Time trends of process and impact indicators in Italian breast screening programmes--1996-2005

Since its establishment in 1990, one of the main tasks of the Italian Group for Mammography Screening (GISMa) has been that of collecting, as systematically and thoroughly as possible, data on the activity of the various mammography screening programmes implemented in Italy. Data are collected in an aggregated way and gathered through a standardised form to calculate process and impact parameters. Analysis of data collected in the period 1996-2006 shows that the number of Italian women involved in mammography screening has constantly increased over time. In 2006, almost 8 out of 10 Italian women in the 50-69 year age range lived in areas where an organised screening activity had been implemented. The geographical distribution of the screening activity in Italy continues to present differences between Northern and Central Italy, where the coverage ranges from 95.7% to 89.6%, and Southern Italy and the Islands, where this value is still lower (45.9%). The discrepancy between annual theoretical coverage and actual coverage still persists (21%). Throughout the period, crude attendance rate reached the acceptable 50% standard, presenting the same North-South trend as the coverage rate distribution. In areas where the centralised management is more established or complete, the participation rate was higher compared with areas without such characteristics, with differences of about 4-9%. The time trends of the other parameters included in the analysis showed a good average performance. For example, benignant/malignant surgical biopsy ratio (B/M ratio) continued to progressively decrease, reaching 0.24 at first screening and 0.14 at subsequent screening in 2005; overall detection rate, detection rate for in situ and small cancers (< or =10 mm) showed a good trend, reaching 6.0 per thousand, 0.7 per thousand, and 1.3 per thousand, respectively, for first screening, and 5.2 per thousand, 0.8 per thousand, and 1.6 per thousand for subsequent screening in 2005. On the contrary, excess referral rate at first screening persists over time. Data comparing activity volume and programme duration were also considered. In programmes with greater activity volume (average test number: 24,596) the referral rate for first screening was higher if compared with programmes with lower activity volume, but still within acceptable standards. Nevertheless, this performance is compensated by better specificity and sensitivity. An improvement in quality with the increase of programmes experience continues to be evident: compared with new programmes, programmes with more than 6 years of activity showed lower recall rates and higher positive predictive values and overall detection rates in first screening. Outcomes related to subsequent screening present analogous performances. These results continue to be consistent with those achieved by other European programmes and reassuring for all Italian mammography screening professionals.