Read Original Story

Study finds computers hinder accuracy in mammography


5 Star

Study finds computers hinder accuracy in mammography

Our Review Summary

This story reports on the first large scale analysis of whether computer aided detection (CAD) improves the interpretation of mammograms.  As was reported in the story, the study found that the use of CAD actually reduced the accuracy of the breast cancer screening.

The story included technical detail about how CAD is used and made important points about the real value of experienced mammographers. 

It covered costs – the additional cost of computer detection and of the machines.   

And it turned to several sources – some involved in the study and some not. 

The story might have helped women with a few more questions.  What should a woman do with this information? Should she go to a facility that uses computer-aided detection? Should she ask that it not be used? Are there other things she should do to make sure that her screening is a high quality as possible?

Overall, though, this story was well done, and in only 638 words.  



Does the story adequately discuss the costs of the intervention?


While the story did not include the overall cost for mammography, it did mention that the use of computer aided detection adds about $20 to the cost of each mammogram.  It also mentioned that this charge is covered by Medicare and most insurers. And it discussed the investment many medical centers have made in the machines – $125,000 – 175,000.

Does the story adequately quantify the benefits of the treatment/test/product/procedure?


The benefits of the technology discussed (or lack of benefit) was reported in absolute number of tumors detected.

Does the story adequately explain/quantify the harms of the intervention?


The harm of treatment – undergoing unecessary biopsy was mentioned.  However – the increased number of biopsies was presented as a relative increase rather than an absolute value.

The story might have mentioned that the use of CAD increased the number of cases of ductal carcinoma in situ or DCIS, for which progression to invasive cancer is less certain – meaning that it may not pose a health risk during the woman's lifetime.  The benefit of cancer treatment for women with this diagnosis is therefore not clear.  Thus enhanced ability to detect DCIS may not affect death from breast cancer. 

Nonetheless, we'll give the story the benefit of the doubt on this criterion.  

Does the story seem to grasp the quality of the evidence?


The evidence that breast cancer detection rates did not differ was presented in terms of the absolute number of breast tumors detected.

The story included the source of the study it was reporting.

Does the story commit disease-mongering?


The story did not engage in disease mongering. 

Does the story use independent sources and identify conflicts of interest?


Multiple sources were used, some directly involved in the study and some not.

Does the story compare the new approach with existing alternatives?

Not Satisfactory

The story described one option to computer-aided detection of mammograms used at one center.  It mentioned one breast center in which every mammogram is read by two radiologists, which was described as more effective at finding cancers than a computer system. It also emphasized the value of experienced radiologists reading conventional mammograms.

Does the story establish the availability of the treatment/test/product/procedure?


The story mentioned that about 25% of mammography centers in the US use computer aided detection.

Does the story establish the true novelty of the approach?


The story did not mislead on the novelty (or lack thereof) of this technology, explaining that computer aided decition is used in about 25% of mammography centers. 

Does the story appear to rely solely or largely on a news release?


Does not appear to rely on a press release.

Total Score: 9 of 10 Satisfactory


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.