GPs' ability to diagnose dyspepsia based only on physical examination and patient history.

OBJECTIVES: To evaluate the diagnostic value of the general practitioner's (GP's) unaided working diagnoses in dyspepsia. To compare the proportions of final diagnoses and the characteristics of patients who would have been referred to gastroscopy or to empirical drug treatment.

DESIGN: Clinical study.

PATIENTS: 400 consecutive dyspeptic patients consulting their GPs.

MAIN OUTCOME MEASURES: Sensitivity, specificity and positive and negative predictive values (PV+ and PV-) were calculated for the GP's working diagnoses in cases of dyspepsia. The outcome of his/her decisions on how to manage dyspeptic patients was also evaluated.

RESULTS: Gastroesophageal reflux disease would have been diagnosed accurately, with a sensitivity of 0.59, specificity 0.83, PV+ 0.43 and PV- 0.90. GPs diagnosed functional disorders with a sensitivity of 0.43, specificity 0.69, PV+ 0.56 and PV- 0.54; peptic ulcer with a sensitivity of 0.37, specificity 0.83, PV+ 0.28 and PV- 0.88; and malignancy with a sensitivity of 0.13, specificity 0.97, PV+ 0.08 and PV- 0.98. Patients who would have been referred to gastroscopy had more often lost weight (p = 0.01), suffered from abdominal pain (p=0.03) and from symptoms partly suggesting irritable bowel syndrome (p< or =0.03).

CONCLUSIONS: The clinical diagnosis of the causes of dyspepsia is unreliable. In selecting patients for gastroscopy, more attention should be paid to risk factors such as age, use of non-steroidal anti-inflammatory drugs and history of previous peptic ulcer.