Read the Full Video Transcript

Ruchika Talwar: Hi, everyone. Welcome back to UroToday's Health Policy Center of Excellence. As always, my name's Ruchika Talwar, and I'm really excited to be joined today by Dr. Kristian Stensland, who's an Assistant Professor of Urology at the University of Michigan. Dr. Sten, thanks for being here with us today.

Kristian Stensland: Hi, Dr. Talwar. Thanks so much for having me. Appreciate the invitation to chat with you all today.

Ruchika Talwar: You have a big background in implementation science, is that correct?

Kristian Stensland: That's correct. Working on it anyway.

Ruchika Talwar: Tell me a bit about implementation science. What exactly is this new emerging field?

Kristian Stensland: Yeah, thanks. Implementation science is the next natural evolution of health services research, and it is kind of taking the baton handoff from what we know works, and trying to figure out how we can adapt what we know works to work in different contexts. We're taking what we know works for some people or works in a clinical trial setting, and then figuring out what works best for whom at exact times and why.

Ruchika Talwar: You wrote recently a really interesting piece in the Journal of Urology outlining how implementation science can be applied in our field in a variety of ways. I loved this piece. I thought it applied to private practice or community urologists in some ways. It applied to the academic setting, researchers in some ways, and it even applied implementation science strategies to solve some of our workforce shortage issues that have been plaguing medicine in general. Why don't we start by going over some of those examples and seeing how we can use implementation science strategies in our field, specifically?

First off, I want to talk a bit about supporting evidence-based androgen deprivation therapy for prostate cancer. Obviously, ADT strategies have been a big discussion in urology, actually both in localized and advanced prostate cancer, because they have been used in high-risk localized disease, as well. Tell me a bit about how you think implementation science can be applied to improve our use of ADT.


Kristian Stensland: Yeah, great question. Thanks. I think this is a great example, and in a way it's kind of the reverse. Implementation science we tend to think of as trying to get people to use something, but we can also look at it as a way to de-implement things or identifying low-value care and really trying to step away from that.

I think one of the areas that we're probably most accustomed to talking about in this setting is in active surveillance for prostate cancer, actually, and then using that as the analogy for ADT where really trying to curb that overuse and identifying strategies to reduce overuse or misuse of certain therapies. This specific range of research is something that one of my mentors, Dr. Ted Skolarus, who's now at the University of Chicago, has really focused his career around, and that's looking at strategies to implement low-value ADT use, things like in the primary setting for primary localized prostate cancer. There's really not a role for androgen deprivation therapy there. We can say that and we can show the evidence that that is the case, that there's really not a benefit and really just harm in there.


But then in identifying ways in which we can get people to stop using ADT in that setting is where the challenge hits. Similar to ways we've tried to identify the barriers and facilitators, so kind of the lingo for that of stopping use of aggressive therapies for very low-risk prostate cancer, we're similarly trying to identify some of those barriers and facilitators to stopping the use of primary androgen deprivation therapy or these low-value settings of ADT. What we've done there is used implementation science frameworks to try to identify why some people might be reluctant to stop those prescriptions or to de-prescribe those. Then we use that to feed forward into strategies to try to de-implement that. Now we're comparing those in a big cluster randomized trial within the VA system.


Ruchika Talwar: All right, that makes a lot of sense. But just to break it down even more, can you walk me through an example of a way you specifically used implementation science to solve this problem?

Kristian Stensland: Yeah, sure. Typically, if you're trying to design just an intervention off the bat, you might completely mistarget something, so you're not really targeting the right behavior. In this example, we were trying to identify specific barriers that someone might have to not de-prescribe. You have someone coming into your clinic and you're saying, "Why would I not stop that ADT for them?" That fed into a strategy of maybe people need support, or they need support from their peers for that. So we said, "Well, what if a good strategy targeting that could be, we have a virtual tumor board kind of a setting, where other people could review that case and then give you the backup so that when you talk to this patient about 'Hey, here are the harms, here are the potential benefits, here's what some of my colleagues said,' as opposed to relying just on your own expertise." That's one way to overcome that barrier.

Whereas on the flip side, if people just aren't identifying that it's a problem, then having something like a hard stop from your pharmacy and a little pop-up, which can be annoying, but having a pop-up that says, "Hey, do you really want to go forward with prescribing this ADT? If so, please provide the justification." That may be just enough to overcome that barrier to de-prescribing. We're specifically targeting those two strategies, and then comparing them to see how they work compared to each other.


Ruchika Talwar: Got it. What I'm hearing is that it really involves digging down to some of the root causes of why these low-value services are being offered, and then coming up with interventions at multiple layers to try to make sure that we can, in this case, I would actually argue maybe de-implement something that is not a high-value care strategy.

Kristian Stensland: Exactly. The advantage over just throwing things up against the wall and seeing what sticks, is being able to find things that are specifically targeted and then acceptable to the providers or whoever we're targeting that behavior change to. Instead of just automatically throwing out the best practice advisory or putting a hard stop at the pharmacy level, we want to know why that might work and try to foresee some of the issues that come up. This has a lot of similarities to a typical kind of a root cause analysis or other techniques you might use in quality improvement. This is just trying to wrap around behavior change strategies and other formal, more rigorous frameworks onto that kind of a QI backbone.

Ruchika Talwar: Yeah, and it can be applied, like I referenced earlier, in a variety of settings. For example, if a practice is seeing increased spend in an area that doesn't correlate with either better patient care or more efficient care, you can apply these same strategies to figure out why this is happening and try to get rid of that wasteful issue in the system.

Another really interesting example that you had in your editorial that you wrote was in the space of clinical trials, which I think is a huge area of need in both basic science targeted clinical trials and new combination therapies for prostate cancer, new agents in urology, in general. We know that although, using prostate cancer again as an example, prostate cancer affects black men at higher rates and they have poorer outcomes, we still have lower rates of that population in our clinical trials. In my mind, this is a big problem we need to fix. Talk to me a bit about how we can use implementation science strategies in this sphere.


Kristian Stensland: Yeah, it's a great question. This is really where most of my research is focused on, is improving clinical trials. I think you bring up some great issues where a ton of clinical trials fail, unfortunately, and it really wastes a lot of the investments we have both of patient time, physician time, provider time, trialist time. Trying to find the best way to improve all of that is very tough without a rigorous way to establish and improve really the science of clinical trials.

A lot of my work focuses on trying to improve the cognitive efficiency and equity of clinical trials by applying these frameworks. The historical ways that we use to try to improve enrollment and try to reach out to, as you mentioned, specifically black men for prostate cancer trials, make a lot of assumptions and then try to address those barriers without really assessing them first. There are a lot of assumptions made. For example, it's often assumed that people don't have trust in the medical field, essentially, and there's a lot of questions about the ethics and trust in clinical trials specifically. That may not actually be the barrier that's preventing people from participating, and trying to identify more systemic barriers.


Of course, the systemic racism inherent in a lot of our clinical trial structures needs to be addressed, and identifying strategies coupled with people who are actually going to participate in trials so we can get to the root of that, and then identify strategies to overcome them is what we're hoping to do. A lot of my prior work has been adapting frameworks from implementation science to be able to really get down to targetable interventions that we can do to both improve enrollment, in general, improve the design and conduct of trials, and reach out to underserved populations who are really underrepresented by clinical trials as they exist now.


Ruchika Talwar: Yeah, again, I could not agree more with you, and I look forward to seeing your emerging work in this space because it's an area of significant need. What is your message to urologists who are interested in entering this research space?

Kristian Stensland: Again, this is a very new field, and I really encourage people to try to engage with it. You can get off into the weeds, into the very minutiae of a lot of these frameworks and things, but I think taking the general approach. There's this great paper on implementation science, it's just implementation science made almost too simple. It describes things as trying to find really the root reason why the things that we're trying to do don't work well, and how do we do those things better?

I think trying to get down to that root cause, whether you use formal implementation science frameworks or not, really trying to assess the context of "Why is this not working in this specific context or for this specific population? Why is this not working for our providers," is going to lead you down a road, I hope, of improved quality delivered to our patients. I think even just starting from there, you can go as deep as you want in there, but I think there's a lot to be done just with context assessment, and trying to tailor better to what our patients or providers need.


Ruchika Talwar: Yeah, and sometimes the answer is a lot more simple than we think. I think really leaning into this framework of breaking down why providers, doctors, NPs, APPs do the things we do in patient care, as well as why patients might have barriers to obtaining their care can often not be as complicated as we think, so I commend you on the work that you're doing. I'm really excited to read more about implementation science strategies that are built into our field of urologic oncology and urology, in general. Thank you so much for chatting with us today. I learned a lot and I know our audience will, as well.

Kristian Stensland: Yeah, thank you so much for your time and the invitation.

Ruchika Talwar: To our UroToday audience, we'll see you next time.