Computer-aided detection ("CAD") technology, which has been around since its approval by the FDA more than a decade ago, is widely used by radiologists performing mammograms as an assist for detecting breast cancer.

There's just one problem with it, according to researchers who have studied the results of 1.6 million CAD-assisted mammograms from scores of medical facilities across the country: They say it doesn't work and, in fact, increases the likelihood of a misdiagnosis being made.

That assertion is truly big news in the medical community, given how widely used CAD technology is. It is also especially troublesome because it results in doctors sometimes recommending additional procedures -- even surgeries -- that are wholly unnecessary and can produce bad patient outcomes.

The study appears in the Journal of the National Cancer Institute and follows up an earlier study that concluded the same thing, namely, that radiologists find tumors at the same stage and size without or without the assistance of CAD software.

That would result in a wash for the technology, except for the troubling implications of the false readings that result in occasional misdiagnoses and improper treatment recommendations. Study researchers say that using CAD increases false positive readings by about half of a percentage point, from 8.1 percent to 8.6 percent.

An accompanying editorial to the study states that economic incentives -- such as using CAD as an additional layer of liability insulation in the event of medical malpractice litigation -- "may stoke its continued proliferation."

Related Resource: Los Angeles Times, "Mammograms: Computer-aided detection doesn't help" July 27, 2011