A recent front-page Washington Post story examines the issue of physician imaging self-referral.  The story highlights a Midwestern urology practice whose imaging orders increased 700 percent in less than a year after they purchased their own CT scanner. 

The practice's attorney told the Post, "The standard of care for a certain category of patients may require a CT scan and the practice may have decided to purchase a CT scan as a result... Any assertion that there is some wrongdoing simply because of an increase in scans is unfounded... (The practice) understands its obligations very well and complies with all applicable standards."

This reminds me of the Upton Sinclair quote: "It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"

In fact, the story told in the Post is consistent with nationwide trends.  According to the CBO, Medicare spending on MRI and CT scans rose at an average annual rate of 17 percent from 2000 to 2006. That amounts to spending two-and-a-half times the amount taxpayers were paying in 2000 compared to 2006.

Besides, as the Post put it, the "billions of dollars in extra taxes and insurance premiums" to finance this variation in practice, is this really bad for patients? Yes. The Post article cites radiation epidemiologist Amy Berrington de Gonzalez at the National Cancer Institute who states, "As many as 1 percent of all cancers in the United States appear to be caused by radiation from medical imaging." 

Most people have had at least an X-ray or two in their lives.  You know how the technician goes and stands behind the lead wall to push the button?  A CT scan is like getting 600 chest X-rays at one time.  That's a lot of radiation. I wrote on this subject in February here. When that CT scan is not clinically indicated, it violates one of the central principles of medicine: first, do no harm.

To be sure, even imaging that is clinically indicated can increase your risk of cancer, and those risks should be discussed with patients as they weigh their treatment options.  What seems to be lost is the ALARA principle -- a simple rule of thumb stating radiation exposure to a patient should be "as low as reasonably achievable."  Although I am not familiar with the individual cases at the Midwestern clinic, a CT imaging rise of 700 percent in less than a year could indicate that some of the patients could have just used an X-ray, and some didn't need a scan at all. 

It's not to say all intense imaging is bad. As Dr. Guy Clifton, a neurosurgeon who was formerly part of the New America health policy team, has written, practice before modern imaging was very, very messy. The difference between practicing then and now is "night and day."  But now he wonders whether physicians nowadays are scanning too much. 

But let's not despair; there are numerous solutions to consider while clinicians, researchers, the press, and the government gather more information about this issue.  One is bundling.  As suggested in the Post story by Georgetown University economist Jean M. Mitchell, and written about elsewhere, Medicare and other payers could use bundled payment. 

For example, let's say an X-ray is reimbursed at $30 and a CT scan is reimbursed at $1,100. Imagine imaging indications for a broad population should be predictable, just as rates for cesarean sections are predictable. Let's say for the sake of argument that a CT scan is necessary for 15 percent of patients, an X-ray is sufficient for the other 85 percent, and we have 100 patients.  A little math tells us that the average reimbursement necessary per patient should be $190.50.  Payers could make it $200 per patient to allow for expected variation. 

(The math: 85 patients x $30 X-ray = $2,550; 15 patients x $1,100 CT scan = $16,500; sum = 19,050; /100 = $190.50; bonus of $9.50 = $200 per patient for scanning purposes.  Note: reimbursement figures in this example are based on real Medicare data from Post article; percentages are made up for the purposes of the example.)

If they prescribe more X-rays, they can keep the savings; if they provide more CT scans, they can either charge patients extra or petition for special treatment, say for local conditions that make CT scans more necessary.  Hospitals do this for outlier cases; this system is essentially what is done with Medicare inpatient prospective payment. 

In April, Modern Healthcare (registration required) published a great piece explaining  the new Medicare bundling demonstration. Participating hospitals and physicians will receive one payment for certain procedures requiring them to work together to coordinate care and share in the savings. This is even more of a reform than my example above.  The MH article explains the difference between bundling and capitation, too.  CQ HealthBeat wrote on it here.

Another proposal is from physician researcher Thomas Bodenheimer, who suggests two different sustainable growth rates (SGRs) to guide Medicare physician payment: one for office visits and another for surgery and imaging.  A third proposal would implement prior authorization in Medicare for imaging, just as all other third party payers use.  The urologists' attorney cites this as a protection against overutilization, however the Wellmark data provided to the Post casts doubt on whether this is effective.  Certainly more research should be done in this area. 

This is not an area where we should make a far-reaching public policy decision based on one anecdote.  But as we learn about this story,  and others such as McAllen, Texas, which illustrate national trends, we should be having a conversation about how to make it easier for physicians to provide appropriate care.

For more on this issue, see Dr. Clifton's post from last summer here and my post from last winter here.