Thirty RCTs (n=12,923) were included in meta-analyses, four of which were unpublished studies. The mean quality score of studies was 35% (range 19% to 79%).
Patient position: Compared with the sitting position, placement of epidural catheter in the lateral position was associated with a significant reduction in epidural vein cannulation (OR 0.53, 95% CI 0.32 to 0.86; six RCTs).
Fluid injection before catheter insertion: Compared with no predistension, predistension of the epidural space with fluid was associated with a significant reduction in epidural vein cannulation (OR 0.49, 95% CI 0.25 to 0.97; eight RCTs).
Multi-orifice catheters: Compared with multi-orifice catheters, single-orifice catheters were significantly associated with a reduction in epidural vein cannulation (OR 0.64, 95% CI 0.45 to 0.91; five RCTs).
Wire-Embedded Polyurethane Catheter Design: Compared with conventional nylon catheters, wire-embedded polyurethane design was associated with a significant reduction in intravascular catheter placement (OR 0.14, 95% CI 0.06 to 0.30; five RCTs).
Epidural catheter insertion depth: Compared with insertion depths of 7cm or more, insertion depths of 6cm or less were significantly associated with a reduction in intravascular cannulation (OR 0.27, 95% CI 0.10 to 0.74; two RCTs).
Anatomic approach (midline or paramedian) (one RCT) and needle or catheter gauge (one RCT) did not lead to a significant effect.
Significant heterogeneity was only observed for the outcome for fluid injection before catheter insertion (p=0.04). Metaregression showed that more recent trials were more likely to show favourable results for fluid predistention (p=0.008).