Medicine requires doctors to constantly manage probabilities

Medicine requires doctors to constantly manage probabilities

“I’d really feel better if we got the MRI,” Ms. James said. “I understand you think it’s a migraine, but I want to know, just in case. Wouldn’t you?”

Ms. James and I sat in her darkened hospital room—the light bothered her eyes and exacerbated her headache. She was a dialysis nurse with many years of experience in the healthcare field, and I was a first-year doctor trying to convince her that she was most likely suffering from a migraine and did not need additional tests.

Ms. James had woken up the morning before with very concerning symptoms. Her head hurt terribly. She got out of bed, but she felt nauseated and had to lie back down. She thought she needed her morning coffee, but she felt too sick to go downstairs to make some. Her headache had worsened, and she began to notice shooting pains in her left arm. She was scared. A few hours later, her daughter arrived to find her mother’s speech was slurred. The daughter called an ambulance.

By the time Ms. James reached the Emergency Department, her speech had improved, but her headache remained. The fluorescent lights bothered her, and the loud noises of the hospital grated her nerves.  A neurology resident was called to evaluate her. He felt that she was most likely experiencing a migraine and recommended that she be given some medication to help with her pain. He thought it was possible that she could have suffered a TIA—a transient ischemic attack, in which the blood supply to a part of the brain is temporarily blocked—but he felt that this was a less likely possibility. He did not think she would need an MRI scan of her brain unless her slurred speech returned.

The craft of medicine requires doctors to constantly manage probabilities.  Indeed, the weighing of likelihood is built into our methodology: we hear patients’ stories, list possible diagnoses, and then rank them according to probability, creating “the differential diagnosis.” We order tests to rule these possibilities in or out and gradually refine the list until one diagnosis remains.  This approach is systematic, but when followed too rigidly leads to unnecessary tests that inflate the cost of care.  Situations occur very frequently in which tests are ordered to eliminate possibilities that are highly unlikely (the “just in case” scenario), or in which the added knowledge of the test would not affect our therapeutic strategy, but we feel a “need to know”.

This latter situation arose with Ms. James. We had two most likely possibilities—a TIA and a migraine—with only an MRI that might help us differentiate between them. However, because Ms. James had additional medical problems like hypertension and diabetes, she was already being treated with the recommended therapies for secondary stroke prevention. In other words, ordering the MRI would probably have no impact on her medical care. It would only satisfy the “need to know”.

The need to know is powerful, on both sides of the doctor-patient relationship. In the October 17, 2012 issue of JAMA, Jutel and McBain point out that our society places a high value on certainty in diagnosis and that this fascination with certainty may have harmful consequences, leading to unnecessary treatments or to tests that do not influence patient care.  Sitting in the hospital room with Ms. James, who was anxious about her diagnosis, I felt the power of the drive for the “need to know”.  It felt as though I was withholding something from her by telling her the reasons an MRI would be unnecessary.  These conversations with patients are not easy, particularly when patients are anxious about their diagnosis, but if we hope to control the cost of healthcare in this country, they are of utmost necessity.

Ultimately, Ms. James got her MRI, which was negative. Her persistence, our discomfort with the uncertainty, and our worry about hard feelings and the omnipresent specter of litigation all played a role.  She felt better knowing. But the decision still nags at me. If we had been able to convince her that fewer tests actually meant better care, perhaps we all could have reached a better outcome.

Robert Fenster is a psychiatry resident and a winner of the 2012 Costs of Care Essay Contest.

Image credit:

Comments are moderated before they are published. Please read the comment policy.

  • NewMexicoRam

    2 things probably led to this frustrating ending for you:

    1. ” the omnipresent specter of litigation,” as you state.

    2. the patient probably did not have a $5000 annual deductible before payment by the insurance kicked in.

    It would be interesting to know how quickly her headache went away after she found out the results of her MRI.

    • Homeless

      3. The doctor didn’t own the MRI machine. It would be interesting to see how the probabilities would change.

  • Suzi Q 38

    Sorry, but if I had Ms. James’s symptoms, I would also request the MRI.
    I only have one body and life. My family history is riddled with former family members (RIP) that died and/or suffered from the effects of heart disease and stroke. I would want to make sure nothing was ready to “burst,” so to speak. Those are fairly troublesome symptoms.
    I guess the rule of thumb would be this: What kind of insurance does she have? Has she ever had symptoms leading up to this incident? What is her family history? Would I request an MRI for my mother, wife, or other family member? If the roles were reversed, would you ask for the MRI for yourself?

    Really be truthful.

    • NewMexicoRam

      Would you have wanted it if you had a $5000 deductible to meet first? I’m a PCP and that’s what my deductible is.
      Cost consideration by patients HAS to be part of the consideration.

      • Suzi Q 38

        Yes, cost is, but when dealing with the possibility of a stroke and its short term or long term life altering effects, I would pay the 5K. I would prefer to pay the 5K than gamble with a stroke at 57.

        For me, I have an aunt that had a devastating stroke at age 45. My grandfather died at age 48 of a heart attack. I have a couple of uncles that died of heart related events in their 50′s. My own father died at 68 after prolonging his life with TWO bypasses.

        If the patients’ family history is fairly good, and the patient has had no prior incidents, then I would appreciate your frugalness with the cost of my medical care.

        Everyone is different. It is difficult to be the physician and have to “make the call.” I like it when the doctor gives me the reasons why he is deciding this and that for me. I like it when he “talks aloud” and says what he is thinking.

        I also admit that I like it when he includes me in the decision making.

        My deductible is $2K. I do not know if it is for both my husband and me, or just myself. I have it saved. Unfortunately, I have already met my deductible for 2013.

        Yours is more than double mine and you may or may not have any family history. Also, the patient has fear. Fear can cripple her mentally. An MRI, even though it was negative, may have been necessary to rule out several conditions. You may now have to think about other reasons for her symptoms.
        The test may not have been for nothing.

        • NewMexicoRam

          It’s something that occurs often–I mention that the patient’s insurance won’t cover something, or will only cover a small portion of the cost, and the patient decides to decline the test or procedure.
          Yes, expensive tests or procedures are needed at times. But the point of this article was that IN THE PROFESSIONAL’S OPINION, the test was NOT needed because the odds were not in favor of finding anything serious. YOU are trying to turn this into an argument that either:
          1. The doctor has a high chance of being wrong.
          2. That “something has to be wrong” (ie, “something” can be done to correct “something.”
          or 3. That fear is the ultimate test by which we determine appropriate coverage for health care.

          Life is not a guarantee. No matter how much insurance we carry.

          • Suzi Q 38

            That is good that you at least offer it to the patient, allowing the patient to decide, after all options are considered.

            Is there another, more cost effective test that can be given to garner more information?

            Doctor, professional opinions are just what you say they are: OPINIONS. I like to deal with medical professionals that not only give me their professional opinions, but give me options.

            The author of this story did not think an MRI was needed, but maybe Ms. James and family knew of family history or a few minor prior incidents that would heighten her concern.
            I vote for this one:
            2. That “something has to be wrong” (ie, “something” can be done to correct “something.
            I would just change it a bit:
            #2 That these symptoms are so troubling to the patient that they are basic “warning signs” that something is seriously wrong that has not emerged yet. If I have had TIA’s and a bigger stroke is coming, I want to know it if possible.

            Yes, you are the professionals. We need you to help us make our medical decisions.

            On the other hand, you are not the one that gets to deal with the aftereffects of a major stroke or aneurysm should the patient have it.
            The day to day difficulty of getting medical, physical therapy, and nursing care.

            The patient and their family has to suffer and deal with that.


          • NewMexicoRam

            pcmd gets it.
            I’m not saying “die without a fight.”
            But many patients’ insistance to do everything WILL bankrupt the American health care system, eventually.
            I have no problem if you want to use your own money for more testing. Even Europe and Canada allow that.
            But insisting that all the other people who pay into your health insurance company have to cover your desire for expensive tests because of a fear factor, is beyond what America can afford anymore.

          • Suzi Q 38

            Fear factor based on the reality of a bad family history.
            But this conversation is good, because I am starting to better understand what happened to me.

            The neuro wanted to save the insurance company, so he didn’t order the c spine MRI. Only the L MRI. The L MRI showed only swelling, so he felt that it was a waste and didn’t investigate further, even with my added complaints of symptoms with every visit, and the additional necessity of a gastro for my urinary and bowels issues, which were new.

            I understand not wanting to give another $5K MRI of the upper spine. But if I am in danger of becoming paralyzed, I think even the insurance company would welcome the diagnostic test.

            Sadly, by the time the figurative BIG RED neon light came on my forehead, “UPPER SPINE” it had to come from the frustrated and angry directive of the gastro, who happened to be Chief of Staff at the teaching hospital.
            I get the need to conserve insurance funds.
            On the other hand, there must be other tests that could rule out TIA, stroke, or heart attack for the patient in the author’s story.

        • Pcpmd

          The underlying assumption here is that cost doesn’t matter as long as the risk of ANYTHING bad happening is > 0. If the risk isn’t 0 yet, you just haven’t spent/tested enough.

          You do realize that this is a recipe for madness, paranoia and bankruptcy on a national scale, correct?

          • Suzi Q 38

            You put persistent and painful headaches, nausea, pain shooting down her arm and slurred speech observed by a family member at “0″ risk? Interesting.
            Maybe she will be O.K. with no prior incidents or family history or heart disease or strokes.

            For me, along with my family history, I would request an MRI. Just because I request one, it doesn’t mean that the doctor will order one.

            I tried it with my inner thigh weakness, and the doctor talked me out of it several times. He only ordered an L spine. I kept on him, but to no avail.

            Finally, I was getting so many other little symptoms that the doctors thought was minor and attributable to other conditions that were chronic, but not serious.

            The gastro gets the prize here. He guessed upper spine.
            He demanded the MRI and all the tests that I had been asking my neuro for the last year.

            Now I have severe weakness and partial paralysis in my legs. I am barely weight baring, and am here to say that this started 2 years ago.

            It isn’t madness and paranoia.
            Maybe it is business as usual for some, but not for the patient.

          • Pcpmd

            1) Strokes/TIA’s do not cause headaches.

            2) Brain bleeds/tumors, if VERY LARGE, can cause a headache and other neurologic symptoms. These do not get better on their own. They worsen.

            If this patient was having a stroke, her neurologic deficits would not have improved within hours. By definition, a stroke is permanent brain damage. At best, she would see gradual recovery over days to weeks.

            A TIA cannot be seen on an MRI – its purely a clinical diagnosis. This is because, by definition, no damage has yet occured (the only thing an MRI can see).

            So the only role for an MRI in this situation would be to rule out

            A) Massive brain bleed
            B) Large brain tumor

            neither are very likely based on her symptoms, but if needed, a CT scan could give you that degree of reassurance, as could simply observing her for 8-12 hours.

          • NewMexicoRam

            Excellent points, pcpmd.

          • Suzi Q 38

            Thank you.
            So maybe giving the above explanation to her and then offering her a CT scan would have assuaged her fears.
            Given your detailed explanation, I would have understood and waited on the MRI, but wanted the CT scan.
            I would have also offered my fears due to family history, so
            you were clear as to what my medical future may be.

          • Susan Czarnecki

            Pcpmd, I think picked the wrong story to defend very valid points.
            Cranial aneurysms present with “the worst headache of my life”, if they CAN speak. Usually, unless a slow leak, they deteriorate in front of your eyes.
            I am wondering how long he has been in practice. It seems a very common statement from the younger physicians. I, too, have a bad, complicated story of my own and worst in my family. I have done CPR on 3 siblings. Telling my new Cardiologists this, his 1st comment was”Life is no guarantee, we cannot save everybody. ” Having heard this from several in his age group, I’m wondering if this is the new world of medicine. The newest phrase in conversations of current health care and reimbursement “is I gave up my 20″s to do this ! All my friends were having fun ” REALLY ??!? It Is business, only business. The PR departments try to pretty it up and bring the image of days past of your physician staying by your side. But they are gone. It costs too much.

          • Susan Czarnecki

            What you are saying is true but I do not agree in this situation ! I think you could have chosen a better post to support your argument. If she were in your office, looking well, no history, c/o headaches, etc, I agree. A different level of concern is reached when an ambulance arrives @ the ER with a woman with excruciating headache, slurred speech. What would you have recommended to her daughter if she called your office before calling the ambulance ? Would you actually say to her are you concerned enough to pay a $5000.00 copay ? Wait a few hours ?


      • Suzi Q 38

        I may have gone one step further. Had a neuro radiologist read my MRI, instead of a radiologist.

  • Ferkham pasha

    MRI would have helped

  • Michael Rack

    My knowledge of vascular neurology is a little rusty, but I think a CT of the brain (which is more sensitive for blood), perhaps followed by a lumbar puncture, would have been more helpful than an MRI. The patient’s symptoms sound more hemorrhagic than ischemic to me- perhaps a small leaking aneurysm?? For the non-medical people reading this blog, I am more worried about bleeding into the brain rather than a blood clot in this patient.

    • Suzi Q 38

      Thank you.

    • Maura69

      Yes, thank you!

  • Susan Czarnecki

    I pray a more Senior resident was involved in this decision making and I think Psychiatry is probably a good choice for you. The irreversible consequences of a cerebral hemorrhage or clot carries more weight then cost effectiveness in the differential I would think. Instead of seeing her risk factors, you are thinking, no worry- she’s treated ? What happened to loss of time is loss of brain ? In the door, A Rapid Response Team, CAT scan (better choice), diagnosis under an hour was the ER rule when I appeared with the same history and symptoms. I would be interested to hear how you handle these situations 4 years from now.

  • Michael Cutting

    Seems like the patient got to choose her care appropriately based upon the options provided and the physician provided the available options while discussing the benefits and outcomes, sounds like a win to me.

Most Popular