How AI Can Improve Treatment & Cost Forecasting Accuracy Within the Medicare Set-Aside Process 

By Kimberly Wiswell, Director of Client Services, Embedded AI, Gradient AI  

Medicare Set-Aside Arrangements (MSAs) are an important part of workers’ compensation programs and assist in earmarking settlement funds to cover future medical expenses stemming from workplace injury claims. However, it can sometimes be challenging to forecast treatments and accurately predict costs.   

The treatment and drugs included in the cost projection tables within an MSA are drawn from medical records, as well as past medical history. Given the number of medical reports associated with older claims, review of these records can be a lengthy process, often taking several hours or more. In the past, however, this was the only method of ensuring all future treatment recommendations were identified. But now, AI-assisted record review has emerged as a game changing approach. 

Challenges in MSA Treatment Forecasting and Cost Estimations 

Treatment Forecasting Inefficiencies - The current MSA process is laborious. First, a specialized allocator physically goes through the injured worker’s medical records, creating a summary of the injury and the medical treatment. For each treatment, the allocator must determine whether the recommended treatment has been provided to the injured worker. Any unfulfilled recommendations must be included in the MSA.  Next, a treatment plan is developed based on physician recommendations for specific treatments such as surgeries, diagnostic testing, hospitalizations, etc., as well as treatment corresponding to CMS’s standardized guidelines. Last, the allocator prices the services and drugs under the treatment plan using state fee schedules and other pricing resources specified by CMS and recommends an MSA amount for the claim.  

All of this can take several days to more than a week, which unfortunately leads to delays in claim closure and settlement funds being paid to the injured worker. In addition, the existing approach often leads to unpleasant surprises for claims managers. Unrecognized treatment recommendations often, consequently, increase the overall MSA amount, sometimes making the settlement impossible.  

Cost Projection Inaccuracies – It is challenging to accurately estimate future medical costs within the current MSA framework. Most services are priced out based on state workers’ compensation fee schedules or, when unavailable, usual, customary and reasonable (UCR) fees for the jurisdiction. As fee schedules are unique to each state, and fees are updated as often as quarterly, determining the correct pricing can be a complex process. Also, as the claimant’s medical condition and health status change over time, often necessitating new and different treatments, the amount allocated in an MSA may not accurately reflect the future funds needed. Compounding this, CMS requires MSAs associated with settlements to have been done within the past six months. After this point, they are considered stale-dated, leading to additional costs for the claims payer, both for multiple MSA updates and increasing treatment fees over the course of lengthy settlement negotiations and proceedings. 

AI Solutions Address These Challenges and Enhance MSA Management 

AI solutions can significantly improve MSA operations by addressing inefficiencies in the identification of future medical treatments. By analyzing vast amounts of medical reports, generative AI can detect treatment recommendations within the records, compare them to structured medical payment data for treatment already provided to the injured worker, and flag treatment yet to be provided. These insights allow allocators to readily identify services and drugs for inclusion in the MSA.  

For example, by analyzing historical treatment data, AI can determine specific treatment frequencies for evaluations, therapy, and surgical revisions, as well as the injured worker’s current drug regimen. This data helps MSA allocators make better decisions when forecasting treatment and drugs, which leads to improved cost savings and more accurate funding of the MSA. 

MSA cost projections can also be more accurate by leveraging AI. Predictive analytics can monitor trends, analyzing information in real time.  In cases where settlement hasn’t occurred, it can automatically recommend changes to the MSA to meet the changing medical conditions and related treatment needs. AI also notes important things that can change, such as evolving treatment plans, how medication is used, and how much things cost. It can also identify and flag escalating cost-drivers such as costly surgeries or procedures early on with suggested actions on how to address each cost driver to mitigate costs.   

As healthcare continues to evolve, AI will become even more critical to streamlining operations to achieve better outcomes and ensuring the long-term success of workers' compensation programs.  

Share This Article: