R version 4.05 statistical software

Comparison of the prevalence of each risk factor between patients with and without each outcome was performed using the χ² or Fisher exact test. Two-sided P values and effect sizes with odds ratios (ORs) are reported. Comparisons of quantitative data were performed using the Mann-Whitney U (MWU) test. Following the initial bivariate analyses, multivariable binary logistic regression analyses were performed. Age on admission and prenatal psychiatric problems with P < .10 in the bivariate analyses were included as explanatory variables. All selected risk factors were simultaneously entered into the respective regression models. Multicollinearity among the risk factors was evaluated by calculating the variance inflation factor for each variable. To check for the robustness of the obtained statistical significances from the multivariable analysis, regardless of the prenatal history of each outcome, sensitivity analyses, after excluding patients with a prenatal history of each outcome, were further performed. For both the bivariate and multivariable analyses, the threshold for statistical significance was consistently set at 2-sided P < .05 based on the exploratory nature of the bivariate analyses.^{25 (link)} R, version 4.0.5 statistical software (R Foundation for Statistical Computing) was used for the analyses.