Diagnosing True Clinical Spinal Stenosis

Andrew Reish, MD

One of the most common radiologic findings on lumbar MRIs in patients 65 and older is “spinal stenosis,” a spinal canal smaller than a trained radiologist’s accepted standard canal size.

In reality, radiographic “spinal stenosis” is used too frequently as an explanation for non-specific low back pain without an effective evaluation for true clinical stenosis, often known as neurogenic claudication. MRIs are frequently ordered and referrals made for surgical evaluation without a true assessment of the patient’s pain through a detailed hands-on examination. Although useful in supporting a diagnosis, asymptomatic patient lumbar MRIs are not paid for by insurance companies or looked at by radiologists. Therefore the assumption is implicit, however incorrect, that all MRIs of the spine reveal sources of pain. 

As physiatrists who specialize in musculoskeletal medicine, it is our duty to assure no patient is treated for a false positive such as lumbosacral dysfunction, facet arthropathy, vascular claudication or other stenosis mimicking diseases such as diabetic, alcoholic or Charcot-Marie-Tooth polyneuropathy, which can lead to progressive weakness with dysesthesias, as can a pinched nerve from lumbar stenosis.  If misinterpreted with an MRI as the definitive evaluation, a poor treatment outcome is inevitable.

In 2007, Andrew Haig, MD, of the University of Michigan, and his colleagues were one of the first groups to look at the lumbar stenosis MRIs of completely asymptomatic patients in addition to those patients with low back pain with additional concurrent history of polyneuropathy, myopathy and other etiologies.  He found that 65 percent of the individuals in his stenosis study had something on isolated imaging that a surgeon could claim “justified back surgery.”

While undergoing my spine medicine fellowship training at the University of Michigan, I learned to trust my focused exam and clinical history far above the mere pictures from an MRI. After reviewing the MRI, I often tell my new patients that I am treating you, not your MRI. Although an MRI is extremely useful in finding cancer or infection, it has a poor track record of showing whether a patient has no pain, is in pain or even whether the patient is alive or not.  As we age, we have more wrinkles and skin color changes, more hand arthritis, etc., so why shouldn’t we have an arthritis of the back called lumbar stenosis.

Like greying hair, stenosis is part of the aging process, not a disease. But we have been conditioned to think of lumbar stenosis as a pain generator rather than a degenerative aging process like other forms of arthritis.  With lumbar MRIs, there is NO published, universally accepted criteria for grading stenosis.  Why would anyone treat based upon an image where radiologists often can’t agree on what the image means? 

In “mild and moderate” cases of stenosis referred to me for management of low back pain “clinical stenosis,” I often find that the story does not fit.  Clinical stenosis is classically diagnosed when mechanical back extension or walking worsens and a patient’s pain radiates into the legs. But I often have a patient sent to me who presents with flexion biased pain or a different arthritic joint pain source that is revealed upon closer evaluation. A thorough physical exam and history is often sufficient to help with the diagnosis, but sometimes when a person has many other diseases that cloud the clinical picture, I will use an electrodiagnostic exam (EMG/NCS). An EMG/NCS test evaluates nerve and muscle function to give me and the patient more information to guide treatment.  Sometimes an EMG/NCS with paraspinal mapping will save a patient with diabetic polyneuropathy from a lumbar surgery for stenosis that would provide no relief of pain.

As a Good Shepherd musculoskeletal physiatrist, I focus on the spine training that I received at the University of Michigan, which taught me to not give a diagnosis of clinical lumbar stenosis until I have done my due diligence by “whipping the hip, whacking the sacroiliac, poking the troch (greater trochanter) and upsetting the facet.” 

Even then, I will exhaust all conservative options through medications, therapy and education before I recommend surgery for a patient in chronic low back pain. It has been consistently shown that radiographic findings alone are insufficient to justify treatment of spinal stenosis and that patients deserve a doctor who will not focus on the MRI alone in treating them.