Note: The I AM GPH podcast is produced by NYU GPH’s Office of Communications and Promotion. It is designed to be heard. If you are able, we encourage you to listen to the audio, which includes emphasis that may not be captured in text on the page. Transcripts are generated using a combination of software and human transcribers, and may contain errors. Please check the corresponding audio before quoting in print. Subscribe now on Apple Podcasts, Spotify or wherever you get your podcasts.
Aman: Folks, welcome back to another episode of the "I AM GPH" podcast. Now, have we ever wondered about interventions, the human behavioral influence on the public health world? Well, today's guest is gonna be a treat and educate us all about it. Professor Linda Collins is a professor at the Department of Social and Behavioral Sciences, and the director of CADIO, the Center for Advancement and Dissemination of Intervention Optimization. It's an amazing center here that is going to be hosting an event soon, alongside the GPH Department of Social and Behavioral Science. And the talk all about it and interventions today, we have Professor Linda Collins. Welcome to the "I AM GPH" podcast.
Linda: Thank you. It's really great to be here.
Aman: Glad to have you as well. Why don't we hop right in talking about interventions? What is an intervention? How does one even describe an intervention?
Linda: Yeah. Yeah, the way I think of interventions, they are a program that is designed to alter someone's behavior, so that it will prevent disease or possibly cure disease. So these interventions are hugely important in public health. When you think of the effect of behavior on health, millions and millions of lives are lost every year because of behaviors like cigarette smoking, excessive alcohol use, failure to engage in enough physical activity, illicit drug use, things like failure to stay in good control of diabetes. I could go on and on. There's many, many, many public health issues that are directly tied to human behavior.
Aman: Okay, so interventions is one side of it, kind of making a change for, if I understand correctly. So I believe this is all about humans at a large degree, right? So where does the human factor come into place? How would you describe that?
Linda: Yes. Well, of course, people want to be healthy, and most people know what they have to do to lose weight or to stop smoking, but it's hard to do those things. People sometimes need help gaining motivation or gaining skills. Sometimes, a pharmaceutical is involved. For example, if someone is depressed. Talk therapy may be very helpful, but also an antidepressant might be used in conjunction with talk therapy. So a lot of interventions are behavioral, but some interventions may also be what we call biobehavioral because they have both a strictly behavioral and also a biological or pharmaceutical component. In fact, most of these interventions are multi-component. They have a lot of moving parts. You don't use just one strategy. You use a number of different strategies that are trying to influence different aspects of the problem.
Aman: So that sounds amazing, right? Like, how we tackle these things. Could you give us some things you have seen and problems that you might have tackled or what we see in the world when it comes to this part of public health? How does one describe it for a person that doesn't know much about it?
Linda: Yeah. Well, I've been involved in empirical development of a number of different behavioral interventions. But one point I'd like to make is that we, in social and behavioral sciences, are very excited about a new approach for doing this. So intervention science has been around for probably 30 or 40 years. People have been taking an empirical approach to developing behavioral and biobehavioral interventions. And of course, behavioral theory is brought into this. A number of different intervention components are developed based on behavioral theory. And these components are put together to form an intervention that is used to treat or prevent disease. What we are excited about is a new approach that solves what is really a worldwide problem with these interventions? Many interventions are developed without regard for whether they're practical. They're developed for what you might think of as potency. So an academic intervention scientist develops an intervention and is really going for the most effective intervention that they can get, which, and you might think, "Well, why not go for the most effective intervention you can get? That sounds great, that's what you should do." But the problem is that the most effective intervention you can get is very often not practical. It might be too expensive, it might be too complicated, it might be too burdensome for staff. And what's been happening for many years is a lot of promising interventions have been developed in academic settings, and they never are implemented. Or if they are implemented, they're implemented in kind of a detuned way. A number of the components that were so important to the scientists who developed the intervention are taken out because they're not affordable or the staff can't do them because they find them too complicated or burdensome. And as a result, the potential of these interventions to reduce morbidity and mortality has just been severely compromised. And this is actually a worldwide problem. We have a visitor from Denmark in the CADIO Center this month, and I was just talking with her a couple of days ago. She's interested in interventions in the area of cancer survivorship that is helping people who have survived cancer to lead better lives after they're passed the chemo and all the awful treatment that goes along with it. And even in Europe, in Denmark, they're having the same problem where academics develop interventions, but then they never get implemented because they're just too costly or too burdensome. And if an intervention could be the most beautiful intervention you could think of and the most potent intervention you could think of, but its public health impact would be exactly zero if it's never implemented. So what we're excited about is a new approach called intervention optimization that draws on ideas from a number of different fields, notably industrial engineering. Now, in industrial engineering, they develop products all the time and they're developing products for sale, right? They want practical products that actually can be used. And so the idea here is you optimize the intervention. And by that, I mean you start by looking at factors like: What's the most this could cost and be implemented? How burdensome can it be for the staff so that they will really do it? How burdensome can it be for the participants? How long can it be to be practical? Questions like this. You start getting that information and then you can use special research methods to investigate the performance of individual components of the intervention, and then you select a set of components that gives you the best expected outcome that will work within those parameters that, for example, won't cost any more than say $500 or won't be any longer than 20 minutes to implement. So the idea is to develop interventions that are both potent and practical, and then those kinds of interventions have the potential for massive public health impact.
Aman: Could you give us some examples of these interventions?
Linda: Yes. So if you attend the event that we're talking about that's coming up, you'll hear some very nice examples of these kinds of interventions. But I've worked on optimizing interventions in a number of different areas. Now, it's still, since it's a very new approach and it hasn't been implemented in that many areas yet, but I can give you some examples. One is smoking cessation. There's been quite a bit of work in this area — in the area of smoking cessation. Weight loss in obese people. And I've worked on optimization of an intervention aimed at reducing risky sex and excessive alcohol use in college students and among others. At the event, you'll hear about an intervention aimed at getting people who are HIV-positive and not engaged in the healthcare system, engaged in the healthcare system and getting their viral load down by taking antiretrovirals properly. So those are just a few examples, but there's possibilities all over the place. I mentioned I was just talking to someone who's doing work in cancer survivorship. I know of people who are doing work in sleep. I know of someone else who is developing a special intervention aimed at people who are heavy drug users and HIV-positive. The people who use drugs often have cognitive impairment. And the issue there is to develop an intervention that can work for people who have cognitive impairment so that these folks can get their viral load down. So that's just a few examples, but there's so many. I mean, this approach applies to every area of public health.
Aman: You mentioned this one thing which stood out to me, right? Industrial engineering was one area. It seems like this part of public health is interconnected to industries we might not even consider that public health is a part of. What stands out to you when it comes to interconnectedness in our globe and public health with different industries, different mindsets?
Linda: Yes. One of the things I love about public health, and I should say I'm pretty new to NYU. This is only my third year here, and this is my first faculty appointment in a school of public health. I was in other schools before I came here. But I really love the interdisciplinarity of public health. This approach for optimizing interventions that I've been working on for a number of years draws on ideas from industrial engineering, behavioral science, health economics, and basing decision analysis. It integrates ideas from all of those areas. And in the past, I've had grants with engineers where we worked on problems together. So I really love doing interdisciplinary work.
Aman: Okay. Engineers. We've had different types of engineers on this podcast as well, and it shows us that how interconnected public health is, the whole mindset around public health is much beyond what the term health might be and how we consider that.
Linda: Oh, yeah. It absolutely is. And one thing about public health, it has a very kind of, "Let's get it done," attitude, which I think is really healthy. And sometimes, to get it done, you have to talk to people who are outside your area. And that's, I think, a very healthy thing.
Aman: Can you tell us more about CADIO?
Linda: Yeah. Yeah. So CADIO, as you said, is a Center for Advancement and Dissemination of Intervention Optimization, and we do a number of different things. Those of us who are faculty in the center are... M any of us are working on actually advancing the science of intervention optimization itself. For example, one of the junior faculty in the center is working on ways of decision-making. When you're looking at the performance of a number of different intervention components, you might actually have a very complicated decision-making problem on your hands. And this may be further complicated if you have more than one outcome that you're considering because the different outcomes may have to be weighted. Sometimes the results will contradict each other over the different outcomes you might be looking at. So it can become very complicated. And she's been drawing on ideas from basing decision science to develop an approach to help investigators who are facing those really complicated decisions. A number of affiliates of the center are applying intervention optimization in some really interesting areas. One faculty member of the center is applying intervention optimization in the area of child maltreatment, for example. And something else we do a lot is educate other scientists about intervention optimization. This is still a pretty new area and it's not taught in very many graduate programs. In fact, it's not taught in any graduate programs at all that I know of, not even NYU, although we are in the process of developing courses and we will soon be teaching courses in SBS. But we have a course on the Coursera platform. It's free to anyone who's interested in it. And we also do that... That's an asynchronous course. We also do synchronous trainings based on that asynchronous course. These are trainings for people who already have a PhD and are interested in retooling to apply intervention optimization ideas in their work. Intervention optimization is considered kind of radical by some people. It's quite different from the same old, same old that was done for many years. But since 2016, the amount of National Institutes of Health funding that's been devoted to intervention optimization has gone up by more than 400%. So there is a lot of interest being generated in this area now.
Aman: What about it has become radical? What makes people say that?
Linda: The MO in intervention science for many years has been to take all the components that go into a behavioral or biobehavioral intervention and then immediately test that package of components as a package in a pretty simple experiment where you would randomly assign some people to get the treatment and some people to get some kind of a control. Now, that is an important thing to do at certain points in the process. But the downside to sole reliance on that kind of an experiment is that you're not looking inside the intervention at all. You don't have any idea which of the components are performing the way you want them to and which are not. So first of all, you don't have any way to make the intervention better because you don't know which are the weak components and which are the strong components. You don't know whether the effect that you observed, if you do, in fact, observe an effect is due to maybe only one out of eight or 10 components. So you have a huge amount of waste in there, components that are doing nothing, but taking up resources and wasting people's time. And then this is, I think, probably the most tragic part. If that experiment, if the results of that experiment suggest that the package doesn't work, you have no idea why. So you're just kind of left with, "Well, this was a failure. And I don't know whether there's any components in there that are worth saving, and so I just have to go back to the drawing board." On the other hand, if you are using an intervention optimization perspective, you're not going to go directly to that kind of an experiment. Instead, you're gonna conduct an experiment, and there are lots and lots of experimental designs that enable this — that enable you to look at the effects of the individual components and also whether the presence or absence of one component might have an impact on the effectiveness of another component. That's an important question. And so you know which components work, which ones aren't working. You don't have to include any components that aren't working, and so you can develop an economical and efficient intervention. And there's a basic principle that we like to talk about, an intervention optimization called continual optimization. And this means that you keep improving the intervention incrementally over time, which you really can't do that using the traditional approach. So I wanna tell you something that's kind of funny. When I speak to intervention scientists, which I do a lot about these ideas, I routinely get people coming up to me after and saying, "You just blew my mind. This is so radical." People have posted memes on Twitter after I talk of exploding heads and stuff like that, "Linda Collins just blew my mind." So that's a reaction I get from intervention scientists. When I talk to literally anyone else, literally anyone else who's not in the field of intervention science, they always say, "You mean that's not how you do it already?" So-
Aman: Wow.
Linda: To everyone else, it just seems like common sense. But to people in intervention science, just because of the way they've been trained, these ideas often seem very radical, but once people get it, they never go back.
Aman: This is huge, right? This is very important for the industry, especially with the things that are happening in the world that the structure or the structure that we have been following are essentially doing more harm than good, even though it was there in the good place. So how do we change the way we think?
Linda: Yes. I agree with you that this has the potential to be a paradigm shift. And in CADIO, we are doing our best to help intervention scientists who wish to do so: to retool, so that they can use these approaches. Most intervention scientists have been trained in the simple experiment I talked about where you compare the package to a control. They haven't been trained so much in optimization trial design. And optimization trials, admittedly, they're more complex, but the scientific yield is enormous because you just learned so much about what aspects of the intervention are working and which are not. And I believe that if we start doing more optimization trials — which of course everything I do is aimed at trying to get people to do more optimization trials — that will start to build a coherent base of knowledge about what works and what doesn't work, which they have in other fields. Other fields have been building coherent bases of knowledge, but we have... I'm not saying we're not doing it now, but it's been very, very slow in intervention science. And I believe that this would really speed it up.
Aman: When reading about intervention optimization and the whole department, I'm curious to know the difference between intervention optimization and implementation science.
Linda: Mm-hmm. Yeah. First, I wanna mention that there's a great center devoted to implementation science in GPH, the Global Center for Implementation Science. So we have a lot of great work going on in that area. The two, I think, are very complementary. Implementation science is about the study of how best to implement interventions, as the name implies. And intervention optimization is about developing interventions that are readily implementable. Now, of course, once an intervention is developed and optimized, so that it's readily implementable, there may still be important things you can do to improve the implementation of the intervention. Also, I wanna mention that, in some cases, you can think of an intervention's implementation — the process of implementing it — almost like an intervention on the intervention. And if you look at it that way, then you can optimize that. So it's possible to use intervention optimization ideas to optimize intervention implementation, if that doesn't seem too meta.
Aman: Yeah, intervention inception, it seems like. And it's very similar to the radical approach that you were talking about, right? It seems like that's kind of what we're doing at this point.
Linda: Yes.
Rachel: So professor, I have one last question because a lot of people listening to this earlier... I want to circle back to this one thing. There are people that might be watching this episode that necessarily don't have a public health background, but are very interested in the concept of public health. They might be engineers, they might be artists, they might be from wherever all over the world. Is there more of a space for someone that might be a generalist? So there's lots of students that are multifaceted or people that wanna come to GPH that have come from culinary backgrounds, the dance backgrounds. We've seen all kinds of guests on this podcast as well. Is this a great area? Why do you think this might be an area for generalists to find their spaces rather than only them thinking it's a specialist?
Linda: I do think public health is a great area for people with almost any skillset. Of course, it depends on exactly what you want to do, but a lot of public health activities take place in community settings and the kind of backgrounds you talked about: the arts, dance, anthropology, for example. Those kinds of backgrounds, I think, can be extremely helpful. Now, to get an MPH, most of the time you have to take statistics, which I see that as a plus. And for me, that was always super fun, but I know that a lot of people find statistics a little scary. So I wanna emphasize that that's a part of it because it's a science. But I have to say, the people I know who I respect the most are really good at integrating ideas from different areas. And so I think that there's definitely room for integration of ideas from, really, pretty much every area of study that you can imagine.
Aman: Love to hear that. I'm sure a lot of people would love to hear that and probably get rolling on understanding what public health means to them. So let's talk about-
Linda: We have a great MPH program, I should say, in GPH. We have a really great MPH program.
Aman: Absolutely, and for those of you folks listening as well, I, myself, am in the engineering school. I host this podcast and I've learned so much hosting this podcast as well. There's room for everyone, spoken to all kinds of people. So if you're interested in something like that in public health, please check out any public health course for that matter. There's room for everyone in this industry. So professor, why don't we hop into talking about the symposium that's gonna be happening in a few weeks since... First off, for those of you that are listening to this one year after the podcast is published, the symposium is over at this point, but it might be happening again, who knows? But let's talk about what the symposium is.
Linda: Yeah, so this is an activity that's organized by the Social and Behavioral Sciences Department and also in partnership with CADIO. And this is something that... Keep an eye on this because we've been talking about doing this every year. We call it our think tank. We did one last year, it was on displacement last year, and this one is on intervention optimization. And there are plans to do one next year, too. So keep an eye on it because I think they're always super interesting. This one, as I said, it's about intervention optimization as an answer to the challenge of behavior change, which, as we started out by saying, is so important because so much morbidity and mortality is directly related to behavior and could be stemmed by the right kind of behavior change. So I'm going to start off this activity by giving a brief introduction to intervention optimization, and then we're going to have a number of different speakers. One will be Marya Gwadz from the School of Social Work; another one will be Dr. Jen Cantrell, who is in Social and Behavioral Sciences; Dr. Stephanie Cook, who is in Social and Behavioral Sciences; and Dr. Kate Guastaferro, who is in Social and Behavioral Sciences. And they're all going to talk about optimization trials that they have done in different areas. Each one of them is working in a different area. There will be time for discussion and questions and answers, and then we'll have a little reception. It'll be about two hours total.
Aman: So from what I understand, everyone should really attend this event if they have any interest, right?
Linda: Oh, yes, definitely. And it's going to be a hybrid event. Any NYU-affiliated folks are welcome to attend in person, but anyone else at all is more than welcome to register to attend by Zoom.
Aman: So for those of you that are listening before this event, the event will take place on April 19th from 2:00 PM to 4:00 PM, the panel discussion happens from 2:00 PM to 3:30, and it's followed by a reception for those of you that will be in person. But it's hybrid, remember? So we'll have the link to the event in the description for those of you that are interested because it's a great way to learn about intervention optimization, case studies. Industry professionals will give their opinions. And you also have Professor Linda Collins over there to share great insights for everyone. Professor Collins, what would you like to leave us with about intervention optimization as we end this podcast?
Linda: Oh, gosh. Just that it's such an exciting area and there's new things happening all the time in this area. Intervention science is just a great area to be in right now. So much is happening.
Aman: Thank you for giving us all your insights and educating us, and hope this event runs smoothly and really well. It seems like it's gonna be an exciting time.
Linda: Thank you so much.
Aman: All right, folks, we'll see you in the next episode. Till next time.