*274. Insulation from Breakdown as a Barrier to Resident Learning in Clinic
C S Smith, Boise VAMC and University of Washington; C Francovich, Boise VAMC; J Gieselman, Boise VAMC; M Morris, Boise VAMC
Objectives: Document and analyze breakdowns for clinic patients followed by residents.
Determine the degree of resident involvment/insulation from these breakdowns and its effect on patient care and learning
Methods: Trained observers collected ethnographic data from clinic waiting areas, work stations and exam rooms over one year. Initial observations were broad and general, followed by targeted specific observations and interviews of key personnel. Real-time field notes from sixty-eight separate observations (total over 130 hours) were immediately transcribed by the observers into 2919 paragraphs, the data set. Two analysts initially coded text units at the paragraph level into recurrent themes using template analysis. The major theme, breakdown, was then fine-coded at the phrase level with the goal of creating a taxonomy of breakdowns. Using this taxonomy, 156 vignettes were selected from the data to represent the full spectrum of breakdown. These were used to iteratively create a structural model of breakdown and learning. Triangulation and member checking were used to assess credibility and dependability of the model.
Results: In the first step of analysis, fifteen themes were identified and applied with 96% interrater reliability. The major theme, breakdown, was seen in 43% of all text units. In the second step, taxonomy was developed for breakdown and applied with 97% interrater reliability. In the final step, a testable structural model of breakdown and learning was developed. This structural model proved credible.
The model consists of three types of response to breakdown. The first is where the personal experience of a phenomenon (the kinesthetic, endocrine and autonomic interactions), along with attached emotional response, lead to automatic behaviors. If breakdown is not solved at this level, then conceptual models are shared among participants to determine an appropriate action. For breakdowns that are more persistent and become obtrusive, participants approach the problem by stepping back and examining their beliefs and conventions. Insulation from directly dealing with these breakdowns isolates the resident from important learning opportunities.
Conclusions: We define breakdown as a disruption in smooth, unreflective activity that requires a more explicit, deliberate and contemplative perspective. We have empirically identified and validated a plausible model for breakdown, feedback and learning for residents in clinic. Learning should occur around breakdown and feedback, but multiple observations demonstrated that resident insulation from breakdown served as a barrier to resident learning. Our model highlights several problems for resident learning such as role overload, lack of continuity, difficulty in conversing with patients while using an electronic medical record, and differences between "what we say" and "what we do" in our training program.
Impact: This model brings into sharp focus several of the tradeoffs seen when the educational and patient care missions are grafted together in clinic. Unintended structural and systems barriers to resident learning and patient care become apparent. The model itself is testable, and it generates several testable interventions.