Skip to main content
. 2021 Mar 8;22(5):2717. doi: 10.3390/ijms22052717

Figure 2.

Figure 2

The overall risk of bias for all the articles in the dataset was a high risk of bias (54.6%) and some concerns (44.8%). The selection of the reported result had a high risk of bias (45.9%) and some concerns (54.1%). The measurement of the outcome had a high risk of bias (12.4%), some concerns (33%), and low risk of bias (54.6%). The missing outcome data domain had a low risk of bias (99.5%) and a high risk of bias (0.5%). Deviations from intended interventions had a low risk of bias (91.9%), some concerns (3.8%), and high risk of bias (4.3%). The randomization process had a high risk of bias (14.1%), some concerns (80%), and low risk of bias (5.9%).