Medical Error and Self‐Reflection: Analysis of Internal Medicine Residents' Adverse Event Self‐Reflective Exercises

1University of Michigan, Ann Arbor, Ml
2University of Michigan, Ann Arbor, Ml
3University of Michigan, Ann Arbor, Ml
4University of Michigan, Ann Arbor, Ml
5University of Michigan, Ann Arbor, Ml

Meeting: Hospital Medicine 2009, May 14-17, Chicago, Ill.

Abstract number: 121


The Institute of Medicine (IOM) report released in 1999 estimated between 44,000 and 98,000 inpatient hospital deaths resulted from medical errors within the United States annually. Although we have made advances during the past decade to reduce medical error through our understanding of human factors engineering (HFE), development of technology {e.g., electronic orders/chart documentation), prevention of iatrogenic complications secondary to therapy, and improvements in the working environment, an area that continues to lag behind these innovations in the education of physicians in training. Here we present results from our patient safety educational program describing our trainees' analytic skills based on self‐reflective exercises in which they analyzed adverse events.


To provide a descriptive analysis of the types of cases analyzed and perception of error preventability and to demonstrate reliability for an abstraction tool used to evaluate our trainees' error analysis.


Since 2006, we have implemented a patient safety curriculum for physicians training in internal medicine and internal medicine‐pediatrics that provides a framework to improve their understanding and analysis of medical error; guided self‐reflection; and hands‐on experience by working on a team‐based patient safety improvement project. As part of the curriculum, trainees are required to analyze an adverse event in which they were involved using a structured tool. Trainees analyzed 2 cases, 1 prior to and 1 following a seminar series that presented key concepts in patient safety, our model for adverse event error analysis; and introduction to principles of HFE. Fifteen paired sets of self‐reflective exercises were analyzed using an abstraction tool that examined the trainees' initial hypothesis; patient, organizational, experience, and team (POET) factors that contributed to the adverse event; and the final hypothesis following their analyses. Each adverse event was mapped to 1 of 7 error domains. In addition, trainees assigned outcomes to each adverse event. Interrater reliability for our abstraction tool was determined using a Spearman correlation.


Trainees described adverse event outcomes that were refatively evenly distributed across categories: no/minor harm (30%), moderate harm (43.3%), or severe harm/death (26.7%). Medication safety (28.3%), diagnostic errors (20%), incorrect identification (18.4%), and communication (16.7%) accounted for more than 80% of the adverse events described. Of all adverse events, 96.7% were identified as preventable. Our abstraction tool demonstrated significant interrater reliability, ranging from 0.55 (initial hypothesis) to 0.89 (POET factors and final hypothesis) and 1.0 (adverse event preventability). Self‐reflection may serve as a valuable and reliable tool to further enhance the education of our trainees in better understanding medical error and adverse event analysis.

Author Disclosure:

M. P. Lukela, none; S. J. Hamstra, none; R. Mangrulkar, none; V. Parekh, none; J. Del Valle, none.

To cite this abstract:

Lukela M, Hamstra S, Mangrulkar R, Parekh V, Valle J. Medical Error and Self‐Reflection: Analysis of Internal Medicine Residents' Adverse Event Self‐Reflective Exercises. Abstract published at Hospital Medicine 2009, May 14-17, Chicago, Ill. Abstract 121. Journal of Hospital Medicine. 2009; 4 (suppl 1). Accessed March 29, 2020.

« Back to Hospital Medicine 2009, May 14-17, Chicago, Ill.