A Systematic Review of Randomized Controlled Trials That Evaluate Strategies to Avoid Epidural Vein Cannulation During Obstetric Epidural Catheter Placement

Abstract

In this systematic review, we evaluated the evidence for seven strategies which have been proposed to minimize the incidence of epidural vein cannulation during lumbar epidural catheter placement in pregnant women. Multiple databases were searched to identify prospective, randomized, controlled trials between December 1966 and October 2007 that evaluated methods to avoid epidural vein cannulation after lumbar epidural catheter placement in pregnant women. Published trials were evaluated using a quality assessment tool, and results were combined to evaluate efficacy to prevent epidural vein cannulation. Of 90 trials screened, 30 trials were included (n = 12,738 subjects). Five strategies reduce the risk of epidural vein cannulation: the lateral as opposed to sitting position (six trials, mean (sd) quality score = 35% [11%], odds ratio (OR) 0.53 [95% confidence interval (CI) 0.32-0.86]), fluid administered through the epidural needle before catheter insertion (8 trials, quality score 48% [18%], OR 0.49 [95% CI 0.25-0.97]), single rather than multiorifice catheter (5 trials, quality score 30% [6%], OR 0.64 [95% CI 0.45-0.91]), a wire-embedded polyurethane compared with polyamide epidural catheter (1 trial, 31%, plus 4 unscored abstracts, OR 0.14 [95% CI 0.06-0.30]) and catheter insertion depth < or =6 cm (2 trials, 47% [11%], OR 0.27 [95% CI 0.10-0.74]). The paramedian as opposed to midline needle approach and smaller epidural needle or catheter gauges do not reduce the risk of epidural vein cannulation. The risk of intravascular placement of a lumbar epidural catheter in pregnancy may be reduced with the lateral patient position, fluid predistension, a single orifice catheter, a wire-embedded polyurethane epidural catheter and limiting the depth of catheter insertion to 6 cm or less. In general, low manuscript quality weakens the strength of these conclusions.