Computational methods for decision support in healthcare aren’t new.1 In the ’70s and ’80s, there were efforts to diagnose conditions and predict outcomes using data.2 But analog records and low-power computing made achieving trustworthy decision support at scale impossible.
Fast forward to today: everything is digital. Ready access to vast data and low-cost, massive computing power ushered in the Artificial Intelligence (AI) revolution. AI, specifically machine learning (ML), ingests and interprets large amounts of relevant data to decide the best action(s) to achieve a given goal. This basic idea has wide-reaching implications for every facet of the healthcare system.
In a clinical setting, the predictive ability of ML is the heart of decision support, helping guide diagnostics and determine care pathways that lead to the best possible outcome. AI, in this context, aims to improve the quality of care and lower costs by reducing unnecessary treatments and making clinicians more productive and satisfied.
In the revenue cycle, the ability to predict the outcome of a claim and show the shortest path to resolution is game-changing. When you apply ML to claims, processing patterns emerge, problems are identified, and results are consistently predictable. With these insights, you can prioritize claims that will get paid faster and more dependably. Fix workflows to boost efficiency and accuracy for more timely billing decisions and both the health system and the patient benefit. In addition to decision support, automations keep repetitive, error-prone tasks from dragging on productivity by saving precious minutes per transaction. As the time to revenue accelerates, gross margins grow.
The health system has a productivity imperative as it faces growing margin pressure. Productivity in administrative functions has improved over the last twenty years. But the gains have been through labor rather than capabilities or process improvement.3 The labor market is tight. AI offers more than a stopgap for a labor shortage. It’s an opportunity to reallocate staff to high-value, engaging work.
AI levels the playing field too. It gives rev cycle management new power over payers’ ever-changing requirements that cause delays and denials. It supercharges daily common workflows, points to a strategy for retooling the revenue cycle, and lays the foundation for continuous improvement.
Effectively deploying automation and analytics alone could eliminate $200 billion to $360 billion of spending in US healthcare.4 US healthcare systems are at a critical juncture. New ways of working, enabled by technology, will be fundamental to righting the ship or continuing to succeed.5 Changing the mindsets of critical stakeholders with a compelling plan to build value through improved productivity will put revenue cycle leadership and their organizations on a path to long-term success.
1 Machine learning in clinical decision making. Lorenz Adlung, Yotam Cohen, Uria Mor, Eran Elinav, Machine learning in clinical decision making, Med, Volume 2, Issue 6, 2021, Pages 642-665, ISSN 2666-6340, https://doi.org/10.1016/j.medj.2021.04.006.
2 Computer aided diagnosis of acute abdominal pain: a multicentre study. Br Med J (Clin Res Ed) 1986; 293 doi: https://doi.org/10.1136/bmj.293.6550.800 (Published 27 September 1986) Br Med J (Clin Res Ed) 1986;293:800
3 The productivity imperative for healthcare delivery in the United States. Mckinsey & Company, February 27, 2019
4 Administrative Simplification and the Potential for Saving a Quarter-Trillion Dollars in Health Care. JAMA,October 20, 2021
5 Setting the revenue cycle up for success in automation and AI. Mckinsey & Company, July 25, 2023
