How often are gut feelings used by primary care physicians, and how often does a sense of alarm predict serious illness?
A diagnosis can be made via 2 different approaches: a slow, logical, step-by-step reasoning process, or (more often) a fast, intuitive approach based on recognition of patterns of illness seen in previous cases. In this study, Spanish general practitioners reported using the latter approach, their gut feeling, in almost all their consultations with patients who presented with a new concern. The physician's sense of alarm correctly predicted cancer or another serious disease 12% of the time, but the lack of it was correct 98% of the time. This is not the first study to show the role of gut feelings in the diagnostic process. 1c
Plan de l'etude:
Outpatient (primary care)
Intuitive sensations, called “gut feelings,” can be used by primary care clinicians when faced with diagnostic uncertainty. To test the value of this tacit knowledge, these investigators evaluated 155 general practitioners in Spain. The participants had an average age of 46 years and 70.3% were women. During at least one working day, participants completed the Gut Feeling Questionnaire for consecutive adult patients with at least one new reason for consultation (N =1487 patients). The investigators followed up with each patient at 2 months and 6 months, reviewing primary care and hospital clinical records for new diagnoses of cancer and other serious diseases. Participating physicians reported having a gut feeling during 97% of visits. Most (75%) of those gut feelings were a sense of reassurance (ie, they felt confident about the patient’s management and outcome even in the absence of a diagnosis). Participants reported a sense of alarm for 22% of consultations; they took additional diagnostic actions more frequently for this gut feeling. Within 2 months, 4.6% of patients were identified with a cancer or other serious illness. A physician's sense of alarm correctly predicted cancer or other serious disease in 1-in-8 patients (positive predictive value 12%; 95% CI 9% - 16%), but the lack of its presence was correct 98% of the time (negative predictive value 98%; 86% - 98%).
Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine