Profiling: Defining the link between quality, cost of care
June 1, 2009
Health insurance payers are rating their providers by the efficiency of the care delivered, and are beginning to steer patients to the most cost-effective care providers. However, cost-effectiveness does not always equate to high quality, and there are a number of potential problems inherent in the insurance companies' methodologies that make this a disturbing trend for physicians.
"Physician profiling is used to determine how effectively or efficiently your practice takes care of patients," Dr. Repka said. "By definition, profiling for health policy purposes is the efficiency attained. In other words, a given level of quality of care is achieved at the lowest total cost. It's a concept for economists and actuaries, not physicians."
The payers typically rate efficiency as the ratio of actual resource use to expected resource use for an equivalent quality of care for a particular diagnosis, a particular treatment course, or a particular contact with a patient, he said. All of those "ors" depend on the particular scenario in which the profiling is occurring.
"Medicare has been trying to take the high road by doing only quality profiling, but there is a great deal of pressure in both the public and private sectors to expand economic profiling as another method of controlling costs of care," he said.
How profiling ultimately is going to affect physicians is not clear, but it already is being used to stratify them from most efficient to least efficient.
"Payers use this information to steer patients to providers who provide cheaper care or deliver the same quality of care at a lower cost," Dr. Repka said.
Although such an approach may seem reasonable from an accountant's perspective, physicians (and patients) have good reason to be wary.
Many concerns exist about the data. First of all, the process of assigning percentage costs to the many doctors who may be involved with the same patient is capricious, based more on software than on wisdom, Dr. Repka said.
"The software may or may not be capable of exploring nuances of the individual situation," he said. "And claims data may or may not actually represent the activity. Claims may not have been filed. They may have been incorrectly assigned. Just look at all your current [explanations of benefits] and you can see all the potential for problems or inaccuracies."
Utilization data generally do not consider subspecialty differences. For example, nothing in the Medicare coding differentiates a comprehensive ophthalmologist from a specialist, such as an oculoplastic surgeon, who by the very nature of the specialty will have higher costs.
Sorting out costs between overlapping and multiple providers is another danger area. Insurance company "grouper software" examines claims data based on a diagnosis and the Current Procedural Terminology codes that have been submitted for a patient. It then aggregates the cost of the patient and assigns costs to different providers based on the coding.
"You obviously have the potential for problems there," Dr. Repka said. "A cataract patient also may be seeing a cardiovascular surgeon, and the ophthalmologist can get charged for the coronary stent because of problems in the software."
Costs allocated to a physician also often include those over which they do not have much control, such as drug and hospital charges.
"We don't have a lot of say when it comes to what our hospitals or medical centers may charge for certain services, but it goes into our cost basis," he said.
Inadequate risk adjustment is another major problem with physician profiling.
"There's really no way to adjust for risk, patient morbidities, and patient non-compliance," Dr. Repka said. "Perhaps of even more concern are diseases in which we care for very few patients. Small sample sizes can give real flaws in the data."
Despite all the potential for inaccuracy, the "actual costs" as tallied by insurers' varying methodologies ultimately are allocated to an episode, such as a patient visit or a surgery. The insurers then calculate what the expected costs for that episode should have been, based on actuarial databases.
The payers then calculate the cost-efficiency measure (CEM), which is the ratio of actual costs to expected costs. Providers' CEMs are compared, and payers look for those who have the lowest ratios. They then begin "directing" patients to those providers, using the incentives of co-payments or co-insurance to drive the referrals.
"In a way, it's an economic step beyond preferred provider," Dr. Repka said. "The 'tiered' or 'narrowed' networks that are arrived at by profiling are based on the cost-effectiveness of care. Insurers then guide patients to the most cost-effective providers in much the same way they use different levels of co-payment on formularies to guide patients to generic drugs."
Dr. Repka said that profiling may present a barrier to care: Too few physicians could be in the "most efficient" tier, because that's where the insurers are encouraging patients to go. It also means that in many cases insurers are encouraging patients to break long-standing physician-patient relationships.
"And, of course, the overall theme here is that the 'preferred' tier-as judged by the payers-must mean quality, safe, effective care, which in fact may not be the case," he said. "The risk is a network filled with providers who can't provide comprehensive care, who don't, won't, or can't take care of the sickest patients."
Economic profiling has been rejected by Medicare, so it is not part of Physician Quality of Reporting Initiative-at least not yet. But for better or worse, it is here to stay and only will become more ubiquitous.
"Commercial plans haven't had the patience to wait for valid cost-of-care measures or valid clinical definition of episodes of care," Dr. Repka concluded. "They've moved along and profiled everybody. Every physician already has been profiled, although not all plans may be using the data."