When to Start RRT in AKI Alexander Usorov, MD 2/24/09 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 RIFLE-AKI 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 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? Dose and Modality • • • • VA/NIH trial vs Schiffl’s trial Ronco Mehta Vinsonneau (Contniuous venouvenous hemodiafiltration vs intermittent HD for ARF in pts with multiorgan dysfunction syndrome. Lancet 2006) 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 author’s past experience • Improvement in mortality, clinical course, uremic sxs Cont 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 BUN<70 and SCr <5 – 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 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 BUN<60 and SCr <5 – Less intensive regimen: BUN and SCr reached 100 mg/dL and 9 mg/dL • Average time from AKI to HD: 5+2 vs 7+3 days • Higher mortality in the intensive HD group 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 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 < or > 60 mg/dL at initiation of therapy Cont. • Early group – CRRT initiated on hospital day 10+15 – Mean BUN of 43+13 • Late group – CRRT initiated on HD 19+27 – BUN of 94+28 • Survival – 39% in early vs 20% in late group • 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) 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 BUN>84, SCr>2.8, or serum K>6, despite medical therapy and regardless of UOP • Remaining 36 pts, CVVHDF was initiated when UOP was <100ml over 8 hrs despite Lasix • Similar demographics and baseline clinical characteristics • Surgery to renal support time was 2.6+2.2 days vs 0.8+0.2 days 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 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 <20 ml/min on a 3-hr timed collection • Late groups: – BUN>112 – K>6.5 – Pulmonary edema present 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) 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