By Sarianne Gruber
Twitter: @subtleimpact
Dr. Kathleen McGroddy Goetz, though she prefers to be called Kathy, describes her career as a journey. A trained academic in healthcare and life sciences with a research perspective, her first aspiration was to do drug discovery. Kathy describes her twenty-three year career at IBM as “fantastic and interesting”. She started in hardware and software, which at the time was on the cusp of joining business and technology, and later moved more into the business development side of research. Now as VP of IBM Watson Health, she is focused on technology and commercialization where “the pace has been very busy”. Kathy has been instrumental in many of new acquisitions and partnerships for IBM Watson Health including Phytel, a population health software company, and Cloudant, a Data as a Service company for Cloud Computing. Here is a gently edited recount of our interview.
How did IBM Watson Health start its move from technology to commercialization?
IBM has 12 research labs around the world, and six of them do research in healthcare informatics. We have a 10 year track record of some really innovative healthcare projects, but we never had a commercial channel to market them other than our services arm. If you rewind to 2011, where we had one of our grand challenge research projects, we built a computer system to play on TV a game, Jeopardy. That was the accumulation of at least ten years worth of research into things like natural language processing and deep Q&A, question and answer technology, and looking at the next generation of computing -this notion of “cognitive”. We built the system and it was aired on television. Immediately, the public started thinking. They started to call us about commercial applications. They asked how technology like this could be used in different areas and industries that are in need of transformation, and have so much information it is difficult to keep track of it all.
How does IBM Watson Health navigate the perception of artificial intelligence?
The way we think of Watson and Cognitive technology, it is not so much replacing, it is more about augmenting intelligence. It helps make humans smarter by providing access to the research, recalling the facts and providing all of that knowledge. I think there are something like 1.8 million scientific journal articles that are published each year. Not very many people, if any, will have a chance to read, digest and remember them, and then even curate and understand which ones should matter more. I think another interesting characteristic that we think about a lot in this space is the notion of “bias”. So if you think about your doctors perhaps who were trained at respected medical institutions. They were trained to do things in a certain way, and will probably continue to practice medicine with that influence and “bias”. As a physician, you may be aware of new things, but you will have sort of a lens in terms of how you practice. So if there is an opportunity to provide other perspectives, insights or new information to someone at the point of care, that they can also consider we believe it has the potential to make people more effective. We see Watson and Cognitive technology as an advisor and an augmenter of their intelligence, not replacing it.
How does IBM Watson Health manage the adoption of artificial intelligence?
What is more important and basic is how you can integrate technologies like this into the work flow. Over the years, we have had technologies that can help improve health care, but actually it is about getting it into the work flow so it is easy to use and fits in with how doctors, nurses and care managers do their jobs on a day to day basis. You make them more effective within the constructs of how they want to do their jobs and what the right way is for them to do their jobs. This type of adoption into workflow lowers the barriers. We also believe in engagement as part of adoption. Engaging the patient and engaging via the physician is really important. We have two new announcements. IBM Watson Health has extended its relationship with Apple, beyond building mobile apps and selling it to enterprises, we plan to use Apple devices along with apps that leverage HealthKit and ResearchKit and provide the back end cloud services to support this. The other, is the acquisition of the population health software company, Phytel, based in Dallas. Ranked in the best of KLAS ratings, its health management capabilities on some levels are beautifully simple. Imagine you are a case manager or a nurse in doctor’s office, and you want to understand how to better reach your patients at a point in time when it matters. So you can say, show me all my type 2 patients Hb1Ac over 9 and who haven’t been in for an appointment in the last 3 months. We can automatically send out an email blast or call them. We can engage with them to make sure they are being followed up and managing their care. We can even proactively screen a population to invite to a cooking class on learning how to better manage their health.
How do you integrate Artificial Intelligence into Population Health Software?
Phytel has an appliance that sits in the physician office or at the hospital collecting the data. With this appliance, they have the ability to aggregate the data, do population health statistics, population comparisons and metrics. Phytel has some analytics, but it is much more rule based. We couple that with our cognitive intelligence and knowledge based system that’s up on the latest and greatest, new insights, new journal articles and things that have come in. We provide the ability to do more sophisticated machine learning, looking retrospectively at longitudinal patient data and assigning the patients to cohorts, looking at the different pathways and outcomes. You can actually get more sophisticated from the analytics perspective. You can then take and feed that into the system Phytel already built that allows you to be that much more effective in terms of whom you target, where you put your energy, and when and how you reach out and engage with those patients. It is connecting the IBM Watson Health cognitive engine and the machine learning abilities, all of this great analytics technology, the big data and analytics piece that IBM has been so good at for so long with this engaged outreach, best in class software. Then leveraging, this best in class engine to the physicians because it is integrated into their work flow. Phytel has done a lot of nice work to understand how to get that integrated, how to get folks using it, how to do the right work ahead of time to understand outcomes and how to measure outcomes in a way that actually makes a difference and can be significant such that it is a no brainer for people to adopt it.
Kathleen McGroddy Goetz Ph.D., is Vice President of IBM Watson Health where she leads the technology commercialization and strategic partner relationships for IBM’s new business unit, which was formed in April 2015. In her 23 years at IBM, Kathy has held a variety of senior business development, technology strategy and product management positions spanning research and software. Kathy holds a B.S. in physics from SUNY Binghamton and a Ph.D. in molecular biophysics from Cornell University. Kathy is an avid runner and triathlete. She lives in Ridgefield, CT with her husband and children. MedCity Converge Speaker Bio September 2, 2015