Reply: Overdiagnosis of breast cancer in Denmark

As observed by Dr Zahl, it is correct that the cumulative risk in breast cancer incidence increased during the observed period. Indeed, breast cancer incidence has increased steadily in Denmark since the 1960s ( Sundhedsstyrelsen, 2003 ). Assuming linearity in the cumulative risk, we performed a simple least-squares estimation of the trend in breast cancer incidence in the period before screening for both Copenhagen and Fyn. Assuming this trend would have continued if screening had not been introduced, the expected cumulative risk for Copenhagen in 1993–95 was 6.6% and the observed cumulative risk was 6.1% (95% CI 5.3–6.8%). For Fyn the expected cumulative risk for 1996–97 was 5.9%, whereas the observed cumulative risk was 6.6% (95% CI 5.9–7.3%). For both Fyn and Copenhagen the expected cumulative risk was within the 95% confidence interval. Therefore, we see no reason to conclude that overdiagnosis is taking place.