Suppose you are experiencing terrible lower back pain. You are offered an MRI scan as a way of diagnosing the problem. Problem area pinpointed, you then are offered surgery to correct the issue. All good, right? Not really, according to a recent study. The greater availability of MRI scans is leading to an increase in back surgeries, surgeries that would not otherwise be performed and may not be helpful to the patients.

The scans often reveal back abnormalities which are assumed to be the cause of the back pain but turn out not to be the cause. Surgery corrects these abnormalities but often doesn't relieve the pain. Lower back MRI scans are new enough that doctors aren't always able to distinguish cause from effect in them. Someday they may be able to. Right now, use of MRI as a diagnostic tool for lower back pain is controversial.

The number of scanners in use more than tripled between 2000 and 2005, rising from 7.6 to 26.6 per million people.

About 90% of all cases of lower back pain show spontaneous improvement within four weeks. This is one reason why the generally recommended treatment is conservative in nature. Surgery is a last resort and rarely employed until time and other treatments have proven ineffective.

So why are more back scans being done? According to a Stanford University study, it's simply because there are a lot more MRI scanners available. The number of scanners in use more than tripled between 2000 and 2005, rising from 7.6 to 26.6 per million people.

If you buy them, they will come.

Particularly troubling is that two−thirds of these additional scans occurred within the first month of pain onset. This is supposed to be the waiting period during which it is seen if the pain will subside without treatment, as it does in most cases.

The researchers focused on Medicare records of patients with new−onset nonspecific lower back pain. These account for more than 80% of low back pain complaints and include back pain thought to be caused by lumbar strains and sprains, degenerative disk disease, spinal instability and, of course, cases where the cause of the pain is totally unknown. Cases of back pain thought to be due to herniated disk or spinal stenosis (narrowing of the spinal canal) were not included because MRI and surgery are more often beneficial in these cases.

They examined Medicare claims data of 666,455 episodes of first time lower back pain that occurred in 1999, 2002 or 2004 and whether these episodes led to MRI scans or surgery within one year of evaluation. They also determined the number of MRI scanners in the metropolitan areas closest to the patients' residence, using survey data from 1999, 2002 and 2004. The metropolitan areas were then divided into four groups (quarters), ranging from high to low MRI availability.

For 2004, they found that if all the patients had lived in the quarter with lowest MRI availability, there would have been 12,277 fewer lower back MRI scans performed and 3,521 fewer lower back surgeries performed on Medicare patients reporting first time lower back pain in the nation as a whole. Greater access to MRI meant more scans and more surgeries.

While all this can be framed in terms of increased costs, it's really about performing procedures that don't improve the health of patients. Since 1994, low back pain guidelines have recommended delaying MRI use for most patients for at least one month. For whatever reason, increased availability of scanners is causing these guidelines to be ignored more and more.

The results of the study were published online October 14, 2009, by the journal Health Affairs. The article is freely available at