Intercurrent mortality of breast carcinoma patients. An investigation of 3857 cases from the Swedish Cancer Register.

The pattern of intercurrent mortality was studied in 3857 cases of breast carcinoma diagnosed in Stockholm county during 1961-1963 and 1971-1973. The investigation was based on the recorded underlying and contributory causes of death in the Swedish Register of Causes of Death. The observed number of deaths due to various main disease groups was compared with the expected number calculated for an age-matched general population. A 21 per cent increase of intercurrent deaths was observed (95% confidence limits: 13-29%, p less than 0.001). Excess risks were found due to circulatory diseases, other tumours, 'accidents, suicides and injuries' and infectious diseases. The excess risk increased with time during follow-up and in the period 10 to 21 years it was 35 per cent (17-55%, p less than 0.001). These findings could be interpreted as an association between breast carcinoma and other diseases but might also to some extent have been caused by diagnostic errors or errors in the ascertainment of underlying versus contributory causes of death. It is concluded, however, that such possible errors are probably not a major source of bias in judging the mortality trend for the disease during the period encompassed by the investigation. This conclusion was based on the circumstance that the observed excess risks were of equal magnitude during the first ten years of follow-up in both the 1961-1963 and 1971-1973 series.