Utilization manager answers July critique

Dr. James H. Thrall's column "Utilization management steps onto center stage" (July, page 29) carries the subtitle "Referring physicians and radiologists must wrestle decision making back from insurers," prompting the question, Was decision making usurped or abandoned? As chief medical officer of CareCore National, one of the aggravating "1-800 programs" Thrall refers to, I present a perspective from the other side.

When I began my residency, a week or two after William Roentgen's interesting discovery, my chief stated that nearly a third of the studies we did were extraneous. No one then was particularly concerned about excessive $25 chest x-rays. Unnecessary $2500 PET scans, however, have garnered much greater attention.

It is clear that imaging expenses are rising at rates that cannot be sustained. Though many factors contribute to this trend, prominent among them is the failure of referring physicians and radiologists to exercise restraint in their decision making. On the one hand, referring physicians are not paying for what they order; on the other, radiologists make their living from these orders.

Referring physicians face pressures that discourage considered judgment and induce overutilization. Clinicians have repeatedly told me that the economics of their practices dictate rapid turnover of office visits. Thus, from their viewpoint, it is most efficient to generally evaluate a patient and then order multiple studies, hoping a clear diagnosis will emerge. A detailed history and physical might have done the job or narrowed the search, but at the cost of expending time that could be spent more profitably seeing one or two additional patients. Similar pressures lead them to accede to patient demands for costly imaging procedures that the caregiver knows are unnecessary. The physician must decide whether to spend time (income) explaining why an MRI of the lumbar spine is inappropriate after two days of low back pain or to say "Okay, let's take a look" and move on to the next patient. There is no inducement to be frugal.

This is not to say that referring physicians order unnecessary tests solely for economic reasons. They often order tests reflexively, because a new technology was recently discussed at a meeting or on television, or because that study is what was ordered when they were in training five, 10, or 20 years ago. How many imaging requests do physicians thoroughly analyze before placing the order, asking themselves the necessary questions: Will this particular study really reduce the uncertainty of a diagnosis? How will treatment change if positive or negative? Do the sensitivity, specificity, and accuracy of this study actually make it valuable in this particular clinical situation?

We see requests every day for MRI exams of the breast in patients with palpable lesions or suspicious mammograms, ordered "to avoid a biopsy if the MRI is negative." Very few physicians familiar with the negative predictive value of breast MRI would agree to such a strategy.

Often cited among the causes of overutilization is defensive ordering. How much defensive ordering relates to fear of litigation and how much to the fear of making a mistake of omission is unclear. Ordering a study that was unnecessary, however, will prove a successful defense against neither fear.

As a radiologist involved in medical imaging for 30 years, I have marveled at the astonishing growth of our field in its capacity to serve the interests of both patients and our own economic well-being. Somewhere in those three decades, we evolved from consultants to very well paid commodity providers. Yet, in that transition, we have nearly abandoned the best utilization management tool: radiologists' expertise. We buy and install the latest, most expensive machines and advertise their enormous capabilities but often downplay their limitations. Is it a surprise that our colleagues and the public demand more and more imaging?

I often ask radiologists whether they turn away cases or call referring doctors if they know that the exams ordered are unnecessary. Almost without exception, they do so rarely. For one thing, they are uninvolved in the process. They seldom know which exams have been performed or why until the images are ready for interpretation. But even when they know in advance that a study is ill-advised, they express reluctance to call a referring colleague to suggest that the study is better omitted. The reason most often cited is that such an action might cause that colleague to send future referrals to a more compliant competing radiologist.

Is it not our responsibility to guide imaging choices to those most likely to benefit our patients? Or do we just do what is ordered and send the bill?

Organized radiology has rightly pointed out that self-referral by other specialists contributes to the problem of overutilization. Sadly, I have barely heard a word about our own form of self-referral-the recommendation for another exam "if clinically warranted." Such suggestions may occur when the reader is unable or unwilling to making a firm decision. But we have seen practices that routinely suggest additional imaging in their reports.

I recently came across a recruiting advertisement for a practice that paid its members on an "eat what you kill" basis. Now, there's a metaphor the specialty can be proud of, and one that suggests these radiologists are unlikely to say no to even the most outrageous request.

How then to deal with the problem? Dr. Thrall suggests "pay for performance," a retrospective analysis of ordering patterns and volumes linked to reimbursement, as one alternative. But retrospective analyses of tens of thousands of doctors, each believing him- or herself to be unique, would prove more intrusive and less effective than expected. The task of collecting and analyzing the data would be formidable. The analysis would have to be primarily statistical, since good clinical data would be unavailable. Those accused of overordering would proclaim that their careful practices were being attacked for purely monetary reasons, their patients were sicker than average, or the care they provided was better. This approach is virtually unworkable.

REAL-TIME SUPPORT

Provision of online real-time decision support seems more attractive. It may be slightly less time-consuming for data entry than a prior authorization program, and it has the aura of promoting education and, thereby, good medicine. It is, I fear, only an aura. Once referring physicians-or, more likely, the office staff-become familiar with the computer entries that bring a high rating for any particular exam, the temptation will be great to use the same entries for all subsequent patients.

Empirical evidence shows that "beating the system" in this manner (i.e., providing a history that meets guidelines regardless of whether it accurately reflects the real clinical situation) is not considered unethical or lying. We see this every day.

Other tools have been proposed to, in essence, restore sound judgment. Privileging nonradiologists to perform only certain types of studies, based on their specialties, has a positive impact by decreasing self-referral. Requiring documentation of quality indicators, accreditations, and certifications also has a positive impact. But these tactics alone will never curb runaway imaging costs.

We agree with Thrall that it is desirable to educate referring physicians. We always have a board-certified radiologist and several other board-certified specialists available to discuss cases with referring doctors seeking advice or explanations. We have posted a brief, informative imaging guideline on our Web site (www.carecorenational.com). It is beyond the scope of an organization like ours, however, to educate tens of thousands of physicians about the appropriate use of medical imaging.

Prior authorization, the approach least favored by Thrall, is (with apologies to Winston Churchill's definition of democracy) the worst utilization management tool except for all the others. Requiring prior authorization produces two measurable effects: First, and harder to quantify, ordering physicians exercise restraint when they know they are subject to some review. Doctors do not like to be second-guessed, so they think twice before ordering. They also do not want to waste time pursuing a prior authorization they know will not be forthcoming. This is the "sentinel effect," and it is real. Second, we deny authorization to a significant proportion of cases. After conversing with us, referring doctors withdraw requests for another significant number. We base these denials and withdrawals upon our application of thoroughly researched, evidence-based guidelines.

At CareCore National, we use McKesson's Interqual guidelines, at no small expense, because we know they are produced by an exhaustive multidisciplinary process. We supplement these with additional guidelines when we feel that the state of the art warrants modifications between Interqual editions or for studies Interqual does not cover. Our network of academic and community-based radiologists, other consulting specialists, and the medical staffs at all contracted health plans develops or evaluates all criteria continually. We estimate that our fully implemented program reduces utilization by 12% to 20% compared with unmanaged rates. This is not a trivial reduction in costs and radiation exposure to patients at the expense of studies that lack sound medical indications.

CareCore National has never had a quota for denials and never will. There is simply no reason to impose one, as more clearly unwarranted requests are submitted than we can eliminate. We immediately transmit all negative determinations to both the requesting doctor and the patient. We include detailed explanations of the reason for the denial and a list of extensive appeal and reconsideration rights.

Thrall is correct that "physicians and radiologists [should] wrestle decision making back from insurers," but this will never happen until they realize that it was not wrestled away by insurance companies; it was abandoned by physicians.

Dr. Komarow is senior vice president and chief medical officer of CareCore National in Wappingers Falls, NY.

Reply from Dr. James Thrall

Dr. Komarow weighs in with several thoughtful comments, including some that reinforce important points made in my article on the subject. We probably both agree that all responsible observers are concerned about rising healthcare costs, and that eliminating excess costs resulting from unnecessary or inappropriate imaging procedures is highly desirable.

Dr. Komarow's characterization of referring clinicians as unconsidered in their ordering habits and radiologists as often self-promoting and unwilling to turn away unnecessary exams is undoubtedly true for some fraction of both groups. Payers and, more broadly, society at large expect all providers to adhere to best practices and to observe the Institute of Medicine's call to "cross the quality chasm." However, adjudicating each unit of service remotely on a case-by-case basis is simply too cumbersome and expensive for providers and too inconvenient for patients.

Moreover, in carve-out utilization management systems, the conflict of interest shifts from providers benefiting from performing services to utilization management companies benefiting from denying services. Whether one believes that decision making was taken away or abandoned, as Dr. Komarow asserts, does anyone really hold the high moral ground?

Dr. Komarow defends his company's approach to achieving cost reductions by reducing unnecessary tests but does not address several important issues. There is no analysis of the direct or indirect financial costs to participating practices. These substantial costs for referring physicians and radiologists arise from increased telephone calls, periods on hold, and the time required to document and retransmit information. Nor does he address the cost to the insurance industry. Money paid to other parties, such as Dr Komarow's company, adds to insurance companies' overhead and subtracts from the funds available to pay for patient care services.

The issue of quotas is open to interpretation. Utilization management companies may not advertise a quota of exams to be denied, but an insurance company would be unlikely to spend more than it anticipates saving, thereby creating a de facto quota of exams to deny. Moreover, utilization management companies must estimate denial rates for potential new insurance company clients, or they would not sign on in the first place.

Billing errors inevitably increase in prior approval systems, due to increased process steps. Among the companies we deal with at Massachusetts General Hospital, the number of alphanumeric digits in authorization codes ranges from eight to 13. By the time the code gets to us in radiology, there is an error rate of up to 5%. Utilization management companies and their insurance company clients win. Radiologists and hospitals lose.

Prior approval systems can be cumbersome and inflexible in identifying appropriate examinations. They are geared toward saying yes or no to a particular request but not toward finding a more suitable alternative, let alone the best possible examination. If, for example, a radiologist realizes that a patient has just developed abnormal renal function and should not be administered contrast media as part of a CT scan, the system breaks down because approvals for contrast-enhanced versus noncontrast CT are not interchangeable. This leaves the radiologist and patient stuck in limbo since the entire approval process must be repeated through the referring physician's office.

The utilization management companies with which we deal have never learned in detail about the technical capabilities of the MGH department of radiology. How can a company be the arbiter of more responsible resource utilization when it has incomplete knowledge of the resources in question? How can a remote company direct patients according to best practices if it does not know what services are available? Will a conversation between a referring clinician and a company representative unearth the subtleties and cutting-edge aspects of that information? Probably not, but a conversation between a referring clinician and a radiologist working in the department surely will.

The topic of "evidence-based medicine" requires context. While we should all strive to use the best evidence and best practices, it is extremely difficult to summarize literature and stay up to date across the board. For example, among the first 50 journal references noted in the appropriateness criteria published on Dr. Komarow's company's Web site (www.carecorenational.com/ forms/clinicalguidelines, accessed Aug. 29, 2005), the articles cited range in age from two to 26 years. This illustrates how daunting the challenge is, especially in a rapidly evolving field like radiology. Even a two-year lapse in updating the evidence at this time leaves out any information about the capabilities and benefits of 64-slice CT. Practicing evidence-based medicine is a worthy goal, but let's acknowledge its pitfalls and shades of gray.

Whether the emerging concept of pay-for-performance (P4P) can find a golden mean between disparate financial interests and behaviors remains to be seen. P4P programs are based on prospective goals and can accommodate quality as well as financial measures. These programs can be applied to individual physicians or groups. Their singular advantage is the latitude given to providers for any particular case. The goals reflect aggregate behavior over time and recognize that healthcare decision making at the individual case level is not always black or white. The occasional hunch is permitted.

It is interesting to consider what characteristics and attributes an ideal utilization management system might have. These systems will have to make work easier, not more difficult, and should never result in denial of rightful access to services or rightful payment because of their cumbersomeness. The ideal system will guide physicians to best practices to improve quality and will provide statistical feedback. It will evaluate the actual yield from a physician's ordering patterns and compare it with other factors such as ordering rates to correct for prevalence of disease in his/her practice. The art of medicine will be respected as physicians retain some latitude in decision making based on their experience and intuition. Data obtained over time will be used to refine best practices by looking at diagnostic yield associated with different combinations of indications and examinations.

Dr. Thrall is radiologist-in-chief at Massachusetts General Hospital.