While this has been a very turbulent year in all manners of speaking, for health technology it has been a rather eventful year and mostly positive. While the world has been dealing with major geopolitical issues posing risk to economic stability and long-term growth, high inflation rates, higher for longer fed fund rate, US elections, and more, transformation of healthcare through the use of modern technologies has been chugging along. This is, in large part, due to the non-cyclical nature of healthcare. For the most part, the pace of the adoption of technology in healthcare is mostly dictated by long-term sector-specific factors such as data fragmentation, FDA regulatory decisions, federal government regulatory moves, pressing business issues for the key customers, and more. As such, even in the rosiest of geopolitical and economic environments, things don’t move fast, or in the most challenging of environments, things continue to move ahead at their usual glacial pace.
For health technology and its promise to make healthcare more accessible, more convenient, with better quality, there was good news this year. That meant a few things, and I will review my top 5 developments of the year in a second, but the headline of the year has to be the vertical applications of Gen AI. What does that mean? It means that there is significant optimism that advancements in the capabilities of AI and to be more specific, Gen AI, will lead to automating in healthcare what hasn’t been possible to automate until now. Provision of healthcare and doing medical research is complex and involves numerous activities in the right sequence to achieve the right outcomes. The information needed to perform each step sits in many different places and sometimes it’s not even recorded anywhere. When I was a practicing physician, I often made my final determination about a patient’s diagnosis and management plan based on verbal or non-verbal input from the patient and/or their families while in the room. This information had not been recorded yet into any systems and thus was not available to any hypothetical decision support systems or other form of software that could take action as a result of it. This has made the automation of those activities challenging.
With the advancements in the capabilities of Gen AI, it can combine data in different formats, including unstructured clinical notes, and reason through what it means and what needs to be done next. While this has always been one of the key promises of health technology, given the chaotic state of data and the complexity of healthcare decisions, we haven’t seen it done to high quality so far. We’re seeing some promising progress in this area and early signs that the latest generation of health technologies can take over some of the mentioned activities and not a minute too soon. This year, ambient documentation has gone from a pilot project to being scaled across large healthcare organizations. This is a breakthrough moment for the industry. If AI can show tangible benefits to healthcare providers, it goes from a buzzword to a useful technology. By listening to the doctor-patient encounter and creating a concise and accurate summary of the visit, it can save hours per day from the work of a physician or nurse or therapist. I never enjoyed writing notes as part of my job as a physician. It adds work to an already busy day and makes you late for the next patient or procedure that you need to engage with. However, by creating an accurate and comprehensive accounting of what happened, you are ensuring that the next provider is fully aware of the latest with the patient and what needs to be done. If AI can listen and create this record, you’re freeing up major time for other activities and providing relief to the already burned out staff.
While the progress of Gen AI’s applications in clinical, research, and administrative realm was a major story in 2024, there were other key developments with significant long-term implications. Here is my rundown of the top 5 developments in health technology of the past year:
- Ambient documentation goes mainstream: As mentioned above, using a microphone to listen to the doctor-patient encounter and creating a document that needs minimal editing is a major boost to the everyday practice of medicine. While the hours saved are critical, it is possible that there will be major ripple effects that can improve patient care and the economics of the medical institutions. First, an always-on and consistent agent that listens and documents can conceivably capture everything that hasn’t always been consistently documented. This can result in better documentation of patient issues and help the other care providers have a more complete picture of the patients’ issues. Also, with more comprehensive documentation, the medical institution can submit better support for their level of billing and receive faster and higher payment. While this is speculative and not proven yet, it could prove to be another driver of the adoption of this technology over the next few years
- Healthcare stakeholders come together to accelerate adoption of AI: While there has been progress in certain areas such as ambient documentation and home testing, the key customers of health AI technology are still struggling with many issues that have to be addressed before they can fully integrate AI into their businesses. These include governance, quality control, selection of the right vendors, risk of medical algorithms, and more. If those institutions are left to their own to figure it out, we will see an uneven and slower adoption of this technology. That is why industry groups are forming organizations that can solve some of these issues for the customers. For example, Coalition for Health AI (CHAI) announced that they will soon release so-called CHAI Model Cards – which the group likens to ingredient and nutrition labels on food products. This means that when you are examining a health AI model, this Card, analogous to a nutrition label, gives you key information for the criteria you’re interested in before adopting a solution. CHAI is also championing the creation of a national network of independent assurance labs for healthcare AI. CHAI is creating a draft framework for certifying the future assurance labs. These assurance labs will test health AI models using the model card and provide a certification that these companies can use when commercializing their technologies to healthcare customers
- Federal Government steps into the AI regulation Arena: In late 2023, President Biden issued an Executive Order that meant to create an umbrella for the national regulation of the development and deployment of AI in various industries, including healthcare. The significant of this EO is that it signals to everyone that the federal government will not be a bystander while the AI battlefield turns into the wild west. By initiating the inaugural set of regulations to ensure safety, equity, lack of bias, and addressing other relevant issues with AI, it aims to create the rules of the road and instill confidence in everyone who will be touched by this technology. Healthcare, given its very nature, is singled out as one of the key industries where there will need to be special regulations and extra caution. While there is fear that this type of regulation can slow down the pace of innovation, I believe this will be beneficial for the adoption of AI in healthcare in the long-run as it will provide confidence to healthcare providers and patients.
- Rare sighting of sizable IPOs for companies using AI to improve healthcare: While this has been a long draught of all types of exits, including M&A and IPOs, there were a couple of large health AI IPOs this year. Two companies that use AI to tackle different parts of the care delivery successfully made their ways into the public markets, Waystar, a company that uses AI for revenue cycle management, and Tempus, a precision medicine company, had successful IPOs. This is great news for the sector since successful exits and distribution of capital to investors and LPs is critical for the long-term health and continued growth of this technology. If investors do not see return on their investment dollars, at some point they will conclude that while there is much promise in the applications of AI in healthcare, the adoption is too slow and their dollars are better invested in other areas. As such, while there is a whole economy-wide draught in IPOs, these two multibillion dollar IPOs in health AI send a strong signal to the players in this ecosystem that good use cases and strong product development can lead to high return on investment.
- Funding for health tech holds steady: While many of the headlines around funding levels in VC are about the steep decline from the 2021-early 2022 levels, a close examination of the data reveals that those vintage years were outliers. If you compare 2023-2024 levels of funding for healthcare technology, it is consistent with steady long-term rise over time that we typically see with these technologies. For example, the funding is very much in line with 2018-2019, pre-pandemic, levels and that is way up from 2013-2016 levels of funding. As such, the comparison with the zero interest rate, euphoric, high liquidity years of the pandemic era is not really appropriate. This means that in spite of high interest rates and the slow exit market, investors see much promise in this space and are continuing to write checks and wait for their moment. This is good news as things will turn at some point and we will see a large number of exits and those investors will be rewarded for sticking it out through the lean years.
Personally, 2024 is the year of the publication of my book, AI Doctor: the Rise of Artificial Intelligence in Healthcare, and it’s been an amazing year traveling around the world promoting the book and making my rounds in the media. My hope is to provide the innovators, customers, and investors with frameworks for evaluating the many use cases and companies. This will lead to the allocation of capital and efforts into the best use cases and companies, and this will in turn lead to an accelerated timeline for all of us to see the many benefits that we keep hearing about.
