New Numbers Should Factor Into the Mammography Equation
On Friday the New York Times reported that surgeons are performing far too many open breast biopsies to evaluate abnormal mammogram results. A new American Journal of Surgery article analyzed data for 172,342 outpatient breast biopsies in the state of Florida. The main finding is that between 2003 and 2008, surgeons performed open biopsies in an operating room – as opposed to less invasive, safer biopsies with needles – in 30 percent of women with abnormal breast images.
I was truly surprised by this should-be outdated statistic, which further tips the mammography math equation in favor or screening. These numbers matter, and should be based in modern medical practice.
When the Annals of Internal Medicine published the since-adjusted recommendations for breast cancer screening by mammography in November 2009, the stated considerations were not about dollars and cents – which were incalculable – but about the number of women needed to be screened to save one life, and the incidence of false positives which cause harm – worrying, needless biopsies, complications of procedures, overtreatment, etc.
In the context of the health care reform discussion, and considering our country’s out-of-the-sky-and-rising medical bills, some (hopefully) well-intentioned economists heard about those trumped-up mammography papers and concluded that we shouldn’t screen women under 50 for breast cancer because it’s harmful and, what’s more, we can’t keep paying for this sort of care because it’s not evidence-based.
Those conclusions were flawed, though, because the data in those papers were old, as I’ve written previously, and didn’t include studies of digital mammography – which is better for detecting cancer in younger women who tend to have denser breast tissue. In December 2009, I noted that it was unreasonable to consider the costs of open needle biopsies in O.R.’s in any calculation of the harms of mammography, as had the Annals authors, because those kinds of procedures are outdated, or so I thought they were.
It turns out I’ve been living, still, in an academic medical enclave. According to the Times‘ coverage by Denise Grady:
The reason for the overuse of open biopsies is not known. Researchers say the problem may occur because not all doctors keep up with medical advances and guidelines. But they also say that some surgeons keep doing open biopsies because needle biopsies are usually performed by radiologists. The surgeon would have to refer the patient to a radiologist, and lose the biopsy fee…
The Times article suggests this pattern of over-doing open-biopsies, as documented in Florida, likely reflects national tendencies, including variation among different types of practices – academic, hospital-based, etc.
According to the article published in the American Journal of Surgery, the costs of a core needle biopsy using imaging guidance is around $5,000, or – if a vacuum biopsy device is used, around $6,000; the costs of an open procedure in the O.R. run in the range of $11,000 or more. The Times article indicates that doctors’ fees for a needle procedure range from $750 to $1500, and for an open, surgical biopsy from $1,500 to $2,500. For a ballpark estimate of the cost difference, say a core needle procedure is $5,500 + $1,000 for the doctor’s fees – that’s ~$6,500; a surgical procedure is $11,000 + $2,000 for the surgeon’s fees – that’s $13,000, an easy double.
So let’s say, for the sake of future calculations on mammography, that 10 percent of breast biopsies really do need to take place in the O.R. (which is a generous over-estimate, I think it should be 5 percent or fewer). But if 10 percent need be in the O.R.: then 20 percent of breast biopsies in the U.S. each year – said in the surgery paper to be 1.6 million per year in the U.S. – are being performed through an unnecessary, costlier technique.
An extra $6,500 x 20 percent of 1.6 million procedures = $2.08 billion additional costs, per year.
Let’s call it an even $2.1 billion, or $2 billion, we should shave off the collective amount we spend on mammography and appropriate follow-up. The last digit doesn’t matter; these are huge numbers. No wonder the Times put this story on the front page.
These results should be factored into any proper calculation of costs in breast cancer screening. Now add (or better, subtract) the implications of the findings of two weeks ago – that full lymph node dissection is usually not necessary in women, even if the sentinel node is found to be positive at the time of definitive surgery for what turns out to be a cancer.
What needs be reassessed by public health specialists and economists who weigh in on these issues – and please help me out here, Task Force members and Dartmouth friends, if you would, because your input affects public thinking and, ultimately, policy – are the legitimate costs of screening (every other year, as opposed to annually), doing needle biopsy procedures (instead of open biopsies) and reducing the costs and long-term complications of surgery by eliminating routine lymph node dissection from the equation.
And then we should assess those numbers relative to the costs of treating a woman with metastatic breast cancer, which still has not yet been determined.