This article is the first in a three-part series on artificial intelligence (AI) in the health information management (HIM) revenue cycle field. Part 2 focuses on considerations for evaluating and planning for AI technology, and Part 3 will focus in-depth on AI use cases present in healthcare today.
Once upon a time, artificial intelligence (AI) was a thing of science fiction. Talking cars that navigate themselves and robot assistants that deliver our needs before we realize we need them were fanciful creations of Hollywood minds. Thirty years later, AI is commonplace and growing more and more pervasive. We use AI in our everyday lives, sometimes without even knowing it. In a recent Pega survey, 33 percent of respondents initially thought they used AI technology. However, when further probed, 77 percent of respondents actually use AI-powered services and/or devices.
From automated vacuum cleaners to self-driving cars and even manufacturing robots, AI technology is being implemented to enhance the timeliness, accuracy, and effectiveness of our human efforts—and it is here to stay.
But what is AI exactly, and how does it affect us in health information management (HIM) and the revenue cycle? Merriam-Webster defines AI as “the capability of a machine to imitate intelligent human behavior.” IBM breaks down the definition even further by explaining “artificial intelligence leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind.”
Studies show that while people are eager and optimistic about the benefits of AI, they are also often confused and uncertain about how it is used today. There are several examples of AI in many of our daily routines. Some examples of AI-enabled technology are the Roomba vacuum cleaner, smartphones with speech recognition in talk-to-text, autonomous driving vehicles, and other vehicle safety features such as self-parking and driving assistance lane departure warnings. Home assistant devices such as Google Assistant and Amazon Alexa also use AI technology to play music, order products and services, make phone calls, activate or deactivate other devices, and many other tasks—and the list is growing. While all of that is great, it still may leave some scratching their heads, saying, “But what is AI really, and how can it help HIM?”
According to a research study conducted by Change Healthcare, two-thirds of healthcare facilities and health systems are using AI to assist their revenue cycle. Of those, 72 percent of respondents are using AI applications for eligibility/benefits verification and 64 percent for payment estimations (likely due to the No Surprises Act). But AI plays a bigger role in revenue cycle management (RCM) than just these areas. From the 2022 State of Revenue Integrity survey published by the National Association of Healthcare Revenue Integrity (NAHRI), other areas with AI-enabled products and functions for RCM include charge description master (CDM) maintenance, charge capture, denials management, payer contract management, physician credentialing, and claim auditing, just to name a few. AI can help predict claim denials by analyzing past denial trends and alerting HIM professionals of the potential denial in advance of billing. This affords the opportunity to review and correct claims pre-bill. Using AI to create dashboards to track denials by type or payer can provide insight to help reduce recurring denials and improve workflows and education as needed. Self-pay patients can benefit from AI by integrating financial assistance technology that checks for any financial aid the patient might qualify for, as well as being used in price transparency.
Provider documentation can also benefit from the use of AI with not only speech recognition to dictate directly into their notes but also with integration into formatting documentation and placing specific dictation into specific areas of their notes. Having AI integrated with provider documentation reduces queries from coding professionals, which speeds up claims submissions and helps to keep account receivable (AR) days within benchmark standards.
Charge capture can be automated with rule-based technology and natural language processing (NLP, a branch of AI) to ensure documented charges are not overlooked prior to claim submission. AI-enabled technologies are also used to establish specific criteria to support a smoother initial claims submission process by establishing various bill hold days to allow providers time to document, device, and supply charge capture to occur, and coding personnel time to ensure documentation queries and record completeness are achieved for cases/accounts prior to claims submission.
Computer-assisted coding (CAC) uses NLP to highlight specific terms within documentation then apply an appropriate ICD-10-CM diagnosis, procedure codes, and CPT codes. In 2013, the Cleveland Clinic reported a 22 percent reduction in the time coding professionals spent coding accounts when they implemented a CAC program. Having the majority of codes already attached to accounts allowed coding professionals to review for accuracy instead of having to manually enter codes themselves or have providers entering them. Some AI-enabled technologies are also used for completely autonomous coding, eliminating the coding effort completely.
AI-enabled technologies can help improve quality and efficiencies where, without it, there would be a need for more people resources that healthcare is already hard pressed to find. Over the last few years, our industry has been faced with challenges recruiting and retaining skilled healthcare personnel. In the health information management profession, associate degree programs have been shuttering steadily across the nation over the past 10 years. We currently have all-time lows in unemployment in our country. In 2018, the US had an unemployment rate of 3.8 percent and currently has a rate of 3.5 percent. We are also facing a steady decline in the US birthrate. A study conducted by the University of Southern California concluded that the “imbalance between children and retirees is growing such that the economic burden on a child born in 2015 will be nearly twice that of a child born in 1985.”
To put it plainly, the rate of births in the US is not substantial enough to replenish our current workforce and had been on a consistent annual decline since 2007. And this isn’t just a concern in the US; this is also a trend globally. The global birthrate is expected to fall below replacement level sometime between 2050 and 2100. One solution to this is to develop and adopt more AI technologies to fill the gap in our labor needs. The concern that computers will take over our jobs can cause fear. The solutions are here today to support us, not replace us, and that will be the continued trend if we take the lead and embrace the evolution in our field.
As an industry, we need to educate ourselves and our healthcare colleagues on the definition, benefits, and use cases for AI. We need to ensure staff comfort with the emerging technology and reinforce that it is not a replacement to our knowledge, skill set, and expertise but to succeed in our ever-changing environment of HIM.
Tami Montroy is a director of central fee abstraction.
Glenda Rakes is an HIM director/privacy officer.
Learn more about CDI and AI in the white paper “CDI: Compliant Technology Adoption and the Role of Clinical Documentation Specialists.”
By Tami Montroy, MS, RHIA, CCS, and Glenda Rakes, MHIIM, RHIA, CHPS
Take the CE Quiz