MD Anderson 2020 Interview Project
Chapter 03: Digital Transformations


Chapter 03: Digital Transformations



Media is loading


Dr. David Jaffray spends this chapter descsribing his role as Chief Technology and Digital Officer at MD Anderson Cancer Center. He speficially describes the initiatives he has been a part of that seek to bridge data science, technology, and oncology. Dr. Jaffray also considers the possible resistance to these intitatives, though emphasizes how proper stewardship of data can assist in the fight towards eliminating cancer. He also reflects on the importance of collobration across the institution, noting the need to redefine how promotion for faculty should be thought of in order to counter inidividualism. Dr. Jaffray ends by meditating on the overall outlook of MD Anderson’s response to COVID-19.


Nina Nevill Yeah, absolutely. Well, it feels as if we’ve covered a good amount of ground regarding the past year, and unfortunately, we’ve done things a bit backwards or in the opposite order of what I had originally planned, so forgive me for being all over the place. But for the sake of having this on the record, since we have you today, again, this feels like going back to the start, but if you could talk a little bit about your specific role and specifically about digital technology, I guess this could also fit into the conversation in terms of what goals moving forward you may have, or the institution may have, regarding the specific position.

David Jaffray, PhD I think I’m going to ask you for just a minute to grab some water and come back. Is that okay?

Nina Nevill Sure, of course.

(break in audio)

David Jaffray, PhD Okay, so you want to ask me again just to remind me?

Nina Nevill Sure. In terms of your specific position in your work with digital and technology, I guess the first part would be what changes have you seen, and the second part would be what direction moving forward do you foresee? And those can obviously blend together as much as needed.

Nina Nevill Yeah, no. This is a very exciting time. And it’s been brewing for a while, obviously, the transformation of medicine through digital technologies, but even more importantly, the transformation with which we do our science and advance the improvement medicine through approaches that use digital technology. And the excitement, in my mind, comes from the fact that new technologies are allowing us to measure things that we’d never been able to measure before, to understand what’s happening to a human, what state they are in, what state their disease is in, and then to target and treat it is one thing, but to measure the response, to understand why is that patient responding or why isn’t that other patient responding, I think, is just as important, so that we can learn. And these changes are bringing about an opportunity for us to really think differently about how we handle data within the organization. And this is a conversation that many organizations are starting to go through. They start with the idea that we can use computers to predict things, we can use AI and machine learning, but by and large, they really are dependent on the data. And so, what focus is, that we move from focusing on the computing side to really, where is the data coming from? What’s the structure? How did it arrive? What’s the quality of that data? And that whole focus really changes the way we think about the flow of information within the organization. And that’s a very exciting direction. We learned a lot of that during the D3CODE initiative, really connecting one end to the other.

And that kind of thinking, I think, is going to be critical for us moving forward. We had great conversations with other cancer centers and hospitals who are dealing with the same issues, but they’re really focusing on getting some technology in rather than thinking about how they structure the work around the flow of data. And so, we’ve launched an initiative called “The Data Governance Version Two Effort” that looks at how do we govern our data, where does this come from, how do we reflect the people’s participation in it, how do we make sure the quality is there, how do we track where the insights came from, and then, how do we know for sure that we have appropriate use of that information? Do we have the rights to use that data for everything we want to use it for, or is it constrained? And that takes us back to this very interesting conversation with the patient, to ask them, “Are you willing to partner with us to help advance our mission? Can we use your information to help cure cancer?” And that’s a dialogue that is very exciting in the future for us. And so, it really shifts our mindset to not about sharing data, but how do we enter into collaborations with others to use data to answer important questions? And sometimes those others maybe industry or maybe other academic partners, but I think the future very much is, they will be our patients. And how do we talk to them about letting us use their data to help answer important questions? And that digital engagement aspect is definitely the future cancer care, and the future of eliminating cancer, as we’d love to. And so, those two are beautifully aligned, and we’re trying to understand how we put our technology and processes and policies and our governance in place to allow us to participate in that in the future. And so, Anderson needs to become pretty strong in engaging with its stakeholders from a digital perspective. This is a very exciting future. It’s kind of a big leap forward, in many ways, to take that activity and to roll it out across the organization. But everybody sees it. Everybody sees it in their daily lives. They see the fact that they’re measuring things all the time, either their cell phones or their Fitbits, or their bike monitors, Strava, and so on, and they’re just learning and sharing the information and getting insights. And so, we have to figure out how to do that. How do we make that process seamless, take all the friction out of that so that we can just surprise people, delightfully, in a way that gives us the appearance that we are anticipating what they need, or we’re predicting for them what the future is?

A big part of that is the building up of data science capabilities. And data science is a paradigm where you just have so much data coming because the cost of measurement has got so low, and we can track so many things. Data science is the process of using communicational approaches, statistics, and mathematics to pore through that data and to extract insights. And those skills are going to be key because using the human mind to look at the data and draw conclusions is just not tenable. There’s just too much data flowing, now. So, how do we build data science technologies across the organization? And so, one of the signature priorities out of our strategic plan, which was another really interesting thing that came out during the pandemic, by the way, which I didn’t talk about which is amazing, the strategic plan focuses on building data science capabilities, not just data science, but also predictive modeling capabilities across the institution in response to the fact that we will engage our patients on the use of their data to advance our mission, and we will assemble a remarkable scale of that data and answer really difficult questions. And so, we need to build up that capability. So, we’re in the process of creating an institute for data science and oncology, which I think would be the largest concentrated focus of data science capabilities on cancer in a single institution in the world.

And then, combining that with the rich datasets, and combining that with the quality of care and the participation of our patients, is a very promising indicator of where this will take us in accelerating our mission. So, it’s very exciting. A very exciting time. And all the pieces of AI—

Nina Nevill This sounds like a huge—

David Jaffray, PhD It’s huge.

Nina Nevill Sorry, go ahead.

David Jaffray, PhD It is huge. And the whole world is going through it. The whole world is realizing that this is the way the future will be. And those organizations that make the decision to pursue this direction, to make the efforts, to make sure the data’s flowing, to build the teams, they will set the pace. And if we’re setting the pace, that’s where people want to come, whether it’s our patients, or bright young talent, want to come and join us and make progress against cancer, which everybody knows is (inaudible).

Nina Nevill Now, in terms of the excitement, obviously it seems like the majority of people would be thrilled with an initiative like this to move forward, but just due to the scale and the amount of change that it would require, has there been any pushback? And if so, from whom? Or is it most widely accepted that this is the direction of the future?

David Jaffray, PhD Well, I think people see it as inevitable at some level, that the way the cost of measurement is falling and we’re using it to characterize people and their disease differently and what their needs are differently and attune that is just a very logical direction. I think some people might be dragged there, and other groups will decide, “Let’s get in front of this. We all see the promise, we’ll get to the try solutions first, and we’ll get the new perspective because we’re there first.” And I think MD Anderson, because of the mission, because of the talent and the people and why many of them are, there I think they see that this is a worthwhile investment. This is something that we need to do for our mission. And I think that resonates across the board. Some of the challenges that come up when you make technology such as that data can move and you can connect dots that may not be obvious to connect is that some people want to make sure that the dots are connecting, that they’re okay with that, that that data that came from that project or that study can be used for that. And that’s a conversation around stewardship of the data. It’s around making sure we’re not letting those resources sit idly. It’s around making sure we have the right person involved in the conversations who helped collect that data, because they’ll make the inside extraction more robust. That’ll bring additional value to that conversation. And that trust architecture I think is very, very important. Having the confidence that, yes, I will let teams access my data, but I expect to be engaged in that process, and making sure that we can keep track of that, I think is really important. And with time, I think the value of bringing all that data together to answer completely new questions that nobody else can, we’ll maybe realize that the protection side of the data is not that important. The benefits of pooling that together to ask new questions with the expertise present, of course, at the same time, that outweighs the protectionist paradigm. And that’s a conversation that’s ongoing across the organization, and I think the world is changing in that direction a little bit. The scientific community is definitely changing in that direction. And the patients are also saying, “We want to be involved in this process.” And so, getting the right technology and processes in place to do that is the right journey. It’s not quick. It’s a cultural change at some level, but I think MD Anderson is at the right point. And MD Anderson, we’re big enough that enough data will come together to answer really important questions. If it was a tiny amount of data that we’re pulling together, then why? But at Anderson, we can really draw together very rich datasets and help answer those questions. It’s pretty cool.

Nina Nevill As someone who is outside of the medical and this science realm, would you be able to give an example of a kind of question that ideally, if we’re thinking dreamland here, would be able to answer with this type of movable data in the future?

David Jaffray, PhD Yeah, for sure. It’d be great to be able to identify, if we could track the outcomes for patients. And maybe they have a specific mutation in their genetics. So, it’s something in their germline. We’ve looked at maybe 10,000 patients like that, through time. Maybe we could say, “Well, did you notice that the outcomes were different between the patients with this mutation versus that mutation?” And that would immediately turn back into a conversation around, “Let’s find out what the mutation is in that person that’s about to be treated and then let them know what the likelihood of outcome is for one treatment or another.” That’s just a simple example. But you need to have large numbers of patients to answer that question because, I mean, there’s a lot of other variation that happens. There are slight differences in treatment, other factors, how long since they started to have their cancer, did we start to intervene, all those factors confound. So, the more that we can have engaged, the more we can work to address those confounders. At the same time, other initiatives that are a little bit more, maybe, practical, day-to-day, I would say, where right now, we’re looking at patients within the hospital who get blood transfusions, and we have a team, a very talented team that watch those patients very closely to watch for reactions.

That’s kind of expensive, having people watch the screens and watch the results of those patients. We could learn from all those patients that we’re watching, and from all the great observations and actions of the teams that are doing that, and we could put together an artificial intelligence approach that watches and learns and does it automatically. It would be great for our patients because then those staff could go work on something else more patient-related. But also, we could take that technology to somebody else across the planet and decrease the risk of an adverse reaction for a blood transfusion, or a blood product transfusion. So, that’s another example. And there are many, many more.

Nina Nevill That’s very helpful, just to have something slightly more tangible or concrete to get an idea of what this could look like in practice. And the fact that it could be to the point of being, then, informing treatment, that’s definitely exciting. And it seems like, from all that you’ve mentioned today and from what little research that I have done, that MD Anderson is also excited about this, has obviously hired you for many great reasons, but in terms of trying to think of whether reflecting on this past year, or just since you’ve been in this line of work, what could the institution do better or differently to advance these goals of getting to this place?

David Jaffray, PhD One of the things that we wrestle with, which is actually not just at MD Anderson, but the whole world, is that the promotion of academic faculty, that whole development of their career, is largely based on their individual performance, which is contrary to the idea of collaborating together and working as a team. And so, one of the things that MD Anderson started as part of our data governance initiative, and this is being led by Caroline Chung and Carin Hagberg and others, is starting to ask the question, “What’s the best way to acknowledge the contribution of our faculty in team efforts? What’s the best way to look at how they’re promoted?” If someone comes in and helps build the infrastructure and shape the processes and everything so that the entire organization can be more productive but doesn’t necessarily get that first authorship on a paper, how do we reward that kind of investment, that kind of thinking? And I think it’s possible to document and record that. I think it is possible to do it. But the current framework of independent academic research, as a solo sport, is just not aligned with that kind of thinking. And as an institution, we’re big enough that we can create an ecosystem where those kinds of contributions are recognized, and that people get promoted. And if you promote those kinds of people, actually, the whole place wins because they work to the benefit of many, many individuals. And that’s very powerful. That’s a great, massive promise that we could tap into here at MD Anderson.

Nina Nevill That would be wonderful to hear about, if that was the case in many institutions across the board in different departments and fields, it seems like collaboration is not always celebrated, especially for those in this specific track. So, it’s lovely to hear that there are some conversations happening, at least to move in the direction of acknowledgement, let alone more than an acknowledgement for communal work.

David Jaffray, PhD Yeah. I mean, we celebrate individuals all the time, and that’s fine. There’s nothing wrong with that. But the way they do their work and the way we measure that contribution needs to be diversified. It needs to be recognized because of the impact in many different ways than just the traditional measures. And there’s lots of great examples of people who have had massive impact but have nothing to do with impact factors in general. Lots and lots of examples. Like Jeff Bezos, for example, and his work, or Musk. He does his work as a team. He has to. Everybody carries a contribution. You wouldn’t want to just recognize Jeff or Elon as the contributors, right? It wouldn’t make any sense.

Nina Nevill Absolutely not. Well, I feel that everything I had hoped to get out of today and more has been fulfilled. There are always more questions and always more things to be asked, but before we come to a close, I suppose is there anything that you hoped that I would ask about, or anything that you feel like could be fitting to talk about in light of anything we’ve discussed so far?

David Jaffray, PhD I’m just thinking about—the focus was really on the period of the pandemic, that’s correct, right? That’s the kind of thinking, is that it’s been a very interesting—a lot of great things were done. Yeah. There’s so many people I can speak about who have made contributions on so many different fronts, and that I could highlight. But I’m sure others will make the same comments in different directions specific to their areas. The IT team has been pretty amazing. Their investments over the past several years, well before I got here, gave us a really sound foundation for the move to work from home and to adapt so quickly to the change. So, I’ve said this in many settings over the past year that we benefit from the foresight and the investments that they have led, and structured, to allow us to deal with this fairly smoothly. And so, that’s one aspect, Chuck Suitor and Craig and many others, Neil and John, just many of them who have helped shape that, and Jay. And so, I think that’s one aspect that’s really important.

Also, the guidance from the finance side, Ben Nelson and others, just steering through how do deal with this, how to hit the brakes, how do adjust, and the dynamic nature of that has been really, really helpful for the whole organization to have confidence that what was happening was well thought through, managing that, and converging out the risk. It’s also been a heck of a year from a cybersecurity perspective, actually, which has been a little stressful, but like I said, I’m pretty calm about these kinds of things. And as long as I know that we’re working away, all the time, to improve our ability to defend and we’ve been very thoughtful in what we invest in to make that happen. If we’re doing that, every day, if you’re making it better, you’re not falling behind. As soon as you’re stationary, you are definitely falling behind on this front. So, that’s been an exciting part of the past almost year, now. It really started to take off last July or August. So, I don’t have anything else to say. I can imagine there’s a ton more that will come out of my head at some other point in time, but if you want to come back, I can probably think of some other things.

Nina Nevill Well, thank you so much. Yeah, if there are any follow-up questions or anything that we need clarification on, I’m happy to reach out again, of course, if you are willing. But it’s been absolutely lovely to chat today and get to hear your perspective and your experiences over this past year. So, we greatly appreciate you taking the time.

David Jaffray, PhD Yeah, no, it’s a pleasure. I presume you’re just going to cut different chunks of this out when you format all this, is a huge amount of work. Is that right?

Nina Nevill Yeah, I think so. Let me just go ahead and pause the—



Publication Date


Conditions Governing Access


Chapter 03: Digital Transformations