Generative AI and large language models like those behind ChatGPT are already being deployed in all areas of healthcare today. Developers are diving into the deep end of the pool to come up with potentially time-saving tools to add to their core technologies, and providers are using the technology for everything from patient engagement to clinical decision support.
But is generative AI ready for prime time? What can it really do at this early stage? What do healthcare IT leaders and employees at healthcare organizations need to know?
Healthcare IT news sat down with Vrinda Khurjekar, senior director of Americas business at Searce, an AI consultancy, to get her expert answers to these and other questions.
Q. What are your thoughts on generative AI? Can it be used reliably in healthcare today?
A. After the internet-led revolution, I believe that generation AI is the next big technological advancement that will fundamentally change the way we work and live. Generative models have been around for a while, but their mainstream adoption will drive these transformations at scale.
Its applicability across different sectors is already visible, so healthcare as a sector is no different. Whether it’s improving the patient experience by introducing conversational processes for booking appointments, or using summaries to enable doctors to view the right information at the right time, or improving overall efficiency in the hospital care system, each of these implementations has profound implications. about the way we think about patients and overall healthcare.
Although it is still in its early stages and many things still need to be done – such as which use cases have the least hallucinations (errors) and which data privacy and sharing rules we will have to define to ensure the ethical nature of the models, among others – the results are promising.
In terms of whether it can be used reliably for every use case you can think of, probably not and certainly not yet. But is there a way to get us to that stage, with the right governance and control, certainly yes.
Q. What are the challenges generative AI faces when it comes to mainstream use?
A.
-
Unclear where to apply. For all companies, the urgency to adopt gen-AI is causing them to choose use cases that do not add value. For example, time and time again we see conversations like: “We need to do something with gene AI, our board wants it.” And in many cases there is no clear answer to the question of what exactly. Another example is when we say to customers, “To do ‘X,’ we need to get your data in order,” and customers say, “Well, that’s not a priority. We can’t do that.” So there are several disconnects that need to be addressed.
-
Hallucinations. Most of these models are hallucinatory. A lot of work is needed to get them into a stable state for relevant business functions. Most companies may not want a slower rollout to iron out these wrinkles.
-
Less visibility on ROI. Currently, most of these models are expensive to use and implementation costs are high due to talent shortages. Companies struggle to make it very clear how high upfront costs can be justified for longer-term ROI. Until economies of scale are achieved, access will remain limited to organizations able to free up budgets, limiting mainstream adoption.
Q. Many healthcare IT vendors, including electronic health record companies, are moving to generative AI functionality. What should leaders and physicians in hospitals and healthcare systems do when confronted with new generative AI tools?
A.
-
Ensure Gen AI controls at your supplier onboarding. Thoroughly evaluate your vendors’ security, compliance, and governance policies. It almost requires a rewrite of internal guidelines to accommodate all the nuances that come with AI technologies.
-
Adopt a zero trust policy. Stricter compliance audits on what is shared between IT providers, how visibility is returned to you and how the data is processed internally.
-
Invest in internal experts. The space is new and everyone is learning as it evolves. Investing in in-house experts and working with the right industry partners who can help you navigate the evaluation and implementation processes at these suppliers will be important to the success of these launches.
Q. Oracle recently made waves with the launch of its Clinical digital assistant generative AI. Please share a little about this tool and why you think it may or may not be the best path forward at this time.
A. It was quite exciting to see Oracle introduce a range of generative AI services for healthcare organizations within its EHR solutions portfolio. In particular, the vision laid out for the Oracle Clinical Digital Assistant on how it aims to streamline administrative tasks with voice commands is sure to be revolutionary in how you will experience clinical visits in the future.
While the vision is great, we’ll have to see the practical rollout of the features in about 8 to 12 months. However, I firmly believe that the investments are in the right areas. When you visit a doctor, you want the doctor to prepare well for your visit, ensure that he or she has to spend less time reviewing the reports, etc.
Moreover, the patients themselves should have the opportunity to get as much information about their conditions, causes, etc. as easily as possible. Having such a digital assistant not only frees up time for administrators and doctors but also helps provide a seamless experience for the retain patients.
Q. Where do you see generative AI in healthcare in five years?
A. Five years from now is a very long horizon as we see newer applications literally every day these days. But in five years I think the hype would be over. We would certainly see updated business processes with gen-AI being mainstream. Many newer solutions and technology players will emerge and dominate the market in very niche areas.
We will see a complete overhaul of compliance and regulatory guidelines to monitor bad actors and protect individual rights.
Most importantly, the world will focus on a lot of preventive care, which will allow us to proactively monitor our well-being, which in turn will improve the life expectancy of every individual.
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.