Any list of the most important medical technology advances of the past 25 years, or almost any one of those years, would include, very near the top, the development of advanced diagnostic imaging modalities – CT, MRI, PET. Access to precise and detailed images of internal organs and systems has provided an incredible improvement in diagnostic capability, therapy selection and procedure planning and guidance. Capabilities that were unthinkable only a few decades ago are viewed as routine today.
For reimbursement policy, there is no technological miracle that doesn’t create a problem. Advanced imaging tests are expensive to provide, with very high upfront capital costs which translate into substantial per-procedure charges – even if they are cost-effective in the context of a total system of care. Like many other new and powerful technologies, they are seductive and habit forming, highly susceptible to overuse. And they are powerful generators of revenue and potential profit for those who perform the tests. This is a recipe for controversy, and the pot has been boiling continuously for a number of years.
I’ll bypass controversies over coverage policy, where millions have been spent by the developers of these technologies to lobby (successfully) for Medicare and private insurer coverage for the use of PET, the newest of these modalities, for ever broader clinical indications. And we needn’t revisit now-ancient (25- to 30-year-old) battles over whether these technologies should be controlled by CON programs or be allowed to migrate out of the hospital into free-standing community facilities and/or medical offices; they are out there and there will be no going back.
The fight today is over two issues – how much will Medicare pay and who can get paid? The outcome is enmeshed with the current debate on broader healthcare reform. The lines of conflict – the positioning of the most critical physician groups – are instructive.
A little background is necessary. Total payment for advanced imaging tests consists of a professional component to pay the doctor for interpreting test results and communicating them to the patient; and a technical component to cover the capital and operating costs of providing the test. The two components can be combined into a global payment made to a single person/organization, or billed separately by a physician and a testing facility. The testing facility can be a hospital, in which case Medicare pays for the TC under the Hospital Outpatient Prospective Payment System; or a medical practice or community-based imaging center, in which case Medicare pays under the Physician Fee Schedule. For CT, MRI and PET, the TC is far and away the larger and more important category. Furthermore, the PC is set through the AMA’s assignment of Relative Value Units – a relatively non-controversial mechanism. For non-institutional tests, Medicare’s TC payment is set by CMS staff using their best understanding of costs – a potentially highly controversial mechanism.
The volume of advanced imaging tests (PDF) has been growing far more rapidly than most other Medicare service areas and the growing cost of the tests has been highlighted as an area of concern for some years. In 2005, responding to data showing that Medicare payment rates for these and other radiological tests were significantly higher under the PFS than under the HOPPS (i.e., the same test had a higher reimbursement if provided in a physician’s office or free-standing imaging center than if provided in a hospital outpatient department), Congress mandated in the 2005 Deficit Reduction Act (PDF) that payment under the PFS for a lengthy list of imaging tests, including all advanced imaging tests, would be capped (effective Jan. 1, 2007) at the level paid under the HOPPS.
Why had the difference in payment levels developed? TC payments under PFS and HOPPS are both based on cost. If there were a real cost difference, wouldn’t testing migrate to the lower cost setting? But testing was migrating to the (apparently) higher cost setting. Two factors would seem to be at work. First, testing was migrating to the higher profit setting, which may or may not have been the higher cost setting. Second, Medicare’s data on costs in hospitals is excellent and highly detailed, being based upon annual audited cost reports filed under penalty of law by every hospital. Its data on community clinic and physician office costs is both rudimentary and anecdotal, being based largely on unsubstantiated assumptions and extrapolations from un-audited fragmentary data.
At the end of the day, there was no good reason to pay more in the doctor’s office than in the hospital and the DRA 2005 provision doesn’t appear to have slowed the growth of test volume. No harm, then, and some apparent gain – no longer paying an egregiously high amount for tests under the PFS.
But the DRA 2005 caps didn’t solve the problem of rapid growth in advanced imaging expenditures – and Medicare payment systems are extraordinarily blind to whether costs incurred in one place may result in net cost savings realized in another (by elimination, for example, of procedure errors and/or unnecessary procedures) or may create substantial quality outcome benefits. A rapidly rising cost center needs to be addressed. There are three distinct initiatives on the table as I write:
- In the PFS proposed rule for 2010 (PDF), CMS raised the advanced imaging equipment utilization factor used to calculate the TC for tests from 50 percent (where it had always been) to 90 percent (which may or may not be the empirically accurate number – but it at least is based on empirical data developed by MedPAC, the independent Medicare Payment Assessment Commission). The utilization factor – the percent of time in a standard work week that equipment is in use – is a major determinant of the estimated cost of a capital-intensive test like MRI, CT and/or PET. Raising the facor spreads the capital cost of the equipment over more tests, thereby markedly reducing the calculated cost of the test; this proposed change would bring payment for some tests in physicians’ offices below the hosital payment level.
- The healthcare reform legislation working its way through the Congress seems likely to include a proposal to establish (for Medicare or any publicly managed program) Radiology Benefit Managers – organizations which would review and approve the appropriateness of referrals for advanced imaging, contract with providers to get the best possible prices and centralize patients’ imaging information to avoid the need for duplicative tests. (Pharmacy benefit management has an established track record of cost savings and quality enhancement in coordination of prescription drugs; the same principles would be applied to imaging.)
- There are separate legislative proposals in Congress to make it illegal for a doctor to refer a patient for advanced imaging tests to a facility in which he/she has a financial interest – i.e., no self-referral for advanced imaging. There are copious data supporting the proposition that doctors with such a financial interest order more tests than those with no financial interst in a test facility. Medicare has long operated under a broad anti-self-referral statute, but radiological testing is one of a small group of areas (ownership of kidney dialysis facilities is another) exempt from the law.
All three of these proposals make sense, but where various actors stand on them seems to be a pocketbook issue.
The ACR, representing the radiology profession, is opposed to the increase in the utilization factor because it will lower the TC payments to the free-standing imaging centers run by radiology practices. The argument is that some low-utilization centers will close and patients will lose access; MedPAC has addressed this issue, pointing out the proximity of hospital-based testing capability in even the most rural areas, but the argument continues.
Radiologists also oppose RBMs – expressly on grounds that they will be administratively cumbersome, delay needed testing and deprive patients of services on purely financial grounds. All of these arguments were raised agains PBMs and all have proven to be false – patients get prescriptions faster and cheaper, with better review of drug interactions and more reliable identification of generics. There is no obvious reason why RBMs would not work as well.
But radiologists do support the self-referral prohibition for radiology – after all, radiologists receive referrals for imaging, they don’t make them. The prohibition would knock out competition from centers run by multi-specialty group practices and leave radiologist-owned testing facilities as the only game outside the hospital – a major share increase in a still lucrative market. Cherry-picking can be fun.