Thirty studies involving 37,066 participants were included: 18 RCTs (4,536 participants) and 12 non-randomised studies (32,530 participants).
Any ACD versus mechanical compression (30 studies).
There was a higher complication rate in ACD compared with control patients when all studies were pooled (OR 1.34, 95% CI: 1.01, 1.79), although there was significant heterogeneity between studies (P<0.001). When only randomised studies were included, there was no difference in complication rate between patients receiving ACD or control (OR 1.30, 95% CI: 0.90, 1.87), with no statistical heterogeneity (P=0.19).
Angio-Seal versus mechanical compression (12 studies).
There was no statistical difference in complication rate between the two methods when all studies were pooled (OR 1.14, 95% CI: 0.72, 1.81). There was significant heterogeneity between studies (P<0.001). The findings were similar when Angio-Seal was compared with mechanical compression for complication rate in randomised trials, or in PCI or diagnostic settings. The subgroup analysis of randomised trials in PCI settings suggested that Angio-Seal was associated with fewer complications, although this was not of statistical significance (OR 0.46, 95% CI: 0.20, 1.04, P=0.062). There was no evidence of statistical heterogeneity in these subgroup analyses.
VasoSeal versus mechanical compression (10 studies).
There was a higher complication rate among patients treated with VasoSeal compared with controls (OR 2.27, 95% CI: 1.35, 3.80), with no significant heterogeneity between studies (P=0.65). This risk was similar among randomised studies only (OR 2.78, 95% CI: 1.51, 5.13), with no significant heterogeneity between studies (P=0.73). This increased risk was seen in PCI settings (OR 2.52, 95% CI: 1.36, 4.65), but not in diagnostic or mixed settings.
Perclose versus mechanical compression (15 studies).
There was no statistically significant difference in complication rates among patients treated with Perclose compared with control, either in any setting or in any one particular setting. Restricting the analysis to only randomised studies, there remained no evidence of benefit or harm from Perclose.
Other subgroup analyses found similar complication rates between ACDs and control among patients treated with glycoprotein IIb/IIIa inhibitors, patients treated with smaller and larger sheath size, or Perclose in the PCI setting.