When it comes to diagnosing most back pain, MRI machines are like Monty Python’s medical machinery that goes “bing.” For back pain, MRI and X-ray are medical machines that make false alarms.
They don’t always make false alarms, of course. MRI is a miracle technology, no doubt about it — the ability to get clear images of soft tissues deep inside the body is valuable, and extremely tempting for everyone involved.12 MRI can shine when it’s actually needed and done well.
But if MRI is a great power, it’s not being used with great responsibility: it’s not needed anywhere near as much as it is actually used, and it’s especially not needed for most low back pain. Spines usually look worse than they are. Seemingly scary spinal degeneration is shown by MRI in high percentages of symptomatic people. Diagnosis based mainly on such findings is usually misleading.
Low back pain is extremely multifactorial, and the spinal glitches MRI reveals are just one ingredient in a rich stew of risk factors. Zooming in on what MRIs show is doomed to diagnostic failure. Radiology reports are usually written without clinical context, and results are often presented to as if a spinal glitch is a diagnosis in itself. But low back pain experts have long understood that you simply cannot reliably diagnose low back pain with MRI or with X-ray in isolation3 — and trying to do so reliably raises false alarms that actually do harm.4 Premature MRI is actually often worse than useless, scaring patients badly and doing real harm.
Wise use of MRI is great, but the common overuse and abuse of MRI is a disaster of over-medicalization.
The over-imaging debacle is official (and it has been for years)
This is an old issue. It has been criticized at least as far back as the early 90s. At that time, one study pointed out that almost half of MRI’s in California were “medically inappropriate” and profit motivated.5 The lameness of MRI was pointed out firmly by the American College of Physicians and the American Pain Society in their 2007 guidelines for the management of low back pain.6
Bizarrely, that message has not yet reached many therapists and doctors. Scientific journals are still publishing (and re-publishing) guidelines like these — perhaps eventually the message will get through!
The most important recommendation was that doctors should avoid giving people X-rays and MRI unless the clinical situation is really rather bad, such as severe and persistent neurological symptoms. The authors labelled this recommendation “strong,” and believe the evidence to support it is “moderate.”
I think that the recommendation should now be “stronger than strong, very strong, really extremely strong,” and the evidence supporting it can be considered “bulletproof.”
That image shows the full history of that persons life. It shows the fresh wounds that may be related to their current pains, but it also shows all of the scars that they’ve accumulated over the years. So in showing the wounds and scars, no radioloist should be giving opinons as to what features correlate with pain until they’ve seen the patient and assessed them and understood their pains. This is a general flaw in medicine. I don’t believe radiologists should write a report without the context of the pain that the person presents with.
Stuart McGill, interviewed by John Childs for Evidence In Motion podcast (2018)
What’s the big deal? Why is it so important to actually avoid using X-ray and MRI to diagnose back pain?
No one wants a false alarm, but what’s the big deal about a few diagnostic red herrings? It’s a worse problem than you might think in (at least) two major ways …
First, X-rays and MRIs genuinely spook people! It strongly reinforces the idea that something might is broken or crooked, a common and extremely misguided idea about back pain (and many, many other problems7). And nothing is worse for back pain than fear. Fear is the “back killer.”8
Second, imaging often just fails to clarify the situation, or it actually muddies the diagnostic waters. A mountain of scientific evidence clearly suggests that back pain correlates really, really badly with these test results. Many people with no pain have all kinds of things “wrong” with their backs, and vice versa. Many problems revealed by scans that seem like “obvious” problems are not. For instance, not only do at least half of “slipped” discs return to where they belong just fine on their own,9 but it’s actually the worst ones that are the most likely to resolve on their own.10 And so the diagnosis and treatment often goes spinning off in the wrong direction. This is a major part of the reason why there are such scary statistics about the economic costs of back pain.
There are exceptions — sometimes imaging finds something important — and that’s why these tests can be appropriate for some kinds of severe and persistent low back pain. But it’s just a generally lousy way to try to figure out why your back hurts.
Looking where the (high-tech) light is
MRI overuse is a classic “streetlight effect” mistake: focusing only where the light is good. MRI makes it easy to medically emphasize what seems important in back pain, but the condition of the spine is one thing, and back pain is another. Consider the results of a major 2015 review by Brinjikji et al: signs of degeneration are present in very high percentages of healthy people with no problem at all. “Many imaging-based degenerative features are likely part of normal aging and unassociated with pain.”11
So spines usually look worse than they are, and seemingly scary spinal degeneration is shown by MRI in high percentages of symptomatic people. Diagnosis based mainly on such findings is frequently misleading.
Quality control issues
There’s expert consensus on this topic because the evidence is quite clear: there are huge quality control problems with MRI, just shoddy work. Consider the hair-raising 2016 experiment that sent the same woman with back pain and sciatica to ten different MRI facilities, producing such a variety of conflicting diagnoses that it would be laughable if it weren’t so tragic.12 The reliability of MRI intepretation in the real world is crazy bad (in America, anyway — it’s definitely less of a train wreck elsewhere).
Or consider what happened when researchers tried to diagnose using only MRI — no clinical information about the patient.13 They looked for patients with stenosis — spinal canal narrowing — based on the assumption that stenosis is probably almost always painful. They failed, because so few patients with that kind of pain actually had narrowed spinal canals! And so many who actually did have narrow canals didn’t have any pain.
It’s unclear whether anything that an MRI can see is useful in predicting how things are going to go for sciatica patients with herniated discs — and, therefore,whether they might benefit from surgery. Barzouhi et al confirmed that MRI is not useful in this way by following 283 cases, half who got surgery early, and the other half who only got surgery is it seemed necessary later.14 No important differences in these groups were detected, and disc herniation size was not linked to outcomes at all.15
The good news: it’s not actually hard to do MRI right (or at least a lot better)
Doing it right mainly means taking fancy pictures of our insides only when it’s actually needed — that is, when there are ominous signs and symptoms, or significant chronicity.
But it also means using better tools for the job as needed, especially positional and dynamic imaging, which can show problems that just won’t turn up on garden variety scans. And metabolic scans, which show inflammatory hot spots, which can highlight a joint that’s in trouble right now (versus one that merely has scars from an incident many year ago). Some of these imaging techniques have greater risks, and so the need must be even greater to justify them — but sometimes these are the only tools that will get the job done.
Most importantly, it means that imaging should be always be interpreted in clinical context. It’s not a “finding” until it connects in some way with a person’s case.
More good news: patients can do a lot about this
Despite all the science and warnings, it is still routine for me to hear from patients who have been X-rayed by their chiropractors and MRI-ed by their doctors in the early stages of back pain. And it’s still routine for those people to be told that what was found on their MRI is conclusive evidence of the cause of their back pain. I have almost never heard of a patient who was told that signs of typical degeneration might be meaningless, even though that’s what every patient should be told.
Take matters into your own hands.
Ideally, better imaging is going to need better doctors and a better system, of course. But improvements can be surprisingly patient driven. With just a little bit of education, patients can…
- Cheerfully refuse premature imaging! If you get back pain, and someone tries to beam rays through you prematurely, just say, “Thanks, but no thanks. The American Pain Society says it isn’t necessary unless I can’t feel my legs.”
- Politely ask if other imaging options are appropriate. For instance, if you are being told that no cause for your back pain can be detected, but you have a clearly repeatable pain with a specific movement or position, ask if — perhaps, just maybe — you should be imaged in that position, or moving through it.
- Privately resolve to take radiology reports with a huge grain of salt, regardless of what the doctors say. You should also try to ask your doctor for a more nuanced interpretation of the results in clinical context, but that’s a lot trickier — many doctors won’t have any better idea of how to do that than you do yourself! The main thing is just to maintain a strong healthy skepticism.
The pros and cons of MRI for back pain (from an educated doctor’s perspective)
This is brilliant:
Comic by Patrick Lyons of Coogee Bay Physiotherapy.
This is a great comic, but not every reader is going to fully appreciate the humour in the doctor’s thoughts, so I’ll elaborate a bit:
What’s a “bottom up understanding of back pain,” and why’s that bad? It’s the idea that back pain comes primarily from backs (bottom up), when in fact we have really strong evidence that back pain severity and chronicity is powerfully by the brain (top down).
“Greater disability scores associated with MRI utilization.” One of the most common ways of measuring the badness of back pain is “disability” as determined by a very carefully designed questionnaire. And disability gets worse (higher scores) when MRI is involved in the assessment of back pain, probably because it “medicalizes” and dramatizes. This is a nocebo effect (opposite of placebo).16 Basically, looking for things wrong with people’s spines makes people fear their spines — which leads to hypervigilance, sensitization, and disability. Which is tragic and ironic.
“Reduced sense of well-being following exposure to MRI.” Very similar to the previous item! “Well-being” can be high even when you have a bunch of back pain … or it can be low. When back pain is over-medicalized — too much fancy diagnosis, scary treatment options bandied about — people feel worse about their situation. More worried!
About Paul Ingraham
I am a science writer in Vancouver, Canada. I was a Registered Massage Therapist for a decade and the assistant editor of ScienceBasedMedicine.org for several years. I’ve had many injuries as a runner and ultimate player, and I’ve been a chronic pain patient myself since 2015. Full bio. See you on Facebook or Twitter.
Have you got chronic back pain? Visit my advanced patient tutorial, Complete Guide to Low Back Pain Or for an interesting philosophical discussion of the general problem, see my article, Your Back Is Not Out of Alignment.
- Digital Motion X-Ray — What’s the risk from the radiation exposure? Is the diagnostic potential worth it?
- When to Worry About Low Back Pain — And when not to! What’s bark and what’s bite? Checklists nd red flags and non-scary possible explanations for alarmingly back pain
- 6 Main Causes of Morning Back Pain — Why is back pain worst first thing in the morning, and what can you do about it?
- “No brain, no pain: it is in the mind, so test results can make it worse,” Lorimer Moseley, TheConversation.com.
What’s new in this article?
May 16, 2020 — Substantial editing, and added an article summary.
2019 — Major upgrade today. Added two encouraging new sections about best practices and patient-driven improvements, plus another featured an excellent comic about the pros and cons of MRI. Described the results of Barzouhi et al. And added a helpful quote from Stu McGill. Added further reading suggestions. And put a fresh coat of editing paint on the whole thing. It’s practically a whole new article.
2018 — Clarified that MRI isn’t always useless for back pain.
2017 — Science update, added a reference to Herzog et al’s remarkable MRI reliability results.
2016 — Science update, added citation to Webster et al clearly supporting the claim that MRI results can actually cause harm with false alarms.
2007 — Publication.
- Trentu.ca [Internet]. Coughlan R. Technology Idolatory: An Exploration of Healthcare's Love Affair with Machines That Go "Bing"; 2004 Jan [cited 15 Mar 17].
“There is something wonderfully compelling about new technology whether it comes in the shape of new cars, kitchen appliances or technological advancements in the modern clinic. I want to briefly explore how our values and beliefs concerning technology may contribute to some problematic aspects of modern medical practice.”
- Swedlow A, Johnson G, Smithline N, Milstein A. Increased costs and rates of use in the California workers’ compensation system as a result of self-referral by physicians. N Engl J Med. 1992;327:1502–6. PubMed #1406882 ❐
From the abstract: “Of all the MRI scans requested by the self-referring physicians, 38 percent were found to be medically inappropriate … ”
Self-referring physicians are physicians sending patients to pain clinics or imaging facilities that they own: that is, referring patients to themselves (“You need to buy more testing from me.”).
- Deyo RA, Weinstein DO. Low Back Pain. N Engl J Med. 2001 Feb;344(5):363–70. PubMed #11172169 ❐
Way back in 2001, Deyo and Weinstein were strongly critical of overmedicalization and excessive imaging and surgery for low back pain, the poor correlation between imaging results and symptoms, and the absence of any clear pathophysiological mechanism to explain most low back pain.
- Webster BS, Bauer AZ, Choi Y, Cifuentes M, Pransky GS. Iatrogenic consequences of early magnetic resonance imaging in acute, work-related, disabling low back pain. Spine (Phila Pa 1976). 2013 Oct;38(22):1939–46. PubMed #23883826 ❐ PainSci #53341 ❐ “Early MRI without indication has a strong iatrogenic effect in acute LBP, regardless of radiculopathy status. Providers and patients should be made aware that when early MRI is not indicated, it provides no benefits, and worse outcomes are likely.”
- Swedlow A, Johnson G, Smithline N, Milstein A. Increased costs and rates of use in the California workers’ compensation system as a result of self-referral by physicians. N Engl J Med. 1992;327:1502–6. PubMed #1406882 ❐
- Chou R, Qaseem A, Snow V, et al. Diagnosis and Treatment of Low Back Pain: A Joint Clinical Practice Guideline from the American College of Physicians and the American Pain Society. Ann Intern Med. 2007 Oct 2;147(7):478–491. PubMed #17909209 ❐ PainSci #56029 ❐
Marvelously progressive, concise, and cogent guidelines for physicians on the treatment of low back pain. These guidelines almost entirely “get it right” in my opinion, and are completely consistent with recommendations I’ve been making for years on PainScience.com. They are particularly to be praised for strongly discouraging physicians from ordering imaging tests only “for patients with low back pain when severe or progressive neurologic deficits are present or when serious underlying conditions are suspected.”
- “Structuralism” is the excessive focus on causes of pain like crookedness and biomechanical problems. It’s an old and inadequate view of how pain works, but it persists because it offers comforting, marketable simplicity that is the mainstay of entire styles of therapy. For more information, see Your Back Is Not Out of Alignment: Debunking the obsession with alignment, posture, and other biomechanical bogeymen as major causes of pain.
The Litany Against Fear
In the classic science fiction novel, Dune, the Jedi-like Bene Gesserit use the “litany against fear” to focus their minds and calm themselves:
I must not fear.
Fear is the mind-killer.
Fear is the little-death that brings total obliteration.
I will face my fear.
I will permit it to pass over me and through me.
And when it has gone past I will turn the inner eye to see its path.
Where the fear has gone there will be nothing.
Only I will remain
~ “The Litany Against Fear” (Dune, by Frank Herbert)
A lot of low back pain patients could benefit from the “litany against fear.”
- Zhong M, Liu JT, Jiang H, et al. Incidence of Spontaneous Resorption of Lumbar Disc Herniation: A Meta-Analysis. Pain Physician. 2017;20(1):E45–E52. PubMed #28072796 ❐ PainSci #53529 ❐
Several studies — like Kjaer 2016, which followed dozens of patients for eight years — have suggested that herniated discs spontaneously de-herniate. This is the first meta-analysis of those studies. The pooled data from eleven studies shows an extremely high overall incidence of disk resorption: a whopping 66% in patients who received conservative therapy (anything but surgery). The number is unreliable due to low data quality, but so high that it’s safe to assume that approximately “lots” of herniations resolve on their own.
- Chiu CC, Chuang TY, Chang KH, et al. The probability of spontaneous regression of lumbar herniated disc: a systematic review. Clin Rehabil. 2015 Feb;29(2):184–95. PubMed #25009200 ❐
Not only do many lumbar disc herniations resolve on their own, or with just a little help from conservative therapy, but the worse the herniation the more likely it is to regress: exactly the opposite of what common sense predicts! This systematic review of 31 studies revealed a strong pattern of better regression from the worst cases: “Patients with disc extrusion and sequestration had a significantly higher possibility of having spontaneous regression than did those with bulging or protruding discs.”
- Brinjikji W, Luetmer PH, Comstock B, et al. Systematic Literature Review of Imaging Features of Spinal Degeneration in Asymptomatic Populations. AJNR Am J Neuroradiol. 2015 Apr;36(4):811–6. PubMed #25430861 ❐ PainSci #53872 ❐
- Herzog R, Elgort DR, Flanders AE, Moley PJ. Variability in diagnostic error rates of 10 MRI centers performing lumbar spine MRI examinations on the same patient within a 3-week period. Spine J. 2016 Nov. PubMed #27867079 ❐
People mostly assume that MRI is a reliable technology, but if you send the same patient to get ten different MRIs, interpreted by ten different radiologists from different facilities, apparently you get ten markedly different explanations for her symptoms. A 63-year-old volunteer with sciatica allowed herself to be scanned again and again and again for science. The radiologists — who did not know they were being tested — cooked up forty-nine distinct “findings.” Sixteen were unique; not one was found in all ten reports, and only one was found in nine of the ten. On average, each radiologist made about a dozen errors, seeing one or two things that weren’t there and missing about ten things that were. That’s a lot of errors, and not a lot of reliability. The authors clearly believe that some MRI providers are better than others, and that’s probably true, but we also need to ask the question: is any MRI reliable?
(See also my more informal description of this study, which includes an amazing personal example of an imaging error.)
- Haig AJ, Tong HC, Yamakawa KS, et al. Spinal stenosis, back pain, or no symptoms at all? A masked study comparing radiologic and electrodiagnostic diagnoses to the clinical impression. Archives of Physical Medicine & Rehabilitation. 2006 Jul;87(7):897–903. PubMed #16813774 ❐
In this study, about 150 people were assessed for back pain in different ways, including MRI, but “radiologic and clinical impression had no relation.” In other words, there was no useful similarity between evaluation of the patient with MRI, and evaluation by examination and taking a history. “The impression obtained from an MRI scan does not determine whether lumbar stenosis is a cause of pain.” Since MRI does in fact identify narrowing of the spinal canal, and this is the whole basis of diagnosing spinal stenosis with MRI, these results also strongly imply that a narrowed spinal canal does not (alone) cause back pain.
- El Barzouhi A, Verwoerd AJ, Peul WC, et al. Prognostic value of magnetic resonance imaging findings in patients with sciatica. J Neurosurg Spine. 2016 Jun;24(6):978–85. PubMed #26871651 ❐ PainSci #53585 ❐
- More detail: No important differences in these groups were detected: regardless of whether they got surgery early or not, only the original degree of nerve root compression and disc extrusion was associated with their outcomes … and only weakly. Disc herniation size was not associated with outcomes at all. In other words, big or small, herniations do not tell us anything about how a case of sciatica will work out.
“Nocebo” is roughly the opposite of placebo: harm powered by belief, instead of relief.
Latin for “I shall harm” (which I think would make a great supervillain slogan). It refers to the harmful effect of … nothing but the belief in or fear of a harmful effect. Give someone a sugar pill and then convince them you actually just fed them a deadly poison, and you will probably witness a robust nocebo effect. A common funny-if-it’s-not-you nocebo in general medicine is the terror of “beets in the toilet”: people eat beets, and then think there’s blood in the toilet, and call 911. Nocebo is a real thing, and not to be messed with. It is one of the chief hazards of excessive X-raying and MRI scanning, for instance: showing people hard evidence of problems that often aren’t actually a problem.