formats

Why the Dutch Girl Cried

Half-way through the medical student seminar, the Dutch girl began to cry.

The seminar, a conversation among medical students and patients, was intended to teach aspects of chronic illness. A young patient advocate from the Netherlands was an invited guest. None of us in the room, including her, had anticipated her tears.

“I was startled,” she told me later. She had thought the American patients would describe personal courage, doctor-patient communication, and personal resource-finding. Instead, she heard an angry litany about roadblocks in our health care system. “I was very upset,” she said, “because ill people should not be spending their precious energy arguing with their health insurance companies, begging them for help. They need that energy to recover and feel as well as possible.” So this day’s lesson was not about chronic illness. It became instead a conversation about, to her, America’s incomprehensible approach to medical care.

She cried because she heard American patients protesting private insurers’ denial of diagnostic tests, of medications, and of rights to consult specific physicians. These, she thought, would have been trivial, uncontroversial, and easily-supported requests in her native land. She explained in a later e-mail: “I did not want to emphasize how happy I was, because I felt guilty as well. I also felt a little embarrassed. I didn't have to write a letter or make a phone call when I needed chemotherapy or a new experimental drug or a prosthetic hip. How lucky I am,” she continued, “that I was born in the Netherlands.” Hers is message Americans should be startled to hear.

I, a teacher, had been taught. I had presented to students what I considered to be common problems in chronic disease management—identifying resources, self-advocacy, maintaining optimism. I now saw that I had unquestioningly accepted the structure of our health care system. I had been oblivious to how crazy it looks to foreign eyes.

I asked her Dutch physician, who had worked for several years in the United States and is my friend, what features of American medicine he would like to import to his home? His answer: “Nothing. I feel that the Dutch health care system is very good… I do not miss many things.” I then thought about patients of mine who, travelling, had fallen ill in England, France, Germany, Italy, Korea, Hong Kong, and Japan. They had found unencumbered access to top flight care and no bankrupting hospital bills, the opposite of what ill foreigners experience here.

My Dutch colleague says, and his patient’s tears show, that the uniform (national) health care services of other lands work better than does our patchwork plan. Americans pay 62% more per capita for health care than do the Dutch. One of every 7 Americans, but not a single Dutch person, has no medical insurance at all. Dutch insurers do not waste resources on armies of functionaries hired to find reasons to deny. Dutch physicians’ offices are not peopled with administrative assistants whose job is to get requests approved or to contest refusals to pay.

In the United States, obtaining approvals from private insurers is an abstruse, argumentative, and unpredictable process—at least two cycles of application-refusal-appeal then a telephone conversation with an arbiter who may or may not be authorized to approve. In contrast, Medicare and Medicaid—our closest approximation to the European way—makes approvals for tests or medications straight-forward, reasonable, and predictable. Medicare offers fair rules, uniformly applied; private insurance does not.

The young Dutch woman saw, more easily than did I, the cruelty in our system. Patients should focus on getting well, she said. Doctors should spend their time with their patients, not arguing with insurers. I agree.

America can do better. The President’s Affordable Care Act (ObamaCare) is a start—a health care system that has uniform rules for all for those who qualify, perhaps for all at a future date. The Affordable Care Act promises to use our resources for direct patient care and not for private insurance bureaucracies. With it, perhaps, we will be able to offer care for every person, well or chronically ill, and not subject the ill and vulnerable to the vagaries of whether they can or cannot be insured.

We should do this now. We should not need a young Dutch woman’s tears to tell us what to do.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

Obamacare – the Best idea since Medicare

I just received an email from my member of Congress Dr. Nan Hayworth in which she wrote: Assuring that all Americans have access to good medical care and to affordable, portable health insurance is a crucial goal for our country, and one that is a mission for me as the only woman physician who is a Member of Congress. The massive new health law passed in 2010 was meant to help us, but, for many reasons, it actually causes harm.

She is totally wrong on the issues and here is what I replied to her:

I could not disagree with you more. Trying to care for people who are uninsured, or who suddenly change insurers at an employer’s whim, is near impossible under the current system when patients have long term (chronic care) needs. The bureaucracy of the private insurer system is infinitely more complex, and much more expensive, than a uniform care system would be. The Obama plan is the best idea since Medicare. You should support it.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

Hospitalists and End-of-Life Decisions

Idée reçue number one:  when hospitalists are the doctors in charge of inpatients, hospitalizations are more efficient, and care is better.

Idée reçue number two:  thoughtful end-of-life planning is a laudable and comforting goal.

What happens when these values—efficient care, thoughtful planning—clash?

This question arose because of a friend’s 100-year-old mother. She was vigorous and alert at her 100th birthday party in August; in November she suffered a stroke. She became less coherent—on transient occasions—incoherent and confused, then she developed some new medical problems. She is now confused and incoherent more often than not, but sometimes she recognizes her family and friends. She is not moribund, but she is not going to improve. During a hospitalization in late January end-of-life care discussions took place.

In-hospital care has changed, so that now hospitalists, not the referring physicians, care for inpatients on the general medical services. I was there as an available friend; her regular physician, according to the new rule, ceded first responsibility to the hospitalist, a perfectly competent physician, but a stranger. The end-of-life conversation was first introduced by the hospitalist. Her private doctor and I answered the family’s questions after the fact. A few days later January became February. Since hospitalists rotate duty call schedules monthly, a new hospitalist came on the scene. Hospitalist number two had different approach to end-of-life care. The rules do not—cannot—specify that the recommendations of the January physician be carried out by the February physician. So the family (and her private physician and I) faced confusing, partially contradictory recommendations from strangers. Is this the best we can do?

The medical details are unimportant. What is important is that it was not clear which set of physicians—the old or the new—should guide, or even introduce, such emotion-laden discussions. What is important is that there is not now, but should be, a process through which hospitalists and personal physicians can interdigitate their different roles.

It’s a hard choice. If hospitalist teams are asked to manage a floor of 40 or so patients, they must be able to act without obtaining other doctors’ consents, and they must be protected against constant interruptions by the 40 patients’ other doctors. Yet if end-of-life decisions are to be kind, graceful, and caring, those who know the patient best should take the lead. The conversations about my friend’s mother were all ad hoc, arranged when the parties (the hospitalist, the family, the social worker, but not me or her private physician) could meet. Maybe hospitalists could have office hours, like college professors, scheduled times to converse with interested parties. Maybe one should stay responsible for a patient throughout his or her entire inpatient stay.

This is a transition time. Definitions regarding which doctor does what and when are changing. Personal and efficient (or cheaper) medical care values are in conflict. In a more ideal world the families, the personal doctors, and the hospitalists would sit down each night and discuss the end-of-life plans, but that is not going to happen, at least in a way that would be affordable to all.

So, do we structure time for more personal conversations? Revert back to each attending physician writing all the orders and making all the decisions for his or her patient? Accept that the personal aspect of medicine is not affordable? Upgrade the hospice care concept in the acute care setting?

No easy solution, here. But at least we can acknowledge the problem and work to get something better than having these discussions among strangers, some of whom disappear at the end of the month.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

New Federal Drug Research Center

Fareed Zakaria in his Washington Post opinion piece Can Obama get it right on the economy? wrote:

 “The Obama administration, concerned about the dramatic slowdown in drug development, is proposing a new federal research center with a $1 billion budget. A good idea, but U.S. officials should look at the regulatory framework surrounding the process of discovery and development, since the private sector spends more than 50 times that sum every year on drug research. The Food and Drug Administration takes twice as long to approve a drug as its European counterparts. As a result, health-care research has been moving offshore, particularly as China and India innovate in every product and process. “

My wife emailed me the article and asked if it was correct about European vs. US drug approval.

The answer is yes, but it’s the wrong question.

The new federal research center is focused on end-of-the-line research – the final product, which is hugely expensive and highly regulated (testing in humans). Everyone seems to think that is what medical research is about. But real progress comes from looking at upstream events – what causes the disease, how does it get a foothold, what processes allow it to progress, the finger in the dike that prevents the system from collapsing. Finding the right drug for downstream events is reconstructing the wall after the dam has burst.

I assume that

(A) the Director of NIH, Francis Collins is reflecting required statements from the White House about their priorities rather than his, since the policy seems to contradict his earlier persona when he was director of the Human Genome Center or

(B) he just wants to use the machines once applied to seeking the genome that are going to waste, now that better technology has taken over.

The drug development program is such an un-NIH thing to do, and 180° away from what the former NIH director, Harold Varmus initially hired Collins to do.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

More stuff from the New York Times that misses the point

On June 10, 2010, Claire Cain Miller wrote that, “When a doctor recommends a test or a procedure, most patients simply go where the doctor tells them to go.”  

Well, not exactly, at least as I see it. It’s more like, When a doctor recommends a test or a procedure, most patients…ask their doctors who does the best test, then find out that their insurer won’t cover that facility or person, then find out that their doctors don’t know the providers of services that their insurer will cover, then find out that the people on the insurer’s list no longer accept the insurance, then do the test with someone no one knows and find that the doctor who recommended the test can’t understand the results submitted because the person who interpreted the test didn’t answer the question asked because his or her equipment was inadequate or he or she didn’t have sufficient experience with the problem to interpret it correctly.  

Right. Comparison shopping—by the insurers, with quality or communication not a priority. Big savings, eh? 

Before I sat down to fume over the first article, Nicholas Wade wrote, on June 12, 2010, that the genetic map, one decade on, has yielded few of its promised cures.  

I did not find this second article to be a surprise—it is evident to any practicing doctor—but my eyebrows did rise at the statement midway through the article: ‘“Genomics is a way to do science, not medicine,” said Harold Varmus, president of the Memorial Sloan-Kettering Cancer Center in New York, who in July will become the director of the National Cancer Institute.’  

Funny. I wrote about the distinction between science and medicine—citing a then-unnamed Harold Varmus—in my 1998 book, Guarded Prognosis (Hill and Wang), chapter 4, page 156 (actually, the whole chapter is about this topic).  

In 1996 Dr. Varmus was Director of the National Institutes of Health, and I was Acting Director of one of its Institutes, the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS). Dr. Varmus and I had argued about the budget for my Institute, and I, following that discussion, had gone home thinking about a patient of mine for whom contemporary science offered no answers and whose plight seemed to be a matter of indifference to the scientists.  

I wrote about that conversation: “Use only the criteria of other scientists, he seemed to say [to set your budget], while I was more willing to use my own judgment…developed over three decades as a practicing physician. We had different priorities, the scientist and the doctor.”  

Clearly, Dr. Varmus said the same thing to Nicholas Wade in 2010 that he said to me in 1996, except then he was trying to tell me that focus of the National Institutes of Health and of the National Institute of Arthritis and Musculoskeletal and Skin Diseases was medicine, not science, if, indeed, there is a distinction. Now he makes the distinction to explain why his priorities for NIH failed to achieve their goals.  

Sad. The United States has a National Science Foundation, which is for science, and which is distinct from the National Institutes of Health, for medicine. If I read the quote from Dr. Varmus correctly, he argues that NIH achieved its goal because it did good science, even though it failed to achieve a useful result for medicine?  

I guess we still have different priorities, the scientist and the doctor.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

The New Medicine—Hospitalists and Telemedicine

So, two things in The New York Times set me off this week. (Actually three. The third was an article about people who start their sentences with the word “so”, but I don’t want to go there.)

The first thing that set me off was an article about a hospitalist (by Jane Gross, May 26, 2010). A hospitalist is a new breed of doctors who work full time taking care of hospitalized patients, a role we used to assign to interns. Hospitalists are the doctors of record for all in-patients. No longer is the primary or specialist physician who sends the patient for admission in charge. The hospitalist is. The hospitalist idea is not very new. For several decades the patient’s physician has ceded responsibility to the intensivist in the coronary care or intensive care or post operative care units, special circumstances that required a certain technological skill in care. Now this ceding of responsibility has grown; it now covers all hospital care.

There are two reasons for the change: one is to improve hospital efficiency (translation: prevent delays in discharge because tests were not done on time, not for the patient’s convenience but to avoid hospital days for which insurers will not pay); the other is to provide more classroom-type teaching time to interns and residents. This pedagogical goal is ordered by the accrediting commission for training; the apprenticeship model is no longer thought valid; the pedagogical goal has nothing to do with patient care.

The goals of having hospitalists are admirable, the practice less so. What one sees in this transition time, as doctors are adjusting to the new rules, is the insensitivity of some hospitalists to long-term goals for patients with chronic illness, an inpatient’s sense that no one physician, including his or her personal physician, is in control, and—from the patient’s point-of-view—a different set of inefficiencies.

An example: a patient I saw this week, at the end of a three-week hospitalization for a very confusing and severe complication of her chronic illness, still disabled and with visiting nurse service planned, was discharged by the hospitalist with instructions to follow up with…a dermatologist, urologist, gynecologist, hematologist, and rheumatologist. The point is that a single physician, whom she knew before, could handle all the concerns. There was no need to make five follow-up visits when one would do. Efficient, yes—to coordinate multispecialty care in the hospital—but not for the patient’s life outside, when the multiple issues had already been resolved. And, yes, while the rapid and multiple consultation process did save money in the hospital (because of its rapidity, but one can argue that most of the in-hospital consultations were unnecessary), the hospitalist’s discharge plan would waste both money and the patient’s energy if the whole package of outpatient and inpatient care is considered together.

So, on to point number two.

In an article in the May 30, 2010, Business Section, by Milt Freudenheim, a Dr. Oscar W. Boultinghouse, speaking of the marvels of telemedicine, says, “In today’s world, the physical exam plays less and less of a role. We live in the age of imaging.” That comment earns from me a disrespectful, “Hunh?” or an ironic “Right!” Tell the young lady now on our hospital floor, who has lost several fingers to blocked blood vessels, that physical examination plays little role. She had been seen by many physicians, all of whom wildly speculated about possible causes that had not been identified despite several inpatient days of extensive laboratory tests and imaging studies (see point one above). A quick examination by an experienced physician who actually took the time to look at and see the skin of her elbows, knees, and ankles, a physical examination that took less than five minutes, identified the cause.

What is the point? That physical examination (old-fashioned doctoring) still plays an important role. Also, you hospitalist and telemedicine advocates, a physical examination is cheaper and more efficient, too.

So?

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
No Comments  comments 
formats

The Butterfly Effect

According to Sharon Begley and Mary Carmichael in their Newsweek cover story Desperately Seeking Cures, a cornucopia of cures for an infinity of diseases is (or would be) available but for (a) a culture that sets the wrong incentives for academic scientists and (b) a stodgy bureaucracy that does not prioritize urgent needs. 

Well, maybe…but posing the question that way misses important points. The article has a handful of examples, but ignores the fact that literally thousands of such examples exist. Scientists, venture capitalists, government, advocacy groups all have their own priorities, with no rational way but the marketplace to choose among them.

The article focuses on today’s Holy Grail, that a single aberrant molecule—a single target—if corrected or blocked, will reverse the course of an illness. Well, maybe that, too—if, as is unlikely, the illness results from a single gene mutation. But the chronic illnesses that I know, and possibly most cancers as well, have a better analogy: their origins are like the flap of the butterfly’s wings in Brazil that sets off a tornado in Texas. Human biology and human illnesses are so complex that it is unlikely that targeting a single molecule will “cure” any but the most narrowly-defined genetic diseases.

I am not nihilistic: I do believe that “cures” will be found. My argument is that the molecular-target-found: cure-imminent concept lacks humility. The butterfly wing concept is a better model of illness. If our goal were not to find molecular targets but rather to ameliorate those symptoms that we can recognize and measure, then we might take more pride in the progress we have made and we might direct our future research toward more attainable goals.

 

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
Comments Off on The Butterfly Effect  comments 
formats

The Havasupai, Privacy and Medical Research

I learned that the Havasupai of the Grand Canyon recently won a lawsuit against the University of Arizona. They claimed that the university had misused DNA samples that had been obtained from them for what they thought was research on diabetes. And, in fact, the researchers had studied the genetics of diabetes using samples from the Havasupai, but they had subsequently expanded their project to study the genetics of schizophrenia and ethnic origins of Native American tribes. In defense, the University investigators pointed out that the subjects of the study had signed consent forms that allowing use of the DNA samples for other studies. However, the consent form was very general. It did not speak to the possibility that the Havasupai people could be embarrassed by the findings or could have their myths of origin contested. The Havasupai people therefore sued the university claiming misuse of the samples, and they won.

I am sympathetic to concerns of the Havasupai, but I do not feel the question is as black and white as it sounds. True, in the United States, we give very high priority to privacy. The problem is with the definition of the word privacy. Our regulations, and our usual sense of the word, means individual, not group, privacy. The distinction is an ambiguity that needs to be clarified.

It is very common these days for medical research projects to bank DNA samples from individuals — genetic origin of disease is the hottest thing going in medical research right now. Investigators should do this work under very tight restrictions that require that "personal identifiers," such as birth dates, addresses, Social Security numbers, and any information that could allow someone outside the study to identify a specific individual blood donor, be kept under the strictest secrecy — locked cabinets, limited access, etc. The results of the study must be presented only in "de-identified" form, that is, no public information can lead back to a specific person.

Investigators have another requirement. If the National Institutes of Health or any other federal agency study funded the study, because tax dollars paid for the research, the investigators must provide de-identified samples to other legitimate researchers. The samples are considered public because they were paid for by public funds. The point is that the samples are very valuable, and cannot be obtained over and over again, so they should not be wasted.

In other words, the investigators who banked DNA for the Havasupai study of diabetes had to provide samples to a legitimate request of another investigator who thinks that the samples will be valuable for his or her research. Or, should other questions arise, the investigators are expected to use the samples already obtained and not go out and try to get new samples. By federal requirement, research projects that involve DNA banks have outside boards of experts to judge the legitimacy and the feasibility of a request and to decide whether or not it is legitimate to use the specimens for another study. The investigators can not make such a decision themselves. The point of the board is to protect the samples and to protect the dignity of the patients who gave the samples.

The researcher who asks for consent from the person donating the blood will usually not know the possible future uses of those samples, so the researcher uses a very general consent form that says something like "and for future studies as we may determine." The alternative is that the researcher can refuse to release samples until he or she goes back to the subject to obtain additional consents as each applications to use the DNA for other purposes arises. This is very cumbersome. Years elapse, people move or cannot be found. The feasibility of doing the second line of research thus becomes nil. This is not to say that researchers should be freed from onerous duties necessary to protect people's privacy, but rather to point out that the more restraints are placed on research, the less research will be done.

The Havasupai complaint introduces a larger question: the concept of individual privacy versus group privacy. The Havasupai were angered because they perceived group (not individual) stigmatization in the studies on schizophrenia. The privacy laws that researchers now operate under focus on individual privacy and do not think very deeply about group privacy. Nor do the federal regulations decide what might constitute an insult and what might not. Not everyone would consider a study on schizophrenia to stigmatize the group. Some people might be thrilled to learn their origins. Several years ago a study documenting that the Jewish priestly tribe (the Cohanim) could be identified through genetic tracking caused great excitement among the people involved.

I suggest that one cannot do genetic studies ("find the gene for a disease") without studying groups. If this suggestion is correct, any study of any gene could potentially stigmatize the group — it depends on what people may be embarrassed by. Should Greeks disallow studies on thalassemia (Mediterranean anemia), or blacks on sickle cell disease ? Or Jews on Tay-Sachs disease? How do you ask why Kenyans are great marathoners without identifying the group? What about the Long Islanders whose concerns led to the finding of the breast cancer gene — not, as they supposed, because of an environmental contaminant but because a group of ethnically related people lived nearby? Is one person's embarrassment another's clue to the origin of a problem? Whose rights prevail? Those who would be embarrassed? Or those who would benefit from the knowledge?

I do not know how to conflate the rules for individual and for group privacy. I don't think they are the same, and I don't know that prohibitions for the former apply to the latter. It seems to me that extreme restrictions on studies of groups, as we already have on studies of individuals, would inhibit research on the nature of disease, or on the nature of humankind. It seems to me that that privacy rules must be different for groups than for individuals. I suspect the distinction is not so clear and that discussions about where we should draw the line are due.

This conundrum is an opportunity for negotiation between researchers and subjects regarding about priorities. Personal privacy is a priority for some. Group privacy is a priority for others. Scientific advance is a priority for still others. These priorities conflict. There must be a negotiable position that distinguishes the rights of individuals from the rights of groups. Studies on schizophrenia, or on distant past origins of groups, are not intrinsically wrong.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
Comments Off on The Havasupai, Privacy and Medical Research  comments 
formats

Lowering Health Care Costs

On March 20, 2010, Lesley Alderman published an article in the New York Times quoting the responses of several doctors to the question, How can we lower health care costs. Their answers were very narrow:  insure catastrophic but not minor illness, change malpractice law, counsel nutrition, rely on evidence (don’t order tests you don’t need), allow for expertise, use “integrative medicine”, pay to treat childhood obesity, stop overeating, restore the humanity of medicine.

 

These are all good ideas, but I find them limited. The doctors mostly responded to issues in their own fields, and did not think of big ideas that can have large impact on costs.

 

Like Ms. Alderman and others, I do worry about the cost of health care. My solutions would be broader than those listed above.

 

The first solution would be to reduce waste: I see waste in the administrative cost of complying with (or arguing against) a multiplicity of insurers that have different rules and different forms, all dedicated to refusing care. The new health care bill will reduce but not eliminate this cost.  A one-payer system, or uniform rules for all patients, would be a tremendous help.

 

I see waste in the cost of pharmaceuticals, in the United States, that can be purchased at half the cost in Europe or Canada—why cannot we regulate pricing, or profit margins, the way that other countries do? I am willing to let drug developers recoup their development costs, but are we subsidizing Europe, or does our lack of price controls permit manufacturers to charge the most they can and profit more than they should?

 

I see waste in overuse of technologies—the repeated CT scans and MRIs, when a recent one will do. Granted, sometimes one machine, or one laboratory, does a better job than another, but mostly not, so this is a training issue: training both doctors and patients in rational use. It is also an abundance issue—if the machines are available, why not use them, the argument goes, particularly when, having been purchased, they must be paid for by maximum use?

 

I have other proposals that some may find unpalatable. Some patients do not accept answers doctors give, and seek other opinions. I do not argue with that right; it is fair and appropriate. Doctors differ in personalities, skills, and judgment, so bureaucratically locking one patient to one doctor is bad policy. On the other hand, this past week I was the fifth academic, professorial rank, highly trained rheumatologist to see a patient who had not heard the answer she wanted overall (I agreed with three of the prior four). It was easy for her to see so many doctors—who, by the way, each repeated all of her tests (I did not)—because her insurance covered all costs, no questions asked. Perhaps requiring her to pay upfront some part of the cost after, say, the third opinion, would reduce the over-doctoring that is waste. Or, as the physician in Ms. Alderman’s article who recommended covering only catastrophic care suggested, perhaps routine items of care, or minor medical problems, could be paid out-of-pocket, a low initial cost reducing the demand for what is (for some) totally free care.

 

The other unpopular solution that I offer concerns the inequity in physician pay. There is a 10-fold difference in income between primary care, non-procedure-oriented medical specialties and the surgical or procedure-based medical specialties. The difference is not based on achieved skill, hours worked, or even physician’s personal responsibility. (Surgeons often turn care before and after the operating room over to physician’s assistants or to their medical colleagues.) The pay scale is based on perceived value and drama of care—fair enough in a capitalist world—but one can argue for a narrower differential among specialties, for salaried rather than fee-for-service payment, and for federal support for training costs so that physicians do not enter practice deeply in debt (and so choose lucrative specialties to recoup).

Saving money is a test of values. Do you want insurance companies to set the rules regarding what tests you can or cannot have? Do you value capitalism so much that you concede price-setting to the companies that make the drugs you take? Do you want to have access to as many exotic technologies, and as many physicians, as you (not your doctor, not medical need) wish, as often as you wish? Do you want your doctor to be an entrepreneur in a fee-for-service system, earning what he or she can charge, or would you prefer that her or his diagnostic and treatment decisions be made without concern about the doctor earning less or more depending on the choice?

 

While I don’t find these questions hard to answer, I know that different answers are equally legitimate. The debate has just begun.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
Comments Off on Lowering Health Care Costs  comments 
formats

Health Care Reform Bill

A patient recently sent me the following question: As a doctor who has been working in the system even before Medicare Medicaid, are you for this final draft of the health care bill?

Here is part of my answer:
To answer your question, yes, I very strongly favor this bill — because I believe that once it is enacted it will progress to a much stronger bill, which I do not think can get passed at this time.

There is absolutely no question in my mind that Medicare and Medicaid made a huge difference in the lives of people who are generally not powerful and not very loud voiced. I actually wrote about some of the period of transition from before Medicare to after in my first book, Guarded Prognosis (now out of print.)

In case you didn't guess, I also have very strong feelings about Roe versus Wade (or about the availability of abortion). I also began my career before abortion was legal, have many (ugly) stories to tell, and I never want to see that time come again.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
Comments Off on Health Care Reform Bill  comments