The Science Behind Telling “Sick” From “Not Sick”

Jeff Kline contributed a very interesting article to the Emergency Medicine Journal last week – offering up a bit of a potential science behind the “gestalt” in medicine. We’ve seen multiple examples where clinician gestalt performance is very similar to carefully-derived, evidence-based, risk-stratification criteria. Specifically, the diagnoses of “acute coronary syndrome” and “pulmonary embolism” have been evaluated in the past – and only the newest attribute-matching tools have offered any promise regarding improving upon simple clinician judgement.

This newest study from Kline, et al, evaluated 50 patients in the Emergency Department and their facial reactions to visual cues. It turned out, the 18 patients from this cohort ultimately diagnosed with significant cardiovascular syndromes displayed significantly decreased expressive variability when prompted with multiple stimuli. The reasonable conclusion, therefore, is patients with serious diagnoses may exhibit measurable, reproducible behavior changes. A small study, to be sure, but hence the idea – there is something encoded in our emotional intelligence helping us evaluate “sick” from “not sick” in the Emergency Department.

Supposing this observation holds up to further scrutiny, the results do not surprise me at all. Part of clinical training in Emergency Medicine involves simple voluminous exposure to as many patients as feasible. The behaviors of each different patient, their clinical features, and their outcomes become encoded in this entity, the clinician “gestalt”. And, what this study reflects is something we all recognize – a patient is not simply a collection of risk factors, or a Revised Geneva Score – the physicality associated with how a patient exists in the examination room provides additional information. The intuition of the experienced clinician, then, may be based as much in reading patients’ faces as it is synthesizing clinical knowledge.

This has interesting implications for other developments in medicine, as well. The time pressures in Emergency Medicine, or in other outpatient settings, that simply cut down on time spent with each patient, may detract from the quality of the evaluation. Telemedicine, another technological advance aimed at diluting and expanding coverage, may suffer as a result of diminished communication of these critical nonverbal cues.

Regardless, this study is quite unique in the spectrum of Emergency Medicine research, and hopefully inspires a follow-up generation of research.  Or, alternatively, what would you say forms the basis of our “gestalt”?

Link: “Decreased facial expression variability in patients with serious cardiopulmonary disease in the emergency care setting






Ryan Radecki