When to Start RRT in AKI

1
When to Start RRT in AKI

• Alexander Usorov, MD
• 2/24/09

2
New Diagnostic Criteria for AKI

• Acute Dialysis Quality Initiative
• Plus several Critical Care Societies
• Equals Acute Kidney Injury Network or AKIN
• The fundamental goal is to improve the outcomes
  for patients who are at risk
• The first AKIN conference was held in Amsterdam
  in September 2005
• Focused on the development of uniform standards
  for definition and classification of AKI

3
RIFLE-AKI
4
Indications for RRT in AKI

• Volume overload unresponsive to diuretics
• Metabolic acidosis refractory to medical
  management
• Intoxication with dialyzable drug or toxin
• Uremic symptoms
• Encephalopathy
• Pericarditis
• Uremic bleeding
• Progressive azotemia in the absence of specific
  symptoms

5
Indications are open to interpretations

• How volume overloaded?
• What should potassium level be?
• How severe for metabolic acidosis?
• What is the definition of diuretic resistance?

6
Dose and Modality

• VA/NIH trial vs Schiffls trial
• Ronco
• Mehta
• Vinsonneau (Contniuous venouvenous
  hemodiafiltration vs intermittent HD for ARF in
  pts with multiorgan dysfunction syndrome. Lancet
  2006)

7
Timing?

• Less data available
• Early literature (1950s-1960s) is significant for
  the concept of prophylactic HD in AKI
• Introduced by Dr. Paul E Teschan
• Observational report using prophylactic HD in 15
  pts with oliguric ARF from Renal Center of the US
  Army Surgical Research Unit
• HD initiated prior to BUN reaching 200 mg/dL or
  uremic sxs
• Comparison was done to authors past experience
• Improvement in mortality, clinical course, uremic
  sxs

8
Cont
9
RCTs

• Conger et al conducted a study on US Naval
  Hospital Ship USS Sanctuary between April and
  October of 1970
• 18 patients with post-traumatic AKI
• Intensive HD arm with pre-HD BUNlt70 and SCr lt5
• Non-intensive regimen with delaying HD until BUN
  approached 150 and SCr approached 10 or if
  clinically indicated
• Survival - 5/8 pts (64) vs 2/10 (20) pts
• Major complications (Gram-neg. sepsis,
  hemorrhage) were less freq in intensive arm

10
Increased Mortality in Early HD

• Gillum et al examined 34 pts at University of
  Colorado in 1986
• Pts were paired and randomly assigned once SCr
  reached 8
• Intensive regimen with pre-HD BUNlt60 and SCr lt5
• Less intensive regimen BUN and SCr reached 100
  mg/dL and 9 mg/dL
• Average time from AKI to HD 52 vs 73 days
• Higher mortality in the intensive HD group

11
(No Transcript)
12
Conventional wisdom

• In the absence of uremic symptoms, start
  hemodialysis if BUN is around 100 mg/dL
• No additional benefit seen with earlier HD
  initiation nor more intensive HD prescription

13
Moving On

• Further studies focused mostly on the timing of
  initiation of CRRT
• Gettings et al published a retrospective analysis
  of 100 consecutive patients with post traumatic
  AKI in 1999
• Early vs late initiation based on BUN lt or gt 60
  mg/dL at initiation of therapy

14
Cont.

• Early group
• CRRT initiated on hospital day 1015
• Mean BUN of 4313
• Late group
• CRRT initiated on HD 1927
• BUN of 9428
• Survival 39 in early vs 20 in late group

15

• Critical points
• Non-randomized, retrospective
• More pts with multisystem organ failure or sepsis
  in late group
• More pts oliguric on first day of CRRT in early
  than late group, leading to suggestion that there
  was a confounding effect (?physician bias)

16
More Retrospective Studies

• Elahi et al reported a series of 64 consecutive
  patients s/p cardiac surgery at a single UK
  center between January 2002 and January 2003
• In 28 pts, CVVHDF was started once BUNgt84,
  SCrgt2.8, or serum Kgt6, despite medical therapy
  and regardless of UOP
• Remaining 36 pts, CVVHDF was initiated when UOP
  was lt100ml over 8 hrs despite Lasix
• Similar demographics and baseline clinical
  characteristics
• Surgery to renal support time was 2.62.2 days vs
  0.80.2 days

17
(No Transcript)
18
Limitations of the studies

• All recent studies are retrospective
• Using BUN as a surrogate measure of AKI duration
  is problematic
• Urea generation varies from patient to patient
• Volume of distribution of urea in critically ill
  patients is variable as well
• Bias by indication

19
How about a prospective study of CRRT timing?

• Bouman et al randomized 106 criticall ill
  patients with AKI to three groups
• Early high-volume CVVHDF (35 pts)
• Early low-volume CVVHDF (35 pts)
• Late low-volume CVVHDF (36 pts)
• Two early groups txt started within 12 hrs of
  meeting inclusion criteria
• Oliguria x 6 hrs despite hemodynamic optimization
• Measured cr clearance lt20 ml/min on a 3-hr timed
  collection
• Late groups
• BUNgt112
• Kgt6.5
• Pulmonary edema present

20
Outcome

• No significant differences in survival were
  observed
• Critical point is that 28-day mortality was only
  27, much lower than in prvsly reported studies
  of critically ill patients with AKI
• Small sample size lead to low statistical power
• Interestingly, 6/36 pts in late group never got
  RRT (2 pts died and 4 pts recovered renal fxn)

21
So When Do We Initiate RRT?

• Inadequate data available to answer this question
• Observational data suggests better outcomes are
  associated with early RRT initiation
• ? If less sick patients are included in these
  early groups
• Also, most pts with AKI are not treated with RRT

22
(No Transcript)