Simulation Based Medical Education (SBME) of Intraoperative Frozen Section Interpretation.
Gregory R Kotnis, Trent A Wilkerson, Christina P Yantsides, Stephen S Raab. University of Colorado Denver, Aurora
Background: A major source of latent error in diagnostic anatomic pathology is secondary to pathologist training. The traditional apprenticeship training model focuses on completion of specifics tasks and not on quality improvement, standardization, and continuous learning. Using standardized SBME modules, we assessed baseline resident performance in intraoperative diagnostic consultation practice as a means to target training for specific weaknesses.
Design: We created a SBME intraoperative consultation program that includes 400 intraoperative cases modules. Slides from each module were reviewed, the original interpretation was confirmed, and a new interpretation in a standard format was generated. A diagnostic difficulty score and a technical quality score, using a five point Likert scale, were assigned. Nine pathology residents representing all five years of post graduate training (PGY) completed the SBME modules to establish baseline competence. The residents were provided the cases using standaridzed answer forms and assigned a technical quality and difficulty score. A total of 30 modules of 5 cases each were completed. The concordance between the technical and diagnostic scores was correlated with the original assessment; the interpretation was examined for error in 2 ways: 1) if the standard and the resident interpretation directly matched, and 2) if the standard and resident interpretation matched in a general, but not a direct sense (e.g. necrotizing granuloma and negative.)
Results: Overall the pathologists in training were 84% correct using the loose criteria and 70% correct using the strict criteria (see Table 1). Segregating the cases by organ system allowed us to track trends, indicating the gynecologic and gastrointestinal specimens had high rates of error, while lymph nodes for metastaic cancer did not. Tracking of technical quality and root cause analysis points to shatter artifact and fat content to be particular reasons for poor technical scoring.
|Total Correct Mean (Loose)||84%||Total Correct Mean (Strict)||70%|
|Correct Mean (Loose) PGY1||68%||Correct Mean (Strict) PGY1||44%|
|Correct Mean (Loose) PGY2||87%||Correct Mean (Strict) PGY2||68%|
|Correct Mean (Loose) PGY3||78%||Correct Mean (Strict) PGY3||66%|
|Correct Mean (Loose) PGY4||88%||Correct Mean (Strict) PGY4||88%|
|Correct Mean (Loose) PGY5||94%||Correct Mean (Strict) PGY5||86%|