2.5. Study quality:
Given that trial inclusion with a high risk of bias may distort the
outcomes of a meta-analysis (26, 27), the Cochrane Collaboration tool
was used for evaluating the risk of bias. The following items were
evaluated: randomization sequence generation; allocation concealment;
blinding of participants, personnel, investigator, and assessor;
attrition rates; and companies financial interest. These items were
given high, unclear, or low risk of bias rating. A study was categorized
as having low, medium, or high-risk according to the key areas of
allocation concealment, reporting of attrition rates, and participants
and assessor (high = all 4 factors rated high, medium = 2 or 3 factors
rated high or unclear, and low = all key areas rated low).