Khang Nguyen, MD, discusses the power of augmented intelligence to enhance patient care and support the clinician workforce
AI and machine learning are at the forefront of transforming health care today — and when applied as value-based health care innovations that enhance the human touch and keep the patient in the center, they’re best positioned to set clinicians up for success.
Khang Nguyen, MD, chief transformation officer of the Southern California Permanente Medical Group, discusses how AI can increase clinicians’ joy in medicine and improve patient outcomes with Chris Grant, chief operating offer and executive vice president of Kaiser Permanente, The Permanente Federation, and host of the Permanente Medicine Podcast.
Dr. Nguyen, also the chief medical officer for care navigation with the Federation, described the innovation priorities at Kaiser Permanente. These priorities include supporting the physician workforce, utilizing data and predictive interventions to improve patient outcomes, and boosting quality and affordability for patients.
One of the current focus areas for AI efforts at Kaiser Permanente has been reducing the administrative burden on the clinician workforce. Contributing to this challenge is the exponential growth of information, an expanding population of older adults, and office visit documentation.
Dr. Nguyen shared the multiple benefits of an AI ambient listening tool — recently rolled out nationally to more than 25,000 Permanente physicians — that transcribes office visits (with patient consent) that can be edited and added to electronic health records.
“When we start to recruit new physicians and we show them how cutting-edge Permanente [Medicine] is, how our leaders have really pushed for this technology to support our physicians, this now resonates with people who are interviewing,” said Dr. Nguyen.
“For us, it’s been the differentiator for some folks deciding to work [at Kaiser] Permanente.”
Related AI scribes story: Analysis: AI scribes save physicians time, improve patient interactions and work satisfaction
Dr. Nguyen also highlighted the potential to utilize artificial intelligence and Kaiser Permanente’s extensive medical database to develop predictive analytic models aimed at enhancing patient care. These models could be instrumental in identifying patients who may benefit from proactive, early interventions.
“Wouldn’t it be great if we were to be able to target patients we knew were going to get a cancer or have highest risk for cardiovascular disease, for congestive heart failure, or the patient that is going to bounce back to the hospital and really pinpoint those patients with some predictive analytics in the future,” said Dr. Nguyen.
Related health care AI podcast: Assessing the impact of AI in medicine
The podcast episode also examined how the value-based care model, which emphasizes prevention, coordination, and patient-centered care, creates a better environment for innovations that improve care quality. Dr. Nguyen explained that having integrated care delivery at Kaiser Permanente means that at every step in a patient’s journey, someone is contributing to delivering the best care possible.
“Every interaction matters and every person matters.”
“Whether you’re the call center agent, the receptionist, [or] the lab tech, you participate and you provide the input into innovation,” said Dr. Nguyen. “Innovation thrives when you get multiple diverse perspectives.”
Connect with Chris Grant by following him on X at @cmgrant or LinkedIn.
Follow us: Subscribe to the Permanente Medicine Podcast on your favorite streaming platform.
Podcast transcript
Transcript is autogenerated. Although edited for clarity, it should not be considered an exact replication of the podcast and may also be updated as needed.
Khang Nguyen (00:03):
This technology allows the clinician to simply look away from the computer, look at the patient, reestablish that humanity in medicine is really what I call it, the joy in medicine, and connect with patient while the machine is documenting in the background.
Chris Grant (00:22):
Welcome to the Permanente Medicine Podcast. I’m your host, Chris Grant, the Chief Operating Officer of Kaiser Permanente, the Permanente Federation. Today we’re diving into what continues to be one of the most fascinating and fast evolving topics in health care, artificial intelligence, and its transformative impact on medicine. Our guest for this episode is Dr. Khang Nguyen. Khang is a dear friend and a close colleague, so I’m so delighted to have him here today. Dr. Nguyen serves as the Chief Transformation Officer of the Southern California Permanente Medical Group and the Chief Medical Officer for Care Navigation with the Permanente Federation. The Southern California Permanente Medical Group is one of the largest medical groups in the nation with more than 8,800 physicians providing care for over 4.8 million members and patients. In addition, Dr. Nguyen co-chairs the SCPMG Virtual Medical Center and brings extensive expertise as a practicing family medicine physician at Kaiser Permanente. He’s the ideal person to discuss innovation and innovation improvement in patient care. Dr. Nguyen, we are so thrilled to have you with us today to share your insight and expertise on this cutting-edge topic.
Khang Nguyen (01:45):
Thank you so much, Chris. It’s such a pleasure to speak to you again. I know we go back quite a ways and that this is a very important topic for all of us to really to talk about and to be educated about.
Chris Grant (01:56):
I think it’s transforming more rapidly than any moment in health care history, the way patients receive care and the way doctors and other clinicians deliver it. But before we explore this exciting topic, I thought we could start with you telling us a little bit about yourself. Perhaps speak about your role in the Southern California Permanente Medical Group and the Permanente Federation and share something about what you enjoy doing outside of medicine.
Khang Nguyen (02:24):
Thank you so much for asking that question. Like most everyone else, I started off really just taking care of patients. I’m a family physician by training. Loved it. Couldn’t imagine doing anything else. When I started over 20 years ago, saw babies being born, used to do circumcisions in clinics, and then taking care of patients as they got more mature in their life, so to speak, and really loved that. But from the very beginning, I really found myself looking for solutions to provide better candidates faster in a way that’s more affordable for our patients. And very soon I found myself trying to help the medical office find problems and solve for things that really were troubling or our MOB at that time. So I became the group pack coordinator and then the physician in charge of my medical office and Mission Viejo. I really enjoyed that, but really had my eye on more systems of things.
And I quickly found myself in a role of access lead for Orange County and where I managed over 300 doctors. I was fortunate enough when Dr. Davidoff got into becoming the executive medical director that he asked me to become the transformation lead for Southern California. And then, so from that point on, I began to look at technology systems, processes for all settings, inpatient, outpatient care at home, a hospital at home. And from there I was asked to help lead the Federation for care navigation. Outside of work. I rediscovered my joy for tennis a couple of days a week. I find I have a network of non-KP friends that I play with, believe it or not, at 4:30 in the morning. And I would remiss not to mention my wife and my family who might love and enjoy time with to travel.
Using AI to reduce the administrative burden on clinicians
Chris Grant (04:00):
Fantastic. Thank you for sharing that. Fascinating background. Let’s shift to our topic here, artificial intelligence. It’s making significant inroads across health care. So how do you see AI enhancing patient care today? Can you share some specific examples of how it’s transforming the way physicians work at Kaiser Permanente?
Khang Nguyen (04:24):
As we begin to look for the future for today, we really are looking at reducing the administrative burden that our workforce is really facing. This challenge has come upon us really because of the digital revolution. Information is coming at clinicians, whether they be doctors, nurses, pharmacists, faster than ever, and there is a transparency that exists around us. So one of the first things that we can do with artificial intelligence is to reduce the administrative burden, ambient listening or clinical documentation through the use of AI. Kaiser Permanente was the world’s largest deployer of this technology. This has been really transformative because this technology allows the clinician to simply look away from the computer, look at the patient, reestablish that humanity in medicine is really what I call it, the joy in medicine, and connect with the patient while the machine is documenting in the background. And that’s why they call it ambient listening.
It’s literally documenting the background and it’s able to summarize it and synthesize the note in a way that a clinician would normally put it, and the clinician is actually allowed to review the note, to edit it, modify it before it actually becomes part of the permanent electronic medical record. Now, this seems very little, but for most of the clinicians that I’ve talked to, this has been a game changer. A lot of clinicians who are on the verge of retirement have basically said, “This has changed it. I love what I do now. I’ve returned to the medicine I knew 20 years ago and I’m going to stay a little bit longer.” So the retention of doctors is becoming really a beneficiary of this AI tool. On top of that, we’ve also noticed that when we start to recruit new physicians and we show them how cutting-edge Permanente is, how our leaders have really pushed for this technology to support our physicians, that this now resonates with people who are interviewing for jobs.
For us, it’s been the differentiator for some folks deciding to work for Permanente or not Permanente. Now, keeping with those themes, there are a couple of things coming out that I think that are very critical to reducing the administrative burden— something called the augmented response technology. And it’s called augmented because it helps to draft a note for patients’ emails to us. Now, this is not going to be perfect from day one, but just getting the head start pulling information, drafting that note is a big deal for our physicians who really have in-basket bloat and really have a lot of what we call pajama time now, where they’re just doing a lot of work after hours. So that is coming, and I know it’s being tested with wild success. The other thing that I would actually highlight is that we have some large language models being used to actually help to coordinate care. So be able to understand why patients are calling us, why patients are going online, and then pointing that patient to the right care, the right place at the right time with the right clinician. Ultimately, we want AI to put our clinicians in the best possible place for success, and that means the patient’s care resolution. So those are just three small examples of things that we are deploying have deployed and are looking at and expanding to really help out our clinicians and our patients.
Chris Grant (07:34):
I think one of the biggest misconceptions of artificial intelligence in medicine is that the human aspect is being replaced by a machine or a computer, but as you point out, it’s really the opposite. Really the biggest benefit is that physician can utilize enormous data sets, but also can lean in with the patient, could look them in the eyes, can spend an extra 10 minutes or 15 minutes, can see the tightness in their jaw where a little tear in the corner of their eye because they’re worried or they’re concerned or they’re nervous about something. And that’s something that a computer can never do. To me it’s just game changing. And I love the fact that you mentioned that we’ve got physicians that are planning now to extend their careers because the joy of medicine has come back with the support of ambient listening and other technologies. What’s the biggest misconceptions physicians and patients have about AI?
Khang Nguyen (08:32):
I think the biggest misconception is that it’s autonomous, right? So we talk about artificial intelligence, but it’s not autonomous intelligence. It won’t go off on its own. It’s actually not even designed to do that at this very moment. It requires human intervention. There will be two types of clinician in the future. There will be clinician that uses AI to help out their practice, and there will be one that won’t basically, and that is the one that that will really be struggling to search through the database, looking for stuff that perhaps the AI can offload and allow them to actually truly practice meaningful relationship-based medicine.
How the value-based care creates an environment conducive to innovation
Chris Grant (09:09):
Yeah, I completely agree. Dr. Nguyen, I’d like to take a step back and put your idea guy hat on here to take a broader take on innovation specifically through the lens of value-based care. The terminology value-based care is now pervasive in medicine and it’s what we’ve been practicing at Kaiser Permanente for decades and decades. How does this model which emphasizes prevention, coordination and patient-centered care create a better environment for innovation?
Khang Nguyen (09:41):
So I think you just nailed it when you mentioned those last couple of words, prevention, coordination, patient-centered care. And the reason it’s because it’s end to end. It’s really 360 as we deliver Permanente medicine, it’s value-based. Value means you’re giving the patient more than what they thought and what they signed up for. When my call center agent speaks to the patient on the phone, they make sure that if they’re due for mammogram, that they get the mammogram. They’re due for cervical cancer screening, that they get that if they are due for a lab test, they get that. The reception does the same thing, the medical assistant does the same thing. Now you noticed I said everyone but the doctor. And the reason for that and why this type of care delivery is so good for value-based care and for innovation is that every step along the way, one has something to contribute to the overall care of the patient.
In other words, every interaction matters and every person matters. Whether you’re the call center agent, you’re the receptionist, the lab tech, you participate and you provide the input into innovation. Innovation thrives when you get multiple diverse perspectives. And so as a result of that, innovations thrive because the innovation is around care resolution. It’s around providing the best value. And I sum it up with one statement, and I always tell people whether my call center agent, the receptionist, I would say, what if this were your mother or your father? What would you do? And they would say, before they saw the doctor, I’d like them to do X, Y, and Z, or I’d like the doctor to do X, Y and Z. And if I asked, the doctor would say the same thing. I’d like the agent to do this, or I’d like the lab tech to do this, or the person covering me at night to do that. But it’s this inclusive 360 view of the patient that is unique to us because that’s what we value and that’s what we care about. And that’s where innovation thrives, when we get multiple diverse perspectives.
Chris Grant (11:37):
Fantastic. And it’s this idea of your doctor, your health care organization is thinking about you all the time, 7 by 24. even when maybe you’re not even thinking about your own health. I was at an ophthalmology appointment a while back and I was checking in and there was a receptionist, her name was Rosie. There’s a big bouquet of flowers and balloons behind her, and they say, congratulations. So I say, “Rosie, what are the congratulations for?” And she said, They said, I saved a life.” And I looked at her and I said, “Say more.” “So when I was checking in a patient a couple of weeks ago, they were automatically flagged to be due for a mammogram. And I convinced her to schedule a mammogram that day, and apparently, she tested positive for cancer. So they told me I saved a life.” And that’s a great example of the exact value-based care that you are talking about, using systems for identifying patients that would benefit from preventative care, for creating access for a same day appointment so that the hesitant patient for whether it’s a mammogram, pap smear or prostate exam or a childhood immunization can get in quickly and get that care.
And who knows, you may just save a life. And to your point, Khang, it may be the physician or the surgeon, but it may be the receptionist or the medical assistant. Everybody in a value-based system contributes to betterment for all patients, and that’s the coolest thing of all.
Khang Nguyen (13:29):
Absolutely. At the end of the day, we walk away and you have a stable job, that’s great, but if you feel good about what you do, you’ll do wonders. And we want that top to bottom.
The importance of maintaining the human touch in medicine while embracing AI
Chris Grant (13:40):
I love it. Dr. Nguyen, here’s a follow-up question. How can physicians successfully maintain that human touch that you so eloquently talked about in patient care while at the same time embracing the advantages of artificial intelligence?
Khang Nguyen (13:57):
It’s more important than ever to work on your interpersonal skills to work on the way that you relate to people because now the patient is actually listening. So there’s a sign that says “The doctor is now in.” Well because of AI, the docker is now in and the doctor is now looking at you. Really what we’re trying to say is the machines are not here to replace this. They’re here to augment us and we still really matter. So that’s what I would say is really understand what we’re trying to accomplish with the AI systems.
Chris Grant (14:24):
And Dr. Nguyen, I hear you time and time again emphasizing that this gets back to the roots of medicine. This actually frees up physicians to lean in and really understand their patient, to have the time and the capacity to do the appropriate diagnosis, but also to listen to them and to motivate them in their own care and wellness. Let’s explore patient care a little more here. What are your top strategic priorities when implementing AI solutions and patient care and how do they align with Kaiser Permanente’s overall patient care goals?
Khang Nguyen (15:07):
Everyone working for this organization should really know its mission and vision. In Kaiser Permanente here, we exist to provide high-quality, affordable health care for our patients in the communities we serve. So if you take that perspective and you say, what are the priorities? It’s any AI that helps support high quality. So increasing mammography rates, getting the right patients to the right care at the right time. So high-quality, affordable things that will actually reduce the cost structure so we can deliver Permanente medicine to more people. So people doing self-service online, helping themselves so that we can re-devote our resources to actually other things and reduce that cost structure for our patients and our communities. Our patients stay with us for just a really long time and longer than any other health insurance. I’ve heard numbers of 15 years on average that our patients stay with Permanente, and that’s because of Permanente Medicine.
By the way. When they connect with the physicians and what we deliver, they stick and they stay with us. So my priorities are basically to help to increase quality, to make things more affordable by getting patients the right place at the first time, having them do self-service and then allowing Permanente care to help our patient and our communities. And I think that’s really critical. If we can take our databases and extrapolate them to our communities so that we can pick out the fringe use cases, so that we can pick out the most common things to do predictive analytics to really benefit all of our communities, that’s really something that I think that we should be moving towards. And those would be the priorities we have within our fingertips. Probably the largest organized database. It doesn’t feel like it sometimes, but in speaking to outside folks, we have the largest organized database anywhere in the world, and it’s about time that we leverage technologies like AI to unleash that to benefit our communities and our patient. Wouldn’t it be great if we would be able to target patients we knew were going to get cancer, or had the highest risk for cardiovascular disease, for congestive heart failure, the patient that’s going to bounce back to the hospital and really pinpoint those patients with some predictive analytics in the future. So those would be my priorities, to unleash our databases, unleash the power of Permanente.
Chris Grant (17:28):
Yeah, that’s fantastic. What I’m hearing you say is that continuity of care, that longevity of care creates information on a patient that improves not only the way we meet their needs immediately, but also supports predictive analytics for addressing potential future health challenges that they might face and heading them off early with intervention to support them. Being a member for a long time has real health benefits.
Khang Nguyen (18:00):
It really does. And a while ago, there were fires and what we were able to do with our databases was really to identify those at highest risk for an asthma exacerbation under these circumstances. And we proactively at that time — this is without AI — using our databases, to get ahold of these patients and ask, “Do you have a refill of your ProAir? Do you have your medication?” Also understanding where pharmacies would not be available because of the catastrophes that were going on. So putting that all together to prevent disasters is at the heart of what we’re trying to do. And then putting that on steroids with AI. I know for a fact that the government or legislators looked to us in Permanente. I felt very humbled that at the time of utmost crisis in our country, in the world, the height of COVID, the State of California asked Kaiser Permanente to help administer all the testing for COVID tests in the state of California because they knew that we had within our written database and our operations or organization to actually execute on a really large scale. So that’s a moment of pride, but that’s also a moment to think about what we have and what we can have for the future.
Ethical considerations and challenges around the responsible use of AI in healthcare
Chris Grant (19:08):
It really is fantastic and almost mind boggling when you think about something as fierce as the pandemic and being able to zero in across 12 and a half million patients and say, these are the 800,000 that are the highest at risk, and we’re going to zero in and advance immunizations as soon as possible, or wildfires that break out in Los Angeles and utilizing data and information and say these are the patients that are at highest risk of severe asthma attacks based upon all of the information and data across many years, and let’s intervene and let’s get them support rapidly. You’ve talked a lot about a lot of the positives, but there’s always those that worry that AI will bring complexity to their work. Tell us, what can health care organizations do to address these concerns, whether it’s from a patient or a physician perspective, and what can physicians do to successfully stay ahead of the curve when it comes to AI and emerging health care technologies?
Khang Nguyen (20:14):
There’s a quote that says, with great power comes great responsibility. So artificial intelligence can do a lot of prediction. It can help us do a lot of things. But is our system and is our organization kind of ready for it, so to speak? And so because you can predict future disease based upon this conversation or recording, the question is should you at this point, right? And so there’s some ethical dilemmas, and that’s why I think that all of us should really have really responsible AI groups that we can have these ethical discussions. And I think these things will get better because as AI begins to improve operations, improve throughput so that we have more capacity to do more things, then you can answer some of the other ethical questions. I think for a lot of our frontline physicians looking at things that help out with administrative burden, it’s really setting up structures for deployment and adoption of artificial intelligence. And that isn’t always easy. Clinicians are oftentimes very conservative by nature because we’re health care clinicians, providers, so it is sometimes daunting. But having systems in place to deploy to early innovators, early adopters, late adopters, and having that process in place to be inclusive of all specialties, all settings, is going to be really important for us to adopt things that help out with the administrative burden of artificial intelligence.
The next five years of implementing AI solutions in patient care
Chris Grant (21:36):
That makes a ton of sense. One of the wonderful things that I loved about your comments earlier, Khang, is yes, this is about what Kaiser Permanente is doing within our fairly large footprint, but this desire to also share our thought process and our innovation externally in order to advance all of health care is very real and one of the foremost individuals in the organization that speaks externally at large conferences and major events in order to share with the rest of the health care leadership across the United States what’s possible. So I really appreciate all of the leadership that you provide within KP, but also externally. This has been a fascinating conversation and I could go on for the next two hours. I think you and I are both technology geeks, but I want to close with a forward-facing question. Where do you see AI making the biggest impact in medicine over the next five years?
Khang Nguyen (22:39):
I think there are a couple of dimensions I can think about, and the notion of returning the humanity to medicine is really important. And if you think about what that means, returning the humanity to medicine, that’s really reintroducing and strengthening that relationship bond, which means things that are administrative need to be further developed, things that are able to pull information outside of the medical records so we’re not hunting for it, need to be developed. And I think that will be the biggest advancements: putting the information in front of the physician at their fingertip so that they can focus on the patient. The second thing is really around patient self-service. Patients nowadays have evolved, so our ability to build and engineer AI systems to help people help themselves is going to be the future. Now, it isn’t just where patients want to go. That’s actually one of the strategies we have for the workforce challenges that we have.
The people are living longer in large part, do the great care that Permanente provides, but that also means that their services that they need will go up. So our workforce needs to evolve, that workforce needs to include the patient. So imagine that 4.8 million patients, 4.8 million team members as part of the team care delivery system. So leveraging self-service is another aspect of AI that we have to develop into the future if we are going to sustain in this path. Care navigation is key. So that’s number three. We provide such a comprehensive integrated care delivery model that the great thing is it’s all here. The bad thing is it’s all here and we have to find it. So pointing people in the right direction sounds like an easy task, but it’s actually very hard as we begin to really segment our services and get real specialized in what we do.
And we get interconnected across the Permanente teams, a specialist in Mid-Atlantic who takes care of a patient in Southern California or a specialist in Southern California taking care of a patient in Northern California. So again, that interconnection and care navigation is going to become important. And then the final thing, and I think I’ve mentioned this, is really looking at predictive analytics in a very meaningful way. And not just looking at disease burden, but really looking at social determinants of health equity and really determine making predictions on which people will be at the highest risk, not because of their clinical information, but also their socioeconomic information, where they go for care, et cetera, et cetera. So I think those are the four lenses that I think that will be the greatest things that we can do in the next five years with the help of artificial intelligence.
Chris Grant (25:15):
Care navigation. Self-care that’s really convenient. Predictive analytics to be thinking about you maybe when you’re not even thinking about yourself and determining care needs. And then the line that I love: returning humanity to health care. This has been absolutely fascinating, and I just want to thank you for all that you do for Kaiser Permanente, and thanks for all that you do for physicians across the country. We really do appreciate your leadership, Khang.
Khang Nguyen (25:48):
Thank you so much. It’s been a pleasure speaking to you and all of our Permanente colleagues.
Chris Grant (25:53):
And thank you once again to our listeners for tuning in. If you’re a new listener and enjoyed this episode, please take a moment to subscribe, like or share a review. We’ll see you next time.
——————-
The opinions expressed on this podcast are those of the speakers and are not necessarily the views of Kaiser Permanente, the Permanente Medical Groups, or The Permanente Federation.
