A picture is not always worth a thousand words…

The saying “a picture is worth a thousand words” is often used to explain how a complex situation, idea, or thought can be conveyed with a single picture. There are many examples of when this is true, such as in the media when a poignant photograph expresses some joyful, or unfortunately more often these days, some heartbreaking situation far better than any article can. 


There are also many examples when pictures can convey information effectively and easily within healthcare, such as the use of an infographic to help disseminate scientific literature like the excellent Yann Le Meur  @YLMSportScience and Chris Beardsley @SandCResearch do very well. There are also many images that can express an idea or message simpler, quicker, easier than a blog or an article can such as my ‘physio treatment pyramid‘ or ‘road to recovery‘ pictures.


However, there are many times when a picture is NOT worth a thousand words and when it does NOT covey complex information very well. Instead, some pictures can oversimplify, add more confusion and end up being more harmful than helpful.

For example images of crumbling bones, glass inside knees, and doughnuts shooting jam out being used as so-called educational aids or to help warn people not to ‘slip a disc’ when they bend over are without doubt not helpful, but harmful, myth perpetuating crap.

There are also many examples of when medical images are not put to good use such as X-rays, MRIs, and ultrasounds. There is a commonly held belief by both patients and many clinicians that an image of a herniated disc, a degenerative joint, a torn tendon can explain why someone hurts or doesn’t functioning properly, and yes at times this can, but unfortunately in a lot of cases it isn’t as simple as this, and the use and dependence on medical imaging needs to be questioned a lot more.

However, I do want to make it clear first that there is no doubt that advances in medical imaging and technology have improved and enhanced modern healthcare hugely. From the first accidental discovery of the X-Ray by Wilhelm Roentgen back in 1895 being quickly put to good use by battlefield surgeons to locate bullets and shrapnel in wounded soldiers, to the development of techniques such as CT, MRI, ultrasound etc, have all helped to identify serious life-threatening diseases, illnesses, and injury’s quicker, faster and saved millions of lives. Simply put modern medical imaging has and will continue to be an invaluable tool… when used appropriately.

But where medical imaging is failing is when it is being used inappropriately and incorrectly to explain why some things hurt. Many patients and clinicians still think that pain and disability are as simple as a structural irregularity seen on a scan, that when something is misshapen, torn, degenerative seen on a scan then it must be a source of pain or problems.


Now at times it can be this simple, break your ankle and it will no doubt hurt due to the broken bones, but what if you haven’t had an accident, what if the broken bone heals and the pain persists, what then? Well, this is where it all gets a little bit more confusing and uncertain.

There is growing evidence that demonstrates many so-called worn out, misshapen, degenerative, and even torn things seen on scans are found commonly in people with NO pain and NO disability. We are beginning to recognise and understand that what we thought to be pathology may just be morphology, or ageology, or normology! I’ve just made those last two words up by the way!

For example, Guermazi et al demonstrated that many if not ALL the common pathologies seen in knee scans such as meniscal lesions, synovitis, and articular cartilage damage are found just as much, if not sometimes more in subjects WITHOUT pain as those with pain.

Next Brinjiki et al showed a high prevalence of so-called pathology in lower back MRIs in well over 3000 people without any pain or issues in ages from 20 to 80 years old, and Nakashima et al showed similar results in over 1200 neck MRIs in subjects again with NO pain or issues.


Then in the shoulder, we have Grisih et al who found a staggering 96% of subjects with no pain or issues had at least one so-called pathology on their US scans. Then Teunis et al shows an increasing prevalence non painful cuff tears with age and with up to 65% being non-symptomatic, or Schwartzberg et al demonstrating 72% of middle-aged subjects had non-symptomatic SLAP lesions, or Le Goff et al showing that over 50% of calcific deposits seen in cuff tendons are non-symptomatic, or Lesinak et al highlighting that nearly 50% of young elite level professional baseball pitchers have cuff tears and SLAP lesions with no effect on their performance.

And this is just in the shoulder, I could go on and on showing study after study conducted on symptom-free subjects that shows so-called pathology in many areas that do NOT cause ANY pain, or ANY dysfunction!

So we can see it is not just as simple as seeing a worn out, misshapen, or torn thing on a scan is causing pain or needs to be treated or fixed. A saying that is often used to explain this, is that many things seen on scans are just signs of ‘aging on the inside’, just like we see signs of ageing on the outside like wrinkles in the skin and grey hair on our heads.

Another way to explain this can be with a picture of a telephone, just because you see this picture of a telephone, you can’t tell if its ringing or not? This is the same for a picture of a torn or worn out structure, we can’t tell if its ringing (i.e painful) or quiet (i.e non painful), that’s why we need to use the scan COMBINED with a full history and of course a physical examination.


And another way to explain these things seen on imaging that I also like comes from David Poulter @Retlouping. He explains that a scan can be like a wedding photograph, just like mine above. By looking at the group photo with everyone looking happy you can’t actually tell who is happy, who is sad, who is angry, or even pissed etc… just like we can not tell by looking at an MRI with all the things reported as problematic which actually are, and which are not.

While we are on the subject of medical imagine I want to discuss a few more issues I come across that are becoming more and more frustrating. The first is that many think scans help reassure patients and so I am hearing more and more clinicians referring patients for scans just to reassure them when there is absolutely no clinical indication.

Not only is this wasteful, and lazy practice, there is little evidence that scans help reassure patients (ref), in fact there is evidence that they can do the opposite and cause more fear, angst, and harm (ref).

To put this as bluntly as I can it is the job of the CLINICIAN to reassure a patient, not a fucking scan!

Next I want to discuss some issues with the reporting that occurs on many scans. Now I have many radiologist colleagues and friends and they are without doubt experts in reading and interpreting medical imaging, and their role in doing this is invaluable to me and the healthcare profession in general.

However, many radiologists do not have the time to take the patient history’s or conduct a full exam of the patient. Yet despite this, I am seeing more and more patients being informed by radiologists what treatment or surgery they need based on what they see on a scan alone.

This drives me bonkers, as more often than not the recommendations are incorrect, however, once the seed is planted into a patients head it is really, really hard if not impossible to unpick it, and when I try to give an alternative it often leads to confusion and mistrust for the patient as they are not sure who best to beleive.

So please radiologists if you are reading this, please stop giving diagnostic and treatment advice based soley on what you see on a scan.

Whilst still on the subject of reporting it should be recognised and understood that although this is a skilled job it is still remarkably unreliable. This recent study by Herzog et al  highlights the lack of consensus in the interpretation of just one patients lumbar spine MRI.

This one patient visited 10 different medical centres and had 10 MRIs within a 3 week period. Each scan was reported by 10 different radiologists and had a total of 49 different pathologies reported. However no pathology, that’s none, zero, zilch, nil, nada were reported consistently by all 10 radiologists, not one! As the saying goes, interpretation really is all in the ‘eye of the beholder’.

Finally, I think if more radiologists were aware of the prevalence of changes seen in the non-symptomatic populations more this would change the way they interpret and report images, which can have a huge impact on patients.

For example, this study here shows that when an image report is put into context with epidemiological data patients are less likely to be prescribed unnecessary medications by their clinician. And this study here demonstrated that by simply rewording reports again with epidemiological data and in a positive light, such as changing the word tear for high signal or defect, it was shown to improve patient understanding and satisfaction.

So in summary we can see that pictures and images used in healthcare have both positive but also negative effects. I think it is all healthcare professionals responsibility to try and maximize the positive and reduce the negative when it comes to medical imaging. Clearly it has a role, but we need to use it wisely and sensibly.

I will leave you with the brilliant acronyms used by Richard Heyward in his editorial on the issues of imaging in the BMJ back in 2003 here. When it comes to medical imaging and scans do not B.A.R.F or create V.O.M.I.T, that is do not use Brainless Application of Radiological Findings and create Victims Of Modern Imaging Technology

Never treat the scan, always treat the human!

As always thanks for reading


17 thoughts on “A picture is not always worth a thousand words…

  1. Well done article. It makes me think that an entrepreneurial enterprise would be to create software that includes epidemiological data of imaging findings to amend the reported findings and reduce the nocebo effect, which, as we all know, is very difficult to undue.

    One suggestion, please amend your last line to, “don’t treat the scan, treat the man, or woman. ” thank you 🙂

    • Thank you for your comments.

      The saying don’t treat the scan treat the man, is merely a metaphor than a literal statement. It’s not meant to be sexist but rather a just a saying. If it’s changed it won’t have the same ring or appeal to it!

  2. I recently had to go see a doctor for lateral epicondylitis and he insisted on Xray for my neck and told me I would “never get better” with my elbow because my neck showed “severe degenerative arthritis” the “worst he’d ever seen”. My neck doesn’t hurt, my elbow is better with strengthening and that doctor is the worst kind of idiot. Enough said.

  3. Great post. We’re dealing with this on a daily basis with the chronic pain population. Over the last decade, prescriptions for opioids, MRI scans, spinal injections and surgery levels for back pain have risen by 20-30%, disability levels for lower back pain have increased by 15-20%, with a chronicity prevalence of 80% and life time prevalence reaching 60-70%! Even more alarmingly, back pain prevalence among adolescents and children have also risen, being 3-4 times higher in Western Europe compared to the rest of the world. 133 million work days are lost annually in the UK with lower back pain listed as the most prevalent cause. That is a not only a huge loss of productivity and cost, but a damning presentation of what could be called a failing model of health care.

    Just like a nocebo effect, patients being told their scans shows “damage” or “crumbling”, or their back is “worn out”, “looks like that of a 70 year old” or is “stiff”, “immobile”, “out of alignment”, often end up with further fear of movement, increasing tension and stiffness, persisting pain, sleeplessness, health anxiety, low mood which again affects their ability to work or stay in employment – affecting financial health and further increasing anxiety in a never ending evil circle. And here we are, we’ve succeeded in transforming our patient into a nice, big VOMIT.

    Only after people have been through packets of medication with ever increasing strength – from painkillers to antidepressants and gastro protectors, then eventually ending in repeated injections of corticosteroids – only then does the patient end up in a pain clinic, where they more often than not are told their pain cannot be taken away and they need to learn how to live with it. Can you imagine the impact that this can have on someone’s self efficacy and rehab potential? Only then they might be introduced to mindfulness, pacing, the actual importance of diet and exercise, sleep hygiene, talking therapy, movement retraining, education and advice about what is actually happening in their nerves, muscles and brain. And whilst it is never too late to improve someone’s symptoms, it is certainly much more difficult once all aspects of the patient’s life have been affected, and significant health care beliefs have been formed.

    Never treat the scan, always treat the man indeed!

  4. I am brand new to your blog and can already tell you are my kind of therapist. I deal with this every day, despite the fact that I work with a great team of doctors and radiologists. My explanations usually get through, but it can be difficult. The latest thing I am seeing a lot is patients coming to therapy because their MRI will not be approved until they have a trial of therapy. While I still believe that insurance companies are evil, I have to say that I am not completely opposed to this. It gives me another tool in my arsenal to explain to the patient that even if the MRI showed the worst, that a trial of therapy would still be the best course of action. And very often, after I get them moving for a couple of weeks, they end up not needing the MRI.

    Great article. Keep it up!

  5. Hi Adam. If you are who I think you are I am saddened, disappointed and extremely angry at your comments. As a radiologist and clinician I find your comments rather u professional. Have you spent 5 years at medical school? Have you practiced medicine and surgery after that and gained a further degree in medicine or surgery? Have trained in CLINICAL DIAGNOSTIC RADIOLOGY? I think not. I am diagnostic radiology clinician. We so take histories and examine. Ita what we are trained to do. It is very dangerous to pick up a prime and say I’m going to learn this by going on a short course ans then claim to be an expert. So you work un the NHS most of the time? I believe not. So you thought ill make money this way. There are good docs and there are not so good docs. A radiologist can provide an examination amd report that is meaningful amd clinical based on their years of medical training and knowledge. Who do you go to when u have a difficult case?

    • Dear Dr Chahil

      Thank you for your comments, however I am a little confused at your angst and slightly upset and disappointed at your allegations. First as my profile clearly states I am both an ESP for the West Hertfordhsire Hospital Trust as well as a clinical lead for a private healthcare company in Hertfordshire. And I do work over 50% of my time within the NHS as a front line autonomous clinician in three busy general hospitals, and if you think I make any money from my blogs you are very much mistaken, and I resent the accusation that this or any other of my blogs are doing this

      Next I am at a loss as why or how you think this blog is unprofessional. I make it very clear if you read again that I have many radiologist colleagues and friends who do an invaluable job and who help me both professionally and personally a great deal. This is not in question. Nor have I stated otherwise.

      However, what is in question is that many clinicians, that is both radiologists, physios, doctors, etc etc still confuse pathology with morphology and think pain is simply explained with what is seen on scans. It is not this simple and the many papers I quote highlight this, and I will not apologize for raising this issue and making more aware of it, as I routinely have to correct patients who have been misinformed by others that their scan shows X and this explains Y and need Z, when this is not the case.

      Also I am trained in radiology, and physiology, and anatomy, and pain science, and history taking, and clinical examination etc etc. but I do not have a degree in medicine or surgery, I don’t need one to know who needs, or doesn’t need onward referral for medicine or surgery.

      Finally your appeal to authority and experience is tiresome and mistaken. Just because you have years of practice, this shouldn’t be confused with equal years of knowledge. I have only a years knowledge, but this has been refreshed every year for the last 15 years.

      I would however be interested in your views on the recent Herzog et al MRI reliability study in the case of the patient with low back pain and the lack of consensus between radiologists.

      Kind regards


  6. Hi Adam, thanks a lot for the blog – a crucial issue and I’m sure a day-to-day sticking point for physiotherapists the world over. I strongly agree with your message however I fear that this can be taken too literally by some clinicians i.e. that a scan is always irrelevant to understanding why someone has pain (beyond the assessment for red flag pathology). For example, you quote the Guermazi et al (2012) study showing a high prevalence of ‘abnormalities’ in asymptomatic knees on MRI. However, it should be noted that patients with any evidence of radiographic OA were excluded in this study. From pooled data which included that same cohort, Neogi et al (2009) actually showed a strong correlation between signs of radiographic OA and knee pain. Serbian (2016) also produced similar findings examining patients with bilateral knee OA, that cartilage loss emerged as an independent predictor of both pain intensity and WOMAC score.

    I feel caution is needed our interpretation of the research here. Neogi et al provide a good discussion of the research and discrepancies between studies in their paper. Pain and function can be closely related to scans in some joints/circumstances and I wonder if we will see replication of the above knee example in other joints such as the foot/ankle. However, as clinicians, we are told not to use terms such as “wear and tear” or “degeneration,” nor to give them negative prognoses, even when the clinical picture suggests such. How then, to communicate these findings honestly in a way that avoids, or at least minimises, distress in the patient in addition to avoiding medical jargon? This is a clinical challenge and I’d be grateful for your thoughts.

    • Hi Mike thanks for your comments and I was not aware of the Neogi paper so thanks for pointing me towards that.

      You make excellent points, and I do try to stress that medical imaging is useful and needed when appropriate and in conjunction with a good history and physical exam!

      As for words to use when we explain these findings and the pain patients experience this is not a simple answer, and think it needs to be done on a case by case basis. Personally I don’t see much wrong in calling arthritis arthritis, as for other terms e.g. Wear and tear, degenerative joint disease etc well these can induce fear and threat in SOME patients so caution with these terms is important!



  7. ” what radiologists are not is diagnostic clinicians”

    You lack the knowledge that the actual job title is a ‘clinical radiologist’, following formal trainingin both medical and/ or surgical specialties prior to radiology training. I would encourage you to go through 5+ years of medical school followed by a minimum of 7 but usually 10+ years of postgraduate training and countless exams to become a consultant. You would then be qualified to offer an opinion on imaging and patient management at the same level. You seem to hold yourself to the same regard as a consultant sports physician/ orthopaedic surgeon etc in terms of your understanding of the ‘clinical picture’. I wish you and your patients all the best.

    • I do not hold myself in any regard Sumeetra. I know there are clinical radiologists however I also wonder how many clinical radiologists do actually take a full history and do a full physical examination including functional testing etc, as well as then have the time to conduct the imaging tests requested?

  8. Well said! So incredibly frustrating. Especially when you combine these image findings with the highly anxious (and obsessed) patient.

Comments are closed.