2.5 Data analysis
All analyses were conducted on R statistical software (version 4.2.2), in addition to the base package, meta, and metafor package was used[15]. We used the random effects model to calculate the pooled incidence rate with the inverse variance approach and the corresponding 95% confidence intervals (CI) for the primary outcomes of interest. The I² value represents the percentage of total variation among studies due to heterogeneity rather than chance. After that, the random effects models were used because of the underlying heterogeneity in the studies, containing types of neuroimmune disorders, vaccines, and doses, and the reported outcomes (I² ≥50% or p < 0.05). We performed subgroup analyses to ascertain if results were effect by the type and dose of vaccine or the different disease types. The forest plot was used to represent the result of conducted subgroup meta-analysis graphically. Publication bias was assessed using funnel plots, visual inspection for funnel plot asymmetry, and Egger’s test of funnel plot asymmetry. A sensitivity analysis was conducted to evaluate the robustness and reliability of the results.