Getting a scan sounds simple and straightforward but it is actually a more complex decision than most people would be aware. Using my own experience working in sports medicine, I order many scans. I am requested to order more than are required from my perspective while patients refuse others I consider necessary. There are a several important issues to consider when having a scan but the two main ones are:-
1) Radiation exposure
2) Utility of the test.
In this blog I will discuss utility of the test.
Will this scan actually help me?
When used properly imaging scans can be very informative and can provide accurate and improved treatment options. But when used improperly can cause harm. As discussed in part 1 some scans expose the patient to radiation. Even imaging studies with no ionising radiation can be harmful indirectly, if the information provided is not used or interpreted accurately. One of the major problems with modern medicine is that tests are over relied upon and often poorly interpreted.
Some scans have poor correlation between pathology and pain
For many areas of the body, imaging studies are less than useful due to the high rate of false positive and negative findings. The correlation between findings of spinal scans and pain are notoriously poor predictors. Disc protrusions are present in about 30% of 20 year olds. This group however do not experience any pain or discomfort. Likewise disc degeneration is present in nearly 40% of those aged 20 and who are also pain free. Shoulder scans are similar. As an example of this at the Cowboys Rugby League side one year there was a situation where two players had injured their shoulders. One scan was normal while the other showed a massive complete 360 degree tear of the labrum. (The labrum is a soft fibrous tissue rim surrounding the shoulder socket and assists to stablise the joint.)
What was interesting to note was the player with no tear required surgery while the player with the massive tear had such good muscle function and control that he has continued to play without any intervention and is still playing 5 years later. In reality shoulder scans are something that I only look at after I have made my decision about what needs to be done. Surgeons definitely need them to decide how they are going to operate but scans alone should have little influence over a decision to operate on the shoulder. Knees and ankles are a problem but not quite as significant as the shoulder or lumbar spine. In older patients the utility of knee MRI is very poor. Cartilage tears in knees are very commonly seen in patients with no pain or functional impairment. In a 2012 study of adults with a mean age 62 it was found that 72% of patients with knee pain had cartilage damage but that 68% of patients with no knee pain at all also had cartilage damage. It’s also entirely possible to have knee pain and entirely normal scan as is often the case with patellofemoral pain syndrome. So seeing a tear on a scan does not necessarily mean that it is the cause of your pain and your pain may actually be coming from somewhere looking entirely normal on the scan. This is where trouble can occur and great care needs to be taken with the interpretation of the scans.
Some scans are becoming too good
A few years ago I had a young woman who had complex regional pain syndrome of her right foot. She was sent for a bone scan and the radiologist read the report incorrectly. He was focussed on her left leg and described a whole host of potential injuries and causes for pain. In reality what he was seeing on the bone scan was a completely normal foot which had been bearing more load recently because she was using crutches and not able to put weight through her injured leg. Modern day scanning is at the point where some normal processes are able to be detected and unfortunately then confused with pathology.
Scans are frequently wrong
The accuracy of a scan can sometimes be an issue. With any scan there is a small possibility that it will miss an injury that is present or alternatively give the appearance of an injury that is not present at all. In medicine this is referred to as sensitivity and specificity. Sensitivity is what percentage of injuries are properly picked up and specificity is what percentage of a noninjury are properly seen as normal. The published data of the sensitivity and specificity of MRI for knee injuries is surprisingly disappointing. For assessing the ACL the sensitivity is 86%, specificity is 95% and overall accuracy is 93%. This means that if you have ruptured your ACL the scan is 86% likely to identify this accurately. If you have not ruptured your ACL the scan is 95% likely to be give you the right answer and overall for assessing ACLs, MRIs are 93% likely to be correct.
Sometimes getting scans confuses the issue. Let’s use rope as an analogy for a second. Ligaments act much like a rope, ligaments essentially attach bones together just as a rope might. If you have a rope that you suspect is torn in two you have a couple of options to confirm this. You can take a photo of the rope or you can pull on the ends of the rope. Taking a photo of the rope is like getting a scan. Pulling on the rope is like performing an examination on the knee. If you pull on the rope and there is no tension then you can be confident that the rope is torn in two regardless of what it looks like on the photo.
Information in the wrong hands can be dangerous
As mentioned earlier, scans can sometimes cause harm in more ways than radiation exposure. With the above examples of lumbar spine, shoulder and knee it is easy to see that if a scan is carelessly ordered and not actively and critically appraised then it is easy to get an inaccurate diagnosis. This may delay appropriate treatment and result in less than an ideal outcome. Usually the problem is overdiagnosis but sometimes a scan will miss a diagnosis and the same problems will occur.
While the above examples are fairly obvious instances of harm resulting from a scan there is a much more pervasive and insidious way scans can harm. Fear avoidance is a massive problem especially with spinal pain. What often happens is this: someone has an episode of back pain and they end up getting a scan. The scan identifies several problems in the spine that are unlikely related to the patients’ pain. The results are given to the patient without adequate explanation and possibly a very common, very expected age related issue and certainly non-life threatening set of results, then becomes the focus of the patient. Unfortunately in many cases these have nothing to do with the current injury or subsequent pain being experienced. Pain, disability and morbidity escalate with the patient worrying unnecessarily. A return to work, pain free movement and recreational activity is delayed further. In a large 2013 study it was found that people with lower back pain who received an early MRI stayed on work disability much longer and ended up costing close to $13,000 dollars more to rehabilitate.
So what should I do?
None of this is intended to scare you out of having a scan. In part 1 it is hopefully clear that the vast majority of scans are extremely safe but that there is a small associated risk. Unnecessary and / or repeated CT scans in particular should be questioned. The purpose of part 2 has been to give some perspective to imaging in the context of the seemingly “gung ho” approach that modern society takes to radiology. Scans can be extremely useful and extremely important but it is never as simple as “just get a scan and find out what is wrong”.