Past and Present Statistics on Public Health and Breast Cancer
I've been reading and listening to the debate on the new screening guidelines from the U.S. Preventive Services Task Force for almost a week now and trying to sort through the various claims.
Two historical figures who have fascinated me for years have stories that seem pertinent to the debate. The first is Dr. Ignaz Semmelweiss, a Hungarian-born doctor who worked on the maternity ward of a hospital in Vienna in the 1840's. Semmelweiss noticed that the mortality rate from childbed fever was three times higher on one ward than another. In those days people didn't think in terms of public health statistics, but Semmelweiss wanted to figure out the difference in mortality. In one ward, midwives took care of the patients; in the other doctors and medical students were in charge. Unfortunately it was the latter ward that had the higher mortality rate.
What if the doctors, who came to do their rounds straight from the autopsy room, were bringing in some sort of agent that was causing the problem? Semmelweiss tested his hypothesis by washing his hands before he examined patients, and he made the medical students who worked under him do the same. The mortality rate dropped dramatically!
Semmelweiss thought other doctors would want to follow his lead, but his findings flew in the face of established medical theory and practice; and Semmelweiss was essentially laughed out of Vienna for his crazy ideas. He went back to Hungary, where he was able to repeat the mortality reduction for post-partum women, but where he couldn't convince any of his colleagues that he was right.
In the 1850's, Florence Nightingale, working in a military hospital in Turkey during the Crimean War, effected a dramatic drop in mortality by keeping wards clean. Nightingale wondered if a connection existed between proper sewers and sanitation and healthier soldiers. What if the army made sanitation as important as marching? Folks laughed at her too. However, she refused to let the condescension of military men towards a meddling spinster with strange notions silence her. She spent the rest of her life pushing for proper cleanliness and sanitation in volumes of reports filled with charts of statistics and mortality tables. To help the officials understand the statistics, she invented the pie chart that shows percentages.
Which brings us to November 2009. All week long people have been arguing statistics and what they mean in terms of public policy.
The U.S. Preventive Services Task Force (USPSTF), organized in 1984, has a mission, "to evaluate the benefits of individual services based on age, gender, and risk factors for disease; make recommendations about which preventive services should be incorporated routinely into primary medical care and for which populations; and identify a research agenda for clinical preventive care." In other words they crunch numbers to decide what works best.
Having a group charged to do this is a good idea. As a matter of public policy, we want to put our time and money where they will do the most good. What if a practice we have followed turns out to be ineffective? What if a simple treatment or test exists to improve public health?
The task force decided that the risks outweighed the benefits of routine screening of women with no known risk factors between the ages of 40 and 49. And then the cries went up. Almost everyone knows a woman in that age group who had a cancer detected by a mammogram, and almost everyone considers that woman's life worth any number of false positives and unnecessary biopsies.
These stories are what statisticians call anecdotal evidence. There is always someone who doesn't fit the pattern who calls the recommendation into question. The doctors who worked with Semmelweiss thought they proved his theory wrong when they cited anecdotal evidence of women who died of childbirth fever in the midwives' ward.
Most women whose cancers have been discovered by mammography feel routine screening saved their lives. For years we have believed that if we could only catch every cancer early, we could stop cancer deaths. Now evidence seems to be mounting that it is more complicated than that. Many of the cancers we are catching early with mammography are slow-growing ones which might have never threatened a woman's life if we didn't know about them.
Our reliance on mammograms to find lumps also leads to a false sense of security. Here's my anecdotal evidence. I had a routine mammogram that found a lump in August of 1997. It turned out to be a cyst, which was drained when it became painful. In December of 1997, the radiologist did another mammogram to make sure that the cyst hadn't been hiding a cancer. In February of 1998 when my breast started hurting and itching, I delayed seeing a doctor at least in part because I was confident that my two recent mammograms would have found cancer if it had been there.
The mammogram I had in March of 1998 did not find a lump, but it turned out that I had inflammatory breast cancer, an aggressive form that often doesn't show on a mammogram. At the time of my surgery, there were two tumors on my chest wall that had probably been there when I had all three previous mammograms, but they couldn't be seen because of their position on my chest wall.
Clearly, the task force is correct that mammography is not a perfect instrument, especially in younger women. However, I think their conclusions miss several important points.
First, the panel overestimated the harm from mammograms when they counted the anxiety women face when they need a biopsy or extra tests as a risk. Every woman I know would much rather be over-treated than have a deadly cancer missed.
Second, while evidence is mounting that finding cancer earlier isn't saving lives the way scientists thought it would, finding cancer earlier does save breasts. I didn't see anything in the panel's report about the value of finding cancer early enough for a woman to have a lumpectomy versus a mastectomy.
Third, while mammograms and breast self-exams are flawed instruments, they are the best we have right now.
Florence Nightingale was so convinced by her pie charts and statistics that when researchers like Louis Pasteur proved that bacteria could cause illness, she refused to believe in the germ theory and died in 1910 convinced that it was wrong.
We women brought up on the importance of mammograms find it hard to let go of our security blanket even as science is finding some holes in it.
Here are some more "what if" questions:
What if we use some of the resources that we currently use for routine mammography for regular physicals with doctors? What if those exams find not only breast cancers, but diabetes, heart problems, and other underlying health risks? Wouldn't our overall public health be better?
What if we spend more time emphasizing to women that they need to check with their doctor for any change in their breasts? Would we catch aggressive cancers like IBC that often has no lump and fast-growing lumps that pop up between mammograms? Those cancers are the ones that are killing people.
What if doctors stop focusing on breast cancer as an older woman's disease and become proactive in diagnosing breast cancer symptoms in younger women? What if we recognize that mammograms don't work all that well on younger women and make sure that they are able to get additional tests like MRI's and biopsies if they show symptoms?
People are asking me if I agree with the task force findings. It makes sense to me to do mammograms less frequently if women are also cautioned to see their doctors for any breast changes between mammograms. An aggressive cancer is just as likely to show up four months after a mammogram as fourteen months after. Every twelve months has always been an arbitrary figure. Despite the problems with mammograms in younger women, I would still urge women to get a base-line mammogram at age forty and at least every two or three years thereafter depending on her risk factors.
One excellent result of the task force's report is that people are talking about mammograms and how to reduce breast cancer deaths. What if women took responsibility for their health and called their doctor when they had symptoms? What if we made good quality health care accessible to everyone?