Hard Choices

Would a cost-conscious physician order this MRI?

A "my patient comes first" attitude may denyexpensive care to patients who need it more, these experts say.

This is the first of three articles exploring ethical dilemmas for today'sphysicians. Some are traditional ethical questions, while others--like theone presented here--arise from the era and aura of managed care. The articlesresulted from roundtable discussions organized by Senior Editor Anne L.Finger. The five participants include:

Willard Gaylin, the panel moderator, a clinical professor of psychiatryat Columbia University College of Physicians and Surgeons and co-founderof The Hastings Center, a bioethics institution.

Jordan S. Busch, an internist, medical director of Beth IsraelHealth Care, in Brookline, MA, and an instructor in medicine at HarvardMedical School.

Daniel Callahan, PhD, co-founder of The Hastings Center, now directorof its international programs and a senior associate for health policy.Also a visiting scholar at Harvard Medical School.

Harvey R. Gross, chief, Department of Family Practice, Englewood(NJ) Hospital and Medical Center, and assistant clinical professor of medicine,Mount Sinai School of Medicine, New York City.

Bonnie Steinbock, PhD, professor and chair, Department of Philosophy,University at Albany/State University of New York, with joint appointmentsin the department of Public Administration and Policy and the departmentof Health Policy, Management, and Behavior.

The case: Linda Rollins is a family physician in a small grouppractice with 40 percent managed care patients. Rollins has been treating14-year-old Sandy Baynes since the child was a baby.

Sandy's parents have brought her in because she has developed a limpover the past few months, apparently not due to injury. Although Sandy hadbeen quite verbal and alert, her mother reports that lately she often seemslost for words and loses her train of thought.

After examination, Rollins suspects an intracranial tumor and decidesto order an immediate MRI, which requires prior authorization by the Baynes'health plan. The insurer has denied several of the physician's past requestsfor MRIs.

Sure enough, when Rollins calls the insurer, the utilization reviewerrejects her request. The reviewer says Sandy's symptoms sound like "growingpains." Concerned that the child may be seriously ill, and eager toallay the fears of Sandy's mother, who's been a dedicated patient for manyyears, Rollins insists on speaking with people in authority at the insurer'soffice. After her staff has spent the better part of two days on the phone,Rollins reaches the medical director, who agrees to authorize the test.

The time Rollins and her staff spend seeking permission disrupts theentire office schedule, causing much grumbling among both the physiciansand the support staff.

Willard Gaylin: When I was in school, medical ethics dealt withwrongdoing in medicine and usually involved bad guys doing bad things togood people. Today, medical ethics is often a product not of medicine'sfailure, but of its huge success. Some of our hardest choices are dilemmasin which people of goodwill must weigh one right against another, or onegood against another.

Such is the case with Linda Rollins. Dr. Gross, tell us about the likelihoodof this happening.

Harvey Gross: This scenario happens frequently in our office.If I really felt my patient needed this MRI--regardless of whether the insurancecompany said Yes or No--I would make sure she got it. Often, we will justget the test--after telling the patient that insurance may not cover thecost--and try to figure out what to do later on. Of course, if the expensewill be borne by the patient, the patient is going to be very upset.

Gaylin: Is the principle here "Do the best for every patientin every case?"

Jordan Busch: No, I think the principle is "Do what is mostappropriate to make the diagnosis for this case." In some cases, theappropriate first test may be a plain film. The test with the most bellsand whistles doesn't always provide the answer you're looking for.

Daniel Callahan: If you have limited money available, you lookat those areas where there is a low probability of a bad outcome, and younibble away, particularly if the tests are expensive. Occasionally, you'llget the messy case where the patient has something bad despite a low probability.Insurers have to work with general probabilities, and if you think yourcase is an exception, you fight for the patient's rights.

But we shouldn't immediately blame the insurers for this dilemma. Healthcare is expensive, and employers reach a limit of what they are preparedto give insurance companies to provide care--and we are stuck with this.

Gaylin: Let's change the scenario to a 14-year-old girl who complainsof severe headaches. It's hard to diagnose the cause of a headache. Dr.Busch, would you order an MRI for this patient?

Busch: If you can't resolve whether you think this is a braintumor or not--if you can't look the patient and family in the eye and say,"I do not believe this is a brain tumor"--you either refer thatpatient to a neurologist who perhaps can make that differential diagnosis,or you order the test. You can't be held hostage by the expense of thistest, because the expense of being wrong would be catastrophic for the patient.

Gaylin: Linda Rollins did not refer her patient to a neurologist,did she? She just asked directly for an MRI. So do you feel Linda Rollinswas being profligate?

Busch: Well, she suspects a brain tumor, and she now needs a diagnostictest. You may argue whether a CT scan or an MRI would be better. Eitherway, it's probably going to require authorization, and you may run intothe situation where a reviewer--meaning well but not understanding the case--willsay it does not seem appropriate.

Bonnie Steinbock: Would the reviewer give a neurologist less ofa hard time?

Busch: Probably. Again, if you or your consultant thinks thereis an appropriate test, and you are being blocked by the insurance company,your job is to explain to the family: "I think you need this test,and we would like to proceed with it. Hopefully, we can get these otherissues sorted out."

We have this problem all the time with patients showing up at their specialist'soffice before the referral gets there. They are always asked to sign a waiver:"If you don't have a referral, you are responsible for the cost ofthis test." Now, that may be a financial hardship and may even be prohibitive.Health insurance has limits, and sometimes patients will ask for thingsthat are clearly outside those limits.

Gaylin: I've heard physicians say they refer to "cost-conscious"specialists. I don't want my neurologist to be worried about the Gross DomesticProduct. I want him to worry about my head and whether I am going to diein six months because he didn't want to spend whatever it takes to get anMRI. Dr. Gross, are you my doctor, or are you part of the public healthprogram?

Gross: I am certainly your advocate, but there are specialistswho will do every possible examination. You can rationalize doing almostany kind of a workup. You have to use some judgment in what you are doing,and you have to have a medical belief that this test is indicated for thissituation.

Callahan: The old tradition is that you think only of your patient,but the larger question is: Are physicians supposed to be socially responsiblecitizens these days? If you really want to serve your patients, you wouldn'tuse your local neurologist if the best specialist is in Arizona; you'd flyher out there, because that doctor picks up things that only one in a millionother specialists picks up.

"Well," you say, "you don't have to go that far."How do you balance your obligation? If you are frivolous about referringto the most expensive physician, there's a good chance that the cost ofhealth care generally will go up, and somebody will have to pay for thatat some point. So I don't know how you draw the line.

Gaylin: Dan has raised a serious dilemma. Is a doctor simply thereto serve the best interests of his patient? Or must he also be concernedabout the general state of health care and be a responsible citizen, recognizingthe problems in cost containment that we now have? Dan, are you suggestinghe should be both?

Callahan: He has to be both. In some cases, it will be a verydifficult balancing act; in other cases, it may not. After all, the careis provided through an HMO. The HMO in turn is paid by some employer. Sothe doctor has to say, "What I do is going to have some impact on theeconomics of this plan--which eventually means an impact on my patientsas well." Even if you'd like to get away from the social consequences,you can't.

Gaylin: Dan is suggesting a new ideal for the physician. If thatis so, should we now start training medical students differently? When theymake decisions, should they consider not only the welfare of their patient,but also their general responsibility for limiting health care costs inthis country--which will influence their patients in the long run?

Busch: At Harvard Medical School, we are trying to train studentsto pay attention not only to the risks vs. benefits of ordering tests, butalso to the costs vs. benefits. Health care is a scarce commodity, and weneed to understand how the money we spend will make a difference to thepatient. That gets back to understanding what the pre-test vs. post-testprobability of a positive result is going to be.

Steinbock: Even just running tests imposes burdens on patients--time,for example, and the unpleasantness of being inside an MRI. So althoughwhat is in the best interest of the patient can sometimes be contrary tothe ideal of fiscal responsibility, in many cases advocacy for the patientand fiscal responsibility are consistent.

In other words, ideally you are not going to order tests that aren'tindicated and won't help the patient. There may at times be a real dilemma,but many times it is not a dilemma, because what is right for yourpatient is not the most expensive thing in the world.

Busch: I am not convinced that tests are being used wisely. Ibelieve many tests are still ordered because the doctor doesn't understandthe disease process. Ordering tests consumes a lot less time than takinga good history and doing a good exam. It's easy to say, "Oh, you haveheadaches. Let's get an MRI. Okay, your MRI is normal; there's nothing wrongwith you." In an environment where doctors have less and less time,it's easier to reach for the high-tech tests, which may quickly alleviatethe patient's fears and solve your problems.

Gaylin: Is that ever going to change? How are we going to getto the socially responsible physician that Dan Callahan wants?

Busch: I believe that managed care has moved physicians in thatdirection. By making doctors partly at risk for the health care resource,it has forced them to say, "Wait a minute; did I really need this test?"

Steinbock: But managed care doesn't reward the lengthy discussionwith the patient that you also mentioned.

Gross: That is exactly the point. Today, doctors get rewardedfor tests and procedures, but not as well for the time and effort they spendwith the patient. So the system we describe here doesn't really help thedoctor be an advocate for the patient.

Gaylin: Is it not true that the physician who's socially responsibleand refuses a mammogram for a 40-year-old who then turns out to have a breastcancer will never, ever refuse a mammogram again for anybody of any age?Will a doctor ever put things in the broader perspective, or would we bebetter off to let the doctors be cowboys, order what they want, and letthe HMOs control things? This is the first discussion I've heard in years,and it's refreshing, where the HMOs' refusal to grant every physician requestis a positive contributor to the social scene.

Callahan: Well, as a sometime patient over the age of 65, I amconstantly getting advice from my physician: "You'd better do this,and you'd better get that test." I have to decide whether I reallywant the test. Another question is, what's a responsible patient these days?

Gaylin: Let's forget the relevant tests and the silly tests. Thirtyyears ago, I had a discussion about the cost of health care with the brightestman I knew in money management. At that time, high technology was consideredthe reason for the high cost of care. My friend said, "Don't worryabout this, Will. It's the nature of technology that costs come down."He picked up a ballpoint pen and said, "The first ballpoint pen cost$40; now people give them to you for free."

It turns out that medical technology is quite different, however. Bonnie,if I could offer you a washing machine that's 5 percent more efficient thanyour washing machine and costs 100 times as much, you would laugh at me.But your child is undergoing a diagnostic test for a brain tumor, and Isay all the evidence indicates there's no tumor. The fact is, however, thatan MRI will pick up 5 percent of undiscovered brain tumors--that's onlyone in 20. It is fabulously expensive. You don't want us to spend all thatmoney for your child, do you, Bonnie?

Steinbock: I had a discussion like that once with a gynecologistwho said, "I can't tell if this is a viable pregnancy, because it'stoo early for a regular ultrasound. There's a new kind of ultrasound doneby only one doctor in the community. Do you want the test?"

After I had the test, I thought to myself: This was probably incrediblyexpensive; was it really worth it? The doctor didn't bring up the cost.When a doctor's talking to me about my child or my own health, what I wantto know is: In your judgment, do we need this?

Callahan: The interesting question is: What kind of patient orparent are you? Are you a risk-taker? You have to ask that about your physician,too. Is this a person who is more likely to let things go, or is this anervous person?

Busch: I agree that part of this issue is the doctor's comfortwith uncertainty and what it is he's uncertain about. If a person couldhave a critical disease and die if you don't make the diagnosis, you'd orderthe test even if there's only a 5 percent chance it will help. But if yoususpect the patient has osteoarthritis of the hip, you might be willingto try less expensive X-rays or empiric treatment, and not order the MRIuntil you were prepared to replace the hip.

I think it's probably not in the high-risk diagnoses, like the one describedhere, where we are doing too many MRIs. More likely, it's where doctorsaren't very concerned about serious illness but do have some degree of uncertainty.And there is always some degree of uncertainty.

Steinbock: One of my children once had chest pains. We broughthim in, and I told the doctor, "He is having recurring chest painsthat seem very severe; he falls down on the ground." The doctor talkedto the kid and examined him, and he said to me, "Well, what do youwant to do with this? Do you want to go for more tests?" I said, "Doyou think it's warranted?" And he said, "I don't really thinkso." I said, "Then let's keep monitoring it.

Busch: That's the most critical point. You have to trust yourdoctor, and your doctor has to be willing to step up to the plate and tellyou what he or she would recommend. The litmus test I use in my practiceis what would I recommend for a family member.

Diagnosing an untreatable brain tumor might not make a difference tothe outcome. But if there will be treatment options, I can't imagine thatI wouldn't recommend doing an MRI under those circumstances, despite the5 percent figure. Would I feel that way at 1 percent, at 1/2 percent? Iprobably don't break things down into those fine numbers. I behave moreon whether I think there's a reasonable chance that I could make a diagnosisthat I would otherwise miss--and whether a delay in the diagnosis wouldlikely affect the prognosis and the patient's quality of life.

Gaylin: I think the majority of doctors, given that 5 percenton a potentially serious condition, will recommend the MRI, and that's whathas driven up medical costs. Despite what you say, Bonnie and Dan, you maybe a minority among the vast majority of parents who say, "I have heardabout this test that picks up cancers other tests miss. Will you do it?"If their doctor won't do it, they'll find another doctor who will.

Steinbock: Patients get antibiotics for viral respiratory conditionswhen there is a zero chance they'll help. I never understood doctors whosay, "If I don't prescribe antibiotics, they won't like me; they won'tcome back." So be a little tough, and don't prescribe them.

Gaylin: The MRI has become the penicillin of modern medicine.

Callahan: Which is exactly why the HMOs and the insurers haveto go after it.

Gaylin: So now we come to the wonderful position that these meanold HMOs are the people who have the public conscience at heart, and that--indeed--theyrepresent social good and cost containment in the face of populations ofpatients and physicians who are cure-oriented.

Gross: I don't agree with that 100 percent. While I like the HMOs,I don't feel they advocate solely for patients, doctors, or good medicalcare. They have their own agendas, doctors have theirs, and patients havetheirs.

One of the interesting issues is what happens if we don't give the MRIto the girl who has a lot of headaches. What is the cost to that familyin terms of anxiety, maybe psychiatric bills, visits to other physicians,and other resources that are utilized because the family couldn't get thatMRI?

Busch: That's a great point: How many emergency room visits doyou need at $700-plus a pop before it would have been cheaper to get theMRI?

Gaylin: It's interesting that we've reached this point in thediscussion, and the two physicians haven't brought up the one thing thatcomes up all the time: Suppose you don't do the MRI, and the patient hasa tumor. Are you concerned about malpractice? The malpractice issue hasbeen the guardian of medical ethics in America for a hundred years. Doesn'tthat enter your mind?

Gross: We'd like to believe that whatever we are doing, we aredoing the best for the patient. But of course, being practical, we are concernedabout a malpractice suit. On the other hand, most of the time when we practicemedicine, we are doing it because of what we know about medicine, aboutthe patient, and what is indicated in this specific case.

Busch: I agree. I think good doctors always try to do the rightthing for their patients before they pay attention to risk or cost.

Gaylin: Then why are they prescribing all this penicillin?

Busch: Again, probably less because of malpractice and more becausethey have not come to terms with their degree of uncertainty, and they haven'talways thought through the process.

Gaylin: They don't know that penicillin doesn't work on a virus?

Busch: Maybe it isn't a virus. An interesting study publishedin JAMA looked at patients who presented with upper respiratory tract infectionswhere the doctors prescribed antibiotics. Eighty percent of the time, itdid not require an antibiotic; it was viral. So doctors overprescribe. Butthat also means that 20 percent of the time, antibiotics might have beenappropriate and beneficial. Suppose your patient comes in and says, "Doc,I've been really sick. I can't go to work. Is there something you couldgive me?" If you said, "I can give you an antibiotic, but it mayresolve your symptoms sooner only 20 percent of the time," well, Imight go for that if I were the patient.

Gaylin: Isn't that the same psychology of the MRI? Fifteen percentof the time, 5 percent, 10 percent, or 20 percent?

Busch: Again, I think it depends on your ability as a physicianto live with uncertainty. If you feel confident that this is not the appropriatetest for your differential diagnosis--that this won't really add anything,that you expect this will be negative--you can look the patient in the eyeand say exactly that. Then you have to get the patient to buy into this.If the patient doesn't buy it, and leaves dissatisfied, you have failed.Every patient also has a different comfort level with degrees of uncertainty.The doctor must be sensitive to this, and work with the patient to reacha plan with which both feel comfortable.

Gaylin: I suspect that in a malpractice suit, if you said "No,I did not order the test, because it was only a 5 to 10 percent chance and,after all, I have a responsibility beyond my patient to the general welfare,"it would not sit well with a malpractice jury. Even if it were a 1 in 1,000chance, and the physician said it's ridiculous to do it in this case: Ifthe child involved had a brain tumor, his lawyer would say, "I canwin that case--whatever the odds are--because jurors are sympathetic."

Busch: This is a classic case where you need to convince the familyinvolved that your plan is the right one. You'd say: "I don't believethis is a brain tumor. I don't think any further testing is warranted atthis time. I take this headache very seriously, and I would like to followthe patient closely."

This isn't a case where you'd say: "This isn't a brain tumor. Whyare you even here? Sounds like growing pains. I'll see you next year foryour annual physical." That is where you are going to run into trouble--perhapsnot with this patient, because she doesn't have a brain tumor, but maybewith the next patient, because he does.

Gaylin: This case is only a sketch, meant to stimulate discussionand to illuminate the kinds of problems that cause ethical difficulties.There are routine requests for expensive treatments that are unnecessary.There are, on the other hand, routine requests for unnecessary treatmentsthat have become institutionalized--at one time, routine chest X-rays, perhapsthe PSA test more recently.

Medical ethics exists on a slippery slope where there can never be certitude.Diagnostics is always an incomplete art. We will always have to deal withnew and more expensive treatments--which, in turn, will lead to new andmore challenging ethical dilemmas for doctors.

. Would a cost conscious physician order this MRI?. Medical Economics 1999;15:62.