Aim: Diabetic patients should understand their disease correctly and be sure of what they know, but certainty is rarely considered by educators. Furthermore little is known about how certainty changes with time after an educational intervention. To clarify this, in 38 patients with type 1 diabetes (0.3-36 years duration) we analysed the effect of a course on insulin use by administering a questionnaire before the course, after the course and 1 and 3 years later.
Methods: Answers, accompanied by a subjective estimate of the degree of certainty, were assigned to mastered knowledge (certainty>or=90%, correctness>or=90%), hazardous knowledge (certainty>or=90%, correctness<or=50%), uncertain knowledge (certainty<or=50%, correctness>or=90%) and residual knowledge (total-[mastered+hazardous+uncertain]). Answers were then counted and changes in distribution among areas were analysed by the chi2 test. We also followed the fate of wrong answers.
Results: The course increased mastered knowledge, while other types of knowledge decreased. With time mastered knowledge decreased, patients losing both correctness and certainty. The loss affected declarative knowledge, based purely on theory, more than procedural knowledge, which concerns the way things are done. Wrong answers, mostly given with high degree of certainty, were heterogeneous since some became correct after the course, some remained wrong, some became wrong after the course, some became mistaken after having been corrected earlier.
Conclusions: The analysis of certainty helps in evaluating patient's knowledge; programmes tending to improve procedural knowledge are more likely to have long lasting effects; wrong answers need to be considered on a individual basis.