Get access

Role of clinical context in residents’ physical examination diagnostic accuracy

Authors

  • Matthew Sibbald,

    1. Division of Cardiology, Department of Medicine, University of Toronto, ON, Canada
    2. Herbert Ho Ping Kong Centre of Excellence in Education and Practice, University Health Network, Toronto, ON, Canada
    Search for more papers by this author
  • Daniel Panisko,

    1. Division of General Internal Medicine, Department of Medicine, University of Toronto, ON, Canada
    2. Division of General Internal Medicine, Toronto Western Hospital, University Health Network, Toronto, ON, Canada
    3. Herbert Ho Ping Kong Centre of Excellence in Education and Practice, University Health Network, Toronto, ON, Canada
    Search for more papers by this author
  • Rodrigo B. Cavalcanti

    1. Division of General Internal Medicine, Department of Medicine, University of Toronto, ON, Canada
    2. Division of General Internal Medicine, Toronto Western Hospital, University Health Network, Toronto, ON, Canada
    3. Herbert Ho Ping Kong Centre of Excellence in Education and Practice, University Health Network, Toronto, ON, Canada
    Search for more papers by this author

Matthew Sibbald, c/o Dr R Cavalcanti, Toronto Western Hospital, University of Toronto, East Wing 8-420, 399 Bathurst Street, Toronto, Ontario M5T 2S8, Canada. Tel: 00 1 416 603 6412; Fax: 00 1 416 603 6495; E-mail: mattsibbald@gmail.com

Abstract

Medical Education 2011: 45: 415–421

Context  Clinical context may act as both an aid to decision making and a source of bias contributing to medical error. The effect of clinical history, a form of clinical context, on the diagnostic accuracy of the physical examination is unknown.

Methods  We randomised internal medicine residents to receive either no history or a short stem suggestive of one of six cardiac valvular diagnoses prior to a 10-minute objective structured clinical examination station assessing cardiac examination skills using a high-fidelity simulator. Clinical performance and diagnostic accuracy were compared using a standardised checklist.

Results  A total of 159 internal medicine residents were enrolled after providing informed consent. Of these, 80% arrived at the correct diagnosis, with diagnostic accuracy varying significantly by valve lesion (49–100%; p < 0.0001). Clinical context was associated with improved diagnostic accuracy compared with no history (90% versus 74%; likelihood ratio = 6.6, p < 0.0001), but was not associated with trainees’ ability to identify and characterise physical findings. Among residents given clinical context, higher diagnostic accuracy was only achieved by those able to correctly predict the diagnosis from the history.

Conclusions  Clinical context is associated with enhanced diagnostic accuracy of common valvular lesions. However, this effect seems linked to heuristic hypothesis generation and may predispose to premature diagnostic closure, anchoring and confirmation bias.

Ancillary