Is it Time for an ACO Check-up?

With school soon to start many parents are scheduling check-ups for their kids to be sure all is well. With CMS’ recently announced changes to the Medicare Shared Savings Program (MSSP) it may be time to schedule a check-up for your ACO. Since the proposed changes to this program are significant, it will be increasingly important to understand and assure high performance for ACOs. Since many ACOs are utilized for more than just Medicare beneficiaries it is mandatory that health systems understand their ACO performance and its ability to impact the cost of and/or trends in resulting health care costs.

What are the proposed CMS changes?

CMS is redesigning the program by changing the number of tracks that ACOs can enter into an agreement with CMS (was three now two). The new tracks permit an initial one-sided risk model (i.e., gain sharing) which will likely transition into a two-sided model (i.e., risk sharing, both gains and losses). CMS is clearly transitioning away from one-sided models. CMS Administrator Seema Verma has been quoted to saying that the ACOs have not generated enough results and something needs to be changed to get additional results. Projections suggest a $2.24 billion savings to the Medicare program over the next 10 years with the two-sided model.

The proposed changes to CMS’ ACO program transfer additional risk to the ACO and its sponsors with the anticipation that additional cost savings will result. Under a two-sided risk model the ACO must be able to deliver on the cost savings or it will be a financial loss to the ACO.

The proposed changes to CMS’ ACO program transfer additional risk to the ACO and its sponsors with the anticipation that additional cost savings will result.

What are the key performance indicators an ACO needs to consider?

Savings just don’t happen. They result from appropriate management and focus of transformational effort. Several performance indicators need to be measured and assessed to determine effectiveness. The most important ones include:

• Care Management Effectiveness¹ (CME)
• Provider reimbursement and incentive mechanisms
• Appropriateness of attribution methodology
• Reasonableness of target utilization and costs in light of current performance levels
• Reasonable risk adjustment process
• Risk mitigation tools built into the program (i.e., stop-loss, reinsurance, etc.)

How is performance assessed and measured?

Each of the above factors can be readily measured and compared to best-in-class performance to help measure overall performance. A check-up should include a review of each. Only when they are performing at or above reasonable levels can the ACO be comfortable knowing it can assume additional risk. The best performing ACOs will be able to corner the market and achieve significant results for its partners.

The following describe some basic approaches that can be used to evaluate performance:

• Care Management Effectiveness: Benchmarking is critical in this evaluation. The comparison of actual performance for the attributed population to best-in-class performance is the most obvious approach. The most common approach is the development of risk-adjusted utilization and cost information in the format of an actuarial cost model. This can be readily compared to benchmarks for a comparable population with performance ratios calculated. For example, if the risk score for the population is 1.25 the benchmarks can be adjusted upward by 25% and the ratio of the actual performance to the benchmark performance calculated. If the benchmark inpatient admit rate was 225 admits/1,000 and the risk score was 1.25 the adjusted benchmark would be 281 admits/1,000. If the actual admit rate was 315 admits/1,000, the performance ratio would be 112% (i.e., 315 / 281 = 1.12) suggesting an opportunity to reduce admits by 12%. This type of analysis could be expanded to include most of the key utilization measures and/or PMPM health care costs. When the ratios are significantly higher than the benchmarks, this is an early warning sign to the ACO that they need to improve their performance. CMS data extracts (i.e., 5% limited data set sample) can be used as alternate benchmarks to show the seriousness of the deviation from ideal benchmarks. One of the key aspects of this performance metric is the actual care management mechanisms in place to impact overall utilization and cost performance. The resulting utilization and cost are outcomes of the care management process in place. Processes need to be carefully reviewed to assure high performance.
• Provider reimbursement and incentive mechanisms: Within the ACO individual providers will be compensated at specific payment rates (i.e., % of Medicare payments level). These levels need to be compared to competitive levels to help quantify cost savings. The incentive mechanism within the ACO needs to be reviewed to be sure that it is effectively motivating provider performance. Payment rates can be compared to market levels to determine reasonableness. The prevalence and level of incentive payments and the proportion of providers receiving some type of incentive payment is important to understand how well the program is working.
• Appropriateness of attribution methodology: When patients are not directly assigned to primary care physicians, the attribution methodology becomes very important. The effectiveness of the provider’s response to the ACO performance results will be highly correlated with how they view the validity of the attribution methodology. If they question its validity it is highly unlikely performance changes will occur. Providers need to be in agreement they have been involved with the attributed patients. They need to take responsibility for the care they provide and learn how it could be improved if needed. Most attribution methodologies are quite similar and generally are reasonable. However, there are some characteristics that have to be closely monitored to assure the reasonableness of the methodology and how it impacts individual providers.
• Reasonableness of target utilization and costs in light of current performance levels: Most ACO incentive programs are linked to a specific performance target over a specified period of time. Sometimes these may be expressed independently of current performance levels. In those cases, it is important to translate how much performance improvement is required. If the required change is unreasonable then little improvement may result. The target program performance needs to realistically project what an ACO can accomplish. This needs to reflect benchmark performance levels and real-time performance of other programs in the area. Performance varies significantly from one health system to another. Under-performing health systems should be required to accomplish more over a shorter time period. High performing health systems should not be required to improve as much as low performing systems but still should be rewarded for their stellar performance.
• Reasonable risk adjustment process: The key to any performance assessment is understanding the relative risk of the individuals being treated. If a population has much higher risk, it is appropriate for it to higher than average utilization and cost. The ACO should not be penalized for the results of its higher risk. However, the risk adjustment process must accurately project the impact on utilization and cost. There are many different mechanisms that can be used to reflect varying risk. In the author’s opinion, the best risk adjustment mechanism is the one that providers have the greatest understanding of and can clinically respond to different risk scores. A clinically intuitive risk adjustment mechanism would maximize the provider’s willingness to accept the results of any performance assessment.
• Risk mitigation tools built into the program: Performance measurement will be significantly impacted by unusual conditions or treatment requirements. Catastrophic cases, even with appropriate risk adjustment, can bias the results. The ACO needs to be sure they are adequately protected through the appropriate implementation of risk mitigation tools. This needs to be accomplished in concert with the choice of risk adjustment methodologies.

How often should performance be assessed?

We recommend ongoing assessments be implemented in each ACO. This would include ongoing comparison to benchmarks. Interim results should be monitored and documented by periodic reporting (i.e., monthly, quarterly, etc.). ACOs should establish practices to develop trend analysis and comparisons to targets to facilitate appropriate improvements in care management processes.

Summary

ACO performance measurement is critical to long term business viability. As ACOs demonstrate an ability to control the utilization and cost of healthcare we can rest assured there is hope to reduce the health care burden we face in the United States. CMS’ recent announcement is encouraging, as it focusing more effort on high performing ACOs.

¹Metric developed by Axene Health Partners, LLC which is used to assess utilization and cost performance in terms of ideal performance. Metric ranges from 0% to 100% where 0% suggests ineffectiveness and 100% is representative of ideal performance.

David Axene, FSA, FCA, CERA, MAAA, is the President and Founding Partner of Axene Health Partners, LLC and is based in AHP’s Temecula, CA office.