X-ray Based Diagnosis Leads To Potientially Unnecessary Surgery

Picture by Balint Botz from Radiopaedia

I've written previously that many patients had back surgery when a doctor sends them for an MRI within the first 6 weeks of an initial visit.

Well, guess what? Taking an X-ray to diagnose knee arthritis may make you more likely to consider potentially unnecessary surgery (Lawford et al, 2025) as well.

Many of my patients who go to a doctor or surgeon for their knee pain end up being sent for an X-ray or even an knee MRI. Many of these patients have osteoarthritis (OA) in their knees. Actually, routine X-rays may not be necessary to diagnose the condition. A skilled and thorough assessment based on symptoms and medical history is good enough to make the diagnosis. 

A huge and common misconception is that OA is caused by 'wear and tear'. Research clearly shows that the structural changes seen in a joint X-ray does NOT correspond with the level of pain or disability a person feels. Nor can X-rays predict how symptoms will change.

In fact, X-rays are NOT recommended in Australia to diagnose knee OA. Nearly half of new patients there with knee OA get sent for a knee X-ray and cost their health system A$104.7 million each year.

Researchers in Australia showed that using X-rays to diagnose knee OA can affect how a person thinks about their knee pain and prompt them to consider potentially unnecessary knee replacement surgery.

Many patients with 'terrible' X-rays have no pain while patients with no damage on X-ray have a lot of knee pain. Hence, X-rays are not recommended for diagnosing knee OA or guiding treatment decisions.

The Australian study had 617 subjects across Australia who were randomly assigned to watch one of three videos. Each video showed a hypothetical consultation with a general practitioner (GP) about knee pain. 

The first group received a clinical diagnosis of knee OA based on their age and symptoms and were not sent for an X-ray. The other 2 groups had X-rays done to determine their diagnoses (the doctor showed one group thier X-ray images but not the other group). After watching their assigned video, the subjects completed a survey of their beliefs about OA management.

The results showed that the group who received an X-ray based on their diagnosis and were shown their images had a 36 percent higher perceived need for knee replacement surgery compared to those who received a clinical diagnosis without X-ray.

What was worse was, they even believed that exercise and physical activity could be harmful to their joint. They were also worried about their condition worsening and were more fearful of movement.

The subjects were slightly more satisfied with a X-ray based diagnosis than a clinical diagnosis. This may reflect the common misconception that OA is caused by 'wear and tear' and the joint needs to be replaced.

The study's finding shows that it may be important to avoid unnecessary X-rays when diagnosing knee osteoarthritis. Changing this can be challenging, since many people still expect or want  X-ray imaging. If we can change this mindset, it will minimize unnecessary concern about joint damage, reduce demand for expensive and potentially unnecessary joint replacement surgery.

In my opinion, we as health professionals should not focus on joint 'wear and tear' since it can make patients more anxious about their conditions and concerned about damaging their joints. There are a range of non surgical, non invasive options that can reduce pain and improve your mobility. Exercise is one of many if you read this Cochrane review.

Our health minister says healthcare spending in Singapore could hit 30 billion a year by 2030 in a Straits Times article just 2 days ago. Perhaps this is an area where we need to be more mindful of unnecessary X-ray imaging and joint replacements to bring healthcare spending lower.

You can read about what actually causes your joints to wear out here if you are keen.

Reference

Lawford BJ, Bennell KL, Ewald D et al (2025). Effects Of X-ray-Based Diagnosis And explanation Of Knee Osteoarthritis On Patient Beliefs About Osteoarthritis management: A Randomised Clinical Trial. Plos One. DOI: 10.1371/journal.pmed.1004537 

Locking In Your Knee May Not Mean A Meniscus Tear

Picture from ScienceDirect

A discussion with our three new physios recently revealed that they were taught that whenever a patient has a meniscal tear on MRI with mechanical symptoms it would usually mean surgery is indicated.

Mechanical symptoms are described as 'locking' or 'catching' in the knee joint that is caused by something being trapped or stuck in the knee. They were taught that it can only be removed by surgery. If the mechanical symptoms present with a tear in the meniscus, that is confirmed by MRI, it is usually attributed to the tear. Hence the rationale for surgery to remove the tear.

However, mechanical symptoms in the knee can fluctuate. Other clinicians and patients may have different variations and definitions of such 'locking' and 'catching' in the knee.

I have come across many patients with locked knees (knees that cannot fully straighten or bend), but they are seldom 'locked' all the time. We always match such symptoms in our clinic with a patient's medical or injury history and with objective orthopaedic and functional tests. 

However a study by Thorlund et al (2019) investigated whether unstable meniscal tears are more likely to cause mechanical symptoms compared to other concurrent knee pathologies like articular cartilage damage, ACL tears etc.

A wide range of meniscal tear characteristics like tear pattern, location, size of tear were included in the study. However, no important relationships were found between any of the included factors and patients reported catching or locking in the knee or inability to straighten their knees.

These results question the logic that mechanical symptoms are caused by specific joint pathologies. The authors also compared the frequency of mechanical symptoms between patients with and without a meniscal tear after knee arthroscopy. They found that half of all patients reported catching or locking. They were also unable to straighten their knee fully. However, these mechanical symptoms were equally common among patients with or without a meniscal tear.

This is consistent with my previous post where bucket handle and complex meniscal tears (both of which are commonly operated on) were found in the patient's MRI, but these patients were asymptomatic.

So if you are currently having a 'locked' knee or cannot straighten or bend your knee fully, it does not necessarily mean you need surgery. Two large scale randomized trials referenced below confirm this too. Please come and see us in our clinic for another opinion.

References

Khan M, Evaniew N, Bedi A et al (2014). Arthroscopic Surgery For Degenerative Tears Of The Meniscus: A Systematic Review And Meta-analysis. CMAJ. 186: 1057-1064

Thorlund JB, Juhl CB, Roos EM et al (2015). Arthroscopic Surgery For Degenerative Knee: A Systematic Review And Meta-analysis Of Benefits And Harms. BMJ. 350: h2747

Thorlund JB, Pihl K, Nissen N et al (2019). Conundrum Of Mechanical Knee Symptoms: Signifying Feature Of A Meniscal Tear? BJSM. 53(5): 299-303. DOI: 10.1136/bjsports-2018-09943

Am I Crazy To Feel Pain In My Knee When My Scans Are Normal?

My patient came in yesterday saying that his knee started clicking even though he didn't feel any pain with the clicking. There were occasional twinges of pain only occasionally. However, he was worried that his knee will get worse. After assessing his knee (which turned out fine) and reassuring him, I told him my own experiences.

After I had my third knee operation (within the space of a year), my right knee started feeling better. Of course I started training again as soon as the surgeon permitted. For a start, I ran almost exclusively on grass (since it was the softest surface I could find), often going multiple rounds, to ensure that I can get my mileage since I was hoping to compete again after my injuries.  

However, I was much more sensitive about my right knee, to the point of being paranoid about every sensation I felt in the knee. Each time I was on the bus or MRT and if someone came close to my right knee, I'd move away and glare at the person for coming too close to me. Does this sound like you?

Well, it turns out this action of pain sensitization is common across other painful knee disorders as well. In patients with knee osteoarthritis, pressure from placing your hands on the knees alone can trigger pain. The good news is that this sensitization for painful knees can be treated.

The following systematic review investigated 52 studies that studied pain sensitization across four different painful knee disorders. The authors found evidence of pain sensitization in people with knee osteoarthritis, patellofemoral pain and post meniscectomy patients. They however found conflicting evidence in patients with patella tendinopathy.

The researchers found that the extent of structural joint damage in the observed knee disorders does not correlate to the severity of symptoms. Hence, pain is not necessarily a 'signal' from a joint or area that is damaged.  Meaning x-rays or MRI results does not correlate with pain. Some patients have no 'damage' on x-ray/ MRI but have a lot of pain, while others with lots of damage on film may have no pain.

Many factors play a role to determine if a person will perceive a stimulus as painful or not since pain can be a complex experience that is associated with memories, belief and social context. Anxiety, depression, fear of movement, viewing their condition considerably worse than it actually is may also play a part.

It is suggested that repetitive stimulation may lead to subsequent sensitization of the nervous system. This include loading of the knee joint, ongoing inflammation at the knee joint or related tissues and altered biochemical markers. These factors contribute and maintain the pain sensitization in the knees.

In patients with knee osteoarthritis and patellofemoral joint pain, pain sensitization can be treated through exercise therapy, mobilization, pharmacological (yes, painkillers) and surgical intervention. Correct exercise is recommended for treatment of painful knees. There is the incorrect belief that exercise may harm the joint cartilage in patients with osteoarthritis.

If doctors, surgeons and physiotherapists focus less on x-rays/ MRI's and more on factors (including psychosocial factors mentioned above) relating to each patient's pain and disability, there will be more opportunities for collaboration and improved treatment outcomes.

Reference

De Oliveira Silva D, Rathleff MS, Petersen K et al (2019). Manifestations Of Pain Sensitization Across Different Painful Knee Disorders: A Systematic Review Including Meta-analysis And Metaregression. Pain Med. 20(2): 335-358. DOI: 10.1093/pm/pny177.

Abnormal Knee MRI, But No Pain

Knee MRI by Becky Stern from Flickr

Here's a piece of surprising news for everyone. In this recently published paper I read, nearly all patients who had abnormalities on their knee MRI were asymptomatic, meaning no pain despite having an "abnormal" MRI.

The main inclusion criteria for this study were sedentary adults. They did not do at least 30 minutes of moderate intensity physical activity 5 days a week or 20 minutes of more intense activities 3 days a week. They did not have any knee pain, no current or previous history of knee injury and surgery.

The authors reviewed 230 knees of 115 uninjured inactive adults (51 males and 64 females). Median age was 44 years (range was 25-73 years) and all the subjects had bilateral MRI's (3.0 Tesla, high resolution) done.

Here's what they found. Brace yourself as you read on. MRI showed abnormalities in a whopping 97% of knees. 30% of knees showed tears in the meniscus. Horizontal tears were most common, while bucket handle tears least common.

Articular cartilage (57%) and bone marrow abnormalities (48%) were common in the patellofemoral (knee) joint. Moderate (19%) intensity articular cartilage lesions and severe (31%) were observed.

Grade 4 means bone rubbing on bone

Articular cartilage injuries is my area of interest since I did my postgraduate research in that area. It is interesting to note that a quarter (or 25%) of the subjects had Grade 4 changes (see picture above) visible on MRI but did not complain of pain. Perhaps this is important to remember when imaging the knee since there seems to be more visible findings here compared to the rest of the knee (articular cartilage wise).

Moderate intensity lesions were found in 21% of knee tendons while there were high grade tendonitis found in 6% of knees reviewed. The patella and quadriceps tendons being the most affected.

3% partial ligament ruptures were found, of which 2% were of the Anterior Cruciate Ligament (ACL).

The authors concluded that nearly all knees of asymptomatic adults they studied show abnormalities in at least one knee structure on MRI. Meniscal tears, articular cartilage and bone marrow lesions in the patellofemoral joint were the most common pathological findings. 

They also reported finding bucket handle and complex meniscal tears (both of which commonly operated on) in asymptomatic knees. This interesting to note as bucket handle tears (as well as complex tears) would often cause 'locking' in the knee and therefore require surgery.

There you have it, the subjects were sedentary adults who did no exercise so no one can say that it was running or exercise that "wore out" their joints. And some these very same adults had "terrible" or abnormal MRI's, but were asymptomatic or did not have any pain.

So don't fret if your MRI is abnormal. You may not need any surgical intervention, especially if you do not have any pain or if that pain is easily treated.

Maybe these abnormalities should be just described as "wrinkles" on the inside. 

Reference

Horga LM, Hirschman AC, Henckel J et al (2020). Prevalence Of Abnormal Findings In 230 Knees Of Asymptomatic Adults Using 3.0 T MRI. Skele Radiol 49: 1099-1107. DOI: 10.1007/s00256-020-03394-z