Effectiveness of a small breast screening programme: 25 year evaluation (25 year breast screening evaluation)

To evaluate mammography screening quality on the Island of Jersey over a 25-year period from Jan 1990 to end March 2015 from females invited between ages 50 to 75 using a 2 yearly screening interval. Jersey had a population of only around 67,000 at onset, rising to around 100,000 at the end of the 25 years.

An analysis was performed of key routinely collected measures that are important to determining if a screening programme is on course to reduce breast cancer mortality such as uptake, recall rates, screen detected cancer and interval cancer rates. Further supporting indicators including grade, stage and comparative deaths from breast cancer in screen detected and not screen detected females were also assessed.

Over the 25-year period 19,768 females were invited to screening and 16,866 attended, giving an uptake of 85.2%. There were 501 screen detected cancers of which 400 were invasive, and 101 DCIS. 125 interval cancers presented outside screening over the 25 years. The annual recall rate over the last 20 years was <6% for prevalent round and 4% for incident round screening. Based on the standardized detection ratio (SDR) and uptake, the estimated reduction in mortality from breast cancer was calculated as 40.2%.

Recommended population sizes for breast units range from a quarter to half a million people. For very small units like Jersey serving smaller populations, rigorous quality control is essential to maintain credibility. Despite the small size of the programme evidence shows a similar detection rate to the UK NHS Breast screening programme was achieved. In small programmes careful monitoring of rates of uptake, recall, cancer detection and interval rates are required over adequate time periods together with supporting information to show that small units can achieve national standards and detection rates necessary to reduce breast cancer mortality.

Running a small breast cancer screening programme is challenging for quality control. The impact on mortality can be predicted for small screening programmes despite their size. 10-year group survival in screen detected invasive breast cancer >90%. Interval cancers are more advanced than screen detected invasive cancers, so high suspicion is still required in breast symptoms after "normal" screen result. Mortality in lapsed/ceased attenders suggest that extending age range could be beneficial.