Skip to site content
New to Availity? Get Started

AI Tools Should Boost Clinical Expertise in UM, Not Replace It  

In Part two of this three‑part series with Lydia Turner, Senior Manager of Strategy & Health Industries at PwC, the discussion shifts from AI supported tools and turnaround times to something providers and members care just as deeply about: accurate, clinically aligned decision‑making. 

Lisa Bowden, Sr. Product Marketing Manager, and Lydia explore how health plans can evolve their utilization management (UM) approach to deliver faster decisions without sacrificing quality or trust. At the center of the conversation is a powerful theme: the future of UM depends on technology that amplifies clinical expertise, not replaces it, and on operating models that break down silos between clinical, operational, and technology teams. 

Key Insights from the Conversation 

Guardrails, not speed alone, are what make decisions defensible and trusted 

Faster decision‑making only creates value when it is paired with accuracy and auditability. Lydia explains that when clinicians help define both decision rules and guardrails, technology becomes a stabilizing force rather than a risk factor. 

Clinically informed guardrails enable: 

  • Consistent, guideline‑aligned determinations 
  • Clear escalation paths for complex cases 
  • Reduced risk of inappropriate approvals or denials 
  • Greater confidence in automated and AI‑assisted workflows

This approach preserves clinical quality while ensuring decisions remain defensible in audits and compliant with regulatory expectations. 

“The most effective approach is to use technology to amplify clinical expertise—not replace it. That’s how you get speed and consistency while preserving clinical judgment and building trust.” 

Lydia Turner, Senior Manager of Strategy & Health Industries, PwC  

Trust is an outcome of consistent, clinically aligned UM 

Trust is rarely measured in UM, but Lydia points to it as one of the most important outcomes. When providers see consistent, transparent, and clinically grounded decisions, friction decreases and collaboration improves. 

Strong clinical‑technology alignment helps health plans: 

  • Reduce provider frustration and second‑guessing 
  • Rebuild confidence in UM determinations 
  • Create more productive payer–provider relationships 
  • Support better member experiences by enabling timely, appropriate care 

Accuracy and consistency don’t improve operations. They directly influence how providers perceive and engage with a health plan. 

Embedded teams outperform siloed operations 

Operationally, Lydia highlights the limitations of traditional models where technology teams sit apart from UM operations. This separation often leads to misinterpreted requirements, rework, and delayed value realization. 

Health plans seeing success are shifting to: 

  • Embedded technology teams within UM operations 
  • Shared accountability for clinical and operational outcomes 
  • Closer collaboration between clinicians, operators, and technologists 

When technologists understand the realities of UM workflows firsthand, technology becomes an enabler of process and decision‑making, not a bottleneck. 

ROI isn’t a single number 

Lydia reframes Return on Investment (ROI) as a multidimensional outcome rather than a standalone metric. The real payoff emerges when speed and accuracy improve together, at scale, and in a sustainable way. 

Health plans adopting this model are seeing returns through: 

  • Lower administrative costs from fewer appeals and less rework 
  • Reduced downstream medical risk from more accurate decisions 
  • Stronger provider relationships from consistent decision-making 
  • Better member experiences through timely access to appropriate care 

ROI, in this context, is the cumulative effect of lower cost, reduced risk, and increased trust across the ecosystem. 

Stay tuned for Part Three, where Lydia Turner breaks down what health plans must do to ensure their medical policies are accurate, current, machine-readable by AI tools, and evidence based, so they can be safely used by automation and AI.