The Evolution of AI in Healthcare and the Revenue Cycle

Its highly transactional nature has made healthcare – and revenue cycle management (RCM) in particular – an area that is ripe for automation and the application of artificial intelligence (AI).

For example, manual and redundant tasks within patient access, coding, billing, and collections are prime for automation, as are billing tasks such as posting adjustments with bad debt and small balance write-offs. Meanwhile, AI can be applied to real-time analytics, prior authorization, workflow prioritization/optimization, and denial mitigation. (Download our white paper here.)

Defining the Technologies

Before we can fully understand the evolution and impact of automation and AI in healthcare, we must understand what they are and how they differ.

AI is the collection of technologies that allow the machine to act at the human level of intelligence, which requires learning from past experiences and self-correction to make decisions and reach conclusions. Automation, on the other hand, can run with little or no human interaction by leveraging specific patterns and rules to perform repetitive tasks.

Because they serve similar purposes, the terms “AI” and “automation” are often used interchangeably. However, the differences are not as subtle as they appear. While automation is about setting up “bots” to follow a set of pre-defined rules, AI is about setting up “cognitive bots” to make their own decisions. While both run on data, automation collects that data while AI learns and interprets it. By using automation in connection with AI, data can be collected, transferred, and understood, and automated actions can be performed based on that understanding.

Over the years AI and automation technologies have evolved to the point that they are now addressing some of healthcare’s biggest pain points, in particular around the revenue cycle.

Impacting Healthcare

AI and automation technologies are improving exponentially, making them more agile and useful for enhancing provider operations. As such, more than 80% of healthcare organizations have already implemented an AI strategy while another 15% are planning to do so.

Decades of experience with AI solutions have also helped decision-makers succeed in focusing on the right objectives, including addressing patient dissatisfaction with current fractured experiences and slowing the acceleration of clinician burnout and turnover. Other important focal points are education and corrective actions for providers and coders, which will improve compliance.

Many of the more recent innovations in AI were driven by the COVID-19 pandemic, with healthcare organizations leveraging AI to help meet capacity challenges, accelerate the search for a coronavirus vaccine, transition to telehealth, address staffing shortages, etc. In turn, these advances have set the stage for future innovation, particularly in several key areas including AI-enabled image enhancement and contactless, passive biometrics to limit healthcare workers’ contact with infected patients.

Impacting RCM

Within the revenue cycle, AI and automation are increasing revenue capture and helping early adopters achieve revenue integrity. For example, AI provides insights to identify predictions around payments and denials by identifying and assessing attributes commonly found in at-risk claims prior to submission.

Another example is payment process automation within the electronic health record (EHR) system, which saves time and helps with human errors. It also aids in document understanding by classifying and extracting information from images and PDFs. Other areas of RCM that are prime for automation include retrieving and updating current claims status, inventory allocation, and charge entry.

In recent years, healthcare organizations have accelerated adoption of AI for RCM in response to the chronic and worsening skills shortage. Doing so also reduces administrative costs, which account for about 25% of total healthcare spend, or about $2,500 per patient. Priority targets are revenue-generating functions like verifying a patient’s insurance eligibility, identifying the right medical codes based on the services provided, submitting claims to insurers, and following up with patients on outstanding bills. These, along with repetitive back-office tasks like onboarding and document digitization can be automated and further enhanced by advances in AI and natural language processing (NLP).

An Ongoing Transformation

AI and automation have already had a significant impact on healthcare and RCM. Expectations are high that they will continue to do so as technology evolves, and broader adoption enhances the body of knowledge around maximizing its application.

The key is to cut through the hype and understand their differing roles so appropriate AI/automation strategies can be designed and implemented based on individual organizational needs.