Evidence-informed case rates.

I’m sure that your P&T committee is grappling with a host of concerns related to the future affordability of health care. No doubt you’ve been briefed on pay-for-per formance, the Physician Quality Review Initiative (PQRI), and, of course, capitation. Perhaps you haven’t yet heard much about “evidence-informed case rates” (ECRs). I’d like to trace the recent history of ECRs, describe how they are constructed, and contemplate where they might take us. A brief released by the Commonwealth Fund in New York City drew my attention to ECRs.1 It was written by Francois de Brantes, MS, MBA, and Amita Rastogi, MD, MHA—names that are familiar to many of our readers. These individuals are the current leaders of the national program called “Bridges to Excellence” (look for a future editorial on this topic). Appropriately, the authors give credit for spreading the word about ECRs to other well-known researchers in health care services, including Jerry Solon, PhD, and Mark Hornbrook, PhD. ECRs are the building blocks of a payment model known as Prometheus Payment, Inc., a nonprofit corporation that is working to develop a new method of improving health care quality, lowering administrative costs, increasing transparency, and supporting a patientcentered health care environment. An episode-based payment model has been created to include global fees to pay for all of the health care recommended by various guidelines for treating patients. These global fees are combined with performance incentives that are designed to promote cost control and improve the quality of health care. Part of the payment is withheld and is redistributed ac cording to the health care provider’s performance in measures of clinical processes, patients’ outcomes, and patients’ experiences. Authors de Brantes and Rastogi, on behalf of Prometheus Payment, have been modeling ECRs to determine the extent to which a “fully priced” episode of care can be assembled in a way that distinguishes the various types of risks involved in the total cost of health care. ECRs seek to separate probability risk from technical risk. Probability risk, the classic form of insurance risk, is caused by the likelihood of a negative event affecting a patient. This is typically the result of the patient’s genes, health status, or any external event not controllable by a health care provider. For example, a patient contracts a virus or breaks his leg. In the Prometheus Payment model, insurers bear full financial responsibility for this probability risk. Technical risk refers to health care that is controllable by health care providers and is a result of their clinical skills. The Prometheus Payment model defines the negative consequences of technical risk as a potentially avoidable complication (PAC). For instance, a patient admitted to a hospital for an acute myocardial infarction (MI) may also experience the complication of a urinary tract infection (UTI), phlebitis, or a stroke. The model implies that health care professionals should be able to prevent an acute UTI, phlebitis, and even a stroke for a patient who is admitted with an MI. ECRs, it is thought, separate technical risk from probability risk in order to hold physicians accountable for the former but not the latter. Thus, there is nothing physicians can do about a patient’s genes, but they should be able to prevent an acute UTI while the patient is in the hospital. Some practitioners might argue that some PACs are not under their direct control but, rather, are under the control of a different health care provider. However, the key goal of ECRs is to pay for all health care access, thereby creating collective responsibility for managing a patient’s condition. With the concepts of technical risk and probability risk, the accountability for results can be appropriately assigned to insurers and health care providers, suggest authors de Brantes and Rastogi. How are ECRs constructed? The authors describe a five-step proc ess: (1) defining the boundaries of an ECR, (2) adjusting for regional variations in practice patterns, (3) analyzing the study population, (4) adding a 10% profit margin, and (5) creating an allowance for PACs. Here’s how that allowance might work. We would sum up the total cost of all PACs as a base, then free a pool for PAC allowances equal to 50% of the total base. This money, in turn, is spread in a fixed and proportional way to each ECR, thus creating an ECR-specific allowance for PACs. This allowance creates an incentive for physicians or health care pro viders to reduce the number of PACs, because they get to keep the difference between the allowance and the actual cost of the PAC. For example, a physician group manages 100 diabetic patients who are covered by an ECR. The average severityadjusted base is $4,000, and the average PAC allowance is $1,800. At the end of the year, if the severity-adjusted cost of typical care averages $4,000 and the cost of PACs averages $1,000, the physician group would collect $80,000 from the portion of the PAC allowance that was not used. The $80,000 comes from the difference between the $1,800 allowance for PACs and their average cost of only $1,000. This means that $800 × 100 cases = $80,000. This method, therefore, provides the incentive for health care pro viders to reduce the number and size of PACs; they keep the difference between the allowance and the actual cost of each PAC. The authors believe that the ECR, as created, starts when an episode of care begins and a base case rate is paid. Addi tional allowances become available for comorbid conditions and procedures that are included as part of the ECR. Money from the pool designated for PACs also becomes available, depending on the severity and complexity of the patient’s condition. In general, the Pro metheus Payment model is supposed to encourage health care professionals to EDITORIAL