CPX Domains: Total Cost of Care
Understanding and managing total cost of care is critical for organizations as they transition from fee-for-service to value-based payment.
AMGA’s Collaborative for Performance Excellence (CPX) will support organizations through this challenge by offering granular benchmarking and tracking reduction in total cost of care, for patient populations for whom participants are receiving adjudicated claims, as part of our Total Cost of Care domain.
Initially, CPX will focus on two common metrics used for analyzing total cost of care (both will be risk adjusted):
- Per Member Per Month (PMPM) Costs
- Cost Indices
Figure 1 compares risk-adjusted PMPM costs at 12 different AMGA member organizations with adjudicated claims data. Across all organizations, average PMPM costs were $447, with a range of $296 to $527 for commercially insured patients aged 18–64. Beyond variation in total cost for covered services, there is even greater variation in cost by type of service. Some organizations have high PMPM for every type of service, compared to peers. They can consider broader approaches to reducing utilization, learning from peers what has and hasn’t worked for them. Others have only a few areas with high cost, where they can focus their attention.
To guide this focus, CPX will explore data by type of clinical episode, type of service, and site of care within your organization. The benchmarks will be trended over time, to track each organization’s progress relative to other high-performing companies.
Figure 2 shows cost data at a single organization with $481 million in annual costs for covered services. Here, cost indices are analyzed overall and in meaningful practice categories (rows) and service groupings (columns). An index over 1.0 indicates that observed costs were higher than expected, based on similar episodes at peer organizations.
The major practice categories with the highest overall cost indices were cardiology and pulmonology, each 1.03, accounting for 27.1% and 6.3% of total cost, respectively. Emergency room costs were 7% higher than expected, with pulmonology-related episodes of care reaching 10% higher than expected. In this example, ER costs amount to only $15 million of the $481 million in total costs, but higher ER utilization could contribute to hospital costs that are 4% higher than expected. This could be a significant opportunity, since hospital costs make up more than one-third of total costs for the population covered by this organization’s adjudicated claims feeds.
The blue bars on the left in Figure 2 enable medical groups to prioritize interventions and show that episodes of care in cardiology, endocrinology, orthopedics, and rheumatology make up more than 60% of total costs. Small proportional deviations in these areas can have substantial impact on overall cost.
Granular benchmarks with cost metrics by type of service, clinical condition, contract, or even site of care can yield valuable information around potential areas of opportunity, where costs are much higher than expected. CPX participants will have access to similar dashboards and benchmarks using cost data from their own organization. With this information, groups can quickly identify areas which account for the highest cost, have significant variation, or are higher than expected.
Tip: CPX benchmarks can highlight areas of high cost within participating organizations at a level of detail that provides clear guidance about key drivers of cost. For example, participants may be aware of outliers in pharmacy costs, but the benchmarks may reveal that pharmacy PMPM remains high, even after outliers are removed. Episode-level analysis may reveal specific patterns in the use of biologics or broad differences in generic prescribing rates. Alternatively, if PMPM is high across the board, groups can analyze costs for specific conditions. Or if there is a large proportion of outliers, they can prioritize earlier identification of high-risk patients.
Tip: Variation across organizations helps CPX participants assess how meaningful deviation from the benchmark in a given practice category may be. If cost is higher than the mean in a practice category where there is wide variation, such as orthopedics and rheumatology, it is likely less meaningful than a similar deviation in a category where the distribution of cost among other organizations is very narrow. This suggests that other organizations have developed consistent practice patterns.
Data source: Optum, AMGA’s Distinguished Data and Analytics Collaborator