What Health Plans Need to Know About Provider Usage of AI

Table of contents

Staffing shortages continue to be a challenge in healthcare, so it makes sense that providers would begin to rely on artificial intelligence (AI) solutions to handle some non-clinical tasks and processes. While AI tools can help improve efficiency and accuracy, there are implications for health plans and others who rely upon the outputs from these tools.

Take a look at the 4 most common ways providers are beginning to incorporate AI into their practices and find out how these developments could impact health plans.

1. Ambient listening and documentation

Ambient listening tools capture conversations between providers and patients, automatically generating medical documentation. This technology helps providers improve charge capture and medical record accuracy.

With better documentation, health plans should see fewer errors during medical record reviews. However, more complete records could mean higher charges on claims, as providers are less likely to omit billable services. This could increase overall medical spend for health plans, even as the accuracy of claims improves.

2. Autonomous coding software

As the name implies, autonomous coding software requires minimal human intervention. It uses AI to assign billing codes to medical services. While autonomous coding may currently be used on a limited scale, as it is further developed questions are arising about its effectiveness and neutrality. If the software is accurate, it could reduce errors and improve the precision of provider coding.

Autonomous coding software should mean the coding on claims is more likely to be correct when submitted to health plans. As a result, health plans may find fewer opportunities for cost savings through edits or medical record reviews. Given the questions that still surround autonomous coding, health plans are encouraged to continue monitoring codes.

3. Denial prediction

Providers and revenue cycle management teams are developing AI models to predict when claims might be denied. These models prompt providers to correct claims before submission, with the goal of increasing the likelihood of approval.

This technology could lead to more accurate claims which should reduce the number of corrections health plans need to make. The potential downside, however, is that coding could be manipulated to avoid denials. Health plans must remain vigilant and watch for potential abuse.

4. Care pathway nudging

AI-powered nudging tools, often paired with ambient listening and practice management systems, suggest care options to providers. These prompts encourage providers to ask specific questions or perform certain tests.

While these tools don’t affect the health plan’s decisions about care appropriateness, they could increase medical spend if providers deliver additional, potentially unwarranted care. Health plans must be aware of this possibility and monitor trends in utilization.

What should health plans be doing now?

AI also provides health plans with many tools to handle increased provider AI usage:

AI can monitor for behavioral anomalies, such as sudden spikes in appeals volume, drops in payment integrity true positive rates and unusual or templated communications that may indicate AI-generated content.
Stylometric analysis of provider communications can help flag automated documentation.
Claims pattern analysis can reveal changes in coding practices or increased consistency that suggest AI-driven optimization.

To prevent losses from provider AI usage, health plans should invest in advanced analytics and machine learning to proactively detect improper payments, fraud and abuse. Automating claims review processes with AI supports accuracy before payments are made, while adapting policies and renegotiating rates can help maintain financial stability in response to more accurate billing.

Most importantly, proactively educating providers on AI-enabled tools and establishing human oversight frameworks for explainable AI are essential for transparency and fairness. Ultimately, collaboration between health plans and providers will be the crucial component to balance innovation with payment accuracy.

About Bill Maly, Senior Vice President, Payment Integrity Product Analytics

Bill has been with Optum for 24 years and has led global teams across operations, growth and strategy, and AI analytics. He has helped the Payment Integrity business transform from emerging capabilities to a product suite that generates measurable outcomes and value for clients.