It is hard to capture the magnitude of HIMSS23. Simple numbers like 35,000+ attendees and 1,000+ vendors don’t adequately capture the excitement in the air. The spirit of innovation was palpable and energizing. The two massive show rooms were packed with vendors (and provided plenty of exercise).
I enjoyed catching up with old friends and making new connections. If I tried to recount everything I learned in those conversations, we would be here all day. However, as I walked the floor and talked to people three things stood out. Not surprising, Artificial Intelligence (AI) was on everyone’s mind and AI vendors were well represented.
Interoperability was also huge. The Interoperability Showcase was one of the biggest sections on the floor, right next to the Federal Pavilion, and it was constantly buzzing with activity. I was also struck by the presence of many consumer companies who had a major presence this year. It is clear that the breadth of companies trying to solve problems in healthcare is only growing.
With the recent news around ChatGPT, it is no surprise that AI was on everyone’s mind this year. Everyone is wondering will AI have an impact on healthcare and, if so, what impact. I have to admit that I entered HIMSS23 as something of a skeptic. I certainly believe that AI will have a massive impact on healthcare, but the way I have seen generative AI and Large Language Models (LLM) applied has yet to impress me. Don’t get me wrong, ChatGPT is a giant leap forward in technology and will undoubtedly lead to many amazing things. My issue has more to do with how it is applied in healthcare. In many instances, ChatGPT has been utilized to address limitations within existing workflows, highlighting opportunities for improvement.
Let’s take the now famous, or maybe infamous, example of using ChatGPT to write a letter to a health plan to appeal the denial of an authorization request. While the potential for various inconsistencies and inaccuracies are concerning, they are not my problem with this example. My concern is simply, “Why does a doctor have to write a narrative letter to a health plan in the first place?” Auth decisions should be based on a logical understanding of the data, not on the doctor’s writing ability.
To really utilize the power of AI, we need to take a hard look at the current workflows in healthcare and start to see how they can be re-envisioned, not just patched. We should look hard at anywhere a narrative is generated in healthcare and ask, is this really the best way to convey this information. I would hazard that most of the time it is not.
However, I was heartened to hear at HIMSS23 about some valuable uses for AI. For example, using LLMs to summarize the massive amount of clinical research available today. In 2020, the amount of medical data in the world was doubling every 72 days. 1 There is no way that anyone can keep up and it is estimated that a doctor will only be able to learn about six percent of the information available in their specialty.1 If AI can help solve that problem, then I think we are on to something. However, I suspect, like narratives, the power of summarization is also limited, and we will have to envision new ways for doctors to interact with this wealth of medical data in the future.
The Interoperability Pavilion at HIMSS23 was very well run and packed with organizations like Fast Healthcare Interoperability Resources (FHIR®) and Business Alliance (FHIRBall). The first night I attended a reception for Health Level Seven International® (HL7®)/Da Vinci. The exhibit hall was bustling with activity, resulting in numerous captivating conversations occurring on the walkway outside. Although the march towards interoperability in healthcare may be a long one, up a steep hill, in a blizzard, I am starting to see real glimmers of the destination. While we still have a way to go, it is clear that the industry is taking it seriously.
With the recent Centers for Medicare & Medicaid Services (CMS)/Office of the National Coordinator for Health Information Technology (ONC) proposed rulings around Payer-to-Payer and Prior-Auth2, we are starting to see real problems being attacked with interoperability. It was very heartening to hear many health plans discuss their desire to move forward on these initiatives even before the final rulings come out. The industry is clocking on to the fact that interoperability can drive real value and solve real problems.
That alone is very exciting, but I was also struck by how ubiquitous FHIR is. One of my friends in the interoperability community commented to me that just a few years ago everyone wanted to know what FHIR was, but this year he only met one person who didn’t already know. While other standards, like Health Level Seven International Version Two (HL7 V2), Consolidated Clinical Document Architecture (C-CDA), and X12, an accredited standards committee, are not going away anytime soon. It is clear that FHIR is not just the flavor of the day.
Industry leaders like Best Buy and Salesforce are two notable examples of the diverse array of organizations that attended HIMSS23. In 2020, healthcare represented 19.7 percent of the United States gross domestic product (GDP)3, so it is no surprise that everyone wants in. However, I think there is a more interesting trend going on. Sometime ago I wrote a post on Health Care 4.0. I was looking at how trends in healthcare can be compared to industry and the industrial revolution:
Industry 1.0: Steam power enabled small scale mechanization of physical tasks. | Health Care 1.0: Care is delivered largely on an encounter-by-encounter basis. |
Industry 2.0: Electricity and the assembly line enabled mass production. | Health Care 2.0: Medical equipment, such as MRIs and new monitoring devices, allow care to extend past single encounters. |
Industry 3.0: Computers and robotics allowed automation of routine tasks. | Health Care 3.0: Computers enable electronic health records (EHRs) to coordinate care across teams within an organization, linking multiple encounters. |
Industry 4.0: Connectivity (The Internet of Things) and Big Data / Artificial Intelligence (AI) have enabled linkage across the supply chain, empowering new manufacturing paradigms. | Health Care 4.0: Interoperability between healthcare systems, wearable devices, and AI will combine to empower care across the encounter and organization barriers, opening the door for patients to become full-fledged participants in their own healthcare. |
Just before HIMSS23 Adam Cole, Senior Solutions Architect and AVP Sales Engineering at Smile Digital Health, and I released a blog post called Empowering FHIR. In it, we discussed how important good quality data and industry leading tooling is to the success of FHIR. I continue to believe that these concerns are critical to the success of FHIR, but also to the success of interoperability and the future of healthcare.
By any definition, HIMSS23 was a massive success. It showed that healthcare IT is looking towards the future, and that future is bright. There are still numerous hurdles for us to overcome, but significant strides have been taken. I can’t wait to see where HIMSS24 takes us.
Sam has been working as a software developer and architect for over 25 years, in several different industries such as finance, human resources, and education. In his role as Principal Architect for Innovation and Strategic Initiatives, he designs next generation architecture, raises awareness on product capabilities, supports sales, and conducts research and development for future initiatives. He is an expert in healthcare data architecture and has extensive experience working with clinical data standards including HL7 V2 messages, C-CDA documents, flat files, QRDA I/III, and FHIR. He is an active participant in industry conferences and regularly publishes articles on healthcare technology trends and best practices. Sam is currently advancing standards for AI/NLP outputs as well as investigating how Availity Fusion and NLP can complement each other to help drive innovative solutions that improve long-term health outcomes and streamline operational efficiencies.
1 Densen, Peter. “Challenges and Opportunities Facing Medical Education.” Transactions of the American Clinical and Climatological Association, 2011, www.ncbi.nlm.nih.gov/pmc/articles/PMC3116346/.
2 “The Federal Register.” CMS Advancing Interoperability and Improving Prior Authorization Processes Proposed Rule, 9 Feb. 2023, www.federalregister.gov/documents/2022/12/13/2022-26479/medicare-and-medicaid-programs-patient-protection-and-affordable-care-act-advancing-interoperability.
3 “NHE Fact Sheet.” Centers for Medicare & Medicaid Services (CMS), www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NHE-Fact-Sheet. Accessed 9 May 2023.