Comments on the paper “Medicolegal implications of accuracy of GP referral letters to specialist breast clinic”

Ahmed and colleagues [1] have described well the dilemmas faced by clinicians when diagnosing women with symptomatic breast disease. However, one important factor was omitted that helps to explain the poor correlation between general practitioner findings and those assigned by the specialist in the breast clinic. The positive predictive value of symptoms, signs and non-diagnostic test results (that is, the likelihood that a patient who tests positive actually has the disease) varies strongly with disease prevalence [2, 3]. The same test with the same sensitivity and specificity will have more false positives in a lower disease prevalence setting (e.g. primary care) than in a higher disease prevalence setting (e.g. hospital) [4, 5]. Thus, a clinician with a certain standard of skills in clinical history taking and clinical examination, if working in primary care, will have more false positives than the same clinician working in a hospital setting.