Background: Acute renal failure (ARF) still bears a poor prognosis with mortality rates up to 70% and the ideal form of renal replacement therapy (RRT) remains controversial. The purpose of this study was to conduct a systematic review and meta-analysis of all randomized controlled trials (RCT) to examine the effect of dialysis modality (IHD: Intermittent haemodialysis; CRRT: continuous renal replacement therapy) on survival of patients with ARF and to also study the effect of each modality on dialysis dependence (DD).
Methods: Using and combining two comprehensive search themes (ARF and RRT), we searched electronic databases from 1969 through September of 2007, supplemented by a manual review of abstracts from nephrology meetings and reference lists of review articles. All RCT comparing IHD with CRRT in adult patients with ARF and with explicit reporting of mortality were included. The primary outcome was the pooled estimate of the odds ratio (OR) of mortality for patients with ARF treated with CRRT versus IHD. The secondary outcome was OR of DD at time of discharge for surviving patients.
Results: A total of 587 studies were identified, 554 of which were excluded on initial screening. Analysis of the nine RCT (1635 patients) showed an OR of 0.89 (0.63-1.24) for survival in patients on CRRT. Limiting the analysis to the seven RCT published after the year 2000, revealed an OR of 0.72 (0.58-0.90). The OR of all the studies before 2000 was 1.06 (95% CI 0.67-1.68), as compared with OR of 0.61 (95% CI 0.50-0.74) for studies post-2000. Four studies showed a significantly lower risk of DD among the CRRT group and none showed higher OR for DD. When analysis was limited to the RCT, the OR for DD was 1.07 (0.47-2.39), suggesting no difference in DD between the modalities.
Conclusions: Similar to previously reported meta-analyses, we did not find a significant effect of CRRT on the OR of survival. The progressive reduction in the OR of survival with CRRT relative to IHD might reflect progressive improvements in IHD. The OR of DD was not affected by mode of RRT. In conclusion, compared with IHD, CRRT does not offer an advantage with regards to survival or DD in ARF. Considering its cost and potential disadvantages, it is imperative to identify the subset of patients with ARF that would potentially derive maximum benefit from CRRT. This will require large, adequately powered studies with sufficient follow-up.