MRI overuse is widespread, and dangerous to patients

MRI overuse and misuse really make me angry.

Did you know that a growing number of doctors don’t even read the tests themselves? Another shocking aspect of this industry is that some doctors will order an MRI because it’s quicker than doing a full examination. Rushing to an MRI can oftentimes erase clinical correlation, that is, connecting the dots between pain and what the MRI shows.

MRI is unnecessarily overused. In a study of 221 patients who had MRIs, the results showed that only 5.9% actually needed to have an MRI done. The remaining 94.1% of the patients sacrificed their time and money. What’s worse is that the use of MRI for screening isn’t as effective as other methods.

MRI needs to be judicious and scrutinized by the physician who ordered it. It can be as dangerous as it is useful. Many patients view the MRI as a security blanket, and will go as far as requesting it. I have numerous cases of diagnostic problems that result from misuse of the MRI. More here on clinical correlation (making sure that what’s seen on MRI is in agreement with the patients problem) and the fact that any test including an MRI must be ordered to confirm a preliminary diagnosis that is already known from the history, exam, and more simple, inexpensive tests such as an x-ray.

If you suspect your doctor is just being quick or using MRI to reach that “aha” moment, then you’re in a bad scenario. When I order an MRI, I am 90% certain about what the results are going to show. Doctors need to have a clear-cut idea on what they can expect to see from the results. Next time you’re told to get an MRI, and your doctor has little clue to your diagnosis, you may want to get a second opinion. Also, be sure to ask the physician if they read the MRI themselves.

Nothing makes me quite as angry as the doctors who can’t or won’t read the MRI themselves.

Have you ever had an MRI? Did you question the necessity of it? Do you have a horror story?

“Angry Orthopod” is an orthopedic surgeon who blogs at his self-titled site, The Angry Orthopod.

Submit a guest post and be heard on social media’s leading physician voice.

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

  • Rob Lindeman

    A modest proposal: ask the consumer to pay for the service he receives (or a significant portion of it). What do you suppose will happen to demand?

  • drdarrellwhite

    I get everything that you are saying and largely agree. But tell me, where is the “danger” to the patient in the MRI? Your title implies a direct danger from the MRI itself, which we both know is not the case (unlike frequent CT, for example). Are you referencing your concerns about the interruption of “clinical correlation” in the diagnostic process? Again, I agree, but we as physicians have a responsibility to use words like “dangerous” with some caution and care, especially in settings where our audience includes non-medical personnel.

    A good piece, maybe weakened a bit by hyperbole in the title, but words of wisdom that, if followed, would dramatically reduce unnecessary medical spending.

    • Kevin

      I create all titles, so the blame should be directed at me, not Angry Orthopod.


      • ninguem

        Understood, Kevin.

        Just want to make sure. Were you using the term “dangerous” by the MRI provoking an unnecessary surgery?

        Or are you saying there is something inherently dangerous about being exposed to the magnetic field?

        Besides the known dangers of the oxygen cylinder flying across the room, or internal damage because of a metallic object in the patient’s body.

      • drdarrellwhite

        Got it. Thanks, Kevin.

  • doc99

    Rotator cuff, yes or no?

  • Marina

    My doctor ordered an MRI for an injury to both feet. An x-ray already ruled out bone damage. There was no opening for an MRI for 2 weeks. When I arrived for the test, I was told that they only had time to test one foot. I would have to come back for the other foot. So I left without getting tested. It felt like a total waste of time and money. I had already spent time using Advil and a walking cast. The injury wasn’t fresh.

    Later, I looked up the professional society’s official standard of care for my injury. An MRI was not recommended.

    • Marc Beson

      Actually, an x-ray (radiographs) cannot “rule out” bone damage. It’s because the sensitivity of x-rays is certainly limited. So, it was good practice to get and MRi if the x-ray cannot explain the signs and symptoms. In contrast, it would have seemed bad practice to stop after the negative x-ray, tell the patient there is nothing and send him home. There are hundreds of cases where x-ray did not show a fracture, but MRI did; where x-ray did not show signs of cartilage damage but MRI did. Plus, abnormalities of synovium, tendons, ligaments, muscle, and more require MRI.

      • Marina

        Indeed, when it says in plain wording that an MRI is not recommended for the diagnosis I was given I must have been mistaken.

  • Paul Weiss

    Medicare studies have shown that those who live in areas where MRI’s are more utilized for low back pain are more likely to undergo complex lumbar spinal fusion surgery. In those areas where more surgeries are done, outcomes are not any better. In fact, they may be somewhat worse.

    While the MRI itself is not harmful, the end result being that the patient has a potentially unnecessary surgery as a result is dangerous.

    One other point that AO did not make is that MRI utilization is heavily influenced by physician ownership. Orthopods who have an ownership interest in a scanner are more likely to order tests than those who do not.

    I’ve seen numerous patients who had an MRI for neck or back pain and the results were misinterpreted. It is not uncommon for these tests to have false positives. I’ve heard through patients of their physicians who were perplexed why the patient had pain or weakness on the right side when there was a disc protrusion on the left. I know of a person who underwent unsuccessful spine surgery for such a scenario. (It turned out that he has ALS.)

    I believe that unnecessary MRI’s can be dangerous within the larger scope of their results being misinterpreted.

    I appreciate greatly the post. I am curious as to the reference for only 5.9% of MRI’s ordered being necessary.

    • Rob Lindeman

      Foot Ankle Int. 2007 Feb;28(2):166-8.
      The diagnostic value of MRI in foot and ankle surgery.
      Tocci SL, Madom IA, Bradley MP, Langer PR, DiGiovanni CW.
      University Orthopedics, Inc., Providence, RI 02904, USA.
      MRI is being used with increasing frequency and seems to have become more popular as a screening tool rather than as an adjunct to narrow specific diagnoses or plan operative interventions. Our hypothesis was that the rising accessibility of this test may be resulting in its overuse.

      We retrospectively reviewed 221 consecutive patients referred over a 3-month period for treatment of a lower extremity problem to determine: (1) the number of patients who presented with an MRI already obtained from an outside source, (2) the number of patients who obtained an MRI from the foot and ankle specialist after referral, and (3) the number of times the foot and ankle specialist used these studies or found them helpful in the care of the patient. Fractures (20) were excluded.

      Of the 201 patients without fractures included in the study, 19.9% (40 of 201) had MRI scans during the course of their treatment; 15.4% (31 of 201) presented to their initial visit with an MRI scan from an outside source, and 4.5% (9 of 201) received MRI scans as ordered by the foot and ankle specialist. Eighty-seven percent (27 of 31) of the pre-referral MRI scans were thought to be unnecessary, and 48.4% (15 of 31) had radiographic interpretations that were considered either immaterial to the patient’s pertinent clinical diagnosis or inconsistent with the specialist’s interpretations. All nine MRI scans ordered by the specialist were useful in the care of the patient. Therefore, of the 221 consecutive patients, the foot and ankle specialist would have ordered MRI scans in only 5.9% (13 of 221).

      This study suggests that many of the pre-referral foot or ankle MRI scans obtained before evaluation by a foot and ankle specialist are not necessary. Further studies need to be performed to determine the role of MRI in the screening of foot and ankle disorders.

      • stitch

        Great. But what if the specialist won’t schedule the patient without an MRI? As the PCP, I’ve had this happen even when I didn’t think the MRI was warranted.

  • AR

    As an untrained eye (compared to a skilled radiologist), and using a scan only as a diagnostic, not for planning a procedure, why is it wrong not to look at the MRI itself?
    By only reading the report, you treat it as a lab value — as an indication of whether or not your diagnosis based on clinical findings is accurate. You won’t be swayed to treat or not treat based on what you might think you see, but on what has actually been found by an expert, which could give scans much more reproducibility and reliability as clinical tools.

  • rezmed09

    Yes, it can be more expense to the patients, and yes, it does waste their time. But the big question, is whether or not the patients were happier because they had the mostly useless test.
    Patients in our system often get what they want no matter how useless or expensive.

    • Paul Weiss

      Then the population in general needs to be better educated.

      Unfortunately, there is no money to be made in educating the general public on tests they do not need. (Unlike how there is money to be made in educating the general public on medications they may not need but have potentially dangerous side effects. I am referring to drug ads on TV.)

  • Dr. Dredd

    The Angry Orthopod can be as angry as he/she wants to be… in the comfort of a subspecialty practice. Any patient questions besides those addressing a bone-related issue can be answered by, “Ask your primary about that.”

    If primary care doctors were actually reimbursed fairly for their cognitive efforts, perhaps they would be more willing and better able to weed out the unnecessary scans. But, as Paul Weiss said above, there isn’t any money in trying to convince someone they don’t need a test. When I’m trying to address an ever-increasing number of “quality” measures in less time for less reimbursement, arguing someone out of an MRI just may not get done.

  • Loretta Craig

    My husband, age 75, had some bad dizziness & vomiting, went to ER, they did an MRI. He was transferred to a different hosp. the same day and was given a 2nd MRI, along with EEG and several others. Diag: Seizures. His large head / neck was jammed into the machine. He’s had trouble with his neck ever since! Saw family Dr., he said dizziness was from little beads in ears that get in the wrong place. He was dizzy for 3-4 weeks and then ok, but State Law in Texas: no driving for 6 months! A few months later he sat down in a lawn chair on a slick floor in a Big Box store and it slid back into a support beam hitting the back of his head/shoulders HARD! MRI #3, neck jammed again. Test wasn’t good for some reason, they wanted him to have another one. We hunted for a place that said they had a newer, larger MRI, made a trip across town, neck still got jammed. MRI #4. Then May of this year, he could only see half the page of the diabetic record book he keeps, I called EMS, MRI#5, diag. was TIA. He was there in ER about 5 hours, the bill was $9,043.00, which included two “Xrays” at $2077@ and one at $376.00. Also, a prescription drug for $1200.00! Our dr. thinks it’s Migraine, wants him to see a Neurologist.
    I see another MRI on the horizon!! He says he will refuse it. His neck is in constant pain, now they said he has Arthritis in his neck.

    Rezmed09 Happier? Are you kidding!?!


    As an assistant professor in a Family Medicine residency program my residents wondered why I was insisting on a working diagnosis prior to ordering any study and then they had to convince me that it would have an impact on their clinical decision making and or treatment process. I have always preached the best and most economical test is another brain (provided you choose wisely.) I have experienced consultants who requested an MRI prior to seeing my patient and make it a point to not refer to thenm.

  • Happy Hospitalist

    I’m confused. If a doctor is 90% certain of the diagnosis, why do they need an MRI to confirm it? How much certainty is required to manage the patient? 100%?

    • pj

      point taken Happy. But if your going to use this logic, why send all surgical specimens to pathology? Show me one hospital where this isn’t done….

      In most cases the Dx is known or not at all thought to be mailgnant- yet anything that comes out of a person goes to the lab.

      ANd what’s wrong w/trying to achieve 100% anyway?

      • The Happy Hospitalist

        Testing should be used to assist in the evaluation and management decisions. If you think an MRI is going to change your management decisions, there is rational for doing it. If not, it should be bypassed.

        Why are all specimens sent to path? Great question. Do you have the answer? Because I don’t.

        Nothing in life is 100% except death and taxes. You know that.

        • pj

          When bone/joint surgery is contemplated, an MRI is reasonable indeed. It could be called defensive medicine, but preop imaging helps reduce intraoperative surprises.

          Overall I agree MRI’s, and many other tests, are WAY overused in the US.

          The answer I was given from a pathologist is that tissue is sent to document, from a third party what the surgeon did.

          Would you agree that 100% accuracy is better than 90%? Not achievable, but simply something to strive for. Wrong Dx’s will occur, but the point is, try to minimize as much as possible.

          • Paul Weiss

            “When bone/joint surgery is contemplated, an MRI is reasonable indeed.”

            When a surgeon has a reasonably certain working diagnosis that has been shown to experience improved outcomes as a result of surgery and the patient has failed a reasonable course of conservative management, then that is sound reasoning.

            However, in the area of spine surgery, many look to the MRI to diagnosis the problem. They order it quickly and before offering conservative management. A false positive is found, the patient’s symptoms are attributed to it, and it is cut on. This is the main reason for failed back surgery.

            In a system where a surgeon can earn more money by referring more patients to an MRI center that they own and then doing more surgeries on patients who are not clearly going to benefit, unnecessary tests and interventions are rewarded.

        • Nothing in life is 100%…

          death, taxes, AND menopause…if you are a woman and lucky enough to live that long.

          great debate and catchy title. kudos.

          • pj

            Paul- I totally agree. Spinal MRI’s in particular seem to be fraught w/danger simply because they can be so misleading.

            I tell my patients, back pain is not nearly as scientific as we want it to be. Don’t live or die by the MRI.

            It would also help if more people, including Doctors, realized that chronic severe back pain is seldom cureable with surgery.

            I know many opioid dependent patients with severe chronic low back pain who are told by well meaning friends and Docs, “Just get it fixed!”

            If only it worked that way….

          • stitch

            @pj – it would also help if patients did not have to pay $20-40 dollars as a copay for each PT session. That’s a huge barrier for many of them to pursue conservative care.

  • philly

    I don’t understand. Do you want the primary care physician to look at the MRI? the orthopod? Usually orthopods or radiologists read MRIs. Why would a primary care physician try to look at an MRI.

    Radiologists KNOW MRIs are over-utilized. They cannot be at fault for the overuse of MRIs when THEY are not the ones ordering the study. Orthopedic surgeons and referring physicians are the ones who need to stop relying on imaging all the time and, as you said, at least have a diagnosis in mind.

  • Paul Weiss
  • Elizabeth Rowe

    A major point missing here is that MRI’s ordered from a hospital owned outpatient facility or a radiology owned facility is being read by a general radiologist. Therefore no physician should rely on their diagnosis of any condition. I agree that only docs who can read MRI’s and know the differential they are interested in should order them, and they should read them themselves.
    Primary care docs should not order MRI’s–the patients should first be evaluated by the relevant specialist, who then decides the need for imaging based on knowledge of the relevant anatomy and clinical situation. And he/she should look at the scans and be able to interpret them, since he/she surely knows this anatomy better than a general radiologist.

    Also, by the way, the idea that physicians who own MRI’s order more than those who don’t is not true-very biased reporting on that, promulgated by the American College of Radiology trying to maintain their monopoly. And radiologists DO order imaging–many (anecdotally more than half, but now being documented) radiology reports suggest additional imaging.

    And one more point: hospital outpatient imaging facilities are paid by both Medicare and private insurers up to 3 times more than outpatient facilities, so cost conscious docs should direct their patients away from the hospital owned facilities. For the same reason, physician owners of imaging facilities save the system money, even if it were true that they were a little quicker to order an MRI than another physican willing to “wait and see” to rule out things.

    • Dr. Dredd

      If a primary care doctor were to refer everyone to a specialist before getting imaging studies, the wait times for the specialists would be astronomical. Give us a little credit for knowing how to start a work-up.

  • Jackie

    Interesting discussion. I do have a horror story, though it is on the opposite side.

    I’ve been having MRIs annually since 1990 when my life-long brain tumor was ‘finally’ diagnosed. It was done just before the surgery after a CT scan had found the mass. That first CT had been delayed for at least 8 months from the time when I had first experienced the excruciating headaches. Switched to another family doctor; demanded three, four referrals, and the 4x5x6.5 cm tumor was finally found. The neurosurgeon who had saved my life told me to have an MRI ‘every’ year.

    I only skipped the annual MRI once – trying to save everyone some money, Unfortunately the bad work environment caused my tumors to suddenly grow rapidly and I ended up having to have a GKRS done after the ‘bi-annual’ MRI the following year.

    Next week I am going to have my 4th MRI in 8 months because one of the tumors, which is located near the thalamus, has been giving me problems. I’m supposed to exercise regularly to reduce the risk of cancer (already have had two BC surgeries and chemo) and improve my overall well-being, but the tumor prevents me from doing anything rigorous. Now the pressure is getting worse and I have to take a 600mg Ibuprofen daily.

    The neurosurgeon doesn’t want to touch it. The raidation oncologist doesn’t want to treat it unless it is absolutely necessary. “You’ve been functioning so well…,” she lamented. But she had to order a new MRI because I’ve got new symptoms – constant running nose on one side.

    The title of the article alerted me because of the phrase ‘and dangerous to patients’, even though I remember MRI is supposed to be pretty safe. From what I understand, It might take 50-100 MRIs to receive the equivalent amount of radation received from a CT scan. So the term ‘dangerous’ has to be referring to the psychological impact such as false sense of security. What would the world become if everyone start to only pay attention to the test result of the machines and totally neglect their own ‘feelings’?

    But I will take that MRI…

    • Penny

      The key part I’m nervous about, Jackie, is that the FDA has black boxed gadolinium, the contrast often given with these tumours. You’re right, MRI’s “SEEM” to be safer than x-rays and CT scans but I don’t care for heavy metal to be injected in my head on a constant basis. In fact I’m terrified of it each time and feel as though it’s a game of Russian Roulette! I’m not sure I believe MRI’s are as safe as they say either. Once when I got one I told the guy it was warm. Right afterward I had 100 red spots on my body (but didn’t feel anything more. He had turned it down. They were all on my legs. Within a couple of weeks they turned into permanent freckles — but no one would either listen to my story or look at the results. When something is promoted as good in the medical field, it’s good, and they don’t want to learn anything contrary. It is true that the MRI pictures are 100% better than x-ray, but I wish the gadolinium wasn’t necessary.

      • pj

        The main danger of an MRI is that it opens a pandora’s box of incidental findings, giving patients and their Docs new reasons for potential concern.

        They can also be very misleading. There can be a very poor correlation between MRI findings (of the spine in particular) with people’s symptoms.

  • Penny

    Yikes. Are these dangerous? Most patients believe they are harmless. I have had about six with gadolinium. The last one was the longest ever – about 30 minutes long!

  • Skeptical Scalpel

    I agree that calling the MRI “dangerous” is wrong, especially because many want to replace abdominal CT scans with MRI.

    I also agree with Happy Hospitalist who wonders why you would order a test when you know what the answer will be 90% of the time.

  • Dee RN

    I began my Radiology career as an RN at a premier NMR imaging facility in 1989. I would not call it “dangerous” to undergo an MR as the headline here states. I will say that MR is unmistakenly superior to any xray when the issue points to muscles,nerves,tendons and not the bone. Are they over prescribed? Probably so but it does get to the cause of the symptoms (most of the time) quickly. As far as an ordering MD “reading” the scan, an experienced Radiologist is best and even the most experienced Neurosurgeon is known to call the Neuroradiologist to discuss even the most basic exams if there is any question.

  • Daniel Altman

    HAHAHAHA… read the MRI themselves. That’s hilarious… and malpractice if you are not a radiologists.

Most Popular