Physician variation in diagnostic testing for low back pain. Who you see is what you get

This study examined patterns of diagnostic test use for patients with low back pain. Three specific questions were addressed: 1) What tests do physicians recommend for patients with 3 common types of low back pain? 2) Do physicians in various specialties differ in the tests they would order? and 3) How appropriate are physicians' choices of tests, based on current medical knowledge and expert recommendations?

A stratified national random sample of 2,604 physicians in 8 specialties was mailed questionnaires asking about the tests they would order for hypothetical patients with acute back pain, sciatica, or chronic low back pain. Physicians were also asked which procedures they generally used to evaluate suspected lumbar nerve root compression. These responses were compared with guidelines that have been suggested by the Quebec Task Force on Spinal Disorders, based on comprehensive evaluation of the scientific literature.

Approximately 1,100 physicians responded to the survey (43% response rate). Magnetic resonance imaging was the most frequently used procedure for evaluating suspected lumbar nerve root compression, although a majority of neurosurgeons would still use myelography. Neurosurgeons and neurologists were twice as likely as other specialists to order an imaging study for patients with acute nonradiating pain or chronic back pain. Physiatrists and neurologists were more than 3 times as likely as other specialists to order electromyograms for acute back pain with sciatica or chronic back pain. Rheumatologists were almost twice as likely as other specialists to order laboratory tests for both acute and chronic back pain. The reported use of imaging and electrodiagnostic tests was generally premature and more extensive than that recommended by the Quebec Task Force.

There is little consensus, either within or among specialties, on the use of diagnostic tests for patients with back pain. Thus, the diagnostic evaluation depends heavily on the individual physician and his or her specialty, and not just the patient's symptoms and findings. Furthermore, many physicians may be ordering imaging studies too early and for patients who do not have the appropriate clinical indications. These results suggest a need for additional clinical guidelines as well as better adherence to existing guidelines.