Are MRI results accurate?

Most physicians and patients think that the MRI scan is one of the more sensitive and comprehensive diagnostic tests.

However, there is significant variability in reading and performing the scans, which makes having it done at a reputable institution more imperative:

Magnetic resonance machines, though, vary enormously, and not just in the strength of their magnets. Even more important, radiologists say, is the quality of the imaging coils they put around the body part being scanned and the computer programs they use to control the imaging and to analyze the images. And there is a huge variability in skill among the technicians doing the scans.

Some are recommending that general radiologists may not be appropriately qualified to read specific MRI scans.

Compounding the problem are two issues. First, MRIs are rapidly sprouting up across the country, because the payment system gives a financial incentive for doctors and hospitals to expand the reach of MRIs.

Next, doctors are increasingly relying on scans instead of the history and physical exam. One major reason is patient demand, as there is a perception that “getting an MRI,” leads to fast, definitive diagnoses.

It doesn’t help that obtaining an MRI is discussed so cavalierly in society, especially in the world of sports, where athletes often receive such scans routinely.

This further perpetuates the routine nature of the test.

Comments are moderated before they are published. Please read the comment policy.

  • Anonymous

    This was a really disappointing article. Like most of Ms. Kolata's medical writing, it's anecdote-driven and fairly off-target. Rather than doing the public the service of explaining that way too many MRIs are done, the message of the article seems to be that you need a more thorough MRI with a fancier, more expensive scanner.

    One implied point that does make sense is that we should actually look at whether there are radiologists operating beyond their training.

    But as a primary care physician the main thrust of this article ("you need an MRI, and even if you've had one read as negative in the past, you need a repeat fancier MRI…") will be a nightmare if my patients read it. The real problem is that the patients in her article (and patients in general) didn't receive a thoughtful and skillful H&P. Come on! Neurological lesions are best diagnosed by history and exam. I'll never forget the neuro prof who taught me that a good neurologist (or well-trained PCP with time to perform an exam) can detect neurological lesions that matter at sizes many times smaller than those seen on MRI.

    Too many MRIs serve primarily to document prevalent conditions that have nothing to do with patients' symptoms…lumbar degenerative disc disease is a perfect example. Then the patients become obsessed with these incidental findings and, as the NYT pointed out in the article below, end up on quasi-fraudulent disability for life. Try adding that to the cost of an MRI!

    http://www.nytimes.com/2008/10/08/nyregion/08lirr.html?_r=1&scp=1&sq=long%20island%20railroad%20disability&st=cse&oref=slogin