38 Chapter 2 updated search, one reliability meeting was held and the percent agreement for eligibility was 94.1% (k = 0.78). For the second updated search, two reliability meetings were held with percent agreement for eligibility being 92.9% (k = 0.62) and 90.9% (k = 0.61), respectively. For the papers identified during screening, full-texts were retrieved through Endnote’s automatic search, a manual search or by contacting the author(s). In- and exclusion criteria were checked again, in the following order: digital game, mental health, RCT design, and age. As soon as one criterion was not met, the paper was excluded and further criteria were not checked (i.e., an exclusion on ‘RCT design’ means that the study met the inclusion criteria for digital game and mental health, but not for RCT design, and that age was not further checked). If necessary and available, additional information (e.g., publicly available video of the game, website of the research group) was consulted to determine whether the game met our inclusion criterion. Disagreements on in- and exclusion were discussed and resolved by the two independent reviewers. For the original search and the updated searches, the percent agreement (kappa statistic) for eligibility of the full-texts were 92.5% (k = 0.85), 90.9% (k = 0.82) and 88.7% (k = 0.77), respectively. To make sure that no relevant papers were missed, a snowballing approach was used in which titles and abstracts were screened from the reference lists of the final included papers and relevant reviews that were identified during the screening phase. In addition, for RCT protocol papers that were identified during the original search in October 2020 and updated search in October 2021, it was checked (in March 2021 and August 2023) whether outcomes had been published in the meantime. References were screened by one of the reviewers and if a paper was deemed possibly relevant, the full-text was retrieved and the in- and exclusion criteria were checked by the same reviewer. If the reviewer thought the reference would match all inclusion criteria, the second reviewer confirmed this (or not) and the article was included (or not). Assessment of Risk of Bias in Individual Studies Risk of bias in the included studies was assessed using the revised Cochrane risk of bias tool for randomised trials (RoB 2.0; Higgins et al., 2016; Sterne et al., 2019). The RoB 2.0 addresses five domains of bias: 1) bias arising from the randomisation process; 2) bias due to deviations from the intended interventions; 3) bias due to missing outcome data; 4) bias in measurement of the outcome; and 5) bias in selection of the reported result. Additional guidance was employed for cluster-randomised trials (Eldridge et al., 2021), adding an extra domain 1b examining bias arising from how individual participants
RkJQdWJsaXNoZXIy MTk4NDMw