Meta-analysis of outcome studies in end-stage renal disease

Studies examining outcomes in patients with ESRD are derived from a variety of sources (single center, regional registries, national registries, etc.) and provide discrepant results as to the impact of modality selection on patient survival. In an attempt to understand and resolve these discrepancies, an exhaustive literature search and meta-analysis were performed to compare survival in hemodialysis (HD) and peritoneal dialysis (PD) patients using published data from both registry studies and non-registry studies. Results of the meta-analysis, based on data from 82 literature non-registry studies, 55 literature registry studies, and two registry reports, censored for modality switches, transplants and dropouts and unadjusted for case mix, were inconclusive because survival and mortality outcomes varied with the data sources and formats for outcomes analyzed. Although limited data suggested that differences in case mix may contribute to differences in survival outcomes, adjustments for case mix (predialysis comorbid conditions), adequacy of dialysis, and other important patient level covariates were not possible because of the paucity of available data. These findings highlight the limitations inherent in current literature reports and underscore the need for more uniform, standardized, and detailed approaches than has hitherto been the case in reporting outcome findings in ESRD.