Robotic Process Automation (RPA) in Healthcare – Current Use-Cases

Niccolo Mejia

Niccolo Mejia covers AI applications across industries at Emerj. He holds a bachelor's degree in Writing, Literature, and Publishing from Emerson College.

Robotic Process Automation (RPA) in Healthcare - Current Use-Cases

There are many possibilities for automation in the healthcare industry outside of AI. Robotic process automation (RPA) technology can serve healthcare companies with various use cases involving data transfer and clinical documentation. Moving important information from the business’ frontend to their deeper business processes is among the most common use cases for RPA in healthcare, and many other solutions emerge from this idea.

Use cases for RPA in healthcare involve moving data between the area where users input and interact with the data and the databases where the information is typically stored. These can benefit healthcare companies by making clinical documentation more accessible, freeing up frontend employees by offering self service and improving payments.

  • Clinical Data Extraction: How RPA could help existing systems extract clinical documents and patient data and route it to the human employees that need to use it
  • Self Service Terminals for Hospitals: An RPA tool could be implemented to turn an employee-driven process into a self-service station that patients or customers can use on their own.
  • Processing Healthcare Credentials and Payroll: Employee credentials, timecards, and other payment information may be processed more efficiently when RPA is used in conjunction with other technologies.

We begin our exploration of the possibilities for RPA in healthcare with the process of clinical data extraction, and how RPA may transfer that extracted data more efficiently than some other systems.

Clinical Data Extraction

RPA tools may help healthcare companies retrieve data from both digital and physical clinical documents. They can automate the process of searching through a database for the correct documents and routing them to the appropriate user within the healthcare company’s network. This type of software usually needs a human employee to supply it with login credentials so that it can access that network or an EMR system.

An employee would need to feed the RPA software a list of patient names along with other credentials the ensure it selects the correct documents. Then, the software could extract the documents of the selected patients and send them to their intended users or carry out some other process with them such as printing.

Healthcare companies may find that even with an AI solution for this type of data extraction, they can still improve and build upon their processes to improve efficiency in ways AI may not be suited for. Natural language processing (NLP) solutions may be useful for digitizing clinical documents and identifying them based on the data within. However, RPA may be able to recognize and transfer them faster and based on fewer credentials.

Data extraction challenges that healthcare companies may still face while utilizing an AI solution include:

  • Automation of further steps of data retrieval such as printing or forwarding the documents for employee uses
  • Speeding up detection times without sacrificing accuracy
  • Utilizing metadata so that the documents will not need to be opened to verify them

RPA software can be trained to detect the metadata of a document, such as the filenames of scanned PDFs, their date created, or specific ID numbers for EMR documents. This type of information is typically found by opening the properties of a file or folder. RPA software can mimic this action once it has tracked a human employee doing it numerous times. This type of automation involves processes typically associated with the desktop and all of its functions.

The RPA vendor UiPath refers to this as “desktop automation,” in order to more clearly categorize it within their products. However the term does work to describe a large about of RPA solutions. UiPath lists an introductory demonstration of it on their youtube channel, and we have provided it below with a list of important sections:

  • 0:00: Starting a new automation workflow
  • 2:28: reviewing and editing the workflow in the RPA application
  • 4:15: How each action is prioritized within the workflow, and how to customize this to make the application complete a more specific task.

UiPath’s example shows how an RPA tool may be used to automate basic desktop functions, similar to right clicking files and folders to view their properties. We spoke to German Sanchis-Trilles, CEO and Co-founder of Sciling Information Technology and Services about the types of tasks RPA can automate. When we asked about what RPA really is and for some examples, Sanchis-Trilles said:

You have one person sitting in front of the computer and copying some data from the CRM to the ERP [for instance], that’s something that RPA can automate. You can save loads of time and effort, and [prevent] loads of errors as well. So RPA is the automation of some [mechanical] processes in an intelligent way.

Sanchis-Trilles sees RPA as a means of automation for tasks that do not require a thorough understanding of the information being used or transferred. It follows that it could be used to further automate processes around an AI solution, such as how UiPath documents their success with their client Health Fidelity.

Health Fidelity is an AI vendor for healthcare solutions that helps their clients with healthcare risk adjustment using NLP to detect patient clinical data and charts. This helps organizations participating in the Affordable Care Act (ACA) more accurately adjust their customers’ payments according to their risk. According to the success story, UiPath helped health fidelity complete any patient records that were missing data in a less time-consuming way.

The company’s technology extracts information from documents such as electronic medical records (EMRs), medical charts, consultation notes, and discharge summaries. A human employee was required every time any of these documents had incomplete information, because they needed to manually search for it and enter it back into that document.  

The success story states that since they integrated UiPath’s RPA tools into their software, their NLP solutions have been able to extract 1.2 million patient records. Additionally, UiPath claims Health Fidelity doubled their customer base without needing to add new software developers to their team.

Self Service Terminals for Hospitals

Hospitals, especially their emergency rooms, have an opportunity for RPA in automating patient check-in. They can set up self-service kiosks with screens that allow the patient to type in their information. There may also be a scanner on the kiosk to take a picture of the patient’s insurance card and ID if necessary.

RPA software can then automate the kiosk’s responses to the patient’s input, which would load their information into a triage system and allow the front desk employees to review each patient’s information.

In order to prepare for this type of change to hospital infrastructure, healthcare companies should consider the following factors that may lead to a more smooth integration:

  • A user-friendly kiosk dashboard software with a clear user interface and easy to use buttons
  • Integration of this dashboard into the healthcare company’s database and any other systems related to patient onboarding
  • Finding the right vendors or contractors to build the physical kiosks and supply the correct RPA software

The user interface of a self-service kiosk is important to ensuring a positive user experience and avoiding friction during check in. This means the healthcare company will likely need to provide a strong third party dashboard from a company that specializes in customer service terminals such as ATMs. Once this is in place, however, the kiosk will need to be integrated into the healthcare company’s network in order to be able to verify and transfer patient information.

Patient data can be sent to human employees in triage so that they may determine which patients are in the most danger and prioritize finding them care over others. This helps the employees decide between less serious issues as they come in, though they would still need to be present in the check-in room in case of an emergency that required immediate attention. This would be logged later once the patient has been sent to receive care, and other patients can continue self check-in concurrently.

This type of integration has the potential to automate numerous jobs across any hospitals that choose to adopt it, though it may take longer for companies to fully automate triage in a legal manner. We had a chance to speak to Martin Ford, futurist, keynote speaker, and the author of The New York Times Bestselling Rise of the Robots: Technology and the Threat of a Jobless Future.

In our interview, we asked Ford about which industries he thought could be most affected by automation in terms of employment. While blue-collar work may see considerable amounts of automation across all departments, Ford added,

Some of the people I think are going to be most dramatically affected by this is your general white collar workers. The people that sit in front of computers that do relatively routine and formulaic things again and again. … [They’re] very likely to be vulnerable to this. … It’s really everywhere. It’s factories, and warehouses, it’s going to be fast food … and it’s also a lot of more skilled jobs that require lots of education.

White collar jobs such as hospital triage and check in are examples of this level of automation within healthcare. Using RPA, healthcare companies can automate the check-in process for patients by using self-service terminals integrated into the facility’s database.

An RPA software would intake this as simple data receiving and entry tasks that it could repeat over and over for each patient. The patient’s ailments and conditions could also contribute to how much of a priority they are, which would allow the RPA software to automatically organize the list of patients by priority as they entered the system.

Blue Prism is an example of an RPA vendor selling self-service and self-check-in solutions to hospitals. They claim their RPA platform is focused on filling the gaps in jobs that people can do, but they would be better suited to a job that takes more judgment and critical thought. Their technology automates tasks that would be purportedly better left to a “robot” and is made to serve human employees to decrease friction in their tasks.

Blue Prism lists the following video on their youtube channel, and it features CEO and co-founder Dave Moss describing how their technology platform works. He describes his RPA “robots” as a “virtual workforce,” or a set of robots designed to automate as much of the department as possible. Then, human employees would fill in the gaps of processes that cannot be automated.

Moss also highlights two concepts of Blue Prism’s integration projects. He calls them “the operation,” and “IT,” and explains how each side needs to be prepared to integrate RPA with minimal friction. He explains this starting at 2:31:

 

Healthcare companies and hospitals may want to make sure that their user interface, kiosk design, and backend integration are as efficient as possible before adopting an RPA solution. This is because if the system is more efficient before the RPA software is trained, it will be able to carry out the intended tasks more quickly.

Blue Prism lists a case study in which they claim to have helped University Hospitals Birmingham (UHB) install self service kiosks in their facilities. UHB wanted to provide their patients with a user friendly and intuitive self-check in process, as well as integrate their system into the National program Patient Administration System (PAS). This would purportedly allow them to access a useful database without needing to update cybersecurity.

The case study states that UHB saw 50% improvement in front desk staff efficiency, as well as 51% of their patients registering at the self check-in kiosk. Blue Prism also claims UHB’s patient flow became twice as fast as it had been before adopting their solution.

Processing Healthcare Credentials and Payroll

Another possibility for RPA in healthcare is the automation of processing healthcare employees’ credentials and payroll information such as timecards and other documentation of hours worked. RPA software can automate the transfer of data from the customer-facing method of information entry, such as an employee mobile app, to the backend verification method.

For example, consider if a healthcare company had a mobile app where they could photograph their time card or transcribe the information digitally and send it to the payroll department. RPA could automate the process of transferring that picture or text to the healthcare company’s payment processing system to be paid or rejected.

Alternatively, RPA could automate the transfer of data for personnel verification based on security or network credentials assigned to each employee. An employee would simply need to enter their employee credentials, such as an ID number and their full name, and the RPA software would route that data to the company’s verification method.

However, healthcare companies will need to prepare their business for RPA integration on this level. Deloitte has published an article detailing the possible benefits of RPA in healthcare along with the challenges companies may face during implementation. They describe how it may affect employment at companies that implement it, and the more important tasks human employees can take on with lesser processes automated for them.

Regarding the challenges and benefits of implementing an RPA solution within a healthcare company, Deloitte states:

While it may be challenging to make the case for the budget needed to build or implement RPA tools, health care and life sciences IT leaders can point to the strategic and competitive advantages and ROI this technology may confer. When organizations implement automation tools, they typically reassess and redesign whole processes to make them more efficient and less costly. Automation tools may also help these organizations to accomplish more work with fewer people—and rather than spending thousands of hours obtaining and sifting through raw data, skilled employees can focus on using RPA-curated information to form insights and make strategic decisions that better support patient safety and care, and related research.

Deloitte claims a successful RPA integration could lead to accomplishing more tasks with less people. The remaining employees could focus on healthcare-related tasks such as patient care, the previously stated hospital triage, and research in clinical facilities. However, RPA solutions for identity verification and payroll may come with their own set of challenges.

We identify two key factors to considering how to properly implement an RPA system for these types of healthcare business processes:

  • Ensuring the company has a reliable backend processing tool for the data routed to it
  • Determining if the customer or employee-facing processes will be affected by this implementation, and how to mitigate any possible friction

Before implementing an RPA solution, it may be important for healthcare companies to review whichever processing tool or software they may be using to verify or otherwise work with incoming data. For some this could be an AI payment processing solution or an NLP application for reading employee credentials.

Healthcare companies may want to work to make these processes as efficient as they can before employing an RPA tool so that is it trained to work with the most recently updated version of company operations.

The user-facing part of these processes usually involves entering information into the system to be routed into the network’s processing software. An RPA solution may require some work between the vendor and the healthcare company to ensure the implementation will not affect how a patient or employee uses the system.

If the system does need to be reworked in order to operate more efficiently with RPA, healthcare companies may want to consider how those changes can be mitigated. This would prevent any drop in user satisfaction and reduce potential friction.

Kofax is a vendor offering RPA payroll entry solutions to healthcare companies. Their RPA platform is made primarily to transfer data between various sources and user endpoints. The company also offers other software tools which analyze and help healthcare companies visualize this data.

Below is a demonstration video from Kofax about how their RPA software automates tasks and documents for their clients. The demonstration begins at 1:24, and it shows how their RPA software can build workflows to process business documents:

This video shows how a human employee needs to prepare the RPA software to recognize specific fields within each document and transfer the data in each one. In this case, the RPA software was working with an optical character recognition (OCR) solution. This type of AI solution likely helped to recognize the letters and numbers within the text and allowed the RPA system to accurately transfer them.

 

Header Image Credit: Health Systems Management

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the ‘AI Advantage’ newsletter:

Subscribe
subscribe-image
Stay Ahead of the Machine Learning Curve

At Emerj, we have the largest audience of AI-focused business readers online - join other industry leaders and receive our latest AI research, trends analysis, and interviews sent to your inbox weekly.

Thanks for subscribing to the Emerj "AI Advantage" newsletter, check your email inbox for confirmation.