Fig. 4From: Intervention effect estimates in randomised controlled trials conducted in primary care versus secondary or tertiary care settings: a meta-epidemiological studyDifference in intervention effect estimates when adjusting on risk of biasBack to article page