Background and objective – High-quality mammography reading is a key factor for the benefit of breast cancer screening programmes and continued training in mammography reading is recommended. However, information about the effect of specific breast teaching interventions on the interpretative performance of radiologists is scarce. The purpose of this study is to evaluate the effect of a 3-day breast mammography teaching course on interpretative performance of radiologists within screening programmes in Switzerland.
Methods – Amongst all radiologists (n=77) who attended a 3-day breast cancer mammography teaching course given by two world experts in 2012 in Lausanne, 16 were accredited by one of the three Swiss regional breast cancer screening programmes of interest. Anonymised records of mammography readings within these programmes were linked to the course participation status of radiologists. Recall, false-positive and breast cancer detection rates were computed and compared for participating and non-participating radiologists per quarter between 2010 and 2015.
Results – Eighty-one accredited radiologists were included in the non-participants group and 16 in the participants group. Decrease in recall and false-positive rates were observed after the course for both groups (quarterly percent change QPC) of - 1.1% [95%CI: -1.6%; - 0.6%] and – 1.2% [-1.8%; -0.7%], respectively), with no significant change in trend before the course. Among course participants, the recall rate decreased markedly for first readers for about 2 years (QPC: -8.4% [95% CI: -13.8%; -2.7%]) and for dense breasts during at least 2.5 years after the teaching course (QPC: -3.8% [95% CI: -6.3%; -1.3%]). The corresponding decreases in recall rates among non-participants were more modest and, for dense breasts, could not be associated with the timing of the course.
Conclusion – Results suggest a likely benefit of a 3-day breast teaching course in terms of reduction in recall and false-positive rates in daily screening programme practice for at least 2 years when interpreting mammograms of dense or extremely dense breasts and for less- experienced radiologists. The low attendance rate of programme-accredited radiologists, particularly of less experienced readers, might underestimate the impact of this intervention.