Search This Blog

Follow by Email

Sunday, November 25, 2012

Mammogram screening--reconsidering the wisdom of saying "No."

Three days ago, on November 22, 2012, an article was published in the New England Journal of Medicine questioning the utility of mammogram screening for prevention of death from breast cancer. The authors were research professor Archie Bleyer MD at Oregon Health and Sciences University in Portland, an oncologist who was chief of pediatrics at MD Anderson Cancer Center and H. Gilbert Welch MD, MPH, a professor at Dartmouth Medical School.  The article examines the ability of mammograms to prevent late stage breast cancer by diagnosing and treating breast cancer early as a result of detection by mammograms. They found that mammograms do detect lots of breast cancer, but when we compare women during the years 2006-2008 when mammogram screening was widely practiced to women during the years 1976-1978, there was no difference in the incidence of the really nasty breast cancers, ones that had spread beyond the regional lymph nodes, and only a small decrease in the less nasty but still significant regionally metastatic cancers. Many, many women were treated for breast cancer but only a few were saved from dying of late stage disease with mammogram screening.

Their evaluation was carefully done, making assumptions about the increase in rates of breast cancer over time that were designed to make the results of having a mammogram more favorable. They concluded, somewhat generously, that " Women should recognize that our study does not answer the question “Should I be screened for breast cancer?” However, they can rest assured that the question has more than one right answer."

Mammograms are x-rays of breast tissue and were first introduced by a German surgeon named Albert Salomon in 1913 when he examined mastectomy specimens with x-rays. Mammograms were used rarely before 1978 when widespread use of mammography was introduced in an attempt to identify and treat breast cancer early to reduce mortality and morbidity. The procedure has been controversial since 1978. The first objections regarded the danger of radiation to the breast. A mammogram is performed by squashing a breast between two plates and passing x-rays through it. A digital mammogram detects the x-rays with digital detectors and creates an image on a computer monitor. A film mammogram creates an actual negative on a piece of photographic film. These techniques are equally sensitive but digital machines are replacing film machines due to the overall convenience of storage and communication of images. The dose of x-rays with a typical digital mammogram is 3.9 milligrays, the same as for a film mammogram, and about 4 times the radiation dose of a chest x-ray. It is a small dose and is probably not a significant danger.

After initial worry about radiation, very reasonable concerns remained about the quality of mammograms and mammogram readings. In 1992 the Mammography Quality Standards Act was passed to assure that facilities that performed mammograms were accredited by the FDA to be of adequate quality. 

Mammograms are very big business. I can't find out how big, but if there are about 40 million women between the ages of 50 and 75 and half of those get mammograms at a cost of $100, that would be 4 billion dollars spent on mammograms alone, not to mention repeat mammograms and other technology to further identify actual breast cancer. If I am off by a factor of 4, that's still 1 billion dollars. It is a big deal to write an article questioning the utility of this test. Any move away from a recommendation of yearly mammograms starting at age 40 or 50 is met with outrage. Still, articles and studies continue to demonstrate that the benefit of screening mammography is limited.

Norway has been extensively studied with regard to mammogram screening because they keep excellent records and have had a staggered approach to universal mammogram screening for women. Over the last 20 years multiple studies have come out of Norway suggesting that breast cancer is overdiagnosed by mammogram. Overdiagnosis is what happens when we find a breast cancer that would never have caused harm if it had not been detected. Some cancers do not kill people and are probably taken care of by a healthy immune system. In an article published last April, a study group evaluated the data and estimated that, of all the breast cancers diagnosed, about a quarter of them would never have caused harm. This week's article concurred, but suggested that number may be as many as 30% and that over a million women have been diagnosed with breast cancer since the inception of screening who would never have been affected had they not had mammograms. 

But, one might ask, is it really a big deal to be diagnosed with breast cancer that would not have hurt or killed you? Yes. It is actually a very big deal. The British Medical Journal reported that 50% of women were depressed in the year after they were diagnosed with early breast cancer, the type of breast cancer most likely to be overdiagnosed. But we don't really need studies to tell us this information. It is clear after treating women with breast cancer that they are profoundly affected by this diagnosis, feeling ugly, self conscious, maimed. Treatment complications include disfigurement, chemotherapy side effects, infection, chronic pain, to say nothing of astronomical monetary costs. 

Still, breast cancer kills, and fewer people die of it now that mammogram screening has become standard. How do we explain this? Some of it is due to mammograms. We are detecting lots of breast cancer early. Some of it we should be detecting early and some we should not. Still, some early breast cancers would have become late, bad, killer breast cancers. The study published this week suggests that there are not many of these, but there are definitely some. The treatment of breast cancer has also gotten much better. People who used to die of metastatic breast cancer now remain on chemotherapy and remain in remission for many years, and some are cured. It is undoubtedly because of the huge increase in breast cancer awareness that therapy has improved, and likely the million women who were overdiagnosed were important in helping to pioneer excellent treatment. 

What is a woman to do? What is a doctor to do? I think it might be good to start with recognizing that a decision not to do mammogram screening is not tantamount to a death wish. We give women who reject the recommendation for regular mammograms a really hard time, and that is neither fair nor evidence-based. Mammograms are quite good for evaluating lumps, especially in older women. They are also good for giving peace of mind, since a negative mammogram suggests (but does not prove) that a woman does not have breast cancer. There may be a subset of women, those at particularly high risk of cancer for instance, who would be very wise to have regular mammograms. There may be technology that can help us identify which breast cancers need treatment and which do not. Tests that detect more breast cancers, such as MRI and PET scanning may not be particularly helpful in this situation unless they can reassure us that some breast cancers are of no significance. If we accept that medical resources should be limited, we might look at places where money now used for universal mammogram screening of women might be more effectively spent.

Thursday, November 15, 2012

Creating dependency--is that what we do for a living?

Lately, it seems, I have been treating quite a lot of people who end up in the hospital as a result of prescription drug abuse. Most of them have chronic pain and have been generously prescribed long acting opiate medications such as methadone and morphine by doctors of various types, have taken excessive numbers of these medications or mixed them with other medications and have ended up being unable to breathe for themselves.

In the beginning of the last decade there was a well intentioned movement to recognize that pain was a real issue and should be treated. Pain is not visible, usually, and can often be ignored. Having lots of pain for a long while or intense pain for shorter periods is bad for us. It causes depression, anxiety, leads to post-traumatic stress disorder, and just generally hurts a lot. Humans view torture, deliberately causing another being to have pain, as vile and unacceptable. Conversely we regard the relief of pain as a great gift. In 1999 the Veterans Administration started a campaign to make a rating of pain the "5th vital sign." That means that, in addition to  measuring blood pressure, pulse, respirations and temperature, nurses were encouraged to rate a person's pain on a scale of 1-10. This tended to cause doctors to be more aware of their patients' discomfort and perhaps to offer pain medications. As a doctor, I always find a pain scale difficult to interpret since, unlike blood pressure, it is very subjective. Not everybody is a number person and not everybody has experienced a truly redline pain level on which to base their 1-10 scale. Also some people recognize that stating a pain number above 5 will result in getting really delightful injectable opiate medications. Opiates are not delightful for everyone, in fact some people absolutely hate them, throw up, hallucinate, get hideously constipated, feel out of control. Still. Some people feel SO much better with opiates. They feel warm inside and nothing bothers them. Anxiety goes away and the world becomes a beautiful place.

Since pain became a major focus in medical care, the use of really potent opiates has increased nearly exponentially, and along with that so has prescription pain medication abuse, overdose and unintentional death. Between 1997 and 2007 prescription of opioids increased more than 600%, and that increase has (as far as I can tell) not slowed significantly.  In 2007, 27,000 people died of unintentional drug overdose, the majority of that due to prescribed medications. In our kindness and empathy we as physicians offer increasing doses of long and short acting pain medications that kill 10s of thousands of people and, possibly more tragic, make hundreds of thousands more just-on-the-edge-of-high-all-the-time people dependent upon us for refills. I have been such a physician at times in my many years of primary practice. I would see a patient, know that they had a legitimate reason for opiate pain medications, definitely they hurt and had been taking these pills for years with some improvement, and refilled or even increased the intensity of the medications prescribed. And I was practicing in a way that was encouraged by experts in the field of internal medicine. If a person needed opiates regularly, they should be on long acting ones, and should have short acting ones for breakthrough pain. To offer less was unkind and old fashioned. Sometimes this approach made my patients more functional but surprisingly often there was no dose of opiates at which the patient was truly happy, functional or their pain was actually tolerable. Occasionally, after trying everything I could think of, opiates, anti-anxiety medications, anti-depressants and anticonvulsants I would be forced to admit defeat and let my patient know that they needed to find help elsewhere because I could no longer ethically offer them refills of pills that I knew were not good for them and which were evidently not actually helping. I would offer these patients supervised tapering of the medications, but it was very rare that they were willing to actually get off of the medications completely.

And even with all of this said, I believe that the products of the opium poppy, in all of their glorious diversity, are some of the greatest remedies that the field of medicine has to offer. For post-operative pain, for a toothache, for a kidney stone, for cancer pain there is just nothing like an opiate. Codeine, a low potency opiate, is one of the most useful drugs I can prescribe, relieving stomach cramps, stopping diarrhea, quieting a cough and gentling a headache. There are pains that have no visible cause but are just as intense as a drill press, and opiates can make these tolerable, sometimes. But, as I remember  it, the world of 1997 did not include more people who were miserable than the world of 2007 when there were 600% more opiates prescribed. Regular use of opiates changes people. This is not something I know because I read a study, but because I have treated so many of them. They are less motivated, more foggy. They are tied to the schedule of my office because they feel very unwell when they don't get their medications. Their memories are worse. Their bowels are usually slow, which becomes an issue in itself. Occasionally they overdose and die. Many patients prescribed opiates share them or sell them which leads to dependency in a whole host of other people.

There have been many government and medical profession based responses to this problem, including closer observation of our prescribing habits, efforts to keep patients from getting huge numbers of pills by seeing multiple physicians, pain medication contracts to make explicit the risks of taking the medications and our expectations of the patient for whom we prescribe them. Pain medication prescribing has become big business and many areas have clinics who advertise themselves as treating chronic pain, but really just write prescriptions for patients who are willing to pay for cursory visits. These clinics are being scrutinized and shut down in many places. All of these are good ideas, but I am seeing no decline in the number of patients who are on opiates and who show up with grave consequences of opiate use and abuse.

The problem is that prescription pain med dependence is just the tip of an iceberg of overall increasing dependency on the medical profession due to inappropriate use of medical technology. Physicians generally have ethical motivations, but regardless of our (mostly) good hearts, it benefits us economically to make sure that our patients continue to need us. We have increasing numbers of patients on hemodialysis, whose kidneys have failed. They absolutely require several hours of artificial filtering of their blood at least 3 times a week at the cost of close to $100,000 a year. They are some of the sickest patients we see in the hospital because hemodialysis is not as good as having functional kidneys and they are at least a little bit sick all the time. We do increasing numbers of orthopedic procedures which often are associated with complications including life threatening infections and failure of hardware. These procedures can be wonderfully effective in the right patients, but they are becoming more standard for patients who are at extremely high risk of complications. Our intensive care is more and more effective at making the very sick survive, but they are not, thereafter, well. They have continuing needs for multiple medication an procedures.

Most people don't like dependency. I have heard people say for years that what they dread most is becoming a burden on others. Often what we offer people is the opportunity to become a burden. When we discuss the risks of procedures of medications, I think it should be standard to bring the concept of increasing dependency into the conversation.

Sunday, November 4, 2012

Hospitalists and the field of Hospital Medicine: why we are sometimes terrible and how we can be excellent

Internal Medicine is the branch of medicine that deals with diseases of the internal organs in adults. It also involves dermatology, minor surgical procedures, general psychiatry and preventive care of well people. It is an excellent field, full of opportunities to think and feel and connect with people, mysteries to be solved and an endless variety of stuff to be learned. Internal Medicine contains the subspecialties of nephrology (kidneys), cardiology, oncology and hematology (cancer and blood), infectious diseases, pulmonary and critical care medicine, endocrinology (glands), rheumatology (joints), gastroenterology (guts and livers), neurology and hospital medicine. The most recently invented of those subspecialties is hospital medicine. Unlike the rest of the subspecialties, hospital medicine is defined by the place it is performed, not the body system it aims to treat.

Hospitalists (the internists who practice hospital medicine) take care of patients who are admitted to hospitals when their own doctors do not. Primary care doctors are less and less often involved in taking care of hospitalized patients because they are so busy taking care of patients in their offices that taking the unscheduled time to go to a hospital as well has become impractical. When there is a doctor at the hospital to take care of a patient who comes in very ill, there is no delay in getting appropriate care and the primary care doctor doesn't have to cancel a waiting room full of patients in order to come in. Hospitalists get really good at taking care of the illnesses that are severe enough to lead to hospitalizations. Conversely, primary care doctors forget how to take care of these problems. Hospitalists are paid to be in the hospital at night, so it is no hardship to be at the bedside in the wee hours when people so often  decompensate. Doctoring in the past was more often a profession that was not compatible with having good family time or enjoyable hobbies. Separating office work from hospital work makes it much easier to be a reliable spouse or friend.

On the other hand...The absolutely best care a person could get would be delivered by a competent physician who had known a patient for years, with appropriate input from other physicians depending on the patient's specific needs. As a primary care doctor, when my patients were admitted to the hospital, I knew what had happened that lead up to their illness, what we had tried before, how the person interacted with their family, what their values were. The patient also knew me, trusted me, and felt better in the hospital with its strange smells and routines just seeing my familiar face.

Hospitalists are shift workers, usually working 12-13 hour days, 7 days on, 7 days off, and they are randomly assigned patients as they come in to the hospital. They usually see 10-20 patients a day, discharging and admitting patients to the hospital and coordinating their care. They have access to the patient's computerized and sometimes paper charts, can call their primary care physicians for more information (if they can get them on the phone) and can often take a pretty good history from the actual patient. There is not a whole lot of time for all of this, though, and with up to 20 of more unfamiliar faces in a day, it is pretty difficult to have the kind of intimate knowledge of a patient that would lead to optimal care. Although most doctors are internally driven to do the right thing for every patient they see, it is more work to take a good history than to just go with what the emergency room doctor who originally saw the patient said, it is more work to discharge a patient than let them stay in the hospital one more unnecessary day, and generally just more work to do the job right. Because the hospitalist has to be efficient in order to get all of the tests ordered, evaluated, documented, patients admitted and discharged, consultations ordered and checked, crises averted or alleviated, they often don't ask the questions that would allow a patient NOT to have a test or procedure, they don't necessarily have the gentle conversation that allows a person to make a decision to forego end of life heroics, they don't have time to realize that the patient is really now well enough to go home if certain arrangements can be made. And since the patient doesn't really "belong" to the hospitalist, will not come back and see the hospitalist again, making a meaningful connection is harder, even though it is the right thing to do. Solving the mystery is still interesting, but not imperative, since chances are good that an unsolved mystery will only come back to haunt a different hospitalist.

And yet some hospitalists are excellent. And some hospitalist systems foster excellence.

I have worked in 5 different hospitals in the last year, in 5 very different hospitalist programs. What makes a good one seems pretty clear. First, although good systems are important, there is nothing that can take the place of good doctors. Some of the doctors I have worked with are outstanding. They are patient, have good senses of humor, work well with nurses and patients' families. They take time to really listen to patients. They love the challenge of severe illnesses and have creative minds. Bad systems, though, can burn out and chase off good doctors. The best programs have slightly smaller work loads than the worst programs. Seeing less than 15 patients in a day makes us much more likely to do a good job, to do that little bit extra that makes a difference, to read about what the experts say. When the patient transitions from one doctor to another, either day shift to night shift, or going away after a week of work, it is very important to actually talk to the doctor who will be taking over. Writing  something doesn't cut it. The back and forth conversation, in which doctors ask each other what is really going on, suggest alternative diagnoses or testing not only is good for patients but counteracts the isolation that leads to burnout. Some small hospitals foster collegiality because doctors from all of the specialties end up at the same nurse's station and conversation happens. One program actually had a meeting of the hospitalists every morning which included nurses and other specialties and really improved communication  and broadened perspectives.

All of the hospitals have had some sort of computer system for  record keeping. A simple, user friendly computer system could free up so much time to spend at patients' bedsides or talking to their primary care doctors. Unfortunately, though each of these hospitals computer systems had some clever ideas, none were anywhere near what I would call simple and user friendly. If it was possible to  cut and paste them together, the Vista, McKesson, Paragon, Meditech, Hero systems, it might be fabulous. The one thing that is common to all of these computerized medical record programs is that most of the people who use them hate them. Interesting.

So I guess it's easy to have an excellent hospitalist program. All you need is great doctors, a simple and user friendly EMR (electronic medical record), moderate work loads and face to face signouts. Since this recipe is difficult to concoct it's not too hard to see why hospital medicine is sometimes not excellent. I would like to blink my eyes and find out that we had gone back to the system in which patients were seen by their own doctors in hospitals but the trend right now is towards more hospitalists, not less.