Sodium bicarbonate-based hydration prevents contrast-induced nephropathy: a meta-analysis

Abstract

Background: Contrast-induced nephropathy is the leading cause of in-hospital acute renal failure. This side effect of contrast agents leads to increased morbidity, mortality, and health costs. Ensuring adequate hydration prior to contrast exposure is highly effective at preventing this complication, although the optimal hydration strategy to prevent contrast-induced nephropathy still remains an unresolved issue. Former meta-analyses and several recent studies have shown conflicting results regarding the protective effect of sodium bicarbonate. The objective of this study was to assess the effectiveness of normal saline versus sodium bicarbonate for prevention of contrast-induced nephropathy.

Methods: The study searched MEDLINE, EMBASE, Cochrane databases, International Pharmaceutical Abstracts database, ISI Web of Science (until 15 December 2008), and conference proceedings for randomized controlled trials that compared normal saline with sodium bicarbonate-based hydration regimen regarding contrast-induced nephropathy. Random-effects models were used to calculate summary odds ratios.

Results: A total of 17 trials including 2,633 subjects were pooled. Pre-procedural hydration with sodium bicarbonate was associated with a significant decrease in the rate of contrast-induced nephropathy (odds ratios 0.52; 95% confidence interval 0.34-0.80, P = 0.003). Number needed to treat to prevent one case of contrast-induced nephropathy was 16 (95% confidence interval 10-34). No significant differences in the rates of post-procedure hemodialysis (P = 0.20) or death (P = 0.53) was observed.

Conclusion: Sodium bicarbonate-based hydration was found to be superior to normal saline in prevention of contrast-induced nephropathy in this updated meta-analysis.

Figure 1

Flow chart depicting outline of the search and selection strategy. NS = normal saline; NaHCO₃ = sodium bicarbonate; NAC = N-acetylcysteine. *The study of Schmidt et al. [13] was not prospective. ^†Shavit et al. randomized patients to NS + N-acetylcysteine versus NaHCO₃.[14]

Figure 2

The Forest plot of odds ratios of contrast-induced nephropathy. Sizes of data markers are proportional to the weight of each study in the meta-analysis. Studies are stratified by year of presentation and/or publication. Horizontal bars, 95% confidence interval. NaHCO₃ = sodium bicarbonate; NS = normal saline.

Figure 3

The Forest plot of odds ratios of contrast-induced nephropathy. This is stratified by studies with elective procedures versus those including only emergency procedures. Sizes of data markers are proportional to the weight of each study in the meta-analysis. Horizontal bars, 95% confidence interval.

Figure 4

Forest plot of stratified analysis by studies using iso-osmolar (iodixanol) versus low-osmolar contrast media. Sizes of data markers are proportional to the weight of each study in the meta-analysis. Horizontal bars, 95% confidence interval.

Figure 5

The Forest plot of odds ratios of contrast-induced nephropathy stratified by publication status. Sizes of data markers are proportional to the weight of each study in the meta-analysis. Horizontal bars, 95% confidence interval.

Figure 6

Cumulative analysis of contrast-induced nephropathy. This figure depicts the summary odds ratios of all trials published (in the literature or as an abstract) up to a time point in chronologic order. The odds ratios increase over time illustrating that earlier trials found more pronounced effects of sodium bicarbonate than subsequent studies, while the confidence intervals have narrowed suggesting greater reliability of the effect estimate. Horizontal bars, 95% confidence interval.

Figure 7

Forest plot of odds ratios illustrating the influence of single trials on the overall analysis. Each row represents an overall estimate of odds ratios when omitting one study (left column represents name of omitted study). The lowest row shows the overall odds ratio when all studies are included. No particular study relevantly influences the overall odds ratio. Horizontal bars, 95% confidence interval. NaHCO₃ = sodium bicarbonate; NS = normal saline.

Figure 8

Funnel plot of contrast-induced nephropathy (with estimated unpublished studies). Trials are depicted by circles with the random-effects log odds ratio shown along the horizontal axis and precision in estimating this effect (reciprocal standard error) along the vertical axis. The Trim and Fill method was used to calculate the true center of the funnel (indicated by the vertical line) after filling in estimates of unpublished studies (depicted with black dots). The empty diamond indicates the original confidence intervals of the log odds ratio; the black diamond indicates the corresponding values when the additional imputed studies are also considered.