Determination of cause of death among breast cancer cases in the Swedish randomized mammography screening trials. A comparison between official statistics and validation by an endpoint committee.

Between 1976 and 1982, four randomized mammography screening trials started in five screening centres in Sweden, involving 282,777 women (156,911 invited and 125,866 controls) with the aim to study if invitation to screening reduced the breast cancer mortality. An overview of the trials was performed to reduce the confidence intervals for the relative risk estimates. All 1,296 deaths occurring in women with breast cancer detected after randomization were evaluated by an independent endpoint committee (EPC), consisting of four physicians who reviewed collected medical information that was blinded regarding mammography screening. If there was disagreement between the EPC members at the initial individual evaluation the final classification was made at consensus meetings. In only 6.9% (n = 89) of the cases was there disagreement as to whether breast cancer was or was not the underlying cause of death. It was also found that 'breast cancer as underlying cause of death' and 'breast cancer as underlying or contributory cause of death' according to Statistics Sweden resulted in relative risk estimates very similar to those based on the classification by the EPC. The study thus supports the use of official health statistics in the evaluation of randomized breast screening trials in Sweden.