Estimating the impact of renal replacement therapy choice on outcome in severe acute renal failure.

BACKGROUND

Mortality in severe acute renal failure (ARF) requiring renal replacement therapy (RRT) approximates 50% and varies with clinical severity. Continuous RRT (CRRT) has theoretical advantages over intermittent hemodialysis (IHD) for critical patients, but a survival advantage with CRRT is yet to be clearly demonstrated. To date, no prospective controlled trial has sufficiently answered this question, and the present prospective outcome study attempts to compare survival with CRRT versus that with IHD.

METHODS

Multivariable Cox-proportional hazards regression was used to analyze the impact of RRT modality choice (CRRT vs. IHD) on in-hospital and 100-day mortality among ARF patients receiving RRT during 2000 and 2001 at University of Michigan, using an "intent-to-treat" analysis adjusted for multiple comorbidity and severity factors.

RESULTS

Overall in-hospital mortality before adjustment was 52%. Triage to CRRT (vs IHD) was associated with higher severity and unadjusted relative rate (RR) of in-hospital death (RR = 1.62, p = 0.001, n = 383). Adjustment for comorbidity and severity of illness reduced the RR of death for patients triaged to CRRT and suggested a possible survival advantage (RR = 0.81, p = 0.32). Analysis restricted to patients in intensive care for more than five days who received at least 48 hours of total RRT, showed the RR of in-hospital mortality with CRRT to be nearly 45% lower than IHD (RR = 0.56, n = 222), a difference in RR that indicates a strong trend for in-hospital mortality with borderline statistical significance (p = 0.069). Analysis of 100-day mortality also suggested a potential survival advantage for CRRT in all cohorts, particularly among patients in intensive care for more than five days who received at least 48 h of RRT (RR = 0.60, p = 0.062, n = 222).

CONCLUSION

Applying the present methodology to outcomes at a single tertiary medical center, CRRT may appear to afford a survival advantage for patients with severe ARF treated in the ICU. Unless and until a prospective controlled trial is realized, the present data suggest potential survival advantages of CRRT and support broader application of CRRT among such critically ill patients.