A new survey has found that despite setbacks created by the COVID-19 pandemic, AI use is common in healthcare organizations, and could be all but universal within five years.
The survey, which was conducted by Black Book Research, connected with 540 HIM executives to determine their view on the use of AI in healthcare organizations. The research looked closely at revenue cycle management issues and how they were being addressed (or could be addressed) by AI tools.
According to the survey, 47% of respondents were using AI in one form or another, and 90% expect to see widespread implementation of AI solutions within the next five years. This was the case despite the huge operational and financial strains imposed by the COVID-19 pandemic.
Almost all (95%) of hospital respondents felt confident that AI technology can streamline document creation and enable clinicians to capture a holistic patient history to not only boost outcomes but also build revenue integrity.
One of the major problems they hope to see AI address was problems with revenue realization. Sixty-eight percent of respondents have committed funds for AI coding improvements in FY 2021, and their plans should tackle revenue cycle management issues which have long been challenging for health systems.
On the one hand, researchers found that health system coder productivity has climbed 68% in hospitals over 250 in Q3 2020 as compared with Q3 2017. At the same time, though, the survey noted that both commercial and public payers are now denying about eight percent of submitted claims, which costs health systems nearly 2.5% of net patient revenue.
To address this problem 94% of providers reported being eager to implement more sophisticated clinical documentation improvement tools using AI to improve the speed, accuracy and efficiency of coders.
The HIM execs reported building out new platforms embedding AI-powered clinical intelligence in both clinician and CDI workflows, analyzing notes and clinical data to find gaps and deficiencies in the documentation before notes are saved to the EHR.
Applying AI technology to RCM problems seems to have a substantial effect. In fact, 97% of hospitals saw documented quality improvement and increases in case mix indexes within six months of AI-enhanced CDI implementation.
Specifically, the average case mix overall improvement in the 145 hospitals surveyed with 150 to 450 beds using AI-enhanced CDI technology added $2.3 million in revenue, a 27% increase between Q3 2019 and Q3 2020.
By the way, an interesting side note to this is that healthcare organizations are now having to look at ways of getting telehealth claims processed.
As a recent piece by Khalid Al-Maskari of Health Information Management Systems notes, to optimize reimbursement for telehealth, providers need enough information on patients to determine which need the most hands-on care.
This seems like just one more area in which AI might be useful in managing the RCM process more effectively. After all, one thing AI is particularly good at doing is recognizing trends that might not be readily apparent to human observers. Perhaps it will be able to help providers tackle these emerging reimbursement issues as well as problems arising from traditional encounters.