ARF (APACHE)

Information on this page applies to the APACHE II ARF dropdown only; for the diagnosis code see any one of the several ICD10 codes for types/causes of ARF.

Data Element (edit)
Field Name:	Ap_ARF
CCMDB Label:	ARF
CCMDB tab:	Apache
Table:	L_Log table
Data type:	number
Length:	single
Program:	CC
Created/Raw:	Raw
Start Date:	1988-07-11
End Date:	2300-01-01
Sort Index:	77

The ARF checkbox is checked/true if patient is in Acute Renal Failure as per the APACHE definition.

• SMW

Legacy implementation right in the table

• Cargo

• Categories

• Forms

Additional Info

Use the following uniform definition for ARF/AKI everywhere, including for APACHE II.

Terminology related to Acute Kidney Injury

• Nephrologists want us to use the term Acute Kidney Injury (AKI).
  • The reason is that this entity, whatever it's called, includes the full range of levels of kidney injury from minor all the way up to complete renal shutdown needing dialysis.
• Some other terms for it are:
  • Acute Renal Failure
  • Acute Renal Insufficiency (ARI)

KDIGO Guidelines for Acute Kidney Injury (AKI)

• We use the KDIGO criteria for defining Acute Kidney Injury (AKI, Acute Renal Failure and Acute Renal Insufficiency) (starting January 1, 2019)
• The main thing here is identifying that the observed problem with kidney function is acute, rather than chronic - and THIS is the reason that identifying AKI requires trying to find a past/baseline value of serum creatinine
• The KDIGO guidelines delineate several different "levels/degrees" of AKI. You'll note that (at its lowest level) AKI is present even with pretty small rises in serum creatinine. While one MIGHT think that such small rises are inconsequential, indeed they are not. As indicated in the paper "Small Acute Increases in Serum Creatinine Are Associated with Decreased Long-Term Survival in the Critically Ill", even rises in creatinine of 27 mcg/L in ICU patients are associated with higher rates of death. Thus in this new schema we are not overcounting those with significant AKI, but before we probably were undercounting them.
  • After a patient first developed AKI (as indicated by a rise in creatinine) it may continue to rise at a highly variable rate. The importance of this is that we should NOT re-code an AKI-related code each time the creatinine rises by 27 mcg/L if the continuing rise is simply part of the original event.
  • It is possible, however, for a patient to have multiple AKI events. While this would be indicated by creatinine rising again after it stabilized or fell (without dialysis), it requires a medical judgement to determine whether the re-rising is really part of the initial episode or represents a new AKI episode. There is no firm rule about how long creatinine should cease rising to say the first AKI episode is completed.
• These criteria will apply everywhere we need to identify ARF/AKI -- including:
  • Kidney, acute renal failure NOS
  • ARF (APACHE)
  • Kidney, acute tubular necrosis (ATN)
  • Kidney, acute renal failure, postprocedural
• But NOT for Kidney, renal failure/insufficiency/uremia, unspecified as acute or chronic - since as stated this code is for kidney failure or insufficiency when you don't know whether it's acute or chronic.

• In order to reduce the workload for identifying ARF/AKI, we will implement a first stage screening process to try and filter out the majority of people, who will NOT have AKI/ARF.
  • We expect that this screening will misclassify a few people who do have AKI as not having it, but we also expect that most of those who are missed will continue to experience declining renal function and their AKI/ARF will be identified in the following days.

First stage - screening

• Assume at admission that the patient does NOT have AKI/ARF if ALL of the following are true:
  • (1) Creatinine <110 for males and <90 for females AND
  • (2) No mention in chart of acute kidney/renal problems AND
  • (3) No mention in the chart of oliguria
• The source used for these threshold values of serum creatinine are population-based surveys of serum creatinine in people without known kidney problems:
  • Serum creatinine levels in the US population: third National Health and Nutrition Examination Survey (PubMed)
  • The effect of age on serum creatinine levels in an aging population: relevance to vascular surgery. (PubMed)
• If ANY of 1, 2 or 3 are false, then go on to the full evaluation in the Second Stage

Second stage - Full assessment

• Acute Kidney Injury (AKI) is present if ANY ONE OR MORE of the following are true (these are the KDIGO guidelines):
• (a) Urine output < 0.5 mL/kg/hour for 6 hours
  • so, obviously, you can't make this determination until there has been at least 6 hours of observation of urine output
  • also you need a weight -- if there isn't one already measured you have the following options: Wait for one to be done; Ask the nurse to do one; Do your best to estimate the weight, remembering that if the person appears to be of average size, then you could use default values based on average values in the Canadian population, i.e. 85 kg for men and 70 kg for women
• (b) Increase in serum creatinine by 27 micromoles/L or more within 48 hours
  • so, while this may happen quickly and thus this criterion be met before 48 hrs, you cannot make a full determination that it is NOT true until you have at least 2 serum creatinine values separated by at least 48 hours
  • in the case that the creatinine rises by >27, say in the first 12 hours, but then declines back down so that at the end of 48 hrs the net rise is <27, THEN THIS DOES QUALIFY AS AKI
• (c) Increase in serum creatinine to 1.5 times baseline or more within the last 7 days
  • this criterion is important because since many people have some degree of CHRONIC renal insufficiency or failure, a solitary serum creatinine can't tell you if the high value is acute or chronic
  • thus, to evaluate this criterion, seek a serum creatinine value at least 7 days old -- use whatever is the most recent value more than 7 days old that is available, even if it's years old
  • if there ARE NO values >7 days old, then you can use the sex-specific normal value as follows:
    • Men: 100 micromoles/L
    • Women: 85 micromoles/L

Coding Guideline

• For critical care only: The ARF (Acute Renal Failure) dropdown for APACHE defaults to "enter"; you won't be able to send unless you change it to one of the following:
• Choose "no" if the pt
  • does not meet the #KDIGO Guidelines for Acute Kidney Injury (AKI) or
  • has a diagnosis of Chronic kidney disease (end-stage renal/kidney disease, ESRD), Stage 5, GFR LT 15 (with or without dialysis)
    • Special Case- If a patient comes to ICU with a past history of Chronic kidney disease (end-stage renal/kidney disease, ESRD), Stage 5, GFR LT 15 and had a kidney transplant during this admission. As per Allan, unless the transplanted kidney was truly working postop (i.e. didn't need dialysis for at least a week or two) then do not code ARF.
• otherwise choose "yes"

Data Use

ARF is one component used to generate the APACHE II score.

Specifically, double points are assigned for the Creatinine score if the patient has ARF . (see APACHE_II_Background#Weighting_of_scores and APACHE Scoring table#Physiological Variables)

Data Integrity Checks (automatic list)

	App	Status
Can't check ICD10 ARF vs APACHE ARF	CCMDB.accdb	declined
Query check ApARF CRF	CCMDB.accdb	retired
Query check ICD10 ESRD vs AP ARF	not entered	retired

Historical

• In original APACHE II there were no criteria given for what constitutes Acute Renal Failure. So, from 1988 until 2018 we used a set of simple criteria based only on the admission serum creatinine and absolute urine output. But, when on January 1, 2019 we moved to ICD10 and CCI coding, we decided to use a
• there were criteria for ARF in the APACHE II user manual 1986, from George Washington University, Ver 1.0, that we applied when we started in 1988, they were:
  • creatinine PLUS oliguria. Oliguria was defined as: urine output of less than 135 cc over a consecutive 8 hr period in the first 24 hrs of ICU admission.
  • Copy of this APACHE II User Manual can be found in the article archives library located in HSC JJ387.