Risk of bias* assessment.

*Low for low risk of bias, high for high risk of bias and unclear for risk unclear.†Assessment of whether method used to generate the allocation sequence should produce comparable groups.‡Assessment of whether allocation could have been foreseen in advance of enrolment by participants or recruitment personnel.Ω Assessment of whether knowledge of the allocated intervention was adequately prevented during the study.††Assessment of whether incomplete outcome data were adequately dealt with, including assessment of attrition rates in included studies.€ State any important concerns about bias not addressed in the other domains in the tool.