2.14 Additional analyses

The confidence evaluation and the reporting bias assessment were conducted by applying The Confidence in Network Meta-Analysis (CINeMA https://cinema.ispm.unibe.ch, assessed on 1 October 2022). The confidence should be downgraded by one level when the item “within-study bias” was a “Major concern” or the other items were “Some concern”. The confidence would be downgraded by two-level when other items were “Major concern” (Nikolakopoulou et al., 2020; Papakonstantinou et al., 2020).

The summarizing risk of bias assessments was conducted according to the “Average Risk of Bias” setting in CINeMA. The weighted average score for each relative effect was estimated according to the percentage contribution of studies at each bias level. Studies of a direct comparison, which had low (arbitrarily assigned a score of 1), moderate (score 2), and high (score 3) risk of bias, had 40%, 25%, and 35% percentage of contribution, and the total risk of bias score was 0.40 × 1 + 0.25 × 2 + 0.35 × 3 = 1.95, which rounded to 2 and would be regarded as “Some concerns” (Nikolakopoulou et al., 2020; Papakonstantinou et al., 2020).