Overused MRIs


Today’s Managing Health Care Costs Indicator is 90%


Orthopedist James Andrews thought there was an epidemic of improbable injuries among his patients –and the common denominator was that magnetic resonance imaging (MRI) appeared to be discovering  apparent injuries that just would never matter.   So he scanned the shoulders of 31 major league baseball pitchers who were all pain-free and apparently healthy, and he found that 90% of them had abnormal cartilage, and 87% of them had an abnormal rotator cuff.  Most of us don't use our shoulders like pitchers do - so the 'false positive' rate of MRIs in mortals is probably lower.  But if major league pitchers can pitch with these apparent MRI abnormalities, I can probably do my daily activities without a surgical intervention even if I have a sore shoulder.

Gina Kolata reported this story in today’s New York Times; I’ve looked through PubMed and can’t find reference to the published article.

Other orthopedists quoted said that they virtually never saw a “normal” shoulder or knee MRI – and one patient narrowly averted knee surgery because an orthopedist felt that the diagnosis from the first MRI was too serious based on the patient’s symptoms.  So that orthopedist did a second MRI.

The US has lower utilization of almost every type of service compared to  other developed countries (fewer hospitalizations, office visits, and prescriptions per thousand).   However, high tech imaging is a place where we have both high prices and high utilization. (The only country with higher MRI and CT scan utilization than the US is Japan, where MRIs cost under $100).

We clearly need to start showing more restraint – and not ordering imaging tests where we could answer the question with a clinical exam. We also need to refrain from ordering tests where the pretest probability is so low that the posttest probability that a positive finding was true would still be low.  (See this post for an explanation of this concept)

Unnecessary MRIs are not harmless – and they can often lead to additional invasive therapy (and incremental cost).