Objectives: To evaluate the comparative effects of different types of contrast media with respect to the risk of developing contrast-induced nephropathy (CIN) by synthesizing the current literature.
Data sources:
We searched for original studies in MEDLINE®, Embase®, and the Cochrane Library through October 1, 2014. We also searched for studies in
Methods: Two reviewers independently reviewed each article to identify randomized controlled trials (RCTs) that reported on CIN-related outcomes in patients after receiving low-osmolar contrast media (LOCM) or iso-osmolar contrast media (IOCM). We included head-to-head comparisons of one LOCM versus another LOCM or of LOCM versus IOCM. (Only 1 IOCM is currently available in the United States.) For each study, one reviewer extracted the data and a second reviewer verified the accuracy. Both reviewers assessed the risk of bias for each study. Together, the reviewers graded the strength of evidence for the comparisons and outcomes of interest. We quantitatively pooled the results of studies that were sufficiently similar, using a 25-percent relative risk reduction as the threshold for a minimally important difference.
Results: We identified five RCTs that compared two or more LOCMs, including two studies of intra-arterial administration, two studies of intravenous administration, and one study examining both routes. We identified 25 RCTs that compared IOCM with LOCM, including 18 studies of intra-arterial administration and 7 studies of intravenous administration. No study comparing LOCMs reported a statistically significant or clinically important difference between study arms, and the overall analysis did not suggest that any one LOCM was superior to another. In a meta-analysis, we found a borderline significant reduction in short-term CIN risk with IOCM compared with a diverse group of LOCMs (pooled relative risk, 0.80; 95% confidence interval [CI], 0.65 to 0.99, p=0.045). When the analysis was stratified by route of administration, the aggregate pooled relative risk was 0.80 (95% CI, 0.64 to 1.01) for intra-arterial and 0.84 (95% CI, 0.42 to 1.71) for intravenous. In studies that investigated IOCM versus LOCM, the outcomes of mortality, cardiovascular outcomes, need for renal replacement therapy, and imaging quality or diagnostic accuracy showed no significant difference between groups. One study comparing different LOCMs investigated the outcomes of death and adverse events, and found no difference between groups.
Conclusions: We found low strength of evidence that the risk of CIN did not differ between LOCMs, and moderate strength of evidence that IOCM had a slightly lower risk of CIN than LOCM. The lower risk was not clinically important and just reached statistical significance.