In this episode of Expert Insights for the Research Training Community, Dr. William T. Riley, director of the NIH Office of Behavioral and Social Sciences Research, shares the advantages and challenges of integrating the social/behavioral sciences and biomedical sciences. He also provides examples of this integration in recent trans-NIH initiatives.
The original recording of this episode took place as a webinar on May 19, 2020, with NIGMS host Dr. Judith Greenberg. A Q&A session with webinar attendees followed Dr. Riley’s talk.
Recorded on May 19, 2020
View Transcript Download Recording [MP3]
Announcer:
Welcome to Expert Insights for the Research Training Community—A podcast from the National Institute of General Medical Sciences. Adapted from our webinar series, this is where the biomedical research community can connect with fellow scientists to gain valuable insights.
Dr. Judith Greenberg:
Good afternoon. I’m Judith Greenberg. I’m Deputy Director of NIGMS, and I’m pleased to welcome all of you to another one of our NIGMS webinars for trainees and for others.
We organized these webinars to give you some interesting perspectives on a variety of science-related topics at a time when we know most of you cannot be in your labs—even though that’s probably where you’d rather be right now.
Before we begin, I want to thank a number of people who’ve made these webinars possible. First of all, the communications team at NIGMS who publicize these so well. To our IT people who’ve done a fabulous job in all the technology and making them work so well. I also want to thank all of you who are participating in this, and, of course, most of all our speakers who’ve taken their time to put together interesting presentations.
Today’s webinar is entitled “Integration of Behavioral and Biomedical Sciences at the NIH,” which we think is an important area and not always well understood.
OK, now to my introduction of Dr. Riley. I’m going to be brief because I know he’s going to tell us something about his own career path that brought him to his current position as Associate Director for Behavioral and Social Sciences Research and the Director of the Office of Behavioral and Social Sciences Research—OBSSR—at NIH, where he has been in that position for the last five years.
He’s also been in other parts of NIH for a total of 15 years. During that time, he was also at the National Institute of Mental Health, the National Heart, Lung, and Blood Institute, and the National Cancer Institute. He holds an appointment at George Washington University in their School of Public Health, and his research interests include behavioral assessment, technology-based interventions for health risk factors, and the application of engineering and computer science methodologies to the behavioral sciences. Dr. Riley holds a PhD in clinical psychology.
And so now let me turn it over to him. Dr. Riley.
Dr. William Riley:
Judith, thank you so much and thank all of you for attending. It’s wonderful to be here and actually quite an honor. It’s very nice for both Jon and Judith to invite me to do this.
I’ve long admired the T32 program at NIGMS, and particularly the efforts in integrating behavioral and biomedical sciences, so I’m looking forward to doing this, and hopefully we can stimulate some discussion as we move forward.
So let me begin with a brief history of both OBSSR and, as Judith said, maybe a brief history of me and my career.
So OBSSR has been open since 1995. This is their 25th anniversary. We coordinate behavioral and social science research across the NIH. Our roles are many, but primarily leading trans-NIH behavioral and social science research initiatives, conducting workshops in emerging areas, supporting training, and, of course, co-funding institute and center grants. OBSSR doesn’t fund grants directly, but we work with our IC partners to do that work.
So a brief history of me. I was reflecting on this and I usually tell people about the fact that I’ve moved from position to position mostly out of opportunities that have arisen, but I have to start with a limitation for how I became a psychologist.
When I went to college, I originally wanted to be an astronomer. I was one of the few guys who took his date out into a field to look at the stars and actually looked at the stars with them when we got out there. I started as a math major, and unfortunately as the whiz kids were doing differential calculus in minutes, it was taking me hours to do the same problems, and I realized pretty quickly that my calling was probably not in mathematics. So I still have quite an interest in that. I just sort of do the work a little slower than others do.
I spent about 15 years in academic medical settings and sometime in the research and development space, mostly in the computer science and engineering area. This was back before the days of iPhones, and we were doing portable handheld devices, mostly to do health behavior change and behavioral assessment types of things in a more automated way.
As Judith said, I spent about 15 years at the NIH, lastly here at OBSSR, and I’ve been here about five years now. So just a backdrop about funding for behavioral and social science research at the NIH. We’ve seen a growth in funding pretty much every year over the last five years, both in basic research and then overall in behavioral and social science research at the NIH. And that portfolio represents a little over 10 percent of the total NIH budget.
Now keep in mind when we’re counting this we’re counting anything that has a behavioral and social science component to it, so it’s not that these are all prototypical behavioral or social science studies and projects. Most of them have some piece or component that’s behavioral in nature, but that still represents a little over 10 percent of the overall NIH budget.
And of the players, the largest ones in terms of—this is from last year—competitive behavioral and social science research funding by institute: NIA and NIMH usually battle it out, but unfortunately this one blocked out the others, so it kind of does every other. But some of those institutes tend to do the larger amounts—NCI, NHLBI, NICHD. Some of those other institutes also fund a fairly large amount. And I will just sort of note that even the ones that are kind of at the bottom are not at the bottom because they don’t care about behavioral and social science research; it’s partly the size of their portfolio overall as well.
We began a couple of years ago looking at sub-categories of the behavioral and social science research that we fund at the NIH, and as you’ll see from this slide, we tend to do a fair amount of research and some basic research in attention and learning and memory—that’s always been a core basic research/foundational science at the NIH—as well as social processes and determinants, healthcare and disease management, and mental health. And then you can see some of the others as you go down the list.
We’ve had a strategic plan for three or four years now, about three years now. So three core strategic priorities for us are scientific priorities for us. One has been that we felt like for many years that the pipeline from basic to applied research in the behavioral and social sciences isn’t as strong as it should be, so we’ve been working to improve the translation of basic research into applied behavioral intervention research over the last few years.
The other is in methods, measures, and data infrastructures, improving the way we go about measuring these behavioral and social phenomena, doing that more precisely and more accurately, as well as improved methodologies for understanding these phenomena.
And then the other thing that’s been a consistent concern for us has been the adoption of behavioral and social sciences research in the community. I’m envious of the fact that the biomedical sciences have the FDA and have the systems in place to be able to implement into hospital systems. Part of the problem for the behavioral and social sciences is that we not only implement in healthcare systems, we also implement in communities and schools and policy makers and all kinds of other groups as well. And the ability to be able to foster adoption of the science in those areas is particularly important.
I’ll talk a bit in a little bit about some of the COVID-19-related issues, but that’s a good example where there is some very clear research in some areas having to do with how you manage pandemics, how you deal with some of these things, how you communicate risk, and that science gets applied somewhat spottily depending upon on who’s doing the talking and how it’s being done.
So in the process of putting together our strategic plan, we identified four transformative opportunities that Dr. Collins and I wrote about a few years back in 2016, and these really do kind of begin to get you thinking a bit about the integration of biomedical and behavioral sciences and other sciences as well.
One of the areas was in integrating neuroscience into the behavioral and social sciences. And that connection, whether it’s behavioral neuroscience, cognitive neuroscience, social neuroscience, have all been, I think, great opportunities to wed those things together more than they previously had been. There have been a lot of transformative advances in measurement science. This includes sensor technologies and some of the things we’re doing now with phones and sensors and wearables and home-based sensors—those types of things—as well as actually just improvements in how we do patient-reported outcomes and self-report work as well. There have been some really nice improvements in that area. The digital intervention platforms, which I’ll talk about a little bit more later, and our ability to be able to do our interventions with reach and scalability that we previously weren’t able to do.
As many of you know, most behavioral interventions are very labor intensive, resource intensive, mostly have been done in the past face to face, in person. We thought we were doing reach and scalability when we had group sessions instead of individual sessions, but now that work has been ported to websites, mobile technologies, and a variety of other ways for us to be able to extend the reach and scalability of our interventions.
And then we’ve had a lot of large-scale population cohorts and data integration efforts across those cohorts that have allowed us to answer questions we previously have not been able to answer. The integration of biomedical and behavioral sciences has actually been a principle OBSSR since its inception.
This is an initial article from Norm Anderson, who was the first OBSSR director during its inception and the early days of OBSSR. And he talked about the discoveries of behavioral and biomedical sciences being equally critical for health, but that knowledge of both of those need to be integrated for us to really advance. And that’s been a core theme of our office since its inception, to move forward and integrate better the biomedical and the behavioral sciences at the NIH.
But of course there are some challenges to integrating this work and doing transdisciplinary work. For one, it’s an interpretation of different languages. We don’t always use the same terminology, and actually the worst is when we use the same terminology but mean different things as we say those. Often when I say “mechanism” with my biomedical colleagues they’re thinking about some process that occurs under the skin, and many times the mechanisms I’m talking about are mechanisms that actually reside in the environment, among social interactions and those sorts of things as mechanisms.
We also have to check our scientific assumptions—we each have our own. I’ve been fortunate in some ways that I have never worked in the psychology department in my entire career. I’ve always worked in medical centers or the NIH, which are predominantly more biomedical in nature and the only time I didn’t work in those areas I worked in a private firm that was primarily computer scientists and engineers. So I’ve never really had the luxury that I think my psychology-department colleagues sometimes have of being able to talk to one another and just know that their scientific assumptions they all agree to already, you have to kind of put those on the table and make sure we’re clear about what those are.
We also have to merge different research cultures. Research standards and accepted approaches are different among these various disciplines. Causal inference and what we mean by causal inference, and what is considered adequate justification or evidence for causal inference differs by these situations. It would, if it weren’t for the ethical constraints, it would be very nice to do a study in which I randomly assigned children at birth to adverse childhood events over the course of their lifetime and see what the impact is on their health, but that’s obviously not something we can do, even though that would give us much better causal inference data on the impact of adverse negative events and the mechanisms of that on health.
So we have to come up with other ways of being able to do that to seek a causal inference in that situation. Smoking cessation literature and smoking literature is actually a pretty good example of that. Not doing RCTs but coming to an accepted sort of standard for causal inference. As best as I can remember over the course of 50 years of smoking research, never really assigned people to smoke or not smoke, but we looked through a lot of ways of being able to control for potential confounds to be able to show a clear relationship between smoking and ultimate disease. And their publication standards are also a little different.
There are Nobel laureates in economics that have maybe 50 publications. Now those publications are quite large, quite voluminous, but 50 publications in a cell biologist’s lab would be the kind of thing you would do in a couple of years. It just has to do with the difference in the size of a research publication, and what it means, and how people tend to think about publications in that work.
Computer science is another good example of that, because their presentations in scientific society meetings actually in a lot of ways carry as much weight if not more weight than it does for publications, where that’s quite different in both the behavioral and biomedical sciences.
And then we, obviously, have to know how to play well with others. Be able to listen to each other, be able to try to understand each other’s perspectives and work well together as we do that. There’s one other piece of this, though, that I want to make sure we cover, and that’s this concept that psychology and behavioral science is common sense. And this is one of the things that I think we often struggle with as we start integrating behavioral and biomedical sciences.
So let me just step back from this and give you a sense of what I’m talking about. Everybody, by evolution and all other factors, are amateur behavioral theorists. You have to be. It’s how you come up with deciding how you’re going to predict the behavior of other human beings, what they’re going to do, how to react to environments and those sorts of things.
And so we build our own theories in our mind and our head about how we function and how other people function. But those experiences are obviously idiosyncratic, only to us, and also cognitively biased in a variety of ways. But everybody has those.
Now what some people, I think, believe is that behavioral scientists, as a result of our career study, we may have a slightly better “common sense” regarding human behavior than some other people do. I tend to reject that concept with possible, maybe, but I think what really sets us apart as behavioral scientists is that we embrace the counterfactual of that commonsense solution. Whatever it is that we think might be right, our first thought is, “Maybe that’s not right. Maybe there’s an alternative hypothesis for why this person has engaged in that behavior.” And then we subject those hypotheses to rigorous scientific tests.
So we tend to be, I think, as I talk to people that I consider to be really good at this field, they’re the people who question everything that they think and every experience that they have to ensure that it’s not an idiosyncratic or biased perspective that they have about human behavior.
I want to give a few examples of some of the transdisciplinary research that’s going on at the NIH, and I’ll spend a little time on COVID-19 since it’s so pressing right now. But first I wanted to mention the BRAIN Initiative, which has been in Phase I of the BRAIN Initiative, focused very much on neurocircuitry and tools and instruments and capabilities to be able to assess that neurocircuitry in ways we’ve previously been unable to. And the BRAIN Initiative has been extremely successful at that work.
As they shift to Phase II, they shift to actually applying that work to combining those approaches into understanding cells and circuits and brain and behavior at the end of all of that. So one of the things I’m really looking forward to in Phase II of the BRAIN Initiative is the ability to be able to use those tools now to be able to understand from very elementary behaviors up to more complex behaviors and be able to map those.
Now, to be able to do that, we’ve, over the course of the last decade or so, have gotten extremely good at precise and accurate and temporally dense neuroscience analysis and neuroscience measurement. We need to be able to step up to the plate on the behavioral side and provide an equal level of temporal density, precision, and accuracy on the behavioral end of that spectrum so that we can map these things together more carefully than we currently are able to do.
The HEAL Initiative is another example of a transdisciplinary research program. This is the NIH HEAL Initiative that last year awarded nearly a billion dollars in funding in a multi-pronged approach that includes behavioral sciences and important research questions in that area. And I listed here some of the recently funded projects. One of the things that we did in the early days of the HEAL Initiative as people were talking about could we develop better analgesics for pain control and could we develop better ways to treat people with opioid dependence, one of the things we wanted to make sure was clear is that there are significant social and behavioral factors associated both with chronic pain and with opioids, and some of the research that needs to move in that area in order to improve our ability to treat opioid dependence and chronic pain as well.
So the recently funded projects of the HEAL Initiative, I think you see a really nice integration of biomedical and behavioral research, looking at things like acute to chronic pain signatures. Discovery and validation of biomarkers and end points. A really nice study looking at the healing communities and looking at opioid dependence more from a community-based sociological perspective and how to address that. Work within the justice community on opioid research and treatment.
Behavioral research to improve our medication-assisted treatments for opioid dependence, and a number of other things like that. So I think this has been a really nice example of integrating behavioral and biomedical research in a transdisciplinary way.
But let me finish by saying a little bit about COVID-19 and some of the things that I think are important here as well in terms of that same integration of biomedical and behavioral research. I said this in a blog recently, and I said it in the perspective of this is a really simple way to think about it.
But with my infections disease colleagues I often say, “You tell me what you need people to do, and I’ll tell you how to help them do it.” What sort of things will get them to do the things that we’d like for them to do? Keep in mind that most of the current mitigation efforts we have right now are social/behavioral interventions, risk communication, hand washing, social distancing.
We need to be able to optimize adherence to those. And every day on the TV we see people who are not adhering to that, and how might we go about improving that adherence to social distancing? Balancing the cost and benefit with economic and social impacts.
And then there’s obviously some significant downstream health impacts from those economic and social impacts, especially in mental health and substance abuse, but also in physical conditions that have a significant psychosocial stress aspect to them as well. And then optimizing testing uptake, and then thinking down the road, optimizing vaccine uptake.
So I won’t go into detail on this slide because I’ve said most of this already. Some of the mitigation efforts that are important and why they are important. Our ability to actually do better modeling with better data, especially around some of the social and behavioral characteristics, economic impacts, and social impacts, that we need to be able to better measure and have better data.
Right now what we’re doing is mostly using data from prior flu epidemics as the base by which we begin to do that. But as we move forward, we are beginning to pick up more data that improves our precision with these models. Let me move on here.
The other that’s been a real, I think, important component of this work that if you think back to if this had happened 15 or 20 years ago, what would be our ability to deliver health care in a remote way, it would be abysmal. But because of the work that’s been done in telehealth and digital health, mobile health, we have abilities to be able to automate a lot of what we do, to offload and remotely provide access to healthcare in ways we’ve previously been unable to do, and to be able to reduce the healthcare disparities as a result of that.
Now, one of the things we have to be careful about is that the same people who are most affected and most disparate in terms of health outcomes right now are also the people with less access to broadband and computer systems and mobile technologies to be able to access these things remotely, so we need to be able to address those as well. Obviously, in the concept of doing testing, there’s an important component here to think about that has again to do with social and behavioral research.
I’m very grateful that my biomedical colleagues are working steadily and hard and fast on improving the speed and the platforms by which we do testing for SARS-CoV-2, but the other part of that is to recognize that testing uptake is important. The assumption that once we have adequate testing that people will, of course, want them is actually a poor assumption.
Back in the 1950s, we had tuberculosis screening buses. They drove from community to community and did X-ray screenings for tuberculosis to do screenings. And people realized after a while of doing that, that people weren’t coming. They weren’t being screened; they weren’t using the facilities. The convenience alone was not sufficient to get people to uptake the screening behavior. And that was the birth of the health belief model and things like perceived susceptibility and severity of illness and perceived benefits and barriers to testing that were important considerations as you try to improve the ability of people to uptake the testing that was going to be provided.
And then once the testing’s occurred, you’ve got health literacy issues you need to deal with with interpretation, the effects of that on mitigation behavior—so for instance if I test negative, does that now mean I feel comfortable breaking all the rules about hand washing and social distancing and other things in order to be able to do that? And then services and hand-off and referral. And again, particularly in health-disparate populations in which those referrals and resources are less available.
So if they end up in a situation in which they test positive, will they have those resources available to them? And if they don’t, why would they then subsequently go through getting tested if it doesn’t matter from their perspective? And then, obviously, like I said, complicated in rural and underserved communities where that’s even more of an issue to deal with.
I want to touch just briefly on psychosocial recovery from COVID-19. There’s a lot of research, obviously, now on the treatment and therapeutics and getting people through to recovery, but there’s also a lot of work that we need to do on what happens to them post recovery and during the recovery process. Not only physical recovery, but also psychosocial recovery.
So we have some literature about post-intensive care syndrome, in which people have difficulty with stress and depression and other sort of mental health factors as a result of coming out of intensive care, and a fairly significant intensive care experience, but that’s obviously exacerbated in the current situation by social isolation and the lack of family contact during those intensive care efforts under COVID-19.
And then, of course, as people recover, issues of stigma and also survivor’s guilt of those who made it versus those who didn’t. So there’s a lot of factors there that we need to be able to consider and be able to address as we integrate the biomedical and behavioral work together.
So just to briefly, just so people are aware of some of the things that are already out there. There’s a number of urgent competitive and administrative supplements in COVID-19 going on right now addressing a lot of the things that I just sort of walked you through. OBSSR leads one of those, but a lot of the specific institutes and centers have their own as well, and so there’s a lot of ways in which people who are existing grantees of the NIH can subsequently do more work in that area.
The other thing I wanted to mention briefly to all of you is that as this pandemic was coming along, we had people beginning to do surveys out in the field, and I was concerned we would have nothing but one-off surveys of everybody asking a slightly different question about social distancing, or handwashing, or any of those other factors, and as a result of that, we would not be able to do any data integration or comparison between survey samples.
So we quickly set up a survey item repository in which people could send in the surveys that they’re fielding for COVID-19-related variables, and then we would post them and make them available so that others could borrow from that and use what’s already out there and what’s already being fielded. And we had those in two places—in the Disaster Research Response (DR2), which is at NIEHS, and with the National Library of Medicine, and then in the Phoenix Tool Kit as well there are COVID-19 protocols. So those are all posted for people to be able to use what’s already out there and available, as opposed to creating their own yet again.
And let me just finish up by saying that I’ve been feeling like I’ve been skating where the puck was, not where the puck is going, and so I think we need to begin to think more forwardly about how we integrate behavioral sciences and the biomedical sciences related to COVID-19, whether it’s the unwinding of the mitigation efforts that are currently going on, or dealing with the backlog of elective care, or helping families manage complicated bereavement in situations in which people have died, whether from COVID-19 or just from other causes and have not been able to go through the natural sociological bereavement process that we typically have. Recovery complications that we talked about.
And then down the road, we’re going to also have to begin to address, once we have a vaccine, vaccine hesitancy and concerns about the vaccine and people who spread misinformation about vaccinations and that sort of thing and being able to address that as well.
So let me stop there. You can find me easily. This is my contact information and information about our office. Happy to respond to people if you have questions or concerns or anything, but I’ll open it now back, Judith, to you for questions.
Thank you, Bill, that was terrific. To start off, Bill, two closely related questions.
You talked a little about the challenge of merging the cultures, the biomedical and the behavioral. The first part of the question is, at NIH, for example, what do we do or what does your office do, to bring together those two sides of research to integrate them?
And let me ask the second part, because it’s probably all together, and that is, for trainees who are trained in the behavioral research area, how do they gain enough expertise in biomedical research to be successful in biobehavioral research?
So maybe you can address it both from the researcher’s point of view but also trainees as they’re
thinking of their futures.
Great questions, Judith. Thank you.
As far as the research efforts and how the NIH does this, I think we’ve been very fortunate at the NIH that at the early stages of OBSSR we had to make the case that behavioral and social science was important at the NIH.
I don’t think we have to make that case now, so I’ve felt very fortunate that Institute directors, NIH leadership, rank and file project officers across the span of the research that the NIH funds, whether they understand behavioral and social science specifically, they certainly understand its importance, they understand its value, they understand the potential for it to be integrated. And that’s been a very useful thing, at least in terms of receptivity to it, that I think has been useful.
And the other thing that I think, and this has been a problem for the behavioral and social sciences for some time, is we don’t act defensively. As an office, we go in and offer the things that we believe are useful to offer, try to be humble about it, not oversell what we think the behavioral sciences can do or can’t do, and try to be very clear about that. And in situations where I don’t think we have much to offer, we say, “We don’t have much to offer.” Places where we think we do, we sort of offer that up, but try to think of it more in a service mode.
And I think that’s been really helpful for the office as we try to integrate some of this work. So I think that’s at least how we’ve been doing it at the NIH in terms of doing that work. For trainees, I think it’s really interesting.
I think it’s important—I mean, this is almost like learning a language. Yes, you can study it in books all you want, but ultimately you’ve got to get in the trenches and you’ve got to work with biomedical colleagues and ask lots of questions to try to understand what they’re doing and why they’re doing it, what’s going on, and I think that’s an important component of what we have to do.
Being siloed in disciplines, I think, is increasingly less likely to happen in most academic settings. But even if you feel a little siloed, it’s important to get out and take classes in other disciplines, learn from other areas, do that kind of work. That said, I still think you have to get up to speed quickly in certain situations.
I remember I felt pretty comfortable when I went to NIMH because I had been doing work for some time in the mental health area and in the health risk behavior area, and I felt pretty comfortable. I’d worked mostly in psychiatry departments. I felt pretty comfortable when I got to the NIMH.
My next move to NHLBI I knew some of the same health risk behaviors were still important, but my understanding of cardiovascular science was next to nothing, and so this will sound like I’m really a geek, but on the vacation I had between NIMH and NHLBI, I took the textbook of cardiovascular science with me to the beach and I spent my week reading about cardiovascular function and cardiovascular diseases and all those sorts of things—not necessarily because I wanted to understand it at the level that my cardiovascular research colleagues did, but at least understand the language and understand the type of things they think about and the things that they worry about.
And I think that’s also important, that you quickly immerse yourself in the work that you have to do with your biomedical colleagues and try to get up to speed. Read their literature, look at research that they would typically look at and read periodically, and I think all of these things are important.
Sticking with training for a minute, what are, or are there, other careers for PhD students who are trained in the behavioral and social sciences, other than research?
Yeah, there’s a lot of areas. And I think we actually helped sponsor, along with NSF, a few years back a National Academies report on graduate training in the behavioral and social sciences that’s worth people taking a look at. It’s an easy workshop report to find on the National Academies page.
And our concern there was exactly what you’re asking about, Judith, which is we continue to train as if we believe we’re cloning more behavioral researchers; they’re just going to go into academia and do exactly what everybody else has done all this time.
In reality, that’s not what happens. We have more and more social and behavioral scientists that go into the private sector now than ever did before, particularly in some of the technology and computer science areas and that sort of thing. So that’s certainly one area where people had been doing this work and actually doing research; it’s just not research that gets published, but research in human factor analysis and lots of other things like that.
There’s obviously people who still do practice in the field—still also very important. I was trained in the Boulder Model, in which research influenced practice and practice influenced research. I don’t know that that model has held up as well as it probably could, but I really do appreciate my colleagues who do both. And there was a time when I actually did both, but it’s a hard thing to do.
It’s hard to be in practice and get the research done that you need to get done as well. But I think that’s a helpful way for people to think about that integration from research and practice also.
And then there’s obviously government work, which I think most people don’t think about, and policy work and those types of things. There’s a lot of people that I know who are behavioral and social scientists by training who are in more sort of government administrative positions, and you would say, “Well, what’s the rationale there?”
But it’s still mostly, policy is really primarily changing human behavior, just at a higher level than at an individual level. And understanding how people function, how they behave, how policies will influence or impact them, what the consequences are, including the unintended consequences, are all things that I think behavioral scientists have the ability to be able to do in that space as well.
We have some COVID-19 questions. One of the questions has to do with what areas in research on COVID-19—and I assume this is behavioral research on COVID-19—do you think have been ignored or under-discussed? And how can people begin to improve data collection from minority communities such as incarcerated people or undocumented immigrants?
Gosh, there’s so many. I don’t know that they’ve been ignored. I think there’s different levels of prioritization as people think about this.
I think for me, one of the things that struck me as this epidemic has moved along and as people have done what, from my perspective, are essentially social and behavioral interventions at the public health level, one of the things that has struck me is that they eventually come to figure out what it is that the research would have told them has been in the research literature for 20-plus years.
So one of the things that I think gets ignored is that we’re slower to make some of the implementations that we probably should make because people really aren’t looking at the literature. They’re still thinking of this as a commonsense solution as opposed to what does the research tell us about how people adhere to hand washing? And what does it tell us about paid sick leave and its impact on transmission rates and reducing transmission rates from the flu epidemic and those sorts of things?
I will tell you one of the things that struck me in the early days of this epidemic, Vice President Pence was saying at one point that he had talked to some governors, and they were talking about the importance of paid sick leave, and I had to keep my head from exploding because there’s truly 20 years of literature on the impact of paid sick leave on reducing transmission rates in the workplace. It’s there.
We have some pretty decent data on how well it’s quantified, those types of things. So when you ask about what’s being ignored, a lot of times I think it’s what’s being ignored is research that’s already out there that’s not being adequately translated into policy and translated into the things that we need to be able to do.
On the other question about the particular populations, it’s an area that the NIH has been really interested in, trying to see how do we do outreach in some of these areas where there have been particularly significant impacts, incarcerated individuals is certainly one. The issues having to do with the increases in rates among minorities, increasing rates of death from COVID-19 and those sorts of things, and trying to better understand…our assumption right now is that’s mostly because of the comorbidities that they carry with them into being infected, but it might also have to do with access to care and how quickly they get in and all of those various things as well.
So understanding that a little bit better, I think, would be important also. But I do think the question about incarcerated groups, if I step back from that a little bit, the thing that I think is important for all of us to think about is the community/sociological perspective about how it is that people get infected, and what goes on, and the networking phenomena that are a part of that. How people interconnect or don’t interconnect. How they live together or don’t live together. All of those factors are important in this transmission rate.
And if you look at most of our models, we use, appropriately because that’s what models do, they simplify all of that into some basic transmission rates, and that’s an area where I think, especially among the places where we’ve had the biggest problem, which is groups that are housed together—nursing homes, prisons and jails, those sorts of things—those are places where, in retrospect, we could have moved a little quicker to stop that infection rate at an earlier time.
Turning to another direction. You mentioned that one of the issues in your strategic plan was improving the pipeline from basic to translational research. So this is a two-part question again. Can you tell us what you mean by basic behavioral research, and how do you propose to improve the pipeline?
In two words or less. Basic behavioral research, all the work that we’ve done in attention and learning and memory is one example. Most behavioral interventions are learning interventions, right?
Understanding better how people learn, how they respond, the type of things…the incentive structures that work or don’t work for them. Recent research in self-regulation and attentional bias. There’s lots of research areas in the basic behavioral sciences that have to do with understanding why behavior occurs and in what context, and what are the factors that lead to people behaving or not behaving in certain ways?
So there’s a large basic behavioral and social science research portfolio at the NIH, even a larger one at the National Science Foundation. So that’s the kind of research we’re talking about.
And in the early days of behavioral interventions, we drew from basic research to generate the behavioral interventions that we now currently do. So when I was coming up in graduate school, I was learning about how to treat phobias and panic and obsessive-compulsive disorder based on basic operant and social-learning factors and basic research in that area that led us to the interventions that we developed as a result of that.
We still do some of that, but my sense is that that translation is not as strong as it could be. So one of the things that we’re trying to do with our applied researchers, you have to kind of do this on both sides, one is to make sure that basic research is thinking about what’s the plausible pathway to health, the health outcome that needs to be addressed, so that we’re targeting our basic research that we fund more clearly toward that pathway.
And that doesn’t mean people always have to do translational research, but they have to, in the back of their minds as basic scientists, at least think about a plausible pathway that will get us to some applied outcome, I think.
And then on the applied side, one of the things that I think we need to do is incentivize researchers more to be able to pull from that basic research that’s going on right now and develop novel and new strategies. Right now I feel like we do too much applied research that’s basically adaptations of existing interventions. We tweak them, we modify them a little, we add this piece, or we subtract that piece, but in terms of a new, novel component that hasn’t been done before, we don’t have as much of that happening.
So we’ve been beginning to work on how do we incentivize that a little bit more, and I think there’ll be more that we’ll be doing coming out in the next year or two that will have to do with that as well.
You mentioned just now NSF, so this is a good time to ask. Does OBSSR work with other government agencies? And if so, how?
We work with pretty much all the agencies that have a behavioral and social science component to them, which is quite a few. We do have a close relationship with the National Science Foundation. I serve as an ex officio member of their advisory council for their social, behavioral, and economic directorate. So we have that connection there, and we work with them on a couple of cross-cutting projects as well, including the graduate training effort with National Academy that I mentioned a while back.
We also have pretty considerable interagency work with the CDC, with HRSA, with AHRQ, with a lot of the more applied and public health end of the spectrum in terms of adoption and improving adoption of some of the research that the NIH funds to try to improve that effort. So we’ve done some of that as well.
And then there’s actually an interagency group of social and behavioral scientists that began under the Office of Science and Technology Policy, but things have changed recently but we’ve kept the group going, which is just representatives from all the various agencies—education, justice, Department of Defense, etc. And in all of those areas there are people doing really interesting social and behavioral research and also applying that research to practice.
In the risk area, for instance, the people at NOAA and the people in the weather service and that sort of thing think about this pretty often. The people in environmental safety think about this pretty often. So those are all areas where I think there’s a lot of interagency interaction in this research space.
Thank you. Question about intramural research in behavioral and social science. Which institutes and what kinds of things are they doing?
There’s a few. I would tell you I wish there were more in the intramural space in social and behavioral
science than we currently have. But I think there’s a good basic group of researchers in the intramural space doing this work. NHGRI has an entire branch looking at social and behavioral aspects of genetic testing and various aspects having to do with gene-environment interactions, epigenetic research, and that sort of thing, so there’s a fairly strong group there.
There’s a strong group at NIDA and NIAAA having to do with addiction and substance abuse and behavioral interventions for that, as well as behavioral assessment of those phenomena. So there’s a good group of people working in that area.
There are some people more in the epi-survey- modeling area, the population health research area, and they’re across the board. NIMHD has a strong group there. Fogarty has some people really doing some interesting modeling work in the behavioral sciences as well.
So there’s a spattering, I guess, across the various intramural groups doing this research. One of the things that our office has been working on is trying to get them connected and integrated with one another a little bit better. I think they all feel a little siloed in their own little institutes or the intramural space that they work in. So trying to connect them better. And then once we have that, again, trying to integrate better with some of the biomedical research that’s going on.
You mentioned a couple of times, the word “modeling” came up. We know that requires strong computational and statistical skills. Can you talk about whether and how behavioral training includes quantitative training? How much of this do students learn in their graduate studies?
They learn a lot, but I wish they would learn more. I will say in a graduate studies meeting we had, one of the people was talking about their quantitative social science program, and nobody remembered from A Few Good Men when it was asked whether it’s grave danger and he said, “Is there another kind?” For me, is there another kind of social science other than quantitative social science?
At the basic level, I think we get left behind if we don’t keep up with some of the more cutting-edge advances in computational, statistical, modeling, AI research that’s going on and using those for social science and behavioral science research. So one of the things I’m always telling trainees is whatever amount of quantitative research background you think you have is probably not enough and learn more.
I feel very rusty because it’s been a long time, but we’ve moved from statistical research, which is still a key base of the statistical analyses that we do, and those, of course, have improved and advanced over time. The computational modeling approaches, I think we’re seeing behavioral and social scientists using that more and more, and that’s great to see, and that also connects us to our computational neuroscience colleagues a little better as well.
So it’s nice to see that work, and that gets us more into the modeling space, and then the machine learning and AI approaches that have been going on have also been sort of a nice space where I’ve seen, particularly in the social sciences, some interesting work doing machine learning.
So in all of that, it’s not that you have to know all of that, it sort of feels like to me the same as when I read a cardiovascular textbook just to make sure I got up to speed. You don’t have to be an AI specialist, but you’ve got to know the language, you’ve got to know what it’s able to do, you’ve got to know what things are its weaknesses and strengths and be able to work with people who have those AI skills to be able to analyze the data.
You mentioned the use of mobile devices in various behavioral applications. Can you tell us a little more about that, especially in the current environment with COVID-19, how the use of mobile devices might expand for behavioral health?
Yeah, an area near and dear to my heart. This is one of the areas I feel like I still have a little expertise in.
After a while you become sort of a generalist and can’t remember what you had your expertise in. But the mobile health space, I think, again broadly defined—so it’s not just smartphones but it’s smartphones and sensor technologies and wearables and that sort of thing as well—has really exploded and really, I think, created almost a paradigmatic change in how we assess behavior so that what we typically did was ask people primarily, or we had to do direct observation of what they did in most contexts.
We now have this remote, unobtrusive, fairly temporally dense way of observing people, but observing them via the sensor technologies and the other things we have available to us through smartphones and that sort of thing. So the ability to be able to sense behavior, sense the context of that behavior, has really, I think, been a major change for the field.
One of the projects that we lead right now is called the Intensive Longitudinal Health Behavior Network, and that network specifically looks at using all the cutting-edge technologies we currently have available to us to monitor behavior and to monitor context in real time as closely as possible and temporally densely as possible to better understand the factors that lead to change within people over time.
Most of our data in the behavioral and social sciences is really individual differences between people and not differences within people over time. And so it allows us to do a much more fine-grained analysis of behavior and its contexts and its mechanisms and the things that change as we move forward. So I think that’s been a really great boon in that area.
And then, of course, the other thing that people mostly think about is mobile health as a remote, scalable intervention, and being able to use it for intervention purposes without having to have someone sitting right in front of you to be able to do that. That has, in some cases, in some situations, had good outcomes, in other situations more mixed and more modest outcomes.
So I think there’s a lot of work yet to do on how we improve the intervention that goes on remotely via mobile devices. But I think there’s a lot of promise that is still there. And the other thing I would just mention about that, marrying those two together does what we call just-in-time adaptive interventions in which we often talk about precision medicine, which is essentially a tailoring at baseline. We determine based on baseline data how we’re going to treat this person.
Adaptive interventions in the mobile space really is we’re collecting in real time data about the person and their behavior and the context that they’re in and adapting our intervention over the course of the intervention based upon the data that we’re collecting. And actually there’s some AI applications of that that have been really interesting to see as well. So it’s the marrying of the intense behavioral assessment we can now do with those devices and then delivering interventions that adapt to that data as well.
I’m going to put you on the spot with this question. In the last several years, what would you consider to be some of the great triumphs of behavioral science?
I think there have been quite a few, if you think through some of the things that have happened. We’ve had sort of the old list. Let me start with a couple of the key ones there. I truly think that what we’ve done in terms of changing smoking behavior over the course of our lifetime has been a critical triumph and continues to be so. The one that’s now coming is can we do a similar thing with e-cigarettes?
In the diabetes prevention space, that’s been around for some time as well, but now out in the field and implemented in some regular basis to do diabetes prevention using some of these behavioral techniques. I think some of the work having to do with interventions in mental health conditions, especially—I mean, we started with anxiety and depression, but some of the research now, even in more severe conditions—schizophrenia and bipolar disorder—some of that research has been really nice work.
And again, I think one of the things that we’ve struggled with when you look at accomplishments is how much better those would be if they were fully implemented, if people actually did what the research suggested you ought to do in those situations. Our work in adherence has been, particularly in HIV, and of course the implementation effort right now in HIV is particularly around implementation of what we already know is effective.
So implementation science would be another area where I think there have been some triumphs both for biomedical interventions and behavioral interventions—how we go about making it more likely that people will implement them moving forward.
I think we have time for one last question, and this one goes back also to a career-related thing having to do with a career at NIH in the behavioral sciences. And by this I’m guessing that the person is talking about extramural kinds of positions rather than intramural research. Can you say something about that?
There’s a really nice cadre and network of social and behavioral scientists that work at the NIH in the extramural setting. They work across almost all the institutes and centers. One of the nice things that you can see from my career path has been that you can bounce from institute to institute because social and behavioral sciences are applicable across a lot of those, so you don’t have to be institute-specific or disease-specific in where you think about potentially landing at the NIH.
And the only thing I’ll mention as well, Judith, is when I left the private sector to come to the NIH, I thought, “This will be a nice sabbatical. I’ll do it for a few years, and I don’t know that it will be that interesting, but I’ll at least learn how they work on the other side and I’ll go back out.”
And here I am 15 years later. I found it particularly challenging and important, and the ability to shape the field and work with the community more broadly has just been a wonderful aspect of what I do on a daily basis.
We’re glad that you’re here and that you stayed. And with that, I want to thank you very much. This was incredibly interesting. I hope everyone enjoyed it. Thank you all very much.
Thank you, Judith.
