Analysing Medical Incident Reports by Use of a Human Error Taxonomy

Abstract

The present paper describes analysis results of incident reports collected from three Japanese hospitals. We developed a medical incident taxonomy by adapting an existing scheme in air traffic management. Its application allowed us to elicit characteristics of adverse events and errors that have occurred in health care. The taxonomy was evaluated in terms of inter-rater reliability by use of raw agreement and a tentative measure of chance corrected agreement between two judges.