Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science


Epidemiology and Biostatistics


Zwarenstein, Merrick

2nd Supervisor

Klar, Neil



Printed educational materials (PEMs) have long been used as a tool for knowledge translation despite the conflicting evidence surrounding their effectiveness. A pragmatic, 2x2 factorial, cluster-randomized controlled trial was designed to ascertain the effectiveness of two distinct formats of a PEM (insert and outsert) at improving adherence to guideline recommendations for the management of diabetes complications among Ontario family and general practitioners. Administrative databases were used to compare patient’s treatment regimens at baseline and one year following PEM mailout to determine whether prescription rates intensified in response to the PEMs. A total of 4,118 practices (4,957 physicians) and 185,454 patients were included. Intensification rates in the four groups were similar and approximately equal to 46%. In intention-to-treat analyses, no treatment effect was found with the insert (OR 0.99, 95% CI 0.96 to 1.02), nor with the outsert (OR 1.01, 95% CI 0.98 to 1.04). Thus, PEMs were not effective at improving physician’s adherence to guidelines for diabetes care.

Summary for Lay Audience

Despite numerous medical breakthroughs, the health status of Ontarians is far from perfect. This disconnect is, in part, due to the poor translation of research findings from the bench to the bedside. To address this problem, many studies have investigated different strategies to bridge the gap between researchers and front-line staff. Printed educational materials (PEMs) consist of any recommendation for clinical care, whether it be a journal article, a magazine insert, or a letter that is delivered in print format to the recipient. PEMs have long been used as a strategy to inform clinicians on evidence-based practices and to persuade them to use these treatments or interventions. However, the literature provides conflicting evidence surrounding the effectiveness of PEMs, yet they continue to be used today. Thus, the present study aimed to investigate the effectiveness of PEMs at changing provider behavior.

The Ontario Printed Educational Message (OPEM) trial was carried out in 2005, a time when the prescription rate of drugs used to prevent diabetic complications was well below guideline recommendations. PEMs were thus developed to highlight several evidence-based recommendations for drug use among individuals with diabetes, and Ontario family and general practitioners’ practices were allocated at random to receive one of two formats of the resulting PEM (a post card sized message or a long article, referred to as the outsert and insert, respectively), both, or neither (as a usual situation comparison group). Health administrative databases were used to ascertain the effectiveness of PEMs by observing whether treatment intensification occurred to a greater degree among patients of physicians who received a PEM, compared to those who did not.

Neither the insert nor the outsert were successful at causing physicians to intensify their patient’s treatment regimen by adding a new drug, increasing the dose of a current drug, or switching from one drug to another drug.

Thus, the use of PEMs to improve physician’s adherence to guideline recommendations for diabetes care is ineffective and should not be encouraged. Further research is required to investigate other strategies to inform physicians on evidence-based recommendations, as prescriptions for diabetes care remain below standard today.