Friday, August 22, 2008

AKI criteria

The study of acute kidney injury has been hampered by a lack of a standard definition.

Some of this is semantics--for instance, the decision to change "acute renal failure" to "acute kidney injury" does nothing to help in our treatment of this high mortality condition, except perhaps for reminding ourselves and the patients that this is hopefully not a reversible process (e.g., that the kidneys have not completely "failed" in many instances).

However, the development of a set of reproducible criteria is essential for research studies which seek to identify potentially new treatments for AKI and to be able to reliably compare results between studies. To this end, two major groups have come up with competing criteria. One is the RIFLE criteria and the other is the AKIN criteria; it remains to be seen which set of criteria (which differ slightly) will be universally adopted.

The RIFLE criteria was put forward by the Acute Dialysis Quality Initiative (ADQI) in 2005. It breaks up patients according to 5 categories (see figure above) using the handy acronym "RIFLE", where R= risk, I = injury, F = failure, L = loss and E = ESRD. There are criteria given based on %GFR loss, how much Cr is elevated above baseline, and urine output, not all of which must be fulfilled in order to classify any given patient.

The AKIN "Acute Kidney Injury Network" criteria were published in 2007 after a meeting in the Netherlands comprised of multiple experts on AKI. They have only three stages of AKI (Stages I-III) which generally correspond to RIFLE stages R, I, and F. The AKIN group claims an advantage over the RIFLE criteria in that it has a lower threshold for definining an individual as Stage I AKI (only 0.3 mg/dL increase OR a 1.5X increase over baseline is required to meet this definition in AKIN; in RIFLE one requires a 1.5X increase over baseline. Both definitions can also be met by urine output criteria as well.) This would seem to be a good idea given epidemiologic studies which show that even small increases in serum creatinine are associated with an increased mortality rate.

Which system will prevail? Only time will tell...

No comments: