Objectives: To verify whether transradial (TRA) compared to transfemoral (TFA) cardiac catheterization reduces the risk of periprocedural stroke (PS).
Methods: We reviewed (CRD42021277918) published real-world cohorts reporting the incidence of PS within 3 days following diagnostic or interventional catheterization. Meta-analyses and meta-regressions of odds ratios (OR) performed using the DerSimonian and Laird method were checked for publication bias (Egger test) and adjusted for false-positive results (study sequential analysis SSA).
Results: The pooled incidence of PS from 2,188,047 catheterizations (14 cohorts), was 193 (105 to 355) per 100,000. Meta-analyses of adjusted estimates (OR = 0.66 (0.49 to 0.89); p = 0.007; I2 = 90%), unadjusted estimates (OR = 0.63 (0.51 to 0.77; I2 = 74%; p = 0.000)), and a sub-group of prospective cohorts (OR = 0.67 (0.48 to 0.94; p = 0.022; I2 = 16%)) had a lower risk of PS in TRA (without indication of publication bias). SSA confirmed the pooled sample size was sufficient to support these conclusions. Meta-regression decreased the unexplained heterogeneity but did not identify any independent predictor of PS nor any effect modifier.
Conclusion: Periprocedural stroke remains a rare and hard-to-predict adverse event associated with cardiac catheterization. TRA is associated with a 20% to 30% lower risk of PS in real-world/common practice settings. Future studies are unlikely to change our conclusion.
Keywords: Stroke; accession site; catheterization; cohort; femoral; incidence; meta-analysis; meta-regression; observational; radial; real-world; trial sequential analysis.