The Challenges and Opportunities in Healthcare Administration System Automation

In the US, healthcare’s administrative costs are the highest of all advanced economies, as a share of GDP. They account for up to a third of total healthcare spending in the country, a proportion that’s far greater than the amount necessary to deliver effective care. These costs come primarily from claims processing and payment, prior authorization and eligibility determinations, and quality measurement, among other areas.

A 2020 policy proposal from Harvard University’s David Cutler identified three key ways to reduce healthcare’s administrative spend: establishing an automated clearing house for billing; harmonizing quality reporting; and enhancing data interoperability.

Modest improvements in these areas can save an estimated $50 billion annually. Automation can potentially support all of those improvements and with strategic investments, reap ongoing savings.

Automation is a real solution, but it has the trappings of a buzzword. What does it mean, really? Simply put, automation means taking rote and repetitive tasks and making them automatic, so that the extra time saved in the process can then be funnelled elsewhere. But there are numerous flavors of automation, ranging from intelligent automation (which involves the use of AI) to pseudo automation (which turns pen-and-paper tasks into digital ones). And, despite the near universal hype, not all automation leads to unequivocal benefits.

Consider the clinical example of imaging automation, wherein algorithms analyze radiology images with greater speed and accuracy than humans. By automating image analysis, medical facilities can save time and other resources while achieving better patient outcomes. While it’s frequently touted as a primary example of the sort of automation coming to healthcare, it’s not perfect for every facility.

Most image analysis applications are still in the developmental stage and must be used in conjunction with humans, therefore cancelling out some operational benefits. The tech is also not yet widely available, and doesn’t always have a seamless integration with existing systems. Finally, the tech needs to be reliably proven before it’s fully deployed.

The above example, while on the clinical side of healthcare, can be used as a litmus test for the automation of healthcare administration systems. Solutions need to create less work, not more. Adequate training, staff buy-in, and the unique needs of a particular healthcare facility all need to be taken into account. Change for change’s sake leads to more waste, but the right solution, for the right facility, for the right staff, at the right time, can be revolutionary.

One of the simplest and most effective tools for reducing administrative waste is in scheduling automation. Streamlining and digitizing scheduling makes a simple but otherwise burdensome process simpler for the patient, provider, and administrator. It also makes healthcare facilities more efficient and more competitive and can be utilized at any scale, from small clinics to major hospitals. Having patients schedule and reschedule via an app increases adherence with pre-set reminders and easily adjustable appointment times. It also frees up front-end staff.

While it’s a small automation, it can have big results: thin operating margins mean that eliminating scheduling inefficiencies can save up to $140 billion a year. At the same time, it’s a minimally intrusive solution that doesn’t disrupt other processes or require significant training and observation of staff.

Inefficiencies and errors in healthcare supply chain management cost an estimated $5 billion a year. Primary culprits of this waste include the overlapping roles and responsibilities of key stakeholders and the manual processes that currently exist within the system.

While medical supply chains are harder to predict than traditional supply chains, automation can be tailored to its specificities. End-to-end automation, in this use case, makes each step in the supply chain more efficient: nurses can use a barcode scanner instead of manually entering information; orders of new supplies becomes automatic and fit-to-need; duplicates and errors are caught by algorithm instead of third-party observer; and all data is captured and stored automatically within a single, auditable system.

In healthcare administration, supply chain automation involves a major restructuring of a healthcare facility’s process infrastructure, but it’s not an impossible endeavor when taken on gradually. Some leading healthcare organizations, such as University of California San Francisco Health, are approaching 90 percent automation in their supply chains. More are expected to follow suit.

According to a 2009 study in Health Affairs, for every ten physicians providing care, seven additional people are involved in billing-related activities. But today, through robotic process automation (RPA), automated medical billing can perform charge entry, claim scrubbing, and remittance through timers and natural language processing.

When coupled with AI and machine learning, it’s possible to automate the process further, extracting data from medical documents for proper coding and billing. This increases efficiency in both time and revenue: a healthcare facility doesn’t have to wait as long for payment if invoices are submitted promptly, and automating ICD-10 codes reduces the number of billing errors. Billing automation can also be linked to insurance companies, and expedite the processing of copays and deductible fees.

The upfront costs of deploying an automated billing system are high, both in finances and in complexity. But administrators at larger medical facilities may see them as a pittance compared to the benefits: billing errors accounted for over $28 billion in improper Medicare payments in 2019.

The amount of data generated by healthcare is growing faster than any other industry. Telemedicine, medical devices, personal health trackers, and other IoT-enabled tech mean the flow is set to increase further. Automated data management exists in a positive loop wherein it further enables automation in other areas, while benefiting from further automation in those other areas at the same time.

Today’s innovations in automated data management involve the transformation of individual data silos into a shared data lake. Data is translated into an interoperable format wherein it can be used across a wide variety of inputs, and machine learning can draw insights from the data that’s stored. Automated data management can also increase regulatory compliance by funnelling information through a series of if-then protocols.

Automated data management is an on-going, iterative process and practically all healthcare administrators are at one point or another in the journey: it begins with the digitization of pen-and-paper records, and its endpoint lies far in the future, with fully interoperable data that feeds on itself.

Automation is a transformational process in healthcare administration and the costs can be high. But it will also empower the industry to reduce its burdensome administrative costs, and focus resources on where they belong: improving patient outcomes.