Open Access Open Badges Study protocol

Changing case Order to Optimise patterns of Performance in mammography Screening (CO-OPS): study protocol for a randomized controlled trial

Sian Taylor-Phillips1*, Matthew G Wallis2, Helen Parsons1, Janet Dunn1, Nigel Stallard1, Helen Campbell3, Sarah Sellars4, Ala Szczepura1, Simon Gates1 and Aileen Clarke1

Author Affiliations

1 Warwick Medical School, The University of Warwick, Coventry CV4 7AL, UK

2 Cambridge Breast Unit and NIHR Cambridge Biomedical Research Centre, Box 97, Cambridge University Hospitals NHS Foundation Trust, Biomedical Campus, Hills Road, Cambridge CB2 0QQ, UK

3 Health Economics Research Centre, Nuffield Department of Population Health, University of Oxford, Old Road Campus, Headington, Oxford OX3 7LF, UK

4 NHS Cancer Screening Programmes, Public Health England, Fulwood House, Old Fulwood Road, Sheffield S10 3TH, UK

For all author emails, please log on.

Trials 2014, 15:17  doi:10.1186/1745-6215-15-17

Published: 10 January 2014



X-ray mammography remains the predominant test for screening for breast cancer, with the aim of reducing breast cancer mortality. In the English NHS Breast Screening Programme each woman’s mammograms are examined separately by two expert readers. The two readers read each batch in the same order and each indicates if there should be recall for further tests. This is a highly skilled, pressurised, repetitive and frequently intellectually unchallenging activity where readers examine one or more batches of 30–50 women’s mammograms in each session. A vigilance decrement or performance decrease over time has been observed in similar repetitive visual tasks such as radar operation.


The CO-OPS study is a pragmatic, multi-centre, two-arm, double blind cluster randomised controlled trial of a computer software intervention designed to reduce the effects of a vigilance decrement in breast cancer screening. The unit of randomisation is the batch. Intervention batches will be examined in the opposite order by the two readers (one forwards, one backwards). Control batches will be read in the same order as one another, as is current standard practice. The hypothesis is that cancer detection rates will be higher in the intervention group because each readers’ peak performance will occur when examining different women’s mammograms. The trial will take place in 44 English breast screening centres for 1 year and 4 months. The primary outcome is cancer detection rate, which will be extracted from computer records after 1 year of the trial. The secondary outcomes include rate of disagreement between readers (a more statistically powerful surrogate for cancer detection rate), recall rate, positive predictive value, and interval cancer rate (cancers found between screening rounds which will be measured three years after the end of the trial).


This is the first trial of an intervention to ameliorate a vigilance decrement in breast cancer screening.

Trial registration

ISRCTN46603370 (submitted: 24 October 2012, date of registration: 26 March 2013).

Breast cancer; Fatigue; Mammography; Observer performance; Screening; Vigilance decrement