We read the report by Tamura et al comparing a single bolus of sodium bicarbonate with saline hydration to reduce the risk of contrast-induced nephropathy with interest. Their study not only adds to the increasing evidence that bicarbonate is an effective preventive treatment; more importantly, it is the first study evaluating the meaningful approach of single bolus bicarbonate application. Previous studies used nonuniform protocols, but most of them were based on 6- to 12-hour infusions before contrast application (and an additional 6 to 12 hours after contrast). This represents a critical drawback; it renders this beneficial treatment inapplicable for acute myocardial infarction (in which the risk of contrast-induced nephropathy is increased) and other emergencies. Also, the several-hour-duration protocol is logistically challenging, even for elective procedures.
The study by Tamura et al has demonstrated an impressive treatment effect with a straightforward single-bolus protocol. Should this study affect our clinical decisions or should we rather interpret the results with restraint/caution because of the small patient sample and that it was a single-center study only?
To further explore this issue, we performed a stratified meta-analysis based on our previously published analysis. One study was excluded because of insufficient data. A random effects meta-analysis was performed. The analyses were stratified by the duration of bicarbonate pretreatment. Studies with bicarbonate infusion >1 hour before contrast application were defined as “long-infusion” and those with ≤1-hour infusions as “brief-infusion” protocols. Contrast-induced nephropathy was defined by either a relative 25% and/or an absolute 0.5-mg/dl increase in creatinine.
Figure 1 illustrates a markedly enhanced risk reduction for bicarbonate (odds ratio 0.37, 95% confidence interval 0.19 to 0.72; p = 0.004) for studies using “brief-infusion” protocols. In contrast, studies with “long-infusion” protocols showed alleviated effects (odds ratio 0.66, 95% confidence interval 0.33 to 1.31; p = 0.233).
