More disturbing news on issues in health-related research.
A new paper from the National Bureau of Economic Research examines the validity of the interrupted time-series experimental design.
An article in Nature magazine points out the dangers of trusting medical research too much.
The case of the non-existent link between certain fats in the diet and heart disease reminds us again how poorly designed and executed much research is, and how dangerous dissemination of the results of such research can be.
The light fades but not our evanescent Potpourri, this week featuring stories on computerized point of entry ordering, the presence of large treatment effects in research, characteristics of patients with readmissions, a survey on Medicare physician reimbursement and a study on family caregivers.