The relationship between real life breast screening and an annual self assessment scheme

Certain performance values arising from mammographic test set readings correlate well with clinical audit

“…Clinical audit, however, faces limitations such as: the inability to define actual total number of cancer or normal cases and hence it is unable to accurately determine the sensitivity and specificity score; and the low incidence rate of breast cancer resulting in a prolonged period of time of several years required to gather meaningful data on the performance of individual screen readers . To overcome these limitations, screening test sets such as Breastscreen REader Assessment STrategy (BREAST) and Personal Performance in Mammographic Screening (PERFORMS) were introduced as a supplementary assessment of reader performance.…”

Maintaining quality in the UK breast screening program

“…Consequently, it is perfectly reasonable to make such an extrapolation. This is supported by studies that show that, whilst the scheme is not real life screening, participants' scheme performance does relate to their real life screening performance 25,26,27 . Additionally, the performance of radiologists and advanced practitioners has been shown to be similar when undertaking the scheme 28 which further supports the role development of technologists to interpret screening mammograms.…”

Correlation Between Screening Mammography Interpretive Performance on a Test Set and Performance in Clinical Practice

“…This is in contrast to Rutter and Taplin, 9 who found moderate correlations for specificity but no evidence of correlation between clinical and test-set sensitivity. Scott et al 8 also found a significant, moderate correlation between clinical specificity and the “correct return to screen percentage” on the test set, but no significant correlation for their measure of sensitivity (correct recall percentage); however, they found significant correlations for other performance measures including positive predictive value, false-negative rate, and cancer detection rate. Taken together, these studies suggest that test sets may reflect clinical practice, but only moderately.…”

“…In contrast, Soh et al 7 found significant, moderate correlations of 0.30-0.57 between several clinical audit measures and two test set measures (location sensitivity and Jackknifing free response operating characteristic figure-of-merit) of 60 cases (20 with cancer) read by 20 radiologists, but no correlation with test set specificity. Similarly, Scott et al 8 found significant, moderate correlations of 0.29-0.41 between several performance measures on the PERFORMs test set and clinical performance among 39 readers in the UK. None of these prior studies evaluated the influence of breast cancer prevalence or lesion difficulty on the strength of the correlations.…”