meta-epidemiological study interventions and outcomes: estimates in controlled trials with different Empirical evidence of bias in treatment effect