Clinical Communication Ontology for Medical Errors

Introduction

The literature indicates the critical relationship between teamwork and safety in healthcare performance. In teamwork, communication failures caused 60% of sentinel events reported by the Joint Commission on Accreditation of Healthcare Organizations. Communication failures negatively affect the quality of decision making and result in decreased healthcare quality. Moreover, the difficulties of communication have been the primary cause of errors leading to patients’ death^[1].

Observational studies in the emergency department and other clinical settings demonstrate that clinicians experience high communication loads. The high communication load in an interruption driven, multitask setting negatively affects clinicians’ working memory causing medical errors. Researchers also identify that clinicians spend 80% of their time in communication, with 30% of all communication events classified as interruptions^[2].

Analyzing the communications as a contributor to medical errors will reveal the pattern and provide information for proper intervention and ultimately improve the communication quality and efficiency thus reducing medical errors. Studies demonstrate that clinicians prefer face-to-face communication with colleagues as the primary method for answering clinical questions. Other communication methods such as technology-mediated conversions, email, whiteboard and so on are also being used by clinicians as secondary methods.

Few communication taxonomies and communication error taxonomies are developed in multiple domains other than healthcare ^{[3, 4]}. The taxonomies classify various types of communications that occur between humans, and between humans and technologies. However, little effort has been made to identify the more complex communication in healthcare. We believe that a full consideration of the other contributing factors such as, cognitive affective process, connection topology should be also included in such a multitask, time critical and life relevant domain.

A communication ontology provides formal definitions, relationships and usages of various concepts relevant to the domain of communication. Taking grounded theory as a theoretical framework, we propose an ontology for examining how different patterns of communication contribute to medical errors and how communication patterns are related to each other. We plan to use this ontology to support the collection, storage and interpretation of communication failures in health care settings. The ontology, as an important component of a larger medical error ontology project, will be integrated with the medical error ontology that we are developing.

Development Approach

The communication ontology contains eight axes which are time line, location, participants, message characteristics, media, task of sender, transit steps, and communication outcome classes prototyped in Protégé 3.11. The ontology is primarily based on the observational results of clinical activities, existing communication taxonomies and communication error taxonomies in other domains. In addition, we supplemented the cognitive factors, communication norm and so on to increase the coverage of the clinical communication.

In our ontology prototype, we used a “Communication Error Case” class to model a clinical communication where an adverse result happened due to the deficiency of communication. When a communication error happens, there must be but not limited to time, location, participant, message and media to describe when, where, who, what and how an event happens.

Ongoing Work

The ontology primarily provides the descriptions of clinical communication, allows users to categorize the clinical communication activities and ultimately reveals the patterns for effective communication among clinicians. The development of ontology is an iterative process, the modification and reconstruction will be continuous during the case coding process. We selected a few clinical communication cases from the case pool we built up and coded the cases using the ontology. We aim at an exhaustive and exclusive ontology in clinical communication which not only includes the description of the clinical communication but also be able to analyze the communication pattern, delegate some tasks to machines, mitigate negative interruptions and eventually recommend the solutions for reducing medical errors.