The most complex question raised by the AQC is how to allocate the global payment among various care providers encountered by a patient. Evaluation of this question requires analysis of methods for data aggregation, a structure for determining the relative value of various services provided to a particular patient and a plan for implementing the change from fee-for service reimbursement to global payments.
The logistics for managing care under an AQC arrangement are significant. Providers are responsible for orchestrating the spectrum of care received by patients, and concurrently optimizing both outcomes and costs. The initial hurdle in achieving these goals is to provide reliable and consistent data to the primary care physician decision makers. The AQC relies on physicians adjusting to new incentives that incorporate costs. Therefore it is essential that physicians who choose where to refer patients are fully informed as to the costs of various referral options. If physicians are blinded from this cost data they cannot be expected to make rational decisions regarding the best locations to refer patients. Secondly, these physicians should be provided aggregated outcome data for referral sites. By pairing aggregate outcomes with the costs of services, physicians will be able to determine where patients will receive the greatest value. The data should be aggregated through the insurance provider, as they are a source of continuity between patients and therefore the only party that is privy to the totality of a particular patient’s care and cost tally. By aggregating this data through the insurance provider and distributing it to primary-care physicians, these physicians will be empowered to make rational decisions on behalf of their patients.
Determining the relative reimbursement for various services is also a major issue within the AQC model. Under the traditional fee-for-service model each service is independently reimbursed at pre-defined rates. However under the AQC model, a significant portion of potential revenue is realized through outcomes based bonuses. Were this bonus to be distributed proportional to a physician’s share of reimbursement for a given patient, the system would again incentivize the provision of high-cost services. Therefore providers should find a mechanism of distribution that links payment to the value of services received by a particular patient, as measured through outcomes. Within the physician practice network various specialties should negotiate their payment rates with the practice administration. This provides internal market forces that allow the individual practice network to determine the relative value of various services. Furthermore, since the primary-care physician is tasked with orchestrating the spectrum of care, they should be entitled to a more significant share of performance based payment than specialty providers.
The transition from fee-for-service to a global payment system creates significant changes in provider incentives, and therefore should be executed in a gradual manner to allow an appropriate market reaction. Initially it would be most reasonable to keep reimbursement rates at existing levels, with efficiencies being extracted through changes in referral volume. Over time in-network physician specialists should be required to negotiate with their practice administration to determine reimbursement rates. Should those services ultimately be priced at too high a level relative to the value provided to patients, one would expect the volume of referred patients to diminish accordingly as a result of changing referral patterns. If provided with adequate information regarding utilization and outcomes, practice administrators should recognize which physicians provide improved value to patients, and can subsequently allocate an increased share of overall reimbursement to these physicians. By leveraging these market forces, a system of continual improvement will be created with incentives for investment in both improved outcomes and efficiencies.
I like Mr. Berlin’s essay the best. The discussion about the need for PCP’s to have as much price and quality information as possible to inform their referral decisions is critical, I think.
ReplyDeleteI would like to offer several thoughts. First, I think we will need to find a way to increase PCP compensation significantly, in part, to allow them to spend adequate time with patients to take an adequate history and figure out how best to treat them rather than just refer them out because they don’t have the time to do anything else.
Second, I don’t think ACO’s will work if they are rewarded for keeping costs below a budgeted target while ensuring quality but they are not penalized for cost overruns. They need to be exposed to risk sharing, especially since, historically, physicians never considered it part of their job to know or care about costs.
Third, while the insurers in Massachusetts can speak to this far more accurately than I can, other insurers have told me that their medical claims costs fall broadly into three main buckets as follows: 40% for hospital inpatient and outpatient care; 40% for physician fees, labs, imaging outside of hospital owned facilities, rehabilitation services and other clinical fees, and 20% for prescription drugs. At most, PCP’s services account for 10% of total network costs or 25% of the 2nd bucket.
Fourth, many people don’t know this but Medicare’s RBRVS system has three components. The first is the physician’s actual work which has a 52% weight. The second is an allowance for practice expenses which carries a 44% weight and, finally, the other 4% is an allowance for estimated malpractice premiums. All are adjusted to factor in geographical differences in overall medical costs. Finally, to convert the end result to an actual fee for service, the relative value units are multiplied by a conversion factor which is the value of one RVU. That value in 2010 was $36.0791. The whole system measures mainly differences in costs for various procedures as opposed to differences in value because the latter is just too hard to determine with any precision. Princeton professor Uwe Reinhardt describes all this in a December 3, 2010 post on his Economix Blog.
Fifth, to attack costs, I think we will also need substantive litigation reform. We need to get medical dispute resolution out of the hands of juries in favor of special health courts or other expert panels and we need to provide doctors with robust safe harbor protection from lawsuits if they follow evidence based guidelines where they exist. We also need administrative simplification of insurance products and procedures.
Finally, to help doctors move toward more conservative practice patterns, patients are going to have to change their expectations especially related to end of life care and to disabuse themselves of the notion that more care is better care. Much of the time, it just isn’t.