Shah 2008.
| Methods | Study design: ITS | |
| Participants | Physicians Clinical speciality: not clear Level of training: fully trained Setting/country: outpatient (e.g. ambulatory care provided by hospitals/specialists)/Canada |
|
| Interventions | The PEM studied in this report was the publication "Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes." New England Journal of Medicine, May 21, 2007. This meta‐analysis suggested an increased risk of myocardial infarction associated with rosiglitazone compared with active comparator or placebo | |
| Outcomes | 1 process outcome: number of new users of thiazolidinedione (rosiglitazone or pioglitazone) | |
| Notes | ‐ | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Intervention independent of other changes ‐ ITS | High risk | Quote, pg. 873: "several other studies of cardiovascular risk with thiazolidinediones were reported throughout 2007, which may have contributed to the overall decline in their use" |
| Shape of Intervention effect pre‐specified ‐ ITS | Unclear risk | Quote, pg. 871: "we sought to determine whether physicians’ choices of glucose‐lowering medications changed in the immediate aftermath of the publication of the meta‐analysis" |
| Intervention unlikely to affect data collection ‐ ITS | Low risk | The intervention (publication of report on rosiglitazone) did not affect either the source or the method of data collection |
| Blinding of outcome assessors (detection bias) ‐ ITS All outcomes | Low risk | The outcome was objective |
| Incomplete outcome data (attrition bias) ‐ ITS All outcomes | Low risk | Quote, pg. 871: "we examined prescription claims in the Ontario Drug Benefits (ODB) programme database, which contains records of all prescription medications dispensed to Ontario residents aged ≥ 65 years. We restricted our analysis to people aged ≥ 66 years (approximate n = 1.5 million), purposefully excluding the first year of eligibility to avoid incomplete medication records" |
| Selective reporting (reporting bias) ‐ ITS | Low risk | All relevant outcomes in the methods section were reported in the results section |
| Other bias ‐ ITS | Low risk | There was no evidence of other risks of bias |