Evaluating chest pain. The patient's presentation style alters the physician's diagnostic approach

Clinical prediction rules rely largely on objective data to estimate coronary artery disease (CAD) likelihood. However, characterization of chest pain, which is central to such prediction rules, depends in part on a physician's subjective judgments. We performed a clinical trial to assess the influence of the patient's presentation style on the physician's approach to evaluating chest pain.

Forty-four internists were randomized to one of three treatment groups. Two groups viewed videotapes of the same actress performing the role of a patient in a scripted physician-patient interview in two distinct styles: one group saw a "histrionic" characterization, the other a "businesslike" portrayal. The interviewer was not seen or heard by the subjects; they saw only words on the screen. The third group read a verbatim transcript of the interview. After their initial CAD-likelihood estimates and impressions of probable cause for the patient's symptoms, which were based on history only, the participants in all three groups were given the same laboratory data and a second CAD-likelihood estimate was made. Finally, recommendations for further workup were elicited.

Initial diagnostic impressions differed dramatically: a cardiac cause was suspected by 50% of physicians viewing the businesslike portrayal but by only 13% of those viewing the histrionic portrayal. Likewise, those viewing the histrionic and businesslike videos provided different CAD-likelihood estimates initially (10% vs 20%). However, after the patient's laboratory data were revealed, the difference in CAD-likelihood estimates was no longer significant. Despite their making a similar risk appraisal after receiving all of the data, internists viewing the histrionic portrayal were far less likely to pursue a cardiac workup (53% vs 93%).

Although physicians may evaluate patients who have the same history word for word and the same laboratory data and whom they regard as having nearly identical likelihoods of CAD, the physician's ultimate diagnostic approach can be profoundly affected by the patient's presentation style.