#KeyLIMEPodcast 338: Do the ‘Eyes’ have it?

Reference

William Wu, Andrew K. Hall, Heather Braund, Colin R. Bell & Adam Szulewsk. 2021. The Development of Visual Expertise in ECG Interpretation: An Eye-Tracking Augmented Re Situ Interview Approach. Teaching and Learning in Medicine, 33:3, 258-269.

Reviewer

Linda Snell (@LindaSMedEd)

Background

I picked this article as the title attracted me – what is visual expertise, eye tracking sounds cool, and what does re situ mean anyway? (the actual situation)

Visual expertise in medicine is complex – combining visual gaze patterns with cognitive processes like decision-making or memory retrieval. Lots of areas of medicine rely on this (dare I call it a) competency… derm, radiology, pathology… and it is thought that visual expertise is domain specific.

Many disciplines interpret ECGs, and there can be a significant error rate in trainees and where there are varied teaching methods. Prior studies have looked at visual aspect (eye movements etc.) and not to their link with decision making, recognition, etc…where more of the errors may lie. Recently mixed methods have been used, e.g. verbal protocols of thinking to understand underlying cognitive processes, along with eye tracking (long fixation on one aspect of an ECG could either show the importance of a specific ECG finding, or could represent uncertainty)

Purpose

1. ‘Qualitatively describe the difference in cognitive approaches to ECG interpretation between medical students, EM residents, and EM attending physicians.
2. Analyzing speed and accuracy of diagnosis while using heat maps to highlight differences in visuo-spatial fixation distribution across the ECGs.’

Key Points on the Methods

Cued retrospective recall protocol, using eye-tracking to provide memory cues; Done post hoc, thought not to interfere with task of interpretation.

Generated heat maps of eye fixation, measured speed and accuracy of interpretation.

3 groups of 10: Med 2-3, EM residents PGY 1-4, EM attendings ~ 18 y experience

Protocol for interpreting and interview well described. Detailed description of thematic analysis of transcripts. Good reflexivity statement.

Key Findings

Predictable group differences in time and correct diagnoses.

Heat maps (very pretty)  also differed by group – students more variability, attendings focused on the abnormality.

Themes from interviews

1. Dual process reasoning: unconscious/ automatic vs conscious/analytic. All groups used both but in different ways. Attendings had confidence in initial impression of the ECG (System 1 approach), would often make a diagnosis within seconds using pattern recognition. They used a more analytical approach (System 2 thinking) only if they did not initially pick up on an abnormality or to double check for specific features. Trainees rarely practice System 1, were less confident in initial impression subsequently relied on a systematic System 2 approach.
2. Ability to prioritize by information reduction, made rapid decisions on relevance; attendings >> students. Students thus overwhelmed. Attendings and residents generated early hypotheses based on clinical stem before even seeing ECG, filtering out distractions
3. Clinical implications more senior built in common approaches like ruling out most life-threatening things first, need to think about ‘big picture’

In summary: ‘more experienced practitioners were more likely to rely on System 1 thinking, demonstrated strategies to reduce information overload, and were more cognizant of clinical factors outside of the ECG, such as treatment implications and clinical correlates. Medical students and residents tended to employ more System 2 thinking, demonstrated fewer information reduction strategies, and were less cognizant of clinical implications.’

Authors raise a number of education concepts… of being ‘unconsciously competent’ in experts, and the ‘expert blind spot’, both of which may hinder teaching effectiveness. Also how information reduction can lead to the development of an illness script, and teaching script