Using AI to streamline workers' compensation settlements
Today, artificial intelligence (AI) is playing a significant role in workers’ compensation. The technology is already transforming policy pricing, risk identification and management, and adjusters’ day-to-day jobs, the latter of which is seeing a move away from error-prone manual administrative tasks to an emphasis on strategic guidance around claim management.
These incipient uses of AI are laying the groundwork for an even bigger makeover of the workers’ compensation claims process. For example, insurance carriers and doctors are using AI to better predict injury recovery time and follow-up care based on the historical record of thousands of similar injuries. Very soon, all stakeholders are going to build on this discrete AI use case in a way that is going to dramatically change how injured parties, employers and the insurance companies handling their claims approach settlement negotiations.
Dump thousands of reports prepared by qualified medical evaluators (QMEs) and associated workers’ compensation claim data into an AI engine, and suddenly insurance companies can more accurately pinpoint the cost of short- and long-term care for each injury, and even detail the costs of a doctor’s case management style and future care tendencies. This newfound insight can help arrive at fairer settlement ranges that satisfy workers and employers alike — not to mention save the excessive cost and headache of protracted negotiations and litigation. Furthermore, insurers will have a clearer picture of how much money they will need to set aside to cover the entire lifecycle of active claims.
More variables, better cost predictions
In the ever-evolving landscape of workers’ compensation, the ability to accurately price claims and set reserves early is critical for both carriers and injured workers. Utilizing high-accuracy historical databases, we can now achieve unprecedented precision in early claim evaluations, leading to more effective management and better outcomes for all stakeholders.
Advanced AI models may now leverage historical data to explore multiple scenarios, providing insights into outcomes for both litigated and non-litigated cases, pre-surgery and post-surgery conditions, and the impact of confounding comorbidities, such as diabetes, obesity and arthritis. These models enable early prediction of settlements and the amortization of long-term costs, offering a comprehensive view of potential financial impacts. Models are currently being created based on game theory platforms, which can use serial predictive models to project the probability of therapy, medications, diagnostic tests and surgeries that can potentially result from them, and settlement consequences with and without approvals.
Additionally, AI systems now allow for the creation of detailed worker profiles that consider a multitude of variables. By factoring in the injured worker’s age, ZIP code, employer, company size, access to medical provider networks, medical provider historical behavior and insurance authorization probability models, we can fine-tune predictions and recommendations. The databases can easily identify medical provider network “deserts” and “oases” across not one but multiple carriers. For example, an injured worker with a gastrointestinal claim has no access to this specialty within 300 miles of her home. A prediction for a cost-benefit analysis can account for the costs of long travel combined with overnight expenses, or the decision could be made to simply recommend a compromise and release. This level of detail ensures that each profile reflects the unique circumstances and potential outcomes for every case and even “frictional maps” of circumstances unique to the case that serve either as barriers or opportunities to management.
For example, by understanding the variations in care quality and outcomes based on the worker’s ZIP code or the size of their employer, we can provide more accurate reserve settings and cost projections. Access to specific medical provider networks and the historical behavior of those providers in terms of treatment success and cost-effectiveness are also critical components in this evaluation. Moreover, incorporating insurance authorization probability models allows us to anticipate and navigate potential delays or denials in treatment, further refining our predictions.
Cost certainty and better financial planning
Furthermore, these predictive models are adaptable for integration into long-term investment strategies and financial systems, ensuring that reserves are accurately set and managed. By providing a clear and data-driven approach, AI can assist carriers to make informed decisions that benefit all parties involved.
The value of these models extends beyond the immediate financial benefits. For injured workers, it means faster, fairer settlements and better management of their care. For carriers, it translates to reduced uncertainty, improved financial planning, and enhanced trust with policyholders. All the while, new claim circumstances can be instantly reoriented, such as when a worker who is in the middle of active treatment decides to move out of state, and a sudden settlement is now required prior to natural maximal medical improvement being reached.
In essence, AI is about to kickstart a virtuous cycle. Insights derived from the instant analysis of thousands of past cases and QME reports will produce a clearer picture of the cost of injury care, which in turn will inform reserve cash allotment. If done correctly, this new paradigm should deliver win-win-win outcomes for insurers, employers and workers.
See more4 ways to get ahead of workers’ comp FAQsWhat role can AI play in workers’ compensation?
