Inconsistencies in breast carcinoma registration. An investigation of 855 cases reported to the Swedish Cancer Registry.

The consistency of the data recorded in the Swedish Cancer Registry was analysed in 855 cases of breast carcinoma reported from Stockholm county during 1961-1963 and 1971-1973. The clinical records revealed errors in the recorded data in 56 cases (7%). Forty-one of these should be excluded from the registry. The most frequent cause of this over-registration was that local recurrences or distant metastases from previously diagnosed breast carcinoma were recorded as new primary tumours. Eight cases were found to be incorrectly recorded in situ carcinomas and should be deducted from the incidence of invasive carcinomas. The registry data were found to be unreliable with regard to patients with bilateral breast carcinoma. Nearly half of the cases recorded as such in the registry were found to be patients with secondary manifestations from a previous breast carcinoma. The frequency of erroneous registration was approximately equal during the two periods encompassed by the investigation.