Seven case-control studies (31,095 VTE cases, 143,472 controls) were included in the review.
The overall odds ratio was 2.39 (95% CI 1.71 to 3.35; Ι²=93%).
Subgroup analyses suggested that the pooled risk was greater in low-potency drugs (OR 2.91, 95% CI 1.80 to 4.71; Ι²=76%) than high-potency drugs (OR 1.58, 95% CI 1.50 to 1.67; Ι²=3%) and greater in atypical antipsychotics (OR 2.20, 95% CI 1.22 to 3.96; Ι²=93%) than typical antipsychotics (OR 1.72, 95% CI 1.31 to 2.24; Ι²=87%).
Egger's test for publication bias was statistically significant (p=0.026) and the funnel plot appeared to be asymmetric, indicating possible publication bias.
Sensitivity analyses were reported to be similar to the primary analysis but the results were not shown.