Continuous versus intermittent renal replacement therapy: a meta-analysis

OBJECTIVE: Patients with critical illness commonly develop acute renal failure requiring mechanical support in the form of either continuous renal replacement therapy (CRRT) or intermittent hemodialysis (IRRT). As controversy exists regarding which modality should be used for most patients with critically illness, we sought to determine whether CRRT or IRRT is associated with better survival. DESIGN: We performed a meta-analysis of all prior randomized and observational studies that compared CRRT with IRRT. Studies were identified through a MEDLINE search, the authors' files, bibliographies of review articles, abstracts and proceedings of scientific meetings. Studies were assessed for baseline characteristics, intervention, outcome and overall quality through blinded review. The primary end-point was hospital mortality, assessed by cumulative relative risk (RR). MEASUREMENTS AND RESULTS: We identified 13 studies ( n=1400), only three of which were randomized. Overall there was no difference in mortality (RR 0.93 (0.79-1.09), p=0.29). However, study quality was poor and only six studies compared groups of equal severity of illness at baseline (time of enrollment). Adjusting for study quality and severity of illness, mortality was lower in patients treated with CRRT (RR 0.72 (0.60-0.87), p<0.01). In the six studies with similar baseline severity, unadjusted mortality was also lower with CRRT (RR 0.48 (0.34 -0.69), p<0.0005). CONCLUSIONS: Current evidence is insufficient to draw strong conclusions regarding the mode of replacement therapy for acute renal failure in the critically ill. However, the life-saving potential with CRRT suggested in our secondary analyses warrants further investigation by a large, randomized trial.