Two of the five review authors extracted data independently and in duplicate, using a customized data extraction form developed in Microsoft Excel (Microsoft Excel). Disagreements during the review process were resolved through discussion between the two review authors. In cases where consensus could not be reached, a third review author was consulted to help resolve the disagreement and make a final decision. The Risk of Bias in Non-Randomized Interventions (ROBINS-I) tool was used to assess the quality of non-randomized studies (table 2). The Risk of Bias (RoB2) tool was also used to assess the quality of randomized studies (table 3). We used the same data extraction form to extract the data. We extracted data including (1) study characteristics (author, year, place, and study design); (2) patient’s characteristics (sample size, and sex); (3) intervention and comparison (sample size and treatment dose); and (4) safety outcomes. All steps stated above were performed independently by two authors.