Artificial Intelligence (AI) is Not Ready for Prime Time in Healthcare, Yet

Sometimes we read what others are saying and, because we want it to be true, we imagine it is true. It’s important to break down false promises and inaccurate information about Artificial Intelligence (AI). Healthcare has not even begun to realize the potential of AI, as we are merely at the infancy stages of applying it in healthcare. The current use case is very limited in scope and practice. AI requires massive amounts of specifically categorized data in order for it to learn the patterns and perform as expected. That’s why millions of us click on all the images that “contain busses” to confirm we’re human.

AI is the next “shiny object,” so naturally vendors want to use the terms and promises, instead of the facts. Like any good build-up of a story, this is one that we can’t wait to read. But before we go all “back to the future” in this tale, let’s tune in to the present day reality.

An AI Workforce Cannot Transform Healthcare Today

Some parents used to say, “choose your words wisely.” The same should also apply to AI in healthcare. AI has the potential to transform healthcare delivery, but it simply cannot be accomplished today. We don’t yet know enough about how AI in healthcare might work. We have ideas, and a limited scope of ways, in which it should work. Ultimately we hope to see AI as an integral part of the healthcare value chain, once there are standards for data quality, access, governance, security, and yes, even data sharing. AI does not currently “empower interoperability in healthcare by connecting disparate systems, or connecting humans to data.” It sounds great and the phrasing is fantastic, but it’s just not true, yet.

The scale of AI solutions remains small. We can’t shift from a “digital workforce” to an “AI workforce,” basically because there isn’t one. There’s Robotic Process Automation (RPA), Intelligent or HyperAutomation (the end-to-end automation of processes and data), and other ways of connecting data from disparate systems (APIs), but none of this is true AI.

AI is Not Currently Empowering Humans in Healthcare

AI is not performing tasks and completing processes. It’s not connecting disparate systems. It’s being used in small situations that may lend themselves well to AI in the future, like radiology, supply chain, and pharmacy operations. It’s not being used for Clinical Decision Support (CDS) because we don’t have the needed standards in place, yet.

Let’s take a look at claims status checks. It’s noted in multiple marketing materials that, “RPA is not able to do what AI can do in terms of actionable insights.” That’s simply just not true. RPA can indeed perform claims status checks, access information from multiple systems, and drive actionable insights, faster and less expensively. We’ve been driving value in healthcare-specific RPA for 30 years. This is why Boston Software Systems is different. We don’t lean on fancy words and phrases, we lean on our experience, our customer satisfaction, and our results.

Global Awareness Does Not Create Interoperability in Healthcare

Global awareness is a capacity that incorporates attitudes, knowledge, and skills necessary to promote the greater good; an ability to understand, respect, and work well with people from diverse cultures. Creating better people in an interconnected world. We should all aspire to be better global citizens. But that’s up to individuals. Global awareness cannot create interoperability in healthcare. These are inaccurate statements. Instead, be “self aware” before spending millions for an AI solution that just doesn’t deliver. Because it can’t do what it is marketed to do.

We care about accuracy because in healthcare that’s of utmost importance. And while we’re excited to embrace the FUTURE of healthcare, and AI, we care about our customers and their patients too much to take the risk of developing it using them as a test case (with their money).

What can we do in healthcare today?

  1. Automation can process most of the tasks in payment processing, inventory control and management, supply chain, lab, and payer processing.
  2. Automation can manage almost the entire revenue cycle claims operation without involvement from humans.
  3. Automation can turn insights into actionable information for providers, identifying potential problems, and in many cases, fixing the problems and closing the loop on the entire process.
  4. Automation can access information on 3rd-party websites, complete forms and portal requirements, and connect data from disparate EHR and legacy information systems, thereby reducing administrative “human” fatigue, and improving interoperability efforts.

It’s automation, and we’ve been doing this for 30 years.

Maybe you’ve had business cards printed in the past and the main piece of advice was “say what you do.” Not what you might do years from now, or what you aspire to do, but what you do presently. In our case, it’s keeping our promises, stating our truths as to what our services can/cannot accomplish and in the process, creating happy customers, 100% of whom would purchase from us again. We are first and foremost for our customers, not our shareholders, that’s why we score so high in customer satisfaction.

Why Boston Software Systems?

If you’re looking at “AI in healthcare,” just know that you’re paying for the words and the story. The words are good, don’t get us wrong. But the truth is better, and AI is just not available in the revenue cycle or in healthcare automation services today. Find out more about our automation solution, Boston WorkStation. With most solutions live in <30 days, savings and efficiencies are right around the corner.

One thought on “Artificial Intelligence (AI) is Not Ready for Prime Time in Healthcare, Yet

Leave a Reply

%d bloggers like this: