Accuracy of general practitioner referrals to a breast clinic.

The large number of referrals for breast disease necessitates that some form of appointment stratification is carried out to try and ensure that those patients likely to have malignant disease are seen as soon as possible. A 6-month prospective study was carried out in which all new patient referral letters were assessed and graded on a three-point scale indicating the perceived likelihood of carcinoma. In all, 496 new patient referrals were assessed and graded. There were 94 classed as 'urgent' (representing a likely carcinoma), 186 as 'soon' (carcinoma unlikely but possible) and 216 as 'routine' (carcinoma very unlikely). The median waiting times to being seen in the outpatient department after referral were 6, 20 and 32 days, respectively, for the three groups. Of the patients, 56 (11.3%) were found to have a carcinoma; 41 (73.2%) of these had been placed in the 'urgent' group, 11 (19.6%) in the 'soon' group and 4 (7.1%) in the 'routine' group. All carcinomas in the routine group were coincidental findings. These results suggest that the information in the general practitioner referral letters may be used to reliably identify most patients with breast cancer, allowing appointment stratification and minimising any psychologically damaging delay before treatment.