2.3 Quality Assessment
Two reviewers (two cardiologist physicians) check studies’ titles and abstracts to identify studies that possibly meet the inclusion criteria mentioned above. Then we obtained the full text of these studies and evaluated by two of the authors (two physicians). Any inconsistencies were solved through conversation with a third reviewer (MA).
In this study, Consolidated Standards of Reporting Trial (CONSORT)” checklist was used to perform data extraction6, to estimate the quality of the RCTs (See Supplementary File .2) and in order to evaluate the quality of the non RCTs Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) checklist was used6 (See Supplementary File .3). CONSORT checklist included 25 items to appraise RCT methodological quality and to identify the level of bias management in its analysis and design, while TREND had 22 items. A point was given to each study if there was a match of every single item; therefore, the CONSORT score for RCT’ methodological quality extended from 0 (low quality) to 25 (high quality). On the other hand, TREND scores ranged from 1(low quality) to 22 (high quality).
The extracted data contained specific details about the populations, settings, study methods and interventions. Two authors (two physicians) extracted data individualistically and any incongruity or ambiguity was fixed through conversation with the third author (MS).