Publication bias and meta-regression
The visual inspection of Funnel plots revealed asymmetries among studies reporting major postoperative outcomes (Supplemental Figure 1). Based on the Egger’s test, there were significant for all major outcomes, including postoperative stroke (p <0.0001), SCI (p = 0.004), acute kidney injury (p <0.0001), bleeding (p <0.0001), and in-hospital mortality (p <0.0001). The univariate meta-regression showed significant relationships between the HCA time and the development of postoperative stroke (p = 0.04) and SCI (p = 0.05). Also, the bubble plots revealed that with increasing HCA time, the effect sizes for postoperative stroke and SCI enhance. The regression line provides a good fit for the data because most studies are relatively close (Figure 4).