The promise of AI in healthcare

December 6, 2019

How artificial intelligence can boost surgery center operations and revenue cycle management.

The recent announcement that the Mayo Clinic and Google are partnering to use advanced cloud computing, data analytics, machine learning and artificial intelligence (AI) to accelerate medical research and treatment generated a lot of enthusiastic healthcare industry buzz. There’s no doubt that digital transformation–a key component of the Fourth Industrial Revolution’s convergence of the physical, digital, and biological worlds–will yield great benefit in patient care.

Though it may be less buzz-worthy, machine-based innovation within heath IT promises similarly dramatic change on the business side of healthcare-and the ambulatory surgery center (ASC) industry is well positioned to be one of the first beneficiaries.

ASCs are one of the fastest growing segments of the healthcare system. According to a report by the Advisory Board, by 2020, 60% of outpatient surgeries are expected to be performed in the ASC setting (compared with 40% in 2005). The number of surgeries in the ASC setting has skyrocketed in the past several years, with a 23% increase in 2017. Looking ahead, the ASC market is expected to increase to a roughly $40 billion industry.

A unique set of factors catapults ASCs to the forefront of the next wave of health system IT transformation.

First, the industry’s health information technology (HIT) hasn’t been forced to evolve at the same rate as other providers and payers. ASCs were exempted over the past roughly 10 years from interoperability requirements and standardization under the government’s Meaningful Use (MU) program. Thus, most ASC software systems operate independently without a “common language” for data input and output; collectively, they also lack interoperability within the larger health information exchange (HIE) between provider, hospital, ASC, and patient.

Additionally, because ASC patient volume and the number of approved procedures have increased substantially over the last decade, insurance claims have increased as well, requiring a dedicated effort to expedite processing and minimize delays and denials. As the Centers for Medicare and Medicaid Services (CMS) updates payment rules to include more outpatient surgical procedures, HIT must have the ability to absorb additional codes, improve claims processing and ensure consistency in the revenue cycle.

As ASC executives consider decisions around their HIT resources and investment, they should consider how AI will potentially impact and benefit their business operations-both now and in the long run.

Already, digital robots (a small piece of software programmed to do specific task on a computer) are helping move medical records to cloud-based storage with the goal of being accessible to clinical fingertips. While current medical records are primarily kept in electronic health records (EHRs), those records are usually stored on servers and incomplete, often because they are incompatible with other data sources, pre-date newer EHRs, or are paper chart-based. Paper charts must be scanned and saved as digital files, which doesn’t easily lend to data retrieval. Absent a way to translate medical histories into information that is easily accessible and actionable, some modern EHRs are simply an electronic version of a paper chart.  

To facilitate a clean transfer from a legacy system to a clinically actionable target system in ASCs (and after manual data extraction in the case of paper records), digital robots can supplant time-consuming administrative tasks normally delegated to staff. The robots, using machine learning, are programed to search and select relevant historical data such as allergies, medications, and previous medical procedures for capture and conversion, ensuring both data integrity and compliance with health record standards.

Within the near future, AI will be the key to facilitating data translation and exchange between conflicting software platforms. AI has the ability to use intuitive algorithms to translate disparate input into a common, interoperable language and output. Everything from ASC preauthorization and patient registration, to coding and clinician documentation, to un-adjudicated claims and resolution has the potential to be automated. This will, in turn, decrease the amount of staff labor time currently dedicated to tasks like scanning, data entry, and custom interfaces, and will help to minimize human error in the process.

As ASCs transition to new systems, software will proactively convert the required seven years of digital records to the new platform, eliminating the need for digital robots to handle these functions separately. Given the market dynamics in ASCs in the U.S., with large numbers of joint ventures, acquisitions and consolidations, interoperability between multiple platforms, as well as between payers and specialty medical practices, is vital to consistency and stability in the revenue cycle. This will become even more critical if MU requirements are expanded to ASCs, as they almost certainly will be in the next few years.

While these AI-driven business improvements benefit the operational and financial side of ASCs, these functions will also complement the clinical end. A host of apps and programs already use machine cognition and computing in screening and diagnosis, imaging analysis, operating room flow, recovery monitoring, improvements to the patient experience, and more. By investing in and improving AI capability and pursuing opportunities to automate functions that are prone to human error, ASCs can ensure they remain on the forefront of patient care while reducing the risks of costly administrative mistakes.

Tom Scott is Chief Financial Officer of HST Pathways, a top-ranked software solutions company for the ambulatory surgery center (ASC) industry.