As healthcare systems succeed in implementing artificial intelligence (AI) tools, especially for administrative use cases, medical professionals and their allies may wonder about several issues, including:
- How will new technologies be applied to the protection of healthcare professionals?
- Can digital transformation prepare medical professional liability (MPL) insurers to stay one step ahead, even with rapid changes sweeping healthcare risk management, underwriting workflows, and litigation analysis?
An MPL insurer’s digital transformation begins with building a unified data platform to inform decisions across business operations, marketing, sales, and interdepartmental processes. This data platform should be reinforced with strong data policies, known as robust data governance.
Once MPL insurers achieve the foundation of a unified data platform, with that platform guarded by robust data governance, only then should they place AI tools on top. The question facing many companies at that point is how to most appropriately utilize AI tools with the right data governance framework in place.
RAISE Digital Preparedness
Medical professional liability carriers can adopt new technologies through a process known as RAISE: Readiness, Alignment, Investment, Solution, and Expansion.
R—Readiness: Assess organizational culture, tech infrastructure, data maturity, skill gaps, and strategic alignment.
To achieve digital transformation, MPL carriers don’t just want to achieve their work goals—they want to improve and streamline processes that speed workflows, eliminating manual processes that impede efficiency.
To effectively combine technology and business value, carriers need to attend to both the staff involved and their collaboration with each other and technology. These processes work on multiple levels, including:
- Individual: Carriers can’t just say, “Team, tomorrow we are moving into this digital world. Now execute.” Innovation includes attention to individual upskilling.
- Organizational: Carriers need high-level sponsorship when attempting large-scale change management.
- Industrial: To succeed, carriers need a good data platform that they can integrate with third parties to optimize productivity.
A—Alignment: Present clear objectives, prioritize use cases, engage stakeholders, and highlight responsible AI practices.
To achieve a unified data platform, a shared lexicon is necessary. Experts in claims, underwriting, and other disciplines can help the technology team identify the different data attributes that point to particular words. That lexicon can continue to be refined over time and across teams. Those dots need to be connected, because if there are no commonalities, teams with different functions will end up on disconnected islands. If that happens, even the best AI tools won’t help—they won’t know what to look at for what purpose. So, it takes human collaboration to achieve digital agility and to make AI work for us.
Along the way, carriers can choose use cases to prioritize. Let’s look at how those choices relate to institutional investment.
I—Investment: Allocate budgets, focus on skill development, and build partnerships for AI capability.
Board members will inquire about budgeting. Skeptics may question whether a digital transformation initiative deserves such substantial investment. One way to demonstrate the potential benefits is to consider nuclear verdicts, which have shot up in both frequency and severity over the last 10 to 15 years. For comparison:
Reminded of these risks, some skeptics may pivot from questioning the scale of investment to asking, how can AI help us optimize our outcomes?
Lifetime value prediction is another use case for digital transformation to facilitate AI adoption. Is it worth selling a specific policy to a healthcare provider? Answering this question more accurately and more consistently promises to yield greater financial stability.
Overall, digital transformation shows great potential to help insurers reduce losses, increase gains, and deliver a satisfying experience to their insureds that aids retention, making a compelling case for up-front investment in developing digital agility now to reap manifold benefits over time.
S—Solution: Select pilot projects to demonstrate value, address concerns, and secure buy-in for larger implementations.
Which problem are we trying to solve? To achieve alignment between language models and tasks, this question needs to be asked consistently.
Building pilot projects creates an opportunity to try things internally to expand the organization’s skillset. But upskilling toward digital transformation actually begins with examining workflows through the lens of existing subject matter expertise, through which project leaders can identify:
- Redundancies: Certain tasks are performed circuitously. Where can tasks be streamlined?
- Inefficiencies: A team has to fight to find certain information—it has to be dug up from multiple locations. Could an AI more easily surface these resources?
- Roadblocks: At places in the workflow where processes get stuck, could an AI circumvent obstacles?
- Loop-de-loops: Extra approvals and reviews are required at certain stages. How can these processes be streamlined?
Underwriting Efficiencies
Even when the results of AI pilots meet expectations, a continuous improvement culture should prevail in areas such as efficiency, profitability, or work experience.
In underwriting, for instance, existing expertise can be upskilled by adding AI to help streamline processes.
Litigation Insights
Carriers can—and should—collaborate with subject matter experts in using digital technologies and AI enablement to extract takeaways from allegations against the insured clinicians. AI can provide insight into patterns by type of claim, type of practitioner, type of practice environment, or by geographical location, and we can apply what we learn to new underwriting.
Ethical AI
Bias in AI is a major concern. Human beings have bias, too, so capitalizing on the best in human capital to bring out the best in AI is a wise course. Carriers must remember that human intervention is critical to the success of digital transformation.
One of the key metrics for AI is transparency. That means carriers must track results and ensure they stay on top of divergent results so that they can evaluate whether expectations and reality are still in alignment. People need to see every step of what AI did to come to its answer because each point of transparency is a form of guardrail.
Documentation and the rationales behind human choices also need to be recorded. Documentation is a guardrail also, because when we document our rules, we can get an answer whenever we ask, “Why is the AI taking that next action?” Documentation for traceability allows for safety, for bias protection, and for fine-tuning, whenever feedback indicates that a model needs revision.
E—Expansion: Monitor KPIs, iterate on outcomes, and scale successful initiatives for continuous value realization.
People worry about their job going away, but it’s likelier to simply change—human input becomes more important, not less, as AI tools are adopted. People are needed to curate the data, understand how people are using the tools, evaluate the results, and then revise and optimize. The question must always be asked as to whether other parameters should be included.
And when the parameters are changed, staff must evaluate that the AI answers still make sense, while keeping an eye on the fit between our datasets and the problems that need to be solved. Understanding that AI has the potential to hallucinate and generate biases means building this recognition into workflows.
A Firm Foundation
New ideas, models, and techniques are emerging daily as carriers embrace tech-enabled insurance solutions that serve the needs of medical professionals and healthcare systems. When carriers build unified data platforms and secure it with sound data governance, digital transformation can be achieved, creating the right foundation to succeed in putting AI to work in MPL.