Purpose. Many medical education interventions improve clinicians’ knowledge but fail to change behavior. The authors exposed this knowledge–behavior gap through standardized clinical interactions, thus allowing in-depth exploration of the contributing factors.
Method. A typical evidence-based educational intervention in one clinical domain (early signs of autism) was administered to family medicine residents at the University of Toronto in 2001–02, and change in knowledge was assessed through a multiple-choice test. Six to eight weeks later, participants’ relevant knowledge was documented, and their clinical behavior was observed during four interactions with standardized patients. Factors producing a knowledge–behavior discrepancy were then explored using semistructured interviews, which were audiotaped, transcribed, and analyzed using grounded theory methods.
Results. Half of participants demonstrated varying degrees of knowledge–behavior gap. Eight main rationalizations (relationships, patient agenda, knowledge deficit, clinical style, means to an end, ideals, autism stigma, and systems barriers) were used to justify choices of clinical behavior, and the same rationalizations were used to justify opposite choices of behavior. Two conditions that promote clinical action based on knowledge (level of certainty and sense of urgency) were identified.
Conclusion. The knowledge–behavior gap was exposed and factors reported to influence clinicians’ decisions about whether to implement new knowledge were elicited. That identical rationalizations were used to justify opposite behaviors implies these factors may not be behavioral determinants. Sense of urgency and level of certainty promote clinical action based on knowledge; focusing on these may increase the impact of education on practice.