As hospitals look toward AI adoption, RCM’s ability to attribute value makes it an ideal initial testing ground.
Hospital executives are increasingly turning their attention to generative artificial intelligence (AI) given its potential to transform healthcare. However, with low hospital margins and sometimes empty vendor promises, many leaders worry about investing in “hype” and thus feel compelled to take a wait-and-see approach to new technologies.
The solution is to use revenue cycle management (RCM) as a testing ground for AI – and to help hospital leaders learn about what it takes to implement AI successfully. RCM is an ideal space because successful implementations have a more direct tie to cost or revenue and processes do not involve convincing clinicians to change workflows.
AI Adoption: A Long Road Ahead
The year is 2007. “Cloud computing” has been launched with Amazon Web Services. Google is so popular that the term is added as a verb in most dictionaries. The iPhone hits the market to mass acclaim.
The same year, 2% of hospitals surveyed report having a fully functional Electronic Health Record (EHR).
Why? Training employees and redesigning clinical workflows to work within the EHR required immense capital and operational investment. And — at least prior to the HITECH act and meaningful use payments — its return on clinical operations was hard to measure.
AI presents a similar challenge. Although using AI to improve medical decision-making and reduce administrative burden has the potential to make healthcare dramatically more efficient, understanding how to incorporate it into clinical workflows without devaluing human expertise remains a challenge.
AI's Current and Future Role in RCM
Instead of clinical workflows, many hospitals have focused on how AI can play a role in RCM, which has the benefit of being distanced from patient care and is unlikely to cause adverse health outcomes. So far, AI has been primarily incorporated into relatively rote tasks (like its use within Robotic Processing Automation), or in relatively narrow domains (like Natural Language Processing for Coding), but its potential for future applications is vast.
One such application is using AI as a “forensic accounting" tool to ensure that all Clinical Documentation Improvement and Coding opportunities have been captured:
- Clinical AI can uniquely take into account tens of thousands of data points per chart and flag potential opportunities for human review. This synergistic relationship ensures that AI acts as a support tool, leveraging its computational speed and consistency to handle large volumes of data, which in turn allows human reviewers to focus on more complex decision-making and verification tasks.
- The use of AI in this context comes with full attribution, meaning that all suggestions made by the AI system are fully traceable and can be audited. When used as a Prebill tool, it comes after existing CDI and coding processes, which ensures that any recommendations are net new and the impact to the bottom line can be clearly defined.
- The financial and operational risks associated with implementing AI in the Prebill process are limited. This is because the AI tool comes after existing coding and documentation frameworks, reducing the need for significant operational changes. Furthermore, the financial upside is realized immediately (net new revenue) rather than requiring staff attrition.
Taken together, the data suggests that the average hospital can recover around $2 million per 10,000 discharges and improve quality metrics through more accurate representation of the care delivered to the patient.
Harnessing the Potential
As the healthcare industry stands on the precipice of this AI revolution, it is critical to approach this transition with a blend of enthusiasm and strategic caution. Integrating AI into healthcare RCM allows for safe, incremental adoption with clear, incremental value. Hospitals can navigate this transformation with a clear vision to fully harness AI's transformative potential.