The proliferation and variety of accountable care designs in commercial insurance, Medicare Advantage and managed Medicaid highlight the need predict the appropriate cost of care, ensure access to needed services and raise quality. But how can value be measured?
Nearly all accountable care structures share one feature: selective provider networks. But without the ability to account for differences in illness burden, it's impossible to accurately assess appropriate cost and utilization and ensure fair payment for value.
Many healthcare organizations recognize the importance of risk adjustment, but their approach sometimes lacks refinement. Many organizations reason that sicker patients will consume more services. These organizations use a provider's overall illness burden based on a cost-based risk model to extrapolate expected service utilization levels, such as hospital admission rates, emergency department (ED) visits, pharmacy costs, and imaging services.
Ambulatory care sensitive condition (ACSC) admissions, ED visits, and outpatient imaging are three critical areas for risk adjustment because they all contribute to the growth in healthcare spending.
ACSC admissions represent a breakdown in ambulatory care management, when chronic conditions turn into life-threatening situations. Examples include asthmatic crises and diabetes with short- or long-term complications. According to a recent report by the New England Health Institute, ACSC admissions account for $31 billion of waste per year.
Nearly all of the most common reasons for ED visits can be managed in the primary care office, where timely care can be delivered at a reduced cost. Conditions most commonly seen in the ED include upper respiratory infection, fever, otitis media, viral infection, acute pharyngitis, cough, noninfectious gastroenteritis, vomiting, pneumonia, and headache. An astounding 56 percent of ED visits are potentially avoidable and ED overuse costs the healthcare system $38 billion per year, according to the New England Health Institute.
Imaging, meanwhile, has overtaken pharmaceuticals as the service category with the highest cost growth rate, and by some estimates according to ACR Select, imaging accounts for more than 10 percent of all healthcare spending. The American College of Physicians notes that excessive testing costs $200 billion to $250 billion annually, tied in part to overuse of imaging studies.
Application of Relative Risk Values
It is common for healthcare organizations to apply a health risk score intended to predict total cost to the utilization experience of a provider's patient panel in order to assess the appropriateness of utilization rates.
For example, if a provider's patient panel has a Relative Risk Score (RRS) of 1.15--meaning that based on age, gender, and clinically documented medical conditions, the panel would be expected to cost 15 percent more than a comparison group--many healthcare organizations make the assumption that utilization would be 15 percent higher across the board, whether you're looking at ED, pharmacy, or imaging. This approach is better than no risk adjustment at all, but it is still a fairly crude determinant of appropriate utilization. But risk models can be attuned to the utilization of specific services that have a high correlation of use with specific medical conditions.
How Do Risk-Based Utilization Models Work?
Risk models designed to predict either cost or utilization are developed using similar principles. Utilizing large national medical claims data sets, an association is established between the intended outcome and each of 394 clinical conditions in the model's disease hierarchy. For cost models, the result is a relative cost for every individual. For utilization models, each service of interest is calibrated into a distinct model: admissions, ED, imaging, etc. The model calculates an expected number of services for each condition present in an individual's clinical profile, resulting in a value for the person. This enables the estimation of an expected utilization rate that is anticipated for a person's age, gender, and clinical risk profile. The actual-to-predicted service ratio can be used as a measure of service appropriateness. These utilization models are designed to assess the rate of service per 1,000 individuals for population groups, such as a PCP's panel--not as an indicator whether or not an individual patient should or shouldn't have received the service.
Benefits of Performance Assessment Models
Risk-based utilization models offer a range of benefits to a healthcare system that increasingly recognizes the importance of both quality and value. They can inform global capitation budgets so that the allocation of funds is grounded in clinical realities, support performance-based bonus structures and contracting that accurately reflect both quality and efficiency, promote provider buy-in for value-based reimbursement, and identify areas of overutilization.
Once primary care practices have confidence that they will receive sufficient funding to conduct comprehensive patient management activities and they are receiving clinically-based utilization feedback, they are more likely to engage in the long-term global goals of accountable care. Importantly, the downstream effect holds the promise of achieving system-wide improvements in cost and quality, in part through improved management of chronic conditions that account for a disproportionate share of medical spending.
Matt Siegel is senior vice president at Verisk Health.