*2.7. Assessment of Risk of Bias in Primary Studies*

The National Institutes of Health (NIH) quality assessment tool was utilized for the risk of bias assessment. Two reviewers (KB and MS) independently evaluated the risk of bias and assigned the quality scores based on the tool's dictionary and guidelines (Appendix A, Table A1). The overall quality score was assigned according to the tool guidelines. In case of disagreements, the consensus among reviewers was built through discussions.

#### *2.8. Measures of Effect and Data Analysis*

The Comprehensive Meta-Analysis Package (CMA version 3.0, Englewood, NJ, USA) was utilized to compute the pooled estimates of psychological outcomes, including anxiety, depression, and other psychological indicators. The effect measure was the proportion of anxiety and depression events. The logit transformation of the proportions was used to meta-analyze the data. The Clopper–Pearson method was used to calculate exact confidence intervals for individual studies. Owing to methodologic differences across studies, a random-effects model was used to extract the pooled estimate [32]. Substantial heterogeneity was defined as I2 > 75% [33]. Subgroup analyses by continent (Asia vs. other), country

(China vs. other), survey tool, study quality, gender, and levels of psychological outcomes were performed. Sensitivity analysis or leave-one-out analysis was also conducted to determine the impact of different weights assigned to each study on the final results. Funnel plot and Egger linear regression test statistics were utilized for publication bias [27,34]. *p*-values less than 0.05 were considered significant.

#### *2.9. Assessment of Evidence*

We assessed the certainty of the overall evidence based on the quality of individual studies and scientific rigor of the methodology used in each study. Two reviewers assessed the quality of the evidence and did not know each other's decision.

## **3. Results**
