showed that KRT was eventually started in 795 individuals slated for delayed KRT. After 28 and 60 days, mortality was reported to be higher in the group where KRT was delayed (48 versus 38 percent and 55 versus 44 percent, respectively). The findings of the study suggested that even though the KRT should not be started early, it is not advisable to delay KRT beyond a certain period. 

It has been shown that the chances of initiating KRT are raised in patients with underlying chronic kidney disease compared to the degree of reduction in glomerular filtration rate (GFR) at baseline. This result was demonstrated in a study comparing the prehospitalization estimated GFR in 1746 patients admitted to hospital who developed dialysis-requiring AKI with that of 600 820 hospitalized patients who did not. 

The AKIKI-2 trial used a more extended period of non-KRT management of severe AKI, as seen by higher BUN values at the time of starting KRT (mean BUN 89 versus 123 mg/dL among those who had KRT delayed). Compared with the STARRT-AKI trial, the BUN values at KRT initiation were lower (mean BUN 63 and 85 mg/dL in the early and delayed groups, respectively). 

Based on the AKIKI 2 trial results and other multiple trials, accelerated commencement of KRT is not linked to clinical benefit in the absence of indications. On the other hand, it may lead to increased utilization of health care resources and an increased risk for delayed recovery of kidney function. Still, it can be said that there is a blurred understanding of the threshold after which KRT should not be deferred to achieve spontaneous recovery of kidney function. 

Reference: 

1. Palevsky PM, Berns JS, Motwani S. Kidney replacement therapy inacute kidney injury in adults: indications, timing, and dialysis dose. May 27, 2021. Accessed on 21/06/2021 from https://www.uptodate.com/contents/kidney-replacement-therapy-dialysis-in-acute-kidney-injury-in-adults-indications-timing-and-dialysis-dose?sectionName=TIMING%20OF%20ELECTIVE%20INITIATION&topicRef=8352&anchor=H3&source=see_link#H3
2. Hsu CY, Ordonez JD, Chertow GM, et al. Kidney replacement therapy (dialysis) in acute kidney injury in adult s: Indications, timing, and dialysis dose. 2008; 74: 101.