Diagnostic Imaging roundtable tackles rationales and realities of efforts to rein in imaging use in radiology
Diagnostic Imaging roundtable tackles rationales and realities of efforts to rein in imaging use in radiology
The healthcare industry is grappling with the rising cost of imaging services, spurred by spiraling increases in utilization, particularly high-end MRI and CT procedures. Private health plans have set their sights on reining in these costs through utilization management strategies that range from preauthorization to steerage. Radiologists find themselves defending the appropriateness of procedures on a case-by-case basis, while across town, other physicians reap the benefits of self-referred practices.
But how much of today's imaging utilization can be tied to inappropriate use and self-referral rather than a lack of accountability? What role should the American College of Radiology, and radiologists themselves, play in policing imaging utilization? Do utilization management strategies work in changing inappropriate behaviors? Last but not least, what motivates utilization management companies? Radiologists might be surprised to learn that UM firms and their clients are less concerned about the impact of self-referral on their profits than on the quality of patient care.
In a dynamic Diagnostic Imaging roundtable, stakeholders from radiology and the UM industry discuss the rationale behind utilization management and the resulting effects on radiology practice. The candid conversation that follows touches on current trends in utilization and offers insights into how radiologists can understand, and shape, the utilization management process.
Roundtable participants were:
- Dr. Gregg Allen, chief medical officer, Medsolutions;
- Dr. James Borgstede, president, the American College of Radiology;
- Brandon Wm. Cady, senior vice president, American Imaging Management (AIM);
- Dr. Thomas Dehn, medical director, National Imaging Associates; and
- Donald Ryan, president and CEO, CareCore National.
In addition, Thomas Greeson, J.D., a partner in the healthcare group of Reed Smith LLP in Falls Church, VA, helped moderate the session.
Diagnostic Imaging: Let's identify the trends in utilization growth and break them out by modality, procedure, and setting.
Thomas Dehn: There's a general increase in utilization primarily involving the high-dollar advanced imaging procedures. We find this more so with MRI and MR angiography than with CT. PET scanning is coming along, but in terms of real numbers, it isn't much. One aspect of imaging that we find particularly vexing is the category of follow-up procedures, including plain films. There's been explosive growth in the use of follow-up procedures, with little consensus on which follow-up films should be done.
James Borgstede: We see some of the same issues. The most recent data are from the Medicare Physician/Supplier Procedure Summary. It shows MR and CT increasing the most in utilization. Obviously, much of this is inappropriate, and a large portion of it involves people providing imaging in their offices for economic gain.
Gregg Allen: Medicare is perhaps just a little bit different, but in the commercial population, we've seen 20%-plus year-over-year increases in costs for MR and CT procedures for about five years now. MR and CT are accelerating at rates that have probably shocked even the radiology community.
Donald Ryan: It's a significant problem. The trends are high-dollar driven and advanced-technology driven. There's a significant impact from the self-referral environment, which is to a certain degree economically driven. All of these factors are creating a situation that most payers as well as the government are very concerned about. And as unhappy as most physicians are at having to engage in prior authorization, putting those speed bumps in the road to get people to think twice about using expensive imaging services has proven its value. Unfortunately, it is probably going to become an even more important component of utilization management in the future.
Brandon Cady: As an industry, we've certainly seen this coming. There was a 2001 article in Health Affairs in which physicians ranked MR and CT as the most important clinical innovations of the past 30 years. They are absolutely key tools in the physician community. The second point is that I think all of us-AIM is in 25 states-are seeing a wide variation in utilization. Some of that is due to self-referral, but most is due to distribution of high-tech imaging equipment.
DI: So you're saying if you build it, they will come.
Thomas Greeson: Does anyone know what percentage of utilization increases, particularly in-office imaging, can be tied to physician self-referral?
Borgstede: Data from the Medicare Physician/Supplier Procedure Summary show that for CT in-office imaging, the radiologist increase in utilization was 26% between 1999 and 2004. For the same period, nonradiologist in-office imaging went up 117%. In MR, radiologists increased 11% and nonradiologists 256%.
DI: I'd like some reaction to that from the utilization managers. You heard the numbers. Is that your experience?
Allen: I think that's reflective, but let me give you a little different perspective. Our data show that in the commercial population, we're seeing large payers having anywhere from 5% to 12% of their total high-tech imaging going into self-referral settings or what we would call in-office imaging. That is probably understated because of the variety of creative arrangements physicians have in how they set-up a self-referral scenario with an imaging center.
DI: Do you have a sense of where the big utilization growth is?
Ryan: Certainly, the advent of multiple manufacturers making equipment available at lower cost has made it easier for urologists to install their own CT scanners or ENT guys to put in their own machines. There's no question that if you look at the rate of growth, it's generally much faster in self-referred environments. The bulk of imaging is still performed outside of a self-referred environment, but self-referral is contributing significantly to the rapid increase in utilization.
When you compare two specialists in the same market, a physician who has a CT scanner in his or her office has a referral rate that is four to five times greater. There's no basis for suggesting that the level of complexity of the patient disease process is any different for either of those physicians. There are a lot of other problems with utilization and many causes for it, but certainly self-referral is a very big deal.
Dehn: I'm going to be singing a little off-key in this choir. Based on nearly a decade of data, I believe that self-referral is a critical quality issue, but if you consider the numbers Gregg Allen mentioned, 20% percent year-over-year, that's a 100% increase over the last three or four years. It's remarkable.
Self-referral has really not increased that much, at least in our data. It's at roughly the same 15% to 20% that we see in the commercial population. If you look at any of your big claim sets-the overall spend-somewhere around 10% will go to nonradiologists and nonhospitals. Not all of that is self-referral but we can assume that most of it is. If so, we're looking at about 10% year-over-year. If you assume that 50% of that is inappropriate, which is probably a little more punitive than it is in reality, you're looking at a 5% problem. Honestly, the health plans we're dealing with are not getting exercised over self-referral on the basis of economics. They are concerned on the basis of quality.
Cady: Our observations at AIM are similar, and we have a very similar client base. Self-referral is an important issue that we're all dealing with, but if you put it in a bucket with the distribution of equipment, advances in technology, and the tendency by physicians to rely on this technology, self-referral is not the top driver.
DI: What is the top driver?
Cady: Distribution of equipment. There's also the secondary market. In many of these markets, what used to be hospital-driven utilization in the high-tech area has become in-office and outpatient setting-driven because of the secondary equipment market. Hospitals are often the primary buyers of new and advanced technologies, but the secondary machine needs a buyer, and those buyers are usually physician groups in an office setting.
DI: Do you concur?
Ryan: Our experience is a little different. There's an explosion of imaging facilities, but I don't believe they are all related to the secondary equipment market. They are growing because there's a demand and opportunity exists to create a profitable business. That is certainly a driver. But in very specific areas, including cardiac imaging, as well as others, self-referral is a big problem.
We created a program for a client that involved prior authorization for PET scanning. We identified a group of physicians, in this case cardiologists, who owned their own PET machine. In the first two weeks of the program, they requested 23 myocardial PET studies, of which we very easily, based on the criteria, denied 21. That's a fairly strong indication that there are real issues.
Borgstede: I think equipment is being distributed the way it is because of self-referrals, so we're still dancing around that issue. The other thing is this business about used equipment. I think that really speaks to the issue of quality. We need to look at who is purchasing that used equipment and see if patients are getting the quality of imaging that they deserve, and that you all are getting the bang for your buck.
DI: What about strategies for utilization management? What is the rationale? Are any of these strategies outside traditional credentialing and preauthorization models?
Allen: Prior authorization has retained relative preeminence. Many health plans are now looking at effective ways of getting information about evidence-based criteria for imaging into the hands of referring physicians, resulting in a more accurate and precise imaging study request. Thus far, the primary tool has been prior authorization. There are other models. But most of us who have been doing this for the better part of 10 years are, frankly, disappointed that these other approaches don't work very well.
Dehn: The preauthorization process stops folks and asks them to be accountable for the $2000 to $5000 buying decisions that they would otherwise make in a matter of seconds. What we've seen increasing over the last several years is inappropriate demand. Whether the best preauthorization program involves hard denial, soft denial, mandatory teaching, or a registration program, I'm not sure. But requiring accountability at the time of decision has legs and is going to be around for a long time. And that could be controversial.
Cady: Prior authorization is perhaps the most important tool we have. It certainly has an effect on the appropriateness of utilization. The key is providing real-time education to referring physicians. We are seeing movement toward credentialing, which is increasing competition for referrals based on quality and, in some markets, price.
Greeson: If you've used credentialing and privileging to weed out unnecessary utilization, what percentage of imaging facilities have been denied the ability to contract with your payer client, and of those, how many of them are physician self-referred facilities?
Dehn: We've had a bit of press over the Highmark initiative, in that 24% of sites that applied for participation in the Highmark program were not ACR-accredited. They would have been bounced had they not agreed to go through ACR accreditation. About 11% to 12% chose not to participate or were asked to leave primarily on the basis of inadequate staffing or inadequate number of modalities.
Ryan: A certain number of providers get excluded when you put a privileging program in place, which is what we have done. This keeps some of the nonradiology facilities from providing services and deals with the self-referral issue. Beyond that, applying accreditation and minimum equipment standards has resulted in a fallout similar to Tom's, in the 9% to 14% range. It's the low-end providers who drop out, either because they have a single modality or because they failed to meet our minimum standards.
Dehn: I'd like to comment on the elephant in the room. If we are talking about self-referral, it's incumbent upon us to recognize that no single specialty owns more equipment than radiologists. Radiologists are in a more effective position to self-refer on the basis of their reports than anyone else.
In-office self-referral, appropriate or inappropriate, pales by comparison to the number of examinations that are requested when a physician says, "Look, I had to order this exam, a radiologist put me in the box." That's an issue that the ACR really has to take a look at. Much of it is well meaning. Some of it is done to cover yourself. And some of it is done because they think that the longer the differential diagnosis, the greater the credibility of the report.
Borgstede: Let me make several comments about the strategies and then I'll come back to Tom's comment. First, we appreciate working with utilization management firms. I think the fundamental conflict is that you are still trying to address a volume problem with a per-unit cost solution. In other words, we're talking about preauthorization on an individual exam basis when we really need to look at solutions that address the whole volume problem. I receive a lot of questions from radiologists about administrative denials and steerage of examinations. They want to know what is the real motivation of the utilization management firm. Is it to improve quality and educate referring physicians, or is it merely to save money?
We are focused on preauthorization rather than privileging. And although I've told some of you that the preauthorization piece is considered quick, easy, and cost-effective, I think we need to be focused on the privileging piece. It would be better for patients.
A good example is preauthorization with only a single code allowed. That really doesn't serve the patients well at all. There are administrative denials based on that. You get a patient who comes in, and the only code that's authorized is an enhanced CT scan. Well, it turns out the patient has a contrast allergy and we have to do a nonenhanced exam, and that code is disallowed. The same thing happens with other procedures and can be a significant problem. I'd like to see more involvement placed on the privileging piece rather than the simple solution, the preauthorization piece.
Getting back to Tom's question: If you're looking at a volume of utilization, certainly radiologists do most of the imaging. But radiologists do less than 50% of all imaging. And I think it's inappropriate to say that a radiologist cannot recommend an additional examination. A patient has a fracture, the radiologist recommends follow-up in six weeks. I don't think you can say that's inappropriate.
Dehn: I appreciate that. But preauthorization is unfortunately neither quick-we hear about that every day-nor is it inexpensive. It's a lot less expensive to implement a credentialing and privileging program. The problem is the bang isn't there. But please don't think that I either implied or stated that radiologists should not make recommendations. What I'm saying is that if we're going to be concerned about self-referral, we have to recognize the fact that radiologists are in a position to self-refer. They are not saints. They have the same kind of financial and clinical motivations that other physicians do, and they ought to be included in identification of the problem, because in many cases we find that radiologists recommend studies that don't have to be done.
Borgstede: I would agree. I've heard a number of comments about the methods we could use to educate physicians. The ACR has developed an electronic version of its appropriateness criteria which we've put into a relational database. Others have done similar things; there's already one implemented at Partners in Boston. There, individuals who don't use criteria similar to the ACR's are flagged. In some cases, it can be as draconian as saying the exam is not allowed or that it's inappropriate, and the physician can override that. If you profile the requesting physicians, you can look at who is deviating from the standard. The relational database is going to be key on the educational side in the future.
Cady: One of the benefits of prior authorization programs is collecting information upfront. It comes well before any sort of claims analysis can tell you what's happening in terms of appropriateness and referral patterns. But many of our clients are looking beyond simply rating the quality of the providers within their network. They are passing that information along to consumers. The type of equipment that's in an office, the quality of the staff, who's reading the images, who owns that equipment-all of those pieces of information become very important when people are making purchasing decisions. And many of the biggest health plans in the country are looking more toward consumer-directed health plans, both in terms of the quality of information as well as pricing information within the network. This is going to become more important as we move into the next generation of radiology management.
DI: What's your take on these quality and pricing issues, Mr. Ryan?
Ryan: These are complex issues; I don't think that either data alone or prior authorization alone will fix them. That's part of the problem. Radiologists have abdicated their responsibilities as consultants in many ways and for a lot of reasons. They say they are not going to tell a referring physician that he can't have his study because if they do so, that physician will start referring somewhere else. I think they believe that they're in too difficult a position to do that.
But the problem is not going to get solved without a broad-based combination of components and tools. The trends, even when people are managing radiology, are still in excess of what one would expect based on general increases in population, aging, or morbidity, and certainly more than one would expect based on simple advances in technology. You can't solve it by simply stratifying physicians based on their performance, because it's not all about a few outliers. It's about a substantial number of referring physicians and imaging providers who are performing one to four inappropriate studies a month or every couple of months across 600,000 or 700,000 physicians.
DI: Dr. Borgstede, that's quite a problem. It puts you in a bit of a difficult position.
Borgstede: Yes, it does, to some extent. As I said earlier, we need to have radiologists acting more as consultants. I would refer back to the approach used at Partners, which I think is an excellent approach. It provides, in essence, a relational database showing the appropriate examination in a particular setting, and allows you to look at who is requesting appropriate exams and who is requesting inappropriate exams based on the criteria. You can apply the same thing to radiologists and look at which radiologists are asking for the largest number of additional examinations, then bring everyone a little bit more into line.
DI: That requires quite a bit of technology. Partners has a fairly advanced system. Do we have technology that can be easily applied? Is that something that makes sense to the utilization managers?
Dehn: I think most of the folks in this discussion are familiar with the Partners system. While it is sophisticated, it is not necessarily more sophisticated than the programs that we use. The kind of profiling and identification that Dr. Borgstede and others are talking about is not terribly difficult. But I'd like to underscore the fact, however cliched this may sound, that it is a lot easier to make good doctors better than it is to make bad doctors good. Those who are out there at two standard deviations are, honestly, more of a quality issue than an economic issue. But if you can take the good doctors and tweak them for those one, two, or three unnecessary exams they do a month, that has a huge financial and quality impact on a community.
DI: Mr. Ryan, you said we have a systemic problem here. What about the ideas that Dr. Borgstede mentioned? You could look at the outliers, focus on them, and deliver systems to do that. Is that happening?
Ryan: We are looking at outliers, as I think all of us in this business are. But the outliers aren't going to solve the problem in and of themselves. They're relatively easy to identify in most markets because they surface very quickly. But that's probably 5% or 6% of what the real problem is.
DI: And the real problem is?
Allen: I think Tom [Dehn] and Don [Ryan] have teed this up perfectly. The issue of inappropriate ordering is not confined to a handful of doctors who have no clue as to how to order a test. They're a very small number. What you have is this broad, expansive network of physicians, chiropractors, nurse practitioners, and physician assistants. All of these people can order tests, and virtually all of them order inappropriately some of the time. That's what makes it so vexing. It's not just a handful of doctors who are ordering poorly all the time.
I want to say one more thing about the Partners program. It is a relatively closed health system. These are, for the most part, physicians and service providers who are tied together with an electronic and even financial systems that are not replicated widely throughout the country. We find that imaging ordering behavior is much more appropriate in staff model HMOs, the few that are still left, and in places like Kaiser Permanente, where the physicians are tightly grouped and managed around some of these issues. But that's not how medicine is practiced in large part across this country.
Borgstede: I'd like to come back to a couple of questions that I asked earlier that I think will be important to your readers who are radiologists. That's the issue of administrative denials, steerage, and single CPT code preauthorizations. I think the perception of the radiology community is that we're really not quality- and patient-motivated here, and that what we're really motivated by is saving money. I'd like to hear some comments about that.
DI: That's a good point, and I'd like to roll in one other as well. One of the biggest complaints from physicians is the hassle factor of the process. I'd like to ask the utilization managers to address the questions that Dr. Borgstede brought up but also the hassle factor.
Cady: Medicine is practiced one case at a time in a physician's office. There is educational material out there, and having that information readily available, delivered in efficient, quick, easy-to-follow algorithms is important. We've done a good job of that. We deliver those educational algorithms across a very broad population, across multiple states, and in many situations across 25,000 physicians, not just the physicians that are confined to one particular office or health system. The more exposure that ordering physicians get to this material and this educational information, the more appropriate the imaging will become. And that's really our motivation.
Dehn: Our particular M.O. involves a matrix. If your patient is approved for a brain MRI, it's in the codes as to whether that exam is with, without, or combined. Usually steerage, unless the utilization manager is at risk, has no particular value other than for quality purposes. All of us who have done steerage in pediatric cases, for example, do so because we like to see pediatric cases go to pediatric radiologists. Otherwise, when it's done, it's at the request of the health plan client. As for the hassle factor, one man's hassle is another man's accountability. You can dress this up in a lot of different clothes, but what we're seeing is that there are physicians who find it disconcerting to be made accountable outside of their own office.
Ryan: We may be the most aggressive of UM radiology benefit managers in this area. While we do issue CPT codes one at a time, we provide crosswalks. If you were to receive approval for a CT scan with contrast, and it was decided that the exam should be provided without contrast for whatever reason, such as an allergy, those are automatic crosswalks. We provide those tools to our clients. That would be an automatic approval.
One of the areas where we take a more aggressive stand with MRIs performed without contrast versus MRIs with or without contrast. Those decisions are driven by the clinical criteria and history provided to our physicians and nurses. There's also an opportunity, if the radiologist believes that a different study is appropriate, to call it in and have it reviewed within two days of providing the service. So it requires radiologists to take some action. It's certainly not an overbearing responsibility. We do it because if you look at the national data, in an unmanaged environment where someone wasn't applying those rules, about 40% to 45% of the MRI studies were being performed with contrast. Those studies are valued at twice the cost of an MRI study without contrast. Where we've applied our rules, and we do allow changes, and the criteria are very specific, we found that utilization of contrast for MRI is down to 25%. We think that is quality oriented. It is the right thing to do. Yes, it does save money, and money is part of the program. With regard to steerage, we're interested in steerage because our clients are interested in steerage.
Steerage is driven by a number of factors. The financial issue is one of them, and it usually involves the difference between steering to a freestanding imaging facility or to a hospital. In most cases, payer contracts with hospitals represent reimbursement for the technical component at some percentage of Medicare. That can be two, three, four, and sometimes as high as 10 times the Medicare reimbursement level. If there is an imaging center close by-and the cost for that same service is somewhere around the Medicare level, and it may be possible to save five or six times the cost, with the same radiologist performing the service with good equipment-we think that's the right thing to do for the patient and for our clients, who are the payers.
Allen: I think Don has laid it out pretty well. We have a system related to assignment of the most appropriate CPT code that is driven by the ACR appropriateness criteria. I've talked with a number of radiologists about this over the years in many parts of the country. We believe it provides an effective filter for requests that are coming in from less knowledgeable referring physicians about whether contrast is indicated or whether a sonogram might be preferred over an MR. Our program has very quick and easy solutions for the radiologists to change those codes once they've seen the patient, either before or after doing the study. Our data suggest that doesn't happen more than about 2% of the time in programs that we operate for those single CPT codes.
Sometimes that aspect of the program is dictated by the payers who have limitations on how they can pay claims for the authorizations that we send them. With regard to steerage, Don has made some excellent points. Actually, the politically correct term now is redirection of care. But it serves a number of purposes, and I think we're going to see more of it, as you hear of plans and employer groups talking about transparency of pricing. The concern is for consumers who pay copayments and deductibles for imaging studies. They are asking questions like, "How come I'm going to have to pay $1500 for this MRI over at St. Joseph's Hospital and $650 for it at the imaging center across the street, which uses the same radiology group?"
Borgstede: Regarding the single code approval issue, which is probably the thing that angers radiologists the most, one of the follow-up questions would be, What about the individual who comes in and is approved for a CT abdomen, when obviously there's a need for a CT abdomen/pelvis? The radiologist does the examination that he thinks appropriate, which is what I gather that Dr. Allen wants us to do, and guess what happens? That's when we at the ACR hear the story. It's administrative denial and there's no payment.
Allen: I understand Jim's point. We've described our procedures in sessions with our participating radiology groups and providers across many markets, many of whom have been working with us for a number of years. I suspect they're not nearly as concerned with this issue as some of the newer providers. Still, most of those cases do get approved, and new authorization is issued, as long as we get some information from the facility about the need for the change.
Borgstede: What we hear is that 95% of those are approved. I don't know whether you agree with that statistic or not.
Allen: I would say it's 95% or higher.
Borgstede: The problem is that the 5% denial is the savings you all have promised the insurance companies. You put this program in, and there's your 5%. The analogy that some of my colleagues have made is that it's like lost baggage at the airport. It may only be 5%, but that is still an awful lot. When it's the expansive CPT codes, the range, it's not such a problem. It really gets to be a problem for us when dealing with the single CPT code. The other point I want to leave with you is that it will be hard for radiologists to buy your assertion that your hands are tied by the insurance companies. I don't think that's going to fly.
Dehn: We made a business decision early on in our company that our approvals were going to point to more than one CPT code. It is a little more costly. We don't make money on limiting CPTs, and that's just our rule of business.
Greeson: I'd like to ask each of the representatives of the utilization management companies whether it's their companies' or their clients' requirement that preauthorization be the responsibility of the referring physician. One of the biggest complaints from my clients is that referring physicians have responsibility for conducting the preauthorization, but they absolutely refuse to do so. They'll actually try to steer business based on which imaging center will take over the preauthorization responsibilities from them.
Cady: Our company also uses a CPT grouper system, so we don't experience the hassle factor at the radiology facilities. In fact, radiologists and radiology facilities are very supportive of our program wherever we are. But in regards to the hassle factor, our clients share our belief that the most important component of the education progress is making sure that education is directed at the physician who is actually ordering the study. We do not like to see the provider of the service, the radiology facility, or any other facility trying to go through the preauthorization process.
Ryan: Our position is similar. We expect the referring physician to request and order the procedure because after all, they generally have the clinical information. Most radiologists tell us they're hard-pressed to get good clinical information from the referring physicians. We have worked to eliminate radiology practices that market to referring physicians by saying that if patients are sent to them, they will obtain the authorization. Not to mention the fact that that may violate state and federal fraud and abuse laws. It is, in fact, providing a service to entice a referral. It's surprising how often radiologists will tell us who their peer offenders are. We go after them aggressively and work with our clients to resolve the issue.
Mr. Hayes is editor of Diagnostic Imaging.