Artificial intelligence is fast proliferating across healthcare, with various applications large and small finding their way into workflows industry-wide.
Whether it’s helping clinicians during telemedicine visits, transcribing entire conversations between doctors and patients, writing notes for nurses in response to patient portal questions, helping patients triage their problems via chatbots, or any number of other applications, AI is proving itself useful to many stakeholders in healthcare.
Narinder Singh has been working with AI for years. He is CEO and cofounder of LookDeep Health, a virtual sitting, virtual nursing and virtual care company. Past roles include working in Accenture’s Center for Strategy Technology, a corporate strategy position in the office of the CEO at SAP, cofounder of Appirio, president of Topcoder, and vice president of engineering at webMethods.
Healthcare IT News sat down with Singh to discuss how AI can help increase capacity in telemedicine, the risks posed by generative AI for hospitals and health systems, how provider organizations can overcome these risks, and the role AI is playing in scribe technologies.
Q. You note that telemedicine of course removes the burden of distance from healthcare interactions. But you say it does not increase the capacity for them. How do you see AI helping here?
A. Let me start with some context for why this is a key question, and perhaps the question for the future of hospital care. Every week we speak with hospitals that note patient acuity is rising or they are staff constrained, and most cite both.
The U.S. population over 65 grew five times faster than the total population from 2010 to 2020 – the fastest rate in more than a hundred years. This is part of a longer-term trend and highlights the rising age and associated acuity of the patients hospitals will care for in the future.
At the same time, we have seen repeated projections from worrisome to disastrous for nursing and other roles in the hospital – and that is independent of financial pressures that make it nearly impossible to expand staffing levels.
We now have had a generation of telemedicine inside the hospital – from eICU to tele-consults and now virtual sitting and virtual nursing. At a project level, there have been many successes, but at a macro level, collectively telemedicine in the hospital has had a very limited impact on care, sans one massive exception – COVID.
During the pandemic, we learned that seamless access via telemedicine creates flexibility that enables a system to adapt. Yet, it did not expand our resource capacity. Tele-capabilities can bridge great distances, but do not change the underlying units of work necessary to provide care.
Now, AI can mean many things, but let’s start with what relates to telemedicine – the ability to expand our observational capacity (rather than how it impacts decision-making). Today, a nurse who covers six patients will be in any given patient’s room for one to two hours. Doctors will be in an individual patient’s room generally only a few minutes per day.
Therefore, the vast majority of the time, a patient is without the watchful eye of a provider. This is despite the fact that so much of what is happening with the patient can only be assessed and understood at the bedside.
Are they less active; trying to get out of bed; does their breathing seem more labored; did the alarm go off because the sensor slipped off their finger or the breathing tube slipped out of their neck; etc.?
One branch of AI, computer vision, can let us have eyes on every patient all the time. This can help allocate the scarcest resource in the hospital – the clinical attention of nurses and doctors – more appropriately.
We have decades of evidence that increasing clinical bandwidth positively impacts patients. Video alone – even in rightly compelling areas like virtual nursing – will simply repeat the disappointments of the past. With AI, we can better leverage the time and expertise of our most significant constraint.
Imagine a world where AI acts as a guardian angel for patients and their caregivers. Identifying potential issues and alerting healthcare professionals before a small problem becomes a big one. This isn’t just about efficiency; it’s about fundamentally changing the way we deliver care.
AI can provide that extra layer of support, ensuring that no patient is left unattended, even for a moment. It’s not about replacing human touch but augmenting it, making our healthcare system more responsive, resilient, and, ultimately, more human.
Q. You caution there are real risks posed by generative AI for hospitals and health systems. What are they?
A. Generative AI can streamline prior authorizations, patient coding, and the intricate interactions between insurance and healthcare providers. However, it could also ignite an epic civil war between them.
This productivity could lead to a faster but more complicated landscape of disputes, ultimately requiring more human adjudicators to resolve disputes. Instead of cutting down on admin work, it might actually increase it. Generative AI could infinitely scale the most cynical stereotypes of over usage and aggressive denial of claims.
AI tools are making strides in reducing the time doctors spend on paperwork, especially outside the hospital. But in a hospital setting, the complexity of care and the lack of defined “visits” mean these tools aren’t as effective yet.
We have had years to learn how difficult and specific the development and application of machine learning algorithms are in hospitals. The allure of a magic approach to remove that tedious hard work and its integration into clinical workflows is tempting, but naive.
“Generative” patterns are relevant to many parts of healthcare operations, but they are not a golden ticket. They do not yet address the need to synthesize defined sets of information and repeatedly draw the same conclusions from them. The predictability of inputs and outputs is crucial to evaluation and certainty in clinical decision making.
Q. How can hospitals and health systems overcome these risks posed by generative AI?
A. On the first point related to the battles between insurers and providers, I see no immediate resolution. You simply cannot afford to have humans attempting to deal with the volume of AI-generated requests or responses, so participation in this arms race is unavoidable.
However, engaging in a way that sets a foundation for evaluating and incorporating generative models into workflows offers leverage for the future. Key steps include securing PHI, ensuring checks and balances on outputs, evaluating models within and outside their scope, and not alienating your workforce with premature claims of replacing their roles for a few dollars an hour.
These are just the beginning.
Already we are seeing insiders like Sequoia and Goldman questioning the hype and benefit of generative AI. We will go through a valley of despair; yet focusing on the pragmatic and not falling in love with the broad proclamation will keep many an innovation team from the cutting block. Hospitals need two antagonistic mindsets.
First, experimentation is essential. Generating non-clinical content (emails, communications), evaluating summarization of EHR context, improving language translation, and transcription – these all are areas where generative AI can be safely tuned and targeted for improvements. These applications can free up valuable time for healthcare professionals to focus on more critical tasks.
Second, hospitals must enforce rigorous evaluation and demand repeatability. For clinical scenarios, you should expect proof of any claims of capability. Even better, have an approach for continuous evaluation of AI capabilities within the solution. Concrete claims must ensure that the same set of inputs produces the same results, maintaining consistency and reliability in clinical decision making.
In other industries, technologists, as Norman Vincent Peale once quipped, “shoot for the moon and settle for landing in the stars.” In healthcare we have seen the disastrous implications of such strategies setting back industries for a decade or more (Theranos for blood testing, Watson for AI for cancer).
You can be pragmatic without being slow – the right leaders will drive that balance.
Q. You have observed more than a half dozen transcription companies raising more than $30 million in the last few years. Why is this? And what role is AI playing in these scribe technologies?
A. There are more than a million doctors in the United States. Their time is incredibly valuable, and a generation of being treated like both experts and entry-level data analysts has driven tremendous burnout.
The math is straightforward, and now the technology is more accessible than ever. The narrative that “the time is now” is not a new one, but it may finally be becoming a reality. It’s a wonderful use of technological advances.
AI is playing a pivotal role in these scribe technologies by drastically improving the accuracy and efficiency of transcriptions. With AI, transcription can be done in real time, with higher accuracy and at a fraction of the cost.
The challenge is that in just the last months, AI advances continue their breakneck pace of advancement – redefining the starting point of building such solutions. It is clear that transcription solutions are not foundational AI models themselves; rather, they are solutions built on top of foundational AI models.
The cost of developing competitive solutions has likely dropped by 95%. Better integration with clinical workflows, exceptional go-to-market models, and innovative derivative solutions remain massively important differentiators. However, the quality of difference between top solutions in the AI aspects of transcription itself will become essentially zero.
As a result, in this future, it is only inertia that will prevent prices from dropping dramatically, which should be great for healthcare providers. Lower costs will make these advanced transcription solutions accessible to more practices, further reducing the administrative burden on doctors and allowing them to focus more on patient care.
The surge in investment in transcription companies is a testament to the transformative potential of AI in healthcare – the risks are that the commoditization of the category results in desperate over-promising in order to keep up with investor expectations.
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.