ACCENTURE: How is AI going to change the healthcare landscape?
DR. RON MOODY: Our current healthcare model is unsustainable. It’s unsustainable as a percent of GDP. It's unsustainable with a population that is getting older and has an increasing disease burden. It's also unsustainable from the financial perspective of an individual.
AI can help support healthcare transformation. Today, we have a huge policy and administrative burden in both clinics and hospitals. AI could do what it's done across multiple industries, which is to transition employees to higher value work with greater impact by automating their underlying rote tasks.
Second, AI will impact the discovery of the new. It will move us toward a data-driven, evidence-based culture for how we support health and deliver healthcare. It can examine not only your healthcare data but also lifestyle factors to provide a more complete picture of the individual and treatment for a condition. Think about suicide prevention: AI can not only look at EHR data but also social activity to identify potential candidates for counseling.
A: What are some of the potential quick wins for AI in healthcare?
RM: The first thing to consider is workforce optimization. Robotic process automation (RPA) can free employees to take a larger role in managing care. AI-driven chatbots can direct patients to the right resources at the right time faster and minimize repetitive encounters. Think what could be accomplished by streamlining everyday repetitive tasks—the number of people answering the phone, processing a claim, coding a chart or recording payments.
In these cases, the potential value of AI is high and the barriers to people accepting it are low. We also have models for applying this, based on what's happened in other industries. For example, RPA is becoming more widely adopted. RPA can reduce costs while improving accuracy, empowering workers to provide more personalized and targeted support.
A: What is Accenture doing with AI in health care?
RM: We have taken an active role in helping healthcare leaders understand the technology and how it can be applied to deliver value and improve outcomes today and tomorrow. And we have focused and invested not just on the technology or a specific solution, but rather, how will people and organizations actually use AI. That's what lies behind our whole push for digital studios, user-centered design, and a focus on outcomes.
We've applied this approach to identify problems in the opioid epidemic, for example. We've done examples of predictive analytics with the U.S. Department of Veterans Affairs to look at where a suicide may occur, without even knowing any data on any specific veteran. We’ve worked with another federal client to enable more intelligent processing and analysis of medical claims.
A: How can healthcare professionals begin to prepare for the coming of AI?
RM: AI requires a balanced approach. AI is a powerful technology, but many use cases and applications are still maturing.
You start with education on what AI truly is and how it's being used. AI is not something you touch, it is part of the solution and process. Don’t limit yourself to federal examples as relevant innovation can be found anywhere, including state and local government, private sector healthcare and even other industries.
Always start by examining the problems that you're trying to fix, in the near term and in the future. Specifically, you want to consider how you are going to leverage the data you are collecting. AI is what takes you from being a data collector to being data-driven in the way that you approach solving or reframing your problems.
Natural language processing, which is a form of AI, has multiple places where it can provide value today. NLP is being used to streamline many aspects of the business of healthcare. In a clinical setting, we’re also seeing a long-term focus on searching the literature to assist with diagnosis, extracting data from EHR notes, decreasing the burden on healthcare providers for data capture, and in claims processing.
By focusing on practical problems, you can avoid a lot of pushback related to AI fears that are often overstated. Most people went into healthcare to solve problems, to help make people’s lives be better. When it comes down to time spent on administrative tasks, time away from your family, and wasted dollars, everyone in healthcare is looking for new solutions.
A: What other industries can healthcare professionals look to for guidance on AI best practices?
RM: Logistics is a huge issue in healthcare as we are one of the least automated markets comparatively. One of the reasons that we have so many delays is the number of resources that we need to orchestrate for every patient encounter. It’s about having the right doctor or nurse, the right medicine and the right equipment available at the right location and time.
Logistics also plays out in effective utilization of operating rooms. On the macro scale, AI can make general logistics in the hospital less costly. For example, we have supply replacement problems. Then we have a huge waste because we have medications that go out of date or out of stock. Now think about other industries. They have just-in-time awareness – Walmart revolutionized the practice, and Amazon put it on steroids.
The other one that's immediately applicable is the payer market, as commercial insurers are taking great steps in automating their processes. Our research has found that they can unlock an additional $7 billion in total value over 18 months through the use of AI.
There is great potential for AI to help power healthcare transformation – helping us unlock value to deliver better, more effective care. Achieving the Triple AIM of healthcare is possible but we can’t fear tech like AI or work with the belief that technology alone (even AI) will solve problems in isolation.