Month: March 2021

The hard part isn’t the technology piece but making sure the experience is right enough for patient engagement

Season 3: Episode #80

Podcast with Kash Patel, VP and Chief Digital Technology Officer, Penn Medicine

"The hard part isn’t the technology piece but making sure the experience is right enough for patient engagement"

paddy Hosted by Paddy Padmanabhan
bigunlock-podcast-homepage-banner-mic

In this episode, Kash Patel discusses his role as the Chief Digital Technology Officer at Penn Medicine and provides an overview of their digital transformation initiatives covering all aspects of the institution, including research, academic programs, and patient engagement.

According to Kash, the hard part with digital transformation is not necessarily implementing technology but ensuring the patient experience is seamless with the technology and they feel positive about it.

While making technology choices, Penn Medicine’s first and foremost preference is to make the maximum use of their existing EHR infrastructure. Kash also describes the governance process that includes their leadership and subject matter experts to make technology decisions about newer digital tools and platforms. Take a listen.

Our Podcast Partners:    

Show Notes

05:04We are creating a roadmap for the journey to digitize our research platform that allows us to do all the prejudgments of the clinical trials, human subject trials, and ultimately manufacturing.
10:40The hard part wasn't the technology piece, but making sure that experience was correct and right enough for engagement.
11:03 We bring a lot of technology to the table and figure out what is the level of detail that a patient will tolerate, adhere to, engage with and feel positive about the experience.

About our guest

Kash Patel is the Vice President and Chief Digital Technology Officer at Penn Medicine. Kash has over 20 years’ experience in technology leadership ranging from startups to multi-national corporations and is a seasoned leader in healthcare with a strong focus on innovation and building great teams.

At Penn Medicine, Kash is leading the world class Pearlman School of Medicine Information Technology team. He supports the areas of research, clinical trials, manufacturing high performance computing, informatics, genomic science and many others. In addition, Kash is responsible for all of bespoke software development.

The team has been credited with several innovations that are streaming the complex business of healthcare. In 2020 there is a large focus on COVID-19 activities, working with Microsoft, Google and Apple to develop novel solutions that reflect Penn Medicine’s outstanding reputation.

After graduating in engineering from Sheffield in the UK, Kash gained valuable experience in building business driven software solutions in various industries. His earlier career started in consulting in the UK. He has managed global delivery teams for fortune 100 companies and started a new venture that created leading edge communications technologies.

In healthcare, Kash was the vice president for population health and analytics at Mount Sinai Health System where he led the technology strategy to support the institutions business shift from fee for service to assuming more risk. In addition, Kash was the IT lead for Mount Sinai’s New York DSRIP Program, involving over 250 partners led by Mount Sinai with over a $100M technology investment plan. He also managed health systems analytics and data engineering functions where he developed analytics as a service using advanced technologies.

About the host

Paddy is the co-author of Healthcare Digital Transformation – How Consumerism, Technology and Pandemic are Accelerating the Future (Taylor & Francis, Aug 2020), along with Edward W. Marx. Paddy is also the author of the best-selling book The Big Unlock – Harnessing Data and Growing Digital Health Businesses in a Value-based Care Era (Archway Publishing, 2017). He is the host of the highly subscribed The Big Unlock podcast on digital transformation in healthcare featuring C-level executives from the healthcare and technology sectors. He is widely published and has a by-lined column in CIO Magazine and other respected industry publications.

Our role is to support community health centers and getting members educated about the vaccine

Season 3: Episode #79

Podcast with Dr. Rich Parker, Chief Medical Officer, Arcadia and Jennifer Polello, Senior Director of Quality and Population Health at Community Health Plan of Washington

"Our role is to support community health centers and getting members educated about the vaccine"

paddy Hosted by Paddy Padmanabhan
bigunlock-podcast-homepage-banner-mic

In this episode, Dr. Rich Parker and Jennifer Polello discuss their vaccination distribution program, challenges around vaccine hesitancy, and how they have overcome this. 

Community Health Plan of Washington is a not-for-profit health plan that caters to the underserved and non-English speaking communities. Their role has been to support the community health centers and getting members educated about the vaccine. One of the most successful vaccination outreach programs has been their text messaging campaign, with a nearly 80% success rate.

Arcadia works specifically with data aggregated from disparate data sources like claims, EHR, lab, state health information exchange, social determinants of health, and others. This longitudinal record of each member helps them risk stratify and identify gaps in care, thereby helping them with their vaccination outreach and distribution priorities. Take a listen.

Our Podcast Partners:    

Show Notes

04:36In terms of getting ready for vaccine distribution, our role really has been to support our community health centers around getting members aware and educated around the vaccine.
09:25 Having access to data, the demographic and EHRs combined with all of that other data, gave us actionable longitudinal picture of the member and enabled us to assist our community health centers for needs they might have.
16:41 We had nearly 80% success rate from our text campaign.
20:09I am very enthusiastic about using data to help push the vaccination rates up as high as possible.

Q. Jennifer, can you tell us a little bit about the Community Health Plan in Washington and the populations that you serve.

Jennifer: Community Health Plan of Washington is a not-for-profit health plan. We were founded by 20 federally qualified health centers almost 30 years ago. Now we’ve got approximately two hundred fifty thousand members through our Medicaid, Medicare, which also includes special needs population, and then Cascade Care, which is a new line of business for us that we started this year. Our network consists of one hundred hospitals across the state and one hundred and seventy-four clinics. We have more than twenty-seven hundred primary care providers and over fourteen thousand specialists. The thing that really makes our organization unique is that we believe in the power of community and our mission is to serve our members and our communities across the state of Washington.

Q. Can you start by telling us how long ago you had to start planning for the vaccine. Tell us about the program itself and what have been some of your challenges in rolling out the vaccine to populations?

Jennifer: The planning around vaccine distribution started well before we had an approved vaccine, in terms of looking at eligibility, what would be available and how we were going to roll this out. Our main emphasis was providing support to our community health centers across our network and trying to help them with any logistic challenges or education and awareness need. That’s where our partnership with Arcadia really took off. In 2020 we had great success using the Arcadia outreach module when the pandemic started for things like what exactly is coronavirus, where to go for testing, which evolved to benefit reminders and connections to our community program teams for referrals to social service needs for things like food security. So, we kind of laid the groundwork for all of this in 2020 when the pandemic started and using all of these resources and distributing that information via the outreach functionality within Arcadia. In terms of getting ready for vaccine distribution, our role really has been to support our community health centers around getting members aware and educated around the vaccine.

Q. Rich, tell us about your analytics and about your involvement in the vaccine outreach effort with Community Health Plan.

Rich: Arcadia is a company dedicated to assisting healthcare networks and to some extent, payers, mostly commercial payers, deal with all the disparate data that is out there. So, getting all the data together from different sources, whether it is claims data, electronic health record data, lab data, state health information exchange data, social determinants data, and aggregating it, cleaning it up, making it useful. And then we have a set of analytics that sits on top of that data for each of our customers, allowing them to succeed in what we call value-based care, basically improving the health of the community.

As for vaccines, I have always been interested in vaccines since I started medical school and fortunately, we have these fantastic vaccines available so quickly to help deal with this epidemic. Arcadia has worked with many customers in helping risk-stratify patients. That is figuring out who is at risk for COVID, educating patients as to when they should ask for care or when vaccines are available, where can they get them? And then also looking at gaps in care. So, if people who should have had a vaccine didn’t get it, we can identify those gaps and help our customers fill them.

Q. Where are you getting all this data from? Can you share a couple of insights that you were able to get from the data that helped you to enable Community Health Plan and Jennifer’s team to drive better outcomes or outreach?

Rich: Our main sources of data are from the electronic health record. We get a download of data that’s extremely up to date and we get it usually on a monthly basis and the claims data. Which means that every time a patient is seen either in a doctor’s office or in a hospital setting a claim is generated. That information comes back to us and we can use that to figure out what’s going on with the population. And so that information, for example, at CHPW where Jen works, it would allow us to understand, which zip codes are doing better with vaccination, which are doing worse, and where do we have to focus our efforts more accurately and intensively.

Jennifer: We relied heavily on our Arcadia Analytics platform during this time. We have 20 cases connected to the Arcadia platform. We have got our data from  all those organizations. We have also added ADT data, which is admin discharge transfer data. So, we get information from the hospitals and we also have a separate lab feed. So, all that data really allows us that longitudinal picture of what is going on with the member. This really allowed us to help the community health centers know, who has got care gaps, who hasn’t been seen, who’s at risk for COVID.

We got almost an 80 percent success rate with our outreach efforts, which is high considering we’re dealing with Medicare and Medicaid members that typically are a little bit harder to reach. And so, having access to that data, the demographic and contact information in the EHRs combined with all of the other data, really gave us that actionable longitudinal picture of the member and enabled us to assist our community health centers in reaching those numbers for whatever needs they might have.

Q. Can you tell us about the insights that you got from the platform and the added tools that Arcadia may have deployed? Anything that came out that surprised you or was in some way unanticipated and helped to really improve the outcomes that you were going after?

Jennifer: We’ve got lots of different registries available to us in the platform. And one of the most utilized is our patient registry and being able to sort that registry by members that are at most risk. So, we can sort by the highest risk members, we can sort by members that have a lot of care gaps or chronic conditions. All that flexibility within the platform allows us to tailor different outreach methods within the CHPW language preferences. That was one thing that came in handy over 2020 because we relied on Arcadia for outreach and translated into lots of different languages, which was helpful and used the accurate contact information and targeted those messages by zip code. So, there was not any one thing that stood out. It was kind of a combination of all the different functionalities within the platform that we were able to tailor to each of our centers’ needs at the time.

Q. Can you tell us about some of the challenges that you have to deal with when you’re pulling all these data sources together and what you’ve had to overcome to make your algorithms and your risk stratification models meaningful?

Rich: There are some countries that have a single health record for everyone in the country. And that, in retrospect, seems like a really good idea. But it’s not what we have in the United States. We have many EHRs and still have some people on paper, but most people are on some computer system now. And since it’s healthcare data, it has to be very accurate. Now, sometimes we have challenges with getting corrupted data or incorrect data that could come in the form of a claims file from a payer that has a problem in it. We have very sophisticated tools where we’re usually able to identify the issue with the data and quickly fix it. Healthcare data is complicated, but we have years of experience doing this and the analytics are only as good as the data source that sits underneath it. We spend a lot of time and effort to make sure the data is correct for each customer.

Jennifer: We have a team that works directly with the Arcadia team to ensure that data quality is up to speed and the integrity of the data is there. We have got lots of different connectors just in the EHRs alone. There’s 20 different data points and data connectors there which are with 20 different organizations and each time they make a change to their workflow, it could impact how the data gets back to Arcadia. So, it’s a constant management of the data with our centers to make sure that data quality and data integrity is first and foremost.

Q. You’re addressing a population that may not be as technologically enabled, especially if you are talking about lower income populations. So, what kind of modalities do you use in your outreach as far as the vaccination program goes?

Rich: One of the big learnings for us is that text outreach is the best way to go. In the old days, we were sending letters, then we were making phone calls, and then a lot of people switched from landlines to cell phones and then a lot of people got a lot of junk calls and stopped answering their phones. But we have learned through experience that most people look at texts on their cell phone. We have now sent over three million text messages out to patients on behalf of our customers, all healthcare related, a lot of it around vaccine education, gaps in care for text messages. The messages are short. And the other thing we have learned about text messages is we can embed a URL. So, for example, if there’s a longer message that wants to get out to its patients, they can put a URL in there and the patient or the customer can click on that and get the website that CHPW wants them to see.

Q. How does that work for you, Jennifer?

Jennifer: It worked exceptionally well. As I mentioned, we had that nearly 80 percent success rate and that was from our text campaign. So, I completely agree with Dr. Rich that text is the way to go. We had very good success rates in reaching our members, and the ability to embed additional information was helpful, because you can send a real short message and then have links for additional information. In terms of vaccine, education, and awareness, we were able to a link to our state’s Phase Finder to help folks understand when it was their turn to get the vaccine. And then over the last year, it was great for driving folks to testing locations and benefit reminders. It was invaluable in terms of directing members exactly where we wanted them to go.

Q. We very often hear about vaccine hesitancy and a significant percentage of the American population do not want to be vaccinated or have concerns. Is that a problem at your institution?

Rich: Absolutely. Vaccine hesitancy is a big problem. Probably right now, about a third of people in the United States are reluctant to get the COVID vaccine. We know that people that are over the age of fifty-five are more likely to accept the vaccine. Younger people are less likely to accept the vaccine. And we can use our data to figure out where the gaps in care are.

Jennifer: I think we we are pretty much in alignment with that as well. We serve the underserved and a lot of those communities are non-English speaking and have different cultural beliefs. So, there’s a lot of education and general awareness that needs to take place. Our strategy has been to reach out to those community leaders as potential models or potential leaders that can help distribute vaccine education and awareness information.

Q. Can you share one best practice for your peers in the industry? One each based on your experience, especially around vaccines.

Rich: I am very optimistic, very enthusiastic about using data to help push the vaccination rates up as high as possible. I would say that without data, you are just operating in the dark. You have no idea what is going on with data. With data, we are not going to get perfect compliance, but we will find out who has been vaccinated, who still needs vaccines, and then we can target our outreach to the people that are still outstanding so we can do the best possible effort to get as many people vaccinated as possible.

Jennifer: Dr. Rich, I completely agree with that. Arcadia platform and access to this integrated data and access to this longitudinal record of our members, the ability to sort by risk and look at care gaps to find out who had the first vaccine, who needs the second one, all those different functionalities really allow us to be in the forefront and at the top end of the curve on reaching our members. I think the key is that we got timely data and data that really connects us back to the needs of our members and our communities.

About our guest

Dr. Parker serves as Chief Medical Officer for Arcadia with overall responsibility for the design and implementation of clinical strategies, input into the roadmap and development of Arcadia’s technology and service programs, thought leadership in support of providers transitioning to value-based care, and strategic advisory work for physician leaders at Arcadia’s clients.

Previously, Dr. Parker was an internist with a 30-year history at Beth Israel Deaconess Medical Center. From 2001 until 2015, Dr. Parker served as the medical director and chief medical officer for the 2,200 doctor Beth Israel Deaconess Care Organization. He oversaw the physician network evolve from a fee-for-service payment system to a nationally recognized global payment pioneer Accountable Care Organization.

Dr. Parker’s other areas of expertise include end of life care, medical malpractice, care of the mentally ill, electronic medical records, and population health management. Dr. Parker served as assistant professor of medicine at Harvard Medical School. Dr. Parker graduated from Harvard College in 1978, and the Dartmouth-Brown Program in Medicine in 1985 Dr. Parker is an in-demand speaker to associations, companies, and academic institutions on the topics of population health management, electronic health records, value-based care, and evolutionary, medical and business impacts of stress.

Jennifer serves as the Senior Director of Quality and Population Health at Community Health Plan of Washington. She has over 20 years of extensive experience across the healthcare continuum in the areas of public health, chronic disease management, quality improvement, health policy, population health management and clinical informatics. She has exercised this experience from several points of view across the health care environment and has demonstrated expert facilitation skills in leading teams of clinicians, nurses and physicians through the transformation process of patient care in the ambulatory setting.

Jennifer has worked on regional health information exchange projects and assisted in the design of a clinical decision support tool for patients with type 2 diabetes. She has also served as an Adjunct Clinical Assistant Professor mentoring PharmD candidates at Washington State University.  

Jennifer is currently leveraging her knowledge and expertise as the Senior Director of Quality and Population Health at Community Health Plan of Washington where she leads the company’s quality improvement strategies, population health and clinical data integration programs across the Network of 20 community health centers that operate more than 130 clinics across the state.

About the host

Paddy is the co-author of Healthcare Digital Transformation – How Consumerism, Technology and Pandemic are Accelerating the Future (Taylor & Francis, Aug 2020), along with Edward W. Marx. Paddy is also the author of the best-selling book The Big Unlock – Harnessing Data and Growing Digital Health Businesses in a Value-based Care Era (Archway Publishing, 2017). He is the host of the highly subscribed The Big Unlock podcast on digital transformation in healthcare featuring C-level executives from the healthcare and technology sectors. He is widely published and has a by-lined column in CIO Magazine and other respected industry publications.

Tech firms must build software that aligns with patient demographics, is usable for them, and delivers outcomes.

Season 3: Episode #78

Podcast with Josh Goode, Chief Information Officer, SCAN Health Plan

"Tech firms must build software that aligns with patient demographics, is usable for them, and delivers outcomes."

paddy Hosted by Paddy Padmanabhan
bigunlock-podcast-homepage-banner-mic

In this episode, Josh Goode, CIO of SCAN Health Plan, discusses their digital programs, the patient population they serve, and how they evaluate digital technologies and deploy it at every stage of the care journey for improved outcomes. 

Being a Medicare Advantage plan, SCAN Health deals with the senior population. They strive to address the digital divide in the elderly by implementing software that aligns with their requirements, is easy to use, and delivers improved outcomes while taking care of patient privacy and data security. The technology considers the social determinants of health by implementing a robust data and analytics program that has helped develop AI models to predict chronic conditions. 

Josh also talks about person-centered design processes and how it helps deploy the right digital technology by looking at the patient’s journey touchpoints. Take a listen.

Our Podcast Partners:    

Show Notes

06:39There's a lot of technology and capabilities out there that we can be deploying. But how do you know you're deploying the right one?
07:25 We are very focused on person-centered design.
10:33 When you're dealing with Medicare, the senior population, you got to design the experience with that population in mind.
17:29I think it's our imperative to use data to help influence care, and help improve the service experience as well.

Q. Can you tell us about SCAN Health Plan and the populations that you serve?

Josh: SCAN Health Plan is a Southern California based Medicare Advantage plan and we have been in this business for about 40 years. SCAN Health started out as a social HMO and now we are the third largest Medicare Advantage plan in California. Although our firm is a regional plan, but we do have a strong national presence. We are a leader in a lot of the Medicare Advantage metrics. Our scores are usually near the top of the industry. As per customer satisfaction, we have a strong performance in the star rating and are pretty active from a policy standpoint. We recently got a new CEO Sachin Jain and he came on board about seven months ago. And when you look at such incoming on-board SCAN Health Plan, we’ve had a good foundation, really stable, solid company to the metrics. And now we’re really looking at how do we capitalize on that? How do we build up on that and expand upon all the good work we’ve been doing for seniors across California?

Q. How many lives you cover today approximately and what kind of digital programs have you rolled out in the last couple of years at SCAN Health?

Josh: We’ve got about two hundred and twenty-five thousand listed.

In the last couple of years, we have been focussing on consumer-facing technologies and how do we improve that consumer experience. When I joined SCAN Health, we were more focused on a technology modernization program, a lot of our core admin systems were outdated. They were no longer supported by vendors. Also, our primary core admin systems that does a lot of our administration was written in a programming language RPG. We were hundred percent on premise. So, I put in place a technology modernization program, replacing all those core systems and moved to a SaaS-based model. As all the systems are now moving into a cloud environment, we pivoted our focus more on the consumer facing technology.

We have also built our self-service capabilities, to try to minimize the amount of phone calls we get. To enable our members and our seniors, we have provided online channel options to use. Also, we’ve been doing a lot around data and analytics, advanced analytics. Interoperability has been something we have been doing a lot around last year. And with the new CMS interoperability rule, we are excited. We are really trying to unlock data sharing and trying to focus on our contact centre, our touch points with our members in driving innovation and using technology and data to support those areas of our company.

Q. On the provider side, as well as on the payer side of the business, for instance digital front doors, can you talk about what kind of specific high impact features or functionalities or solutions you’ve launched? How is it making a difference, how do you pick what to deploy, and how do you track whether it’s working?

Josh: There’s a lot of technology and a lot of capabilities out there that we can be deploying. So how do you know you’re deploying the right one? The answer is we like to do journey mapping. Looking at what is that member experience, that constituent experience and looking at what are those touch points that we have, what are the areas that have pain points. We call them – the moments that matter and we look at how we can apply the technology to help solve the issue.

Also, we are very focused on a person-centred design. We use our member advisory committees to give us guidance on things that we need to be working on and have our members, our seniors inform us on the things that we think we need to be working on and focusing on. Then we always have a heavy focus on caregivers and really increasing their abilities has been an area of focus for us.

The other thing that had come up around our members was multifactor authentication on our member portal that is protecting their data. To look after this concern, for us, it starts with that journey mapping of looking at what is the experience and what are those pain points that we can solve with technology.

We have a fully integrated broker portal and we look at their experience of interacting with SCAN Health Plan. By streamlining that, we are making it easier to do business with us from a broker perspective. With our providers we are trying to provide them with better experience as well. So, to sum up, our providers, and our brokers, both can provide our members with a better experience.

Q. You are serving multiple constituencies and your population might not be ready for some of the digital technologies and tools. Can you tell us about one or two unique things about your population that you had to take into consideration while designing these solutions and experiences?

Josh: It’s really not a one-size-fits-all. When you’re dealing with the Medicare and the senior population, you got to design the experience with that population in mind. How do you make it simple? How do you make it easy to utilize those technologies? And that’s something we strive to do with our website and any of our touch points with our members, whether it be telehealth, whether it be doing a virtual visit and even getting that virtual visit invite over to our members.

One of the things we’ve seen in particular with COVID-19 is digital adoption skyrocketed even among the senior population. We saw our member portal registrations go up over 30 percent and these are not one-time registrations. We have also seen our virtual visits, our video conference visit dramatically increased. When the pandemic hit, me and my team brainstormed for solutions and tools to help solve the digital divide with seniors and quickly rolled out a member technology support line. We were able to get it up and running for about three weeks but then after the pandemic really kind of took hold as everybody started going into our virtual environment. And the success out of that is when you look at what happened with the pandemic, you and your members were really thrust to get into a digital environment and some of them were ready for it. You can’t generalize the senior population, some are very digitally savvy and some were not ready for it.

I will never forget our first call which was a forty-five-minute call and was very impactful. It was with a 92-year-old member who was calling. His provider health system had sent him a text to do a virtual visit and you need an email address to register to do it. He never used email and we helped him set up an email account and walk him through how to do that virtual visit. We’ve had a number of stories since then, but it’s something that we’ve been proud of to help solve that digital divide.

Q. In the immediate wake of COVID-19, everybody saw a spike in virtual visits. But now all the data points to the fact that virtual visits are flattening out as patients start going back to clinics and hospitals and there’s a slight of pent-up demand. What are you seeing?

Josh: Something as an industry we have been able to demonstrate through the pandemic is that we can operate in a virtual fashion, but obviously not all care can be delivered virtually. But we have built that trust with our members and our patients, that there are effective ways to service those individuals virtually. And so, we are seeing that it is still running at high levels, but not as high as the peaks that we’ve had in pre-pandemic. So, as we start to open more, those numbers will continue to drop. But what we’re seeing and hearing from our member patient population is that they will continue to use virtual services for select interactions.

Q. Can you talk about your data and analytics program and how you’ve harnessed some of the social determinants of health?

Josh: Data and analytics is something that has always been one of the strengths for us. Under my purview, we manage the architecture, the data infrastructure, the tools and the healthcare informatics department. When you look at our role as a payer, we’re really a data aggregator where we’re getting all the data and we’re using that data to help influence care and help improve the service experience as well.

Also, we’ve got a centre of excellence that we run and enable all of our different business departments around the company and give them the tools to develop their own analytics. So, they can have data at their fingertips to make decisions and serve it up to the leaders in their departments. Also, more recently, we’ve moved into advanced analytics, leveraging AI, machine learning, where we’ve been selective on the use cases we target with AI. It takes more care and feeding as compared to traditional analytics. To make sure you’re focusing on the right use cases for AI and having the right processes in place, we need to be very focussed on our use cases to really improve our ability to leverage data and gain insights on data.

Social determinants of health have always been a priority focus of ours. So, starting out as a social HMO, we were really focused on SDOH. We use a bit of the external sources but have always maintained a good history of information and had the ability to collect that information directly from our members through a variety of means. We’ve been able to develop a pretty rich repository of SDOH data that we’ll leverage across the board. Those clinical models are very effective in looking at predicting some chronic conditions and potential clinical outcomes. We’ll be able to improve the care we deliver by using that data and then coupling it with the robust analytics program.

The last thing I would say on analytics is something we’re really focused on is real time analytics. But with the advancements of technologies, the replication technologies and with the CMS interoperability rule, we need to more tightly integrate with our provider network. This is because we are getting real time data straight from our health systems provider network. And we’re able to take that data and feed it across a rich and robust analytics program to really drive more outcomes as well.

Q. What is your advice to the tech firms and start-ups who are looking to be a part of your journey?

Josh: My advice to those start-ups and tech firms around is, make sure you’re building software that is aligned to your demographic population. As a Medicare Advantage CIO, I see it all the time that we present the software that is not geared towards the senior population. So, make sure you’re engaging that demographic using person-centred design, organising workshops with them, getting their feedback etc. Also, make sure you are building the software that’s going to be usable for them and is going to deliver outcomes.

The last thing I would say is, as a healthcare CIO, we’re all under attack from a cybersecurity standpoint. And even today, you still see a lack of adoption around information security. In the near future, if you’re not a high trust certified vendor, you’re going to have a tough time operating in the market. So, make sure you have a security focus as well.

Q. How did your consulting background prepare you for the CIO role? And what advice do you have for others in the consulting world who want to make a transition?

Josh: So, my background before becoming a CIO was exclusively working in the consulting industry. What really made me a well-rounded individual in my career is learning strategy work, which helped me understand how I need to develop strategies for organizations. Also, I learnt a sizable amount of system implementation work, leading a large system implementation and designing the operating model.

The thing that the consulting background really prepares you for is having that mindset of being able to design an operating model where you can put people in place to be successful and allow them to be able to execute on the strategies that you developed.

About our guest

JOSHGOODE-profile-pic

Josh Goode is Chief Information Officer at SCAN. He provides leadership, direction and support to the company’s information technology (IT) areas including Digital Strategy, Business Intelligence, IT Infrastructure, Project Management, Electronic Data Interchange (EDI) and Application Development. Under Josh’s guidance, SCAN is leveraging its technology investments to meet the individual needs of seniors now and in the future.

Prior to joining SCAN in 2013, Josh worked for Accenture, a multinational technology and management consulting firm. During his 15 years at Accenture, he worked with several health plans throughout the United States, including PacifiCare, CIGNA, Express Scripts and UnitedHealth Group.

His experience includes analyzing, planning and implementing a variety of technological improvements and leading large technology programs, such as systems implementations and IT transformations.

Josh holds a Bachelor of Science in Business Management from the University of Tennessee.

About the host

Paddy is the co-author of Healthcare Digital Transformation – How Consumerism, Technology and Pandemic are Accelerating the Future (Taylor & Francis, Aug 2020), along with Edward W. Marx. Paddy is also the author of the best-selling book The Big Unlock – Harnessing Data and Growing Digital Health Businesses in a Value-based Care Era (Archway Publishing, 2017). He is the host of the highly subscribed The Big Unlock podcast on digital transformation in healthcare featuring C-level executives from the healthcare and technology sectors. He is widely published and has a by-lined column in CIO Magazine and other respected industry publications.

Voice Technology in Healthcare: Enhancing Patient Interaction

Season 3: Episode #77

Podcast with Dr. Stephanie Lahr, CIO of Monument Health and Peter Durlach, Chief Strategy Officer of Nuance Communications

"Voice Technology in Healthcare: Enhancing Patient Interaction"

paddy Hosted by Paddy Padmanabhan
bigunlock-podcast-homepage-banner-mic

In this episode, Dr. Stephanie Lahr, CIO of Monument Health, and Peter Durlach, Chief Strategy Officer of Nuance Communications, discuss voice recognition technology, its demand and adoption level in the marketplace, emerging use cases, and the next stage of improvements in the technology.

Dr. Lahr believes that voice recognition technology is the best way for patients to interact with their providers as opposed to any other user interface in the future.

Nuance is eliminating the burden on clinicians by powering virtual consults with voice-enabled clinical documentation. According to Peter, other aspects such as scheduling appointments, preparing for a doctor’s visit, medication adherence, and other clinical and non-clinical use cases can improve patient experience through personalized voice-based interactions. Take a listen.

Our Podcast Partners:    

PP: Tell us about the voice recognition technology marketplace and the environment for adoption of the technology. What have been the broader learnings so far and what are the next round of improvements likely to be in voice recognition?

Peter: The move from directed voice where you say exactly what you want of a more intelligent listening system, that we call ambient, is really the next generation of voice technology. In healthcare, long before we had the COVID pandemic, we have had the pandemic of administrative burden, overwhelming clinicians, as you probably know, they spend roughly twice as much time taking care of administrative requirements as seeing patients. This has resulted clinicians feeling burnt out, depression going up, and people retiring. It has really been a crisis for healthcare. What they all want as part of the solution is why cannot they just focus on the patient, have a conversation with the patient and use that conversation in the encounter, whether it’s in a physical setting or a virtual setting, and then have the technology to create the documentation for them and do other things like coding and things that they see as taking away from taking care of patients. We have a solution called the Dragon Ambient Experience that we launched right before COVID that does exactly that. We are still early in the journey around ambient, but it’s a very exciting area. In terms of the learning so far, I would say that different physicians have different requirements. So, we roll this out by specialties because you must build these ambient models by specialty. And we have learned that for some clinicians the technology fits in extremely well. For others, they may be looking for the technology to morph in a little way based on the workflow that they have. So as time goes on and as the technology matures, I think you are going to see more and more physicians across a wide degree of specialties really adopt. The results so far in general have been quite good. We are generally, on average, seeing a reduction in clinician burnout for those who are using it. Dropping from about 72% to 17% and freeing up about six minutes per encounter for an average panel size of twenty to thirty a day. That’s a couple hours a day. So, as the technology gets better and better, faster and faster, and needs less and less human involvement in reviewing the note, I think you’re going to see higher adoption and also the development of more automated things coming out of the note like coding and quality abstraction.

Stephanie: I will just add a couple of comments. This is one of those fun times when as a CIO and CMIO, I get to deliver a tool to my clinical colleagues that they are excited about and really makes their lives and their interaction with the patients better. I mean, it’s unfortunate, but true. A lot of the things that we have done and had to do over the last decade or so add to that burden that Peter referred to. And again, this is one of those times where I have a waiting list of people who are excited to try this. I do think that there are certain specialties where this is going to lend itself more to, at least early on, because there are some workflow elements of this. We see our colleagues really being able to take this and fly. And that is great for them, because to Peter’s point, if at the reduced time and documentation and some of those other things, gives them more time to do procedures and be in the operating room and do those things that really impact the improvement of the lives of the patients. The other thing I would say is we are still helping our patients get used to this construct. Most patients are very accepting of this idea of the conversation happening and then the note being created. They are accepting it because they’re anxious and excited to get the intimacy of their relationship with their provider back. But it is still an education point that we must help our patients understand kind of what this is about and what it means and what it does not mean.

Q. So, for the non-native English-speaking populations, there could be an issue with the technology. That is certainly the case with a lot of personal tech. Is that something the technology is beginning to overcome and what are the pros and cons?

Peter: Yeah, that is great. So, the core product in this space for the non-radiology clinician is a product called Dragon Medical One, which is a cloud-based dictation system which approximately 60 percent of all physicians in the U.S. use. The product already supports over twenty-five languages with incredibly high fidelity. It’s sold worldwide today on that front. In terms of the ambient part, as we do these more colloquial conversational stuff that’s earlier in the journey. So, we’re in the process now of morphing that into support. Multiple languages today that is focused on English. But we have a lot of demand, as you would expect, to start penetrating other languages like Spanish, et cetera. So that’s on the roadmap today. So, depending on which product you’re talking about, we either have wide coverage of that or we’re on the early stages for the newest ambient piece.

Q. What are some of the high impact, high value non-clinical use cases for voice technology in healthcare?

Peter: One of the hot areas which crosses both clinical and non-clinical is in the patient engagement space. As you know, like many industries, healthcare is now taking this idea of using digital technologies to redefine what is often called – the digital front door. So, how do you access care and how do you follow up on care using technology? From Nuance perspective, we are a leader in what’s called omni-channel virtual assistant technology. We power these sophisticated applications for companies like FedEx, Disney, American Airlines, when their consumers interact over our telephone line for an interactive voice response system or a chatbot or on a social system like Facebook Messenger. So, we have started to bring all that into the healthcare arena for healthcare providers as they look to do things like manage booking your appointment. Did you prep for your clinical visit? Are you taking your meds? All of these are both clinical and non-clinical use cases.

We are the leading provider of voice biometrics technology for user identification and prevent fraud. This has generally been used in the banking financial services industry for obvious reasons, and we’re bringing that into healthcare now. So that’s a non-clinical thing that we’re seeing. And the last case, which is clinical but is exciting, is there is a whole set of companies that are building technology to use voice to help with diagnosis of clinical conditions. There are companies that are using voice for clinical depression screening, for example. You may have seen recently some COVID screening. So, this idea of using the acoustic signal to predict or screening is at the early stages but is something super exciting for us. We are looking at expanding the capabilities of what voice can do beyond the core use today, which is really for documentation of the clinical encounter.

Stephanie: Authentication passwords and the security of our systems is one of those things that a CIO does keep you up at night. There are just too many systems we all have to be in and out of and so we take shortcuts on the utilization and how we reuse and those kinds of things with passwords. So, I love the idea of voice for authentication on the clinical delivery side. On the patient side, for example, in their homes, as we are looking at that, breaking down of the boundaries of where care occurs and trying to identify that best location to help some of our patients in their home. The best way for them to be able to interact with us will be through voice technology as opposed to any other user interface that may just not be conducive based on some of their limitations. We’re already seeing things where people can set up medication, reminders and things like that, but taking it to the next step of really almost having an attendant at home, a healthcare attendant at home and leveraging voice in that interaction. I think these are some of the exciting pieces in addition. We are starting to utilize the elements of IVR and texting to help improve some of the patient experience elements that are high volume and allow us to be more efficient within the utilization of our in-person resources. And we really see it as a blend. Again, to maintain that intimacy of the relationship. We could start off with some of these automated tools. If they are voice driven, that is then more personal than something else. And then we can hand off to a real person when we get to some of the more complex related things. So, lots and lots of exciting opportunities, I think, with voice.

Q. Peter, you’ve been on an acquisition spree, and most recently you acquired this company, Saykara. Tell us a little bit about where that fits into your overall product roadmap, generally in your acquisition strategy?

Peter: Recently, we had a new CEO come in about three years ago. Historically, we had done a lot of acquisitions and we slowed that down a little bit. But we did acquire Saykara and its really interesting. So, Saykara was founded by a guy named Harjinder Sandhu, who is a close friend of mine. He used to be the healthcare CTO at Nuance about a decade ago, a very sharp guy. And after he left Nuance he went off and started a patient engagement company with a partner of his and then kind of came back to his roots, which is really around the clinical documentation space. They started a company called Saykara to really try to do things like what we were doing with our Dragon Ambient Experience. So, we have been keeping in touch with our agenda over the last couple of years. And recently as they were about to go for their next round of financing, we discussed with them together what happens if we try to combine forces because we really all have a passion for solving this really big problem to help our clinician friends and clients, which is this idea of taking a colloquial conversation and turning that into a highly accurate, structured summarize note with a set of extracted data using a language that often wasn’t even discussed in the conversation explicitly, is a really hard technical task. There are very few people in this world that actually have experience trying to solve this problem. Harjinder’s team had relevant experience. So, for us, it was really an attempt to take our incredible team that we have with DAX and supplement it with the great team that Harjinder had put together and combine that into one journey together with a common mission. That’s how we came together to do this. And so, all the Saykara team are going to be working on our combined DAX effort and we’ll look to integrate components of their technology where appropriate and really try to attack this really important moonshot that we’re all after this in this ambient world. So that’s really what the purpose was of that acquisition.

Q. How has COVID-19 impacted the demand or the adoption for voice-based solutions?

Stephanie: It’s been a really interesting journey over this last year. I think from a technology perspective, one of the silver linings of this pandemic has been the rapid deployment and adoption of a variety of different technologies, some of which was telehealth. We all saw this massive uptick in telemedicine. And we did it in a constructive way. Most of our organizations, for a variety of reasons, didn’t have a ton of experience or a deeply embedded telemedicine infrastructure before that. It really is a different way of delivering care to a patient. It takes practice and experience and a little bit of a different format in order to have a high-quality telemedicine experience with a patient. So, one of the things that we saw was that we had a provider who was sitting in front of a computer, sometimes in front of two computers or multiple screens. Depending on whether the video was integrated, they still needed access to the EHR. They wanted to look at the patient. They needed to create their documentation. And the patient was sometimes looking for additional assistance and kind of how to maneuver through this. It was overwhelming at times to be able to figure out how and what to concentrate on. So, for example, with DAX, it was a great use case. We already knew that it was going to be amazing to take this ambient technology and have a conversation in the background in an in-person interaction, because we want to solve the problem of the documentation burden. But the documentation burden was compounded in this telemedicine environment where we did not have a good way to be able to look at the document, talk to the patient and use the technology at the same time. If the documentation was writing itself while we were having the conversation and I was managing the technology with the patient, what a huge win that was. So, definitely we saw that in telemedicine. And then the other piece, I think was huge was we saw really rapid and sometimes very difficult to predict changes in demand. For example, our nurse call center would at times get a hundred calls in a day and then the next day, with the same staffing plan in place, would receive seven hundred calls in the day. We don’t know exactly which day is it going to happen. One hundred or the seven hundred. It was variable depending on everything else that was happening in the environment, what was coming out in the news, all of those kinds of things. So we began to see that maybe automation with voice was a tool set that we could use to help us get through these high-demand periods. Again, allowing the people who needed to do the in-person work to be able to focus on the highest and most complex elements of DAX and let the voice and other elements maybe be able to help patients who didn’t need that higher level support. And so lots and lots of use cases started to come out around where we could leverage voice to get through this high demand situation where none of us had enough resources.

Q. What have you been seeing in the rest of the healthcare ecosystem?

Peter: On the voice side, specifically, the two big things that have exploded are exactly aligned with what Stephanie did. One is that the clinicians, docs, nurses, et cetera, are not in great shape from a burnout perspective before COVID. And obviously it has been absolutely overwhelming for them. So the demand for anything that could help them get through the day has really exploded, whether it was DAX, our dragon ambient product, or even with our dragon cloud moving more expansively of that. I mean, we had things where a lot of these field hospitals, we work with Epic and others to stand up a whole voice enabled system for field hospitals and a few days in multiple cities. So, there was that whole sort of tools for clinicians to try to reduce as much of this other stuff as possible while they were trying to take care of patients. Stephanie also said the inbound flux of patient requests around prepping for a telehealth visit, trying to log into their portal now about getting a vaccine, have just overwhelmed the [health systems. Most health systems don’t have the infrastructure to deal with it. And as Stephanie said, they certainly don’t have the dollars to fund people to do that. So, our digital patient engagement technology that allows you, just like the website or text channels provide these automated systems to do a lot of the basic lifting for them, both on inbound interactions and outbound. We just also signed deals with several large, major pharmacy chains in the country that are rolling out the covid vaccines now and doing centralized virtual assistant front ends to their scheduling system so people can call up, find out what vaccine are they eligible that actually book an appointment. So, all of that sort of exploded both on the provider side, but also on the health plan and retail pharmacy side as a result of COVID.

Q. What is holding us back from a faster adoption of voice recognition technology in healthcare? What do you think are the big challenges that we need to overcome?

Stephanie: I think the documentation is amazing. We want to do that and help improve that area of the satisfaction of our providers. But now we got to go further than that. One of the challenges that we still need to overcome is the amount of medical information to understand, digest and then utilize in the care of patients is increasing exponentially. We need tool sets that can help us access the relevant information or even provide reminders. So, I really want to see us go beyond documentation and doing things. For example, the relationship between our EHR vendors and the voice recognition side of things so that we can completely eliminate any kind of user interface with our providers that requires a keyboard that is the ultimate goal. Get rid of the keyboard altogether and let me have an interaction, a voice based interaction with the patient and a voice based interaction with the technology. If I need to know when the last CT scan was, let me ask it instead of typing and looking it up. If the system is listening and thinks there’s a piece of relevant information that I should know about as I look into placing an order or creating a plan, tell me about it proactively or alert me that there is potentially a clinical decision, support information that I need in order to make those things happen. The integration between the EHR which now sits at the center of these tool sets and the voice recognition side is an absolute requirement.

Peter: I think Stephanie really hit it on the mark. She really touched on two critical points. So, in terms of driving adoption, obviously the core technology from folks like us has to be good enough to be use number one. I think we have done a good job of that. There’s obviously room for improvement. And certainly, we’re still early on the journey of ambient world. But the early indications are positive. As Stephanie mentioned, there’s two other key points. One is we integrate into other systems. So, the more integrated and more natural the integration is, the better the adoption will be. We’re working with folks like Epic, Cerner, Meditech and others. All these virtual assistant technologies they’ve launched under the names like Hey Epic, Hey Cerner, Hey Meditech are all powered by Nuance technology. So, we’re working with the major EHR vendors to integrate that. So we can do exactly what Stephanie said, which is to basically have a virtual assistant. Every physician and all of us would love to have an intelligent virtual assistant or a physical assistant that works with us. What we’re trying to do is we’re not going to mimic everything a human can do, obviously. But there are a lot of tasks that if you had an assistant, you would really be much more productive. So, this idea of being able to ask what’s the latest CT scan or queue up in order or send a follow up note to the primary care physician, they should be able to do that very seamlessly by voice, as if they were telling their assistant to go do that. Number two, this idea of turning the system into an intelligent system. And for clinicians, generally, what intelligence means is you don’t have to ask it for something. It’s going to recognize something and tell you. So, Stephanie’s example of clinical decision support is a clear one if we’re listening to the conversation of the patient and we know their clinical history, we know what their primary diagnoses are and we hear something, why shouldn’t the system be able to say, oh, it looks like they may be discussing “X’ based on the clinical indicators. This might be a good thing for you to talk to the patient about. That is what’s going to come here as more AI gets deployed to predictive analytics and predictive to clinical decision making. That is really the holy grail here, which is you’ve got a virtual system that can do tasks for you, but it also provides you clinical advice as if they were like a resident or a fellow working next to you.

Q. If there is a best practice that you’d like to share with you as an industry, we’re listening to this podcast. What would that be?

Stephanie: I think the key here is think about the challenges that you are facing organizationally on the front end and the back end. There are so many use cases for this and ask yourself if voice could be a solution for that to help in the efficiency cetera of solving the problem. There are a lot of places that we could leverage voice driven technology that is going to be different than our typical construct. I think it will be effective if we’re willing to be open minded and ask ourselves that question, could voice be part of the solution? And I say part of the solution as my second best practice, which is if voice can lend to the improvement of a situation, it does not mean it has to be the complete answer. So I think the example that I gave earlier about an IVR, having an initial interaction, authenticating a patient, confirming what it is that they’re asking for, and then potentially handing that off to a live person for the more complex parts of what might need to be done is a way to be able to move forward in increments and really start to see progress, knowing that we don’t have to solve the entirety of the problem with one solution, but it might be one of the building blocks. So, think about voice, but don’t expect that to have to be all end all of everything. It may be a component.

Peter: Pick problems that are important to solve and then really get clinician owners and champions inside the organization. In healthcare, there’s a lot of technology promises. And often if you don’t set yourself up for success, no matter how good the tech, it’s not going to be successful. So, I think best practices is having alignment internally, goals we are trying to achieve, metrics we’re trying to drive. Some of this stuff also involves some business process changes and integration with the how to optimize these things. I think having both the supplier of the technology and the internal stakeholders aligned on the objectives, the internal alignment of what we’re trying to drive through and then sort of end-to-end view of how we’re going to solve this, including things that may not be voiced, is really critical so that when you do that, we can absolutely drive meaningful outcomes. We have thousands of examples and material improvements in physician and nursing, satisfaction reduction of administrative time, better financial outcomes, better patient experience. So, if you do it right and you focus on a problem that we can solve with the tech, we can make a difference.

Show Notes

02:36The move from voice to a more intelligent listening system that we call ambient is really the next generation of technology.
08:30 Healthcare is really now taking this idea of using digital technologies to redefine what's often called in healthcare, the digital front door.
11:47 We're already seeing things where people can set up medication, reminders and things like that on voice technology. The next step is having a healthcare attendant at home and leveraging voice in that interaction.
16:19I think from a technology perspective, one of the silver linings of this pandemic has been the rapid deployment and adoption of a variety of different technologies, some of which was telehealth.
28:59In healthcare, there's a lot of technology promises. Often if you don't set yourself up for success, no matter how good the tech is, it's not going to be successful.

About our guest

Stephanie-Lahr,-MD-profile

Stephanie Lahr, MD is the CIO and CMIO at Monument Health, a South Dakota healthcare provider operating clinics, regional hospitals, senior care, surgical units, institutes, and acute care. Stephanie works on the strategy, implementation and management of technologies within the health care system.

Stephanie graduated from The University of Texas Medical Branch at Galveston, completed an Internal Medicine residency, and is board certified in Internal Medicine and Clinical Informatics with an additional certification through CHIME as a Certified Healthcare CIO.

Peter Durlach is the Chief Strategy Officer at Nuance Communications. Peter holds a pivotal role in advancing the portfolio of healthcare solutions to align with industry pressures and shifting needs of healthcare clients. He helped create the Healthcare division and drive significant growth between 2006 and 2011 and then briefly left the company to act as the Entrepreneur in Residence at the University of Pittsburgh Medical Center.

Prior to Nuance, Peter worked as a consultant, president of Unveil Technologies, Inc. and vice president of marketing and business development at Lernout and Hauspie. He graduated from the University of Vermont with Summa Cum Laude honors where he received his B.S. in Business Administration.

About the host

Paddy is the co-author of Healthcare Digital Transformation – How Consumerism, Technology and Pandemic are Accelerating the Future (Taylor & Francis, Aug 2020), along with Edward W. Marx. Paddy is also the author of the best-selling book The Big Unlock – Harnessing Data and Growing Digital Health Businesses in a Value-based Care Era (Archway Publishing, 2017). He is the host of the highly subscribed The Big Unlock podcast on digital transformation in healthcare featuring C-level executives from the healthcare and technology sectors. He is widely published and has a by-lined column in CIO Magazine and other respected industry publications.

The pace of innovation and development of AI tools is outrunning the FDA and other regulators’ ability to stay on top of AI innovations

Season 3: Episode #76

Podcast with Casey Ross, National Technology Correspondent, STAT News

"The pace of innovation and development of AI tools is outrunning the FDA and other regulators’ ability to stay on top of AI innovations"

paddy Hosted by Paddy Padmanabhan
bigunlock-podcast-homepage-banner-mic
In this episode, Casey Ross, National Technology Correspondent at Stat News, discusses his recently published report on FDA-approved AI-enabled tools. These are Software as a Medical Device (SaMD) tools that work as decision support tools to supply patients’ data to physicians and help them diagnose and treat the patients. Data is the core ingredient that AI tools use. As per Casey, one of the major issues prevailing in the industry today is that there are inadequate disclosures on data sets used by many medical devices and algorithms approved by the FDA. To improve healthcare outcomes, transparency and disclosure in date sets must be the central agenda in future. He further states that the pace of innovation, development, and building process of AI tools is outrunning the FDA and other regulators’ ability to stay on top of the AI innovations. Take a listen.
Our Podcast Partners:    

Q: Can you talk about the report you recently published highlighting the possibility of racial bias in some of the FDA approved AI enabled products and devices?

Casey: I built a database of all the FDA cleared AI algorithms to date. As a reporter, I’m always getting press releases from companies talking about the clearances that they’ve gained from the FDA. But there is no real systematic way to look at those products. There is no database that identifies them to look in totality about what has been approved. So, I took a step further after identifying the products and looking at the level of validation that was done on them, like what was the size of the validation sets? What were the methods used? What is in those data sets? How diverse are they by race, by gender? Where were the data sets gained to get a sense of what level of information was disclosed? What is publicly available? And what I found was that it’s really all over the map in terms of the sample sizes that are used to validate these algorithms. And there’s also really very little information about the demographics of the data sets in a way that raises questions about the ability of these products to generalize across populations. And I found that variation happening even within products that are designed to do the same thing, like assess patients for intracranial hemorrhage or stroke or even things like breast cancer.

Q: What kind of products we talking about here? Are these medical devices, software products, and how many of them did you really scrutinize?

Casey: The category is sort of software as a medical device. These are used as decision support tools that supply data to physicians on patients, that helps them make decisions and helps them diagnose and treat those patients. There were 161 products that I identified within specific product codes. You can search the FDA’s databases to try find these and figure out what validation was done on them. I have read medical studies that suggest there is up to 220 of these products and these are all deep learning AI products. So, it is machine learning technology which have all been approved. We see a vast amount of innovation going on in that area over the past six years.

Q: On your reports you focus on the breast cancer related products. Can you talk about that?

Casey: Yeah, that was an area where I’m especially interested in looking at because diversity really matters, and breast cancer varies so widely among patients. And it’s particularly important to have diversity in those data sets so that any AI system that might be advising a doctor or a physician on how to care for these patients sees enough patients and can give good advice so that its conclusions can be generalized to broader populations of patients. What we’ve seen over time with a lot of medical products and algorithms that have made their way into the market is that they’re not tested on diverse groups of people. And instead, their recommendations, their reliability mainly exists within European Caucasian populations, which shouldn’t be acceptable to patients or medical providers.

Q: So there is reason to be concerned about the lack of a standardized validation process and a lack of disclosure specifically around the data that is being used to develop these algorithms and there is a real potential for racial discrimination. Is it correct?

Casey: I think that’s right. It’s the lack of standards there and in particular, disclosure of the contents of the data sets that is troubling from that point of view.

Q: Based on all your reporting, do you think the challenge lies in the quality of the data or maybe even the sufficiency of the data? Or is it more to do with the deficiencies in the algorithms or is it both?

Casey: I think the biggest issue is the quality of the data and the access to the data such that you can have really, truly representative data across populations and have enough of it to be able to train an algorithm to adequately perform the task you’re asking it to perform. There have been some studies done that suggest that the vast majority of data supplied for AI research comes from institutions in three states in California, New York, and Massachusetts. That’s missing a huge part of the places that we sit in now. So many people in so many communities end up getting excluded from that. This is the major hole right now that this ecosystem needs to figure out how to remedy.

Q: You make a very provocative statement at the beginning of one of your reports – ‘AI is now a lawless frontier in medicine.’ Some people might say maybe it’s just a little bit harsh, perhaps because it has had some success in other areas in healthcare, like administrative functions, revenue cycle operations, claim management, fraud and abuse, or even in chronic disease management. What would you say to those who feel that?

Casey: I’m making a comparison to sort of frontier development, like the development of the American West. I’m sort of making that comparison because I’m trying to crystallize the notion that the sheriff isn’t in town yet, that the pace of innovation, the pace of development, the pace of that building process is outrunning the ability of the FDA and other regulators to stay on top of the questions that innovation is raising. That is a big concern right now. I think the FDA is trying very hard, but I think it’s under-resourced and it can’t keep up with the very important questions that this is raising. The other part of that metaphor that is worth diving into is, does that mean that there are a bunch of bandits out there that are a bunch of evildoers who are trying to gather data and do bad things with it? By and large, from the companies and the people that I’ve talked to, I would say no. I would say that most of them are very well-meaning and altruistic. But there is still the issue of unintended consequences that may arise from the use of products that are not fully and carefully vetted. I think once that process begins to fully mature and catch up with the innovation, everyone will be better for it.

Q: You made a comment a little bit earlier about not been enough data available to do a rigorous training of the algorithms. There is a vast amount of data available in the form of images, more so than other forms of healthcare data. What can we do with the large amount of data available, especially the data sitting in our systems, for instance, in hospitals?

Casey: It’s very hard for researchers to come by to aggregate that data to do anything meaningful with it. EHR data is notoriously siloed and kept in environments where it’s just very difficult to access the data and make use of it for meaningful research and purposes that could really benefit people. I think it’s very difficult to harness that data, even though there is so much of it. And about the imaging data, I think a big question for the industry and a big problem right now, is the issue of transparency. Where are those data sets from? What is in them? We need to know the ingredients of these algorithms. We need to know who these people are, where they come from. We don’t need to know their identities. I don’t mean to suggest that, but we need to know how these algorithms are being built on what data so that there can be some confidence in these products, that they can generalize and do what the developers intend.

Q: What are you hearing from policymakers and industry executives, especially tech firms, on how they’re wrestling with the ethical use of data and how they’re moving forward with this?

Casey: Over the past six months a lot of companies are realizing that this is an issue and they’re bringing it out into the light and wanting to talk about it at industry conferences and on virtual gatherings and so forth, to be able to set forth, OK, well, you know what? This is an issue for us in terms of optics. We want to be inclusive companies. We want to emphasize that. And you’re seeing a lot of those companies’ fund research and hold events to talk about it. But there isn’t yet sort of a consensus that emerged on the best way to accomplish this. What are the set of practices that ought to be used to ensure that these products are inclusive and don’t unintentionally discriminate against certain groups? So, I think there’s kind of a recognition that these issues need to be addressed. But how to do that really has not been agreed upon, there really aren’t any clear best practice standards that have been identified. There is just a process that’s beginning to confront those issues.

Q: Is this a question for the FDA or is this more for the industry to self-regulate and self-governance and come up with the best practices and hope that the outcomes are good? What is your thought?

Casey: That is really the big question right now. Whose responsibility is that? Where should that vetting process take place? Should it take place at the FDA before these products get onto the market? That is not happening right now. Some of the people I have talked to, executives of companies say, the FDA clearance, the 510 K clearance that’s granted to most of these products has never really filled that role for any kind of product. So, usually what happens is there are follow up studies done at conferences and by clients of these products to bear out their efficacy. And there is a process that takes place normally in the private market to verify that these products are the best things for patients. The responsibility lies on the health systems to adopt products that are really going to benefit the people. Data is the main thing that these products use in order to deliver services, to help inform physicians to provide care to patients. You wouldn’t say to somebody – ‘you should just take this drug. Don’t worry about it. We don’t need to talk about the ingredients or where it came from or what’s in it. Just take it, OK? It’s fine.’ You would tell them the ingredients. It would be studied rigorously. You would know who is in those validation data sets, you would be able to analyse it in all the different cohorts and how it affects different racial subgroups. That’s done now in public at the FDA for drugs. Now, drugs have a different risk profile. Hence, the data analysis should be rigorously done and must have transparency.

Q: We’ve recently seen some initiatives, especially the one where several health systems come together and formed Truveta, that is going to pool patient data from several leading health systems and use it to analyze it for insights and help improve healthcare outcomes. There are also some other initiatives like the synthetic data challenge that the ONC has come up with. All are looking to address the same problem that there isn’t enough data for us to really analyze or train the algorithms and come up with some kind of heuristics or benchmarks for us to drive the outcomes. Would you care to comment on these initiatives? And is that an alternative? Is this a viable alternative that is taking shape?

Casey: It’s a timely question. I’ve been talking to the executives and stakeholders that founded Truveta over the past week or so to talk to them about that initiative. I think it is interesting in something that the industry, by and large, has just failed to do to date, and that is aggregate a large amount of data that comes from health systems all over the country and not just health systems that are on the coast. Those 14 health systems that are gathered in Truveta represent patients who are spread throughout 40 states all over the country. So, I think that’s really exciting and potentially provides a really great resource that researchers can tap to be able to gain access to large amounts of representative patient data. There still are a lot of questions though with that because we all know about controversies that have arisen from, say, given the hospital system, working with a tech company and sharing their data with that tech company because of all the privacy questions and questions of economic exploitation that might arise from that. It is like you’re using data from the patients that got care at your institution. Then you are selling that data to another entity to do research on it to build a product that that entity will profit from and not necessarily the patient. So, there are issues of consent that get raised in that. There are questions that should be raised and talked about so that there can be a consensus or at least an open public discussion about how to get access to that data, who does it benefit, how to do this in a way that respects the patients and all of the stakeholders?

Q: What are the top two or three items of the unfinished agenda in harnessing data for us to really make a difference in healthcare outcomes? One is interoperability. Can you share your thoughts on this?

Casey: I think interoperability is a key issue and that issue is part of developing data sets at scale, large enough data sets that can be used by researchers and companies to be able to build meaningful and generalizable AI products that will benefit everybody. I think the biggest issues in my mind about that are really transparency, disclosure and some of those regulatory questions. I think it’s really important to think about the nature of these products, which are machine learning. It’s a computer that is able to comb the contours of a data set to form conclusions on its own without being explicitly sort of programed. I think when you have a system like that where it might be somewhat of a black box about how it is reaching the conclusions that it’s especially important for people to know what is going into those training sets. How is it being tested on what data is it being validated? Are these things at the end of the day going to improve care or are they just going to layer on top of care an additional level of cost without providing the benefit that they advertise? And I think that process just must unfold in a meaningful way so that, before we start paying for these things, before they get into the market and start providing care for people, we know that they are fair. We need to know that they are safe. We need to know that they stand some chance of improving care to people. So, I think those are the things that sort of need to be front and center questions that are addressed over the next few years.

Q: To sum it up in one word, would that be transparency?

Casey: I would say that would be the word I would choose as the one word that the industry needs to sort of focus on in the next couple of years.

Show Notes

04:52It's important to have diversity in data sets so that any AI system advising a doctor / physician on how to care for these patients can give good advice and conclusions that are generalizable to broader populations of patients.
09:07 The pace of innovation, the pace of development, the pace of that building process is outrunning the ability of the FDA and other regulators to stay on top of the questions that innovation is raising.
12:07 A big question for the industry and a big problem right now is the issue of transparency in data sources.
21:42The biggest issues while harnessing data are transparency, disclosure, and interoperability.

About our guest

casey-ross-profile-pic

Casey Ross is a National Technology Correspondent at STAT and co-writer of STAT Health Tech, our weekly newsletter on the growing digital health industry. His reporting examines the use of artificial intelligence in medicine and its underlying questions of safety, fairness, and privacy..

Before joining STAT in 2016, he wrote for the Cleveland Plain Dealer and the Boston Globe, where he worked on the Spotlight Team in 2014 and was a finalist for the Pulitzer Prize. A Vermont native, he now lives in Ohio with his wife and three children. When he's not with them, he's in his cornfield, cultivating some of the sweetest bicolor in the Midwest.

About the host

Paddy is the co-author of Healthcare Digital Transformation – How Consumerism, Technology and Pandemic are Accelerating the Future (Taylor & Francis, Aug 2020), along with Edward W. Marx. Paddy is also the author of the best-selling book The Big Unlock – Harnessing Data and Growing Digital Health Businesses in a Value-based Care Era (Archway Publishing, 2017). He is the host of the highly subscribed The Big Unlock podcast on digital transformation in healthcare featuring C-level executives from the healthcare and technology sectors. He is widely published and has a by-lined column in CIO Magazine and other respected industry publications.

The Healthcare Digital Transformation Leader

Stay informed on the latest in digital health innovation and digital transformation.

The Healthcare Digital Transformation Leader

Stay informed on the latest in digital health innovation and digital transformation

The Healthcare Digital Transformation Leader

Stay informed on the latest in digital health innovation and digital transformation.