No abstract

No abstract

Objectives To estimate the total proportional incidence of interval breast cancers in a two-yearly mammography screening programme, and to perform subgroup analyses by woman's age, screening centre-specific recall rate and screening round. Methods Using unconverted electronic data-sets from the 13 screening centres in the EmiliaRomagna Region of northern Italy (540,450 women aged 50 -69 years), a database of 919,538 mammography records was created. Of these, 655,175 eligible single-mammography records (1997 -2002) from 379,318 women were record-linked with the regional Breast Cancer Registry. In the two-year inter-screening interval, a total of 1,022,694.3 woman-years at risk were accumulated, with 695 interval cancers observed and 2428.3 expected. The observed number of interval cancers was divided by the expected number to obtain the proportional incidence. Results The total proportional incidence of first-and second-year interval cancers was 0.18 (95% CI 0.15 -0.20) and 0.43 (0.39-0.47), respectively. Woman's age was inversely associated with proportional incidence in both interval years, with a cut-off point at age 60. A screening centrespecific recall rate greater than the regional average of 5% was associated with a proportional incidence of 0.14 (0.11 -0.17) versus 0.20 (0.17 -0.24) in the first interval year, and of 0.36 (0.31 -0.41) versus 0.50 (0.44 -0.56) in the second. The proportional incidence remained unchanged between the first and subsequent screening rounds. Conclusions The results were in line with the previous Italian data and with the recommended European standards. The inverse effect of woman's age and of recall rate was expected.

The impact of the organised cervical cancer (CC) screening programmes implemented in Europe since the 1990s has been insufficiently evaluated. We investigated the changes in CC incidence following the introduction of a screening programme in the Emilia-Romagna Region (northern Italy). The study period was 1988-2013. The programme, targeting women aged 25-64 years (1,219,000 in 2018), started in 1998. The annual incidence rates that would be expected in 1998-2013 in the absence of screening were estimated, first, by analysing the annual rates in 1988-1997 with a log-linear model and, second, by analysing the annual rates in 1988-2013 with an age-period model in which the period effect was enforced to be linear. Cervical adenocarcinoma incidence trend over the entire period was used to validate both estimates. Observed annual rates were compared to the two series of expected ones with the incidence rate ratio (IRR). Incidence remained stable during 1988-1997, peaked in 1998 and then decreased until 2007, when it stabilised. The two series of expected rates were virtually coincident and their trends roughly paralleled the stable adenocarcinoma incidence trend. After 2007, the median IRR was 0.60 (95% confidence interval, 0.45-0.81) based on the log-linear model and 0.58 (95% confidence interval, 0.34-0.97) based on the age-period model. Thirty-six to seventy-five CC cases were prevented annually for an average annual frequency of 6.5 per 100,000 women in the target population. In summary, consistent circumstantial evidences were obtained that the organised screening programme brought about a 40% reduction in annual CC incidence after 10 years.