Breast cancer data collection for surgical audit.

Data are available about the mortality, the incidence, the stage, the survival, and the treatment of breast cancer. In this country mortality data are published by the OPCS and HIPE and evidence exists to show that there is a considerable shortfall in both these sources. The incidence of breast cancer is recorded by the Regional Cancer Registries and published by OPCS. The registries supply OPCS with a minimum data set which does not include the stage of the disease, which is important, and does include the quadrant, which is not. Audit shows marked regional variations in completeness of registration. International comparisons with Scandinavia are unfavourable and show what can be achieved. Patients who develop breast cancer but do not die from it may be cured. The discrepancy between incidence and mortality, which varies both geographically and historically, should therefore provide a valuable insight into changing survival, but does not do so because the data are unreliable. Wide variations in survival figures in the medical literature are probably due to variations in staging conventions rather than different treatments. It is clear that cancer stages carry a more favourable prognosis, but it still cannot be proved that earlier diagnosis in a particular individual prolongs the life of that individual. This unproven hypothesis is the rationale for mammographic screening. Data on treatment in hospitals do not link diagnosis to operation, so that it is impossible to separate operations for benign breast disease from those for cancer. The OPCS classification of operations is complex with many open-ended choices. Doctors do not participate in the coding process and clerks cannot make up for this. Radiotherapy and chemotherapy data are not collected nationally and the four regional registries who do collect it also rely on their clerks' interpretation of medical notes. Data on the use of tamoxifen are consistent and of high quality.An extensive data gathering mechanism is in place for breast cancer. A huge body of information exists about the disease which should form a valuable database. The poor quality of this information reduces its usefulness. It is the responsibility of doctors to agree on a data set and to ensure its collection. They do it in Scandinavia.