Breast cancer is the second leading cause of cancer death in Denmark, killing 1,100 women annually. Breast cancer has had the highest incidence among women for several years and incidence rates continue to rise, and as a result, there is an apparent need to end this trend. Screening programs are a great way to identify breast cancer early, but false positives can create a large burden of disease. Specificity is a measurement of the accuracy of a test to diagnose a disease-free individual disease-free. By looking at the specificity of the Copenhagen Regional mammography program we’ll be able to identify the effectiveness of the program.

The Copenhagen Mammography program has been running since 1991. Every other year women ages 50-69 are invited to receive a free mammography screening and check for any developing cancer. Independent double readings are done on all projections taken to ensure their accuracy. Previous studies have shown that false positives from screening programs increase the burden of disease, and are associated with higher levels of anxiety. This research analyzes the specificity of the mammography program over time and amongst different age groups to find the effectiveness of the screening program at accurately diagnosis women with breast cancer.

Data collected from the Copenhagen Regional screening program was transferred into a data set that guaranteed the anonymity of the participants of the study. The data set was analyzed using RStudio. These values were then plotted on a graph and the R squared value and the 95% confidence interval are also found to better see the relationship between each data point.

Excluding the data from 1993, the specificity of age group 1 has stayed fairly constant with the 95% confidence intervals all overlapping. This same trend can be seen in age groups 2, 3, and 4. The majority of the specificities for each age group are over .94 indicating that the Copenhagen Mammography Screening Program is accurately identifying individuals that are negative for breast cancer as negative around 94% of the time overall.

Other members of my group calculated the sensitivity and false positive and negative rates. A chi squared analysis was then done to see if the values we calculated fit a distribution that could be expected.

A p-value of less than .05 indicates a significant result and that the values that are found are not due to chance. From this analysis we can see that as a woman ages she is more likely to be correctly diagnoses.

This research has had two main findings. First, within each age group, there is little change in the average specificity regardless of the year being looked at. Second, when comparing the effectiveness of the screening program to accurately diagnose individuals with breast cancer as positive and those without as negative, as the age group increases so does the accuracy of the screenings.

To learn more about this research or other projects that have been done on this topic look at studies done by my research advisor My von Euler-Chelpin