Challenges of teaching EBM

I am just finishing a master's degree in medical education, and my thesis work is closely related to the series on learning and teaching EBM.¹ I have used the name DIETS (for “Disease management Interactive Teaching Script”) for my system, which incorporates up-to-date EBM in teaching the management of chronic disease to primary care doctors and nurses. The DIETS system includes presentations on topics such as hypertension and diabetes, which take the form of mini-lessons or “slide sets.” In contrast to the teaching tips articles,² which a reader can dip into and out of, each DIETS talk must be presented in its entirety, as a reflection of the longitudinal nature of primary care. Another difference is that feedback in DIETS is formalized, appearing in the third slide of each set, whereas feedback is less structured in the teaching tips. An important similarity between the scripts used in the DIETS system and those in the EBM series is that the teaching of particular points conforms with the suggestions of Peile and colleagues,³ who discussed the notion of threshold concepts — key concepts that, once understood, allow learners to understand, interpret or view something in a different way. Until that threshold is crossed, the student will struggle to progress.

I have used part 1 of the teaching tips series² to analyze the synopsis⁴ (published recently in CMAJ) of the Collaborative Atorvastatin Diabetes Study (CARDS) and can suggest 2 additional exercises related to this study. First, in CARDS, the reduction in the death rate for diabetic patients who received atorvastatin was from 5.8% to 4.3%, an absolute risk reduction of 1.5% over 4 years (1.5%/4 = 0.375% annually), which yields a number needed to treat of 100/0.375 or about 275. If atorvastatin costs about £20 per month (about Can$47), how much did it cost to save one life? The answer is 20 х 12 х 275 or £66 000 (Can$154 500). Is this cost-effective? Second, from the CARDS data, we can calculate the 10-year risk of coronary artery disease for diabetic patients not receiving statin treatment. The overall rate of cardiovascular events in the placebo group was 13.4%, and median follow-up was 4 years. Therefore, the annual rate would be 13.4%/4 (3.35%) and the 10-year rate would be 33.5%. The rate of coronary artery disease was approximately three-quarters of the total cardiovascular risk^5,6 or about 25%.

Gerry E. Burns Family Doctor National Health Service Belfast, Northern Ireland