A trial looking at Computer Aided Detection (CAD) to find out if it could be used in the NHS breast screening programme (CADET 2)

Cancer type:

Breast cancer

Status:

Results

Phase:

Phase 3

This trial looked at the use of the Computer Aided Detection (CAD) in the NHS breast screening programme. This trial was supported by Cancer Research UK.

Most mammograms in the NHS breast screening programme are checked by two specialists in reading mammograms. It is more accurate if two specialists examine the mammogram, rather than one. But there is already a shortage of specialist readers and their workload is increasing.

A possible solution is to use a CAD system. An earlier study (called CADET I) looked back at mammograms taken 10 years ago. The mammograms were re-examined by one specialist reader, using CAD. The results showed that a single specialist reader using CAD is just as good as two specialist readers.

This trial looked at the use of CAD for women who were having mammograms at the time. The researchers wanted to find out if CAD is just as good in the real life setting of the screening centre.

Summary of results

The research team found that it was just as accurate for mammograms to be read by one specialist reader using CAD as it was by two specialist readers.

The trial recruited over 30,000 women who were having mammograms.

  • 1,182 women had their mammograms read by two specialist readers
  • 1,152 women had their mammograms read by one specialist reader using CAD
  • 28,204 women had their mammograms read by both two specialist readers and by one specialist reader using CAD

The people reading the mammograms were all specialist readers who read at least 5,000 mammograms every year and do a yearly self assessment, as suggested in the guidelines.

The research team looked at the number of cancers that were found, and at the number of women who were recalled for further tests.

They found that very a similar number of cancers were detected. Two specialist readers detected 199 out of 227 cancers (87.7%) in the 28,204 people in group 3. And one specialist reader using CAD detected 198 out of those 227 cancers (87.2%).

A few more of the women who had their mammograms read by one reader using CAD were recalled for more tests. This was 39 per 1,000 women, compared to 34 per 1,000 women who had their mammograms read by two readers.

The research team concluded that a single expert using CAD could detect cancer in mammograms as reliably as two experts. But they point out that the decision on whether to start using CAD for routine mammograms lies with the NHS Breast Cancer Screening Programme.

We have based this summary on information from the team who ran the trial. The information they sent us has been reviewed by independent specialists (peer reviewed) and published in a medical journal. The figures we quote above were provided by the trial team. We have not analysed the data ourselves.

Recruitment start:

Recruitment end:

How to join a clinical trial

Please note: In order to join a trial you will need to discuss it with your doctor, unless otherwise specified.

Please note - unless we state otherwise in the summary, you need to talk to your doctor about joining a trial.

Chief Investigator

Professor F. Gilbert

Supported by

Cancer Research UK
National Institute for Health Research Cancer Research Network (NCRN)

Other information

This is Cancer Research UK trial number CRUK/06/049.

Questions about cancer? Contact our information nurses

Freephone 0808 800 4040

Last review date

CRUK internal database number:

Oracle 845

Please note - unless we state otherwise in the summary, you need to talk to your doctor about joining a trial.

Deborah wanted to help other breast cancer patients in the future

A picture of Deborah

“Deborah agreed to take part in a trial as she was keen to help other cancer patients in the future. "If taking part in a trial means others might be helped then I’m very happy with that."

Last reviewed:

Rate this page:

No votes yet
Thank you!
We've recently made some changes to the site, tell us what you think

Share this page