“…In one study, risk of bias assessment was based only on data presented in a conference abstract without contacting the corresponding authors [23]. Overall, eight studies were considered low risk [22,24,28,29,31,[34][35][36], four raised some concerns [23,30,33,37] and three studies were deemed at high risk of bias [25,27,38] due to concerns about the randomization process [25], high attrition rate [38] and inclusion of a participating practice after randomization [27].…”