Artificial intelligence is in full swing across the healthcare industry, and the trend shows no sign of slowing down, says Brian Robertson, co-founder, CEO, VisiQuate, a pioneer of intelligent process automation in healthcare financial management for practices, hospitals and health systems. Providers of all sizes during the pandemic have been leaning on A.I., and some are starting to use A.I. chatbots (similar to Alexa or Siri) to help their medical groups optimize key revenue cycle management processes to improve efficiencies and reduce costs.
Instead of spending hours finding data, aggregating it manually, and building reports and dashboards, users of the technology can now just tell their chatbot what they need, and there it is. In the following conversation with Daniel Williams, Sr. Editor and host of the MGMA Insights podcast, Robertson explores how A.I.-driven chatbots are injecting much-needed change to revenue cycle processes.
Daniel Williams: Healthcare IT, like the rest of the industry, experienced tremendous change last year. Tell us about VisiQuate and your primary focus during that time?
Brian Robertson: First, we are all about analytics. We help medical practices and hospitals become data-driven organizations at a granular level to optimize growth and costs at every stage of the revenue cycle and predict where they are headed. Our analytics platform brings eligibility and claims data to a central location and offers insights on top problems. We present the data through a dashboard that allows users to focus on their most essential tasks quickly. The pandemic obviously had a significant impact on providers. We are grateful that we were able to structure our business model in a way that has allowed us to help them make a seamless transition to virtual workspaces. Our team was able to help provider organizations’ revenue cycle staff – big and small – move to home-based offices quickly and set them up with a daily workflow solution through our SaaS cloud platform and solutions.
Williams: What are the most significant claims challenges practices currently face?
Robertson: Addressing payer denials is one of their most critical issues. As organizations work appeals, they quickly find themselves using several technology systems to get one claim exception resolved. They need to be able to see all the data in one location to understand the size of the problem and spot anomalies quickly.
Williams: How can A.I. chatbots help group practices optimize revenue cycle performance?
Robertson: Like Amazon’s Alexa or Apple’s Siri, healthcare chatbots can perform multiple revenue cycle functions related to billing and claims, including helping revenue cycle leaders and staff perform tasks more efficiently by looking for patterns and problems. Revenue cycle teams can interact with A.I. chatbots to do everything from monitoring cash flow during the day to something as simple as identifying the organization’s “Top 10 Medicare Denials” for the month. The technology can even retrieve that information while you're driving to a meeting or taking calls.
Chatbots are also designed to think critically and will take the question one step further to provide deeper insight. For example, if you ask, “What are the top 10 denials for Aetna,” a chatbot will present that information through a dashboard and will also ask, “Is it helpful if I attach all related Aetna appeal forms?” Over time, as chatbots help to resolve each claim, they become smarter. They can determine how a problem was resolved and fix problems at scale, such as taking costs out of the system.
Williams: How does A.I. think like a human, and where is chatbot technology headed in revenue cycle management?
Robertson: Creating a chatbot that thinks like a human begins with an A.I. intent model, which is like an operating system. For example, a chatbot designed to optimize revenue cycle processes might specifically target billing and claims problems that humans attempt to solve. In this case, the A.I. intent model would start by teaching the computer about job role basics and vocabulary such as “payments” and “claims.” For example, the chatbot can be taught that a payment is good and that a denial is bad. Chatbot technology then uses natural language processing and replicates the part of the human brain that can solve binary problems and learn correlation. Today, chatbots can look at events and replicate tasks. For example, the chatbot can be trained to know that a denial is bad and that it can fill out a form to appeal it. It can eventually learn to ask, “Did I get paid the full amount?” As chatbot technology advances, it can figure out ways to see and solve more of these patterns upstream in the revenue cycle.
Williams: How do A.I. and chatbots help specialty practices resolve inefficiencies?
Robertson: A chatbot’s intent model can be trained to understand the nuances around individual specialties. It can provide a rundown of key metrics, such as census volume and daily cash flow, for each group through a practice management dashboard. If a medical group specializes in orthopedics, the A.I. intent model will be specific to that domain and learn about its universe of procedures and diagnosis coding, along with typical problems.
Williams: How can medical groups maximize the intersection of A.I. and human ability in revenue cycle work?
Robertson: A.I. chatbots and revenue cycle staff can work in concert quite nicely, because they bring different necessary attributes to ensure a smooth billing process. At the end of the day, a chatbot is still a piece of software. The technology can think very fast and crunch a lot of data on many different use cases, but it doesn’t have human judgment or intuition. When it comes to claims processing work, a chatbot, in some cases, can help an individual triple the number of claims processed while automating repetitive tasks in their workflow. This works particularly well in an employee gainshare model in which staff is rewarded for speed, accuracy and results, like in a practice setting.
Instead of spending hours finding data, aggregating it manually, and building reports and dashboards, users of the technology can now just tell their chatbot what they need, and there it is. In the following conversation with Daniel Williams, Sr. Editor and host of the MGMA Insights podcast, Robertson explores how A.I.-driven chatbots are injecting much-needed change to revenue cycle processes.
Daniel Williams: Healthcare IT, like the rest of the industry, experienced tremendous change last year. Tell us about VisiQuate and your primary focus during that time?
Brian Robertson: First, we are all about analytics. We help medical practices and hospitals become data-driven organizations at a granular level to optimize growth and costs at every stage of the revenue cycle and predict where they are headed. Our analytics platform brings eligibility and claims data to a central location and offers insights on top problems. We present the data through a dashboard that allows users to focus on their most essential tasks quickly. The pandemic obviously had a significant impact on providers. We are grateful that we were able to structure our business model in a way that has allowed us to help them make a seamless transition to virtual workspaces. Our team was able to help provider organizations’ revenue cycle staff – big and small – move to home-based offices quickly and set them up with a daily workflow solution through our SaaS cloud platform and solutions.
Williams: What are the most significant claims challenges practices currently face?
Robertson: Addressing payer denials is one of their most critical issues. As organizations work appeals, they quickly find themselves using several technology systems to get one claim exception resolved. They need to be able to see all the data in one location to understand the size of the problem and spot anomalies quickly.
Williams: How can A.I. chatbots help group practices optimize revenue cycle performance?
Robertson: Like Amazon’s Alexa or Apple’s Siri, healthcare chatbots can perform multiple revenue cycle functions related to billing and claims, including helping revenue cycle leaders and staff perform tasks more efficiently by looking for patterns and problems. Revenue cycle teams can interact with A.I. chatbots to do everything from monitoring cash flow during the day to something as simple as identifying the organization’s “Top 10 Medicare Denials” for the month. The technology can even retrieve that information while you're driving to a meeting or taking calls.
Chatbots are also designed to think critically and will take the question one step further to provide deeper insight. For example, if you ask, “What are the top 10 denials for Aetna,” a chatbot will present that information through a dashboard and will also ask, “Is it helpful if I attach all related Aetna appeal forms?” Over time, as chatbots help to resolve each claim, they become smarter. They can determine how a problem was resolved and fix problems at scale, such as taking costs out of the system.
Additional resources:
- Virtual assistants, chatbots and digital humans: Health practices discover surprising advantages in conversational A.I.
- Re-"AI"-magining patient access now and for the future
- Insights: Using analytics and A.I. to automate and streamline your revenue cycle
Williams: How does A.I. think like a human, and where is chatbot technology headed in revenue cycle management?
Robertson: Creating a chatbot that thinks like a human begins with an A.I. intent model, which is like an operating system. For example, a chatbot designed to optimize revenue cycle processes might specifically target billing and claims problems that humans attempt to solve. In this case, the A.I. intent model would start by teaching the computer about job role basics and vocabulary such as “payments” and “claims.” For example, the chatbot can be taught that a payment is good and that a denial is bad. Chatbot technology then uses natural language processing and replicates the part of the human brain that can solve binary problems and learn correlation. Today, chatbots can look at events and replicate tasks. For example, the chatbot can be trained to know that a denial is bad and that it can fill out a form to appeal it. It can eventually learn to ask, “Did I get paid the full amount?” As chatbot technology advances, it can figure out ways to see and solve more of these patterns upstream in the revenue cycle.
Williams: How do A.I. and chatbots help specialty practices resolve inefficiencies?
Robertson: A chatbot’s intent model can be trained to understand the nuances around individual specialties. It can provide a rundown of key metrics, such as census volume and daily cash flow, for each group through a practice management dashboard. If a medical group specializes in orthopedics, the A.I. intent model will be specific to that domain and learn about its universe of procedures and diagnosis coding, along with typical problems.
Williams: How can medical groups maximize the intersection of A.I. and human ability in revenue cycle work?
Robertson: A.I. chatbots and revenue cycle staff can work in concert quite nicely, because they bring different necessary attributes to ensure a smooth billing process. At the end of the day, a chatbot is still a piece of software. The technology can think very fast and crunch a lot of data on many different use cases, but it doesn’t have human judgment or intuition. When it comes to claims processing work, a chatbot, in some cases, can help an individual triple the number of claims processed while automating repetitive tasks in their workflow. This works particularly well in an employee gainshare model in which staff is rewarded for speed, accuracy and results, like in a practice setting.
Explore Related Content
More Insight Articles
