Automatic referral of patients with back pain to MRI makes neither economic nor clinical sense

March 10, 2008

Days spent sitting in lecture theatres and seminar rooms, long hikes from scientific sessions to the exhibition hall with a briefcase under one arm and a weighty conference bag on the other shoulder, evenings hunched over a laptop putting the final touches to the next day’s presentation… It could be argued that attending ECR is a recipe for backache.

Days spent sitting in lecture theatres and seminar rooms, long hikes from scientific sessions to the exhibition hall with a briefcase under one arm and a weighty conference bag on the other shoulder, evenings hunched over a laptop putting the final touches to the next day's presentation... It could be argued that attending ECR is a recipe for backache.

Back pain is an extremely common complaint at general practitioners' offices and a frequent reason for people being absent from work. In many cases, the cause of the pain will be mechanical, rather than a serious underlying medical condition. Approximately 80% of people presenting with acute back pain for the first time will be pain-free within six weeks, whatever treatment is prescribed.

Not all back pain sufferers can be dismissed so easily, though. Joint stiffness, spasms, and ongoing discomfort may be due to infection, malignancy, or degenerative disease, and not to bad posture or poor lifting technique (or attendance at a conference). It is this group of patients that warrants further investigation on imaging, with a view to providing a rapid and accurate diagnosis.

Given the wide spectrum of spinal disorders, deciding which patients to refer to imaging is no simple task. Monday afternoon's European Excellence in Education session, "The imaging management of back pain," explained how this can be achieved. The interactive nature of the session was intended to reveal whether the audience is ready to put that theory into practice.

When faced with a patient complaining of back pain, doctors may be tempted to simply refer them for an MRI examination. After all, MRI can illustrate the anatomy of the spine with exquisite detail. Why not take advantage of this as a general, first-line screening tool? Such a strategy is neither cost-effective nor clinically effective, according to Dr. Victor Cassar-Pullicino, radiologist at the Robert Jones and Agnes Hunt Orthopaedic and District Hospital in Oswestry, U.K.

A patient's clinical symptoms and medical history should be the starting point when considering an imaging referral, he said. Certain "red flag" signs are particularly important. For example, if a patient who had previously been treated for breast cancer presents with acute back pain - or if the patient has reduced power in one leg, has lost a lot of weight, or is suffering from sphincter problems - then he or she should be investigated urgently. Children who complain of back pain should also be taken seriously.

The next decision is which imaging modality to use. Once again, clinical information should act as a guide. For example, if the patient's symptoms point to a possible osteoid osteoma, then a bone scan might be the best starting point. A patient with suspected mechanical back pain is most likely to be referred for a plain film x-ray, if anything. But if the medical history indicates spondylolysis is more likely, then it would be more productive to begin with a CT examination. Subtle spinal stress fractures are unlikely to be visible on x-ray.

"All modalities have some limitations, so you need to have a plan," Cassar-Pullicino said. "It is all about knowing what you are looking for in the particular patient you are dealing with, and then selecting an examination that will be able to diagnose it."

Certain spinal disorders may require an interventional approach to make a definitive diagnosis. Image-guided injection of contrast, for example, can identify the precise source of discogenic pain in mechanical back pain sufferers. Interventional techniques may also be used to take tissue samples or administer pain relief.

Whatever approach is taken, it is critical that radiologists be able to link what they see on imaging to the patient's symptoms, he said. Detecting an abnormal signal on MRI, for instance, is only the first step. Determining whether that abnormality corresponds to degeneration, infection, or past trauma is another matter altogether.

The relationship of imaging signs to clinical symptoms is likely to become increasingly important as the number of elderly patients in doctors' waiting rooms grows. An MRI of an 85-year-old's spine will almost certainly reveal extensive degeneration, but these signs of wear and tear may not be the root cause of a patient's pain.

"This is really where the radiologist comes in," Cassar-Pullicino said. "It is not so much about identifying whether there is or isn't disease, it is more about correlating those imaging signs with clinical symptoms. Because the last thing you want to do is to recommend surgery for these patients simply on the basis of an imaging abnormality."

Audience participation was also required at the European Excellence in Education session, "The imaging management of stroke." Early assessment of stroke on imaging is acknowledged to be critical, and such examinations can be performed using a range of different modalities. The question is: which to choose? Delegates were presented with the benefits and drawbacks of CT, CT angiography, CT perfusion, multimodal MRI, and digital subtraction angiography. Criteria used when referring patients for further treatment was also discussed - and then tested.

Another European Excellence in Education session focused on the imaging of cirrhosis and portal hypertension. A selection of case studies was prepared so that the audience could be quizzed on likely diagnoses. Attendees learned about the pitfalls to be aware of when making such assessments in clinical practice, for example, the misidentification of fibrosis as a tumor.