Charles Ryan: Hi, I'm meeting today with two winners of our Pfizer Health Equity Global Challenge Award. These are domestic winners from the United States, Dr. Annie LaVigne and Dr. Sara Alcorn. They're radiation oncologists, and they're going to discuss their topic, which is radiotherapy deserts in prostate cancer, impacted race, poverty, and rural urban continuum on density of providers and utilization of radiotherapy in the United States. A hugely important topic. Dr. Annie LaVigne is a chief resident at Johns Hopkins University School of Medicine in the Department of Radiation Oncology. And Sara Alcorn is an associate professor in the Department of Radiation Oncology and the director of palliative care in the Department of Radiation Oncology at the University of Minnesota Medical School in Minneapolis, Minnesota. Welcome to both of you doctors. It's such a pleasure to have you here and congratulations on winning this award. So I think I'm going to turn it over to you Annie and you're going to do the presentation.

Annie LaVigne: Well, good morning everyone, and again, thank you for the invitation to present our work here today. Over the past decade and a half, our awareness and understanding of existing oncologic disparities in the diagnosis, treatment and outcomes has only continued to grow. We have a rich body of literature that really speaks to the complexity of potential underlying ideologies with investigation of factors such as race, socioeconomic status, and geographic proximity to providers and treatment facilities. But it has become increasingly clear that each of these factors cannot solely account for the persistent disparities that are seen, suggesting the existence of a dynamic interplay between them. And it is precisely this interplay that has yet to be comprehensively explored and what we set out to unpack through this project, and importantly doing so with a meaningful level of granularity at the US county level. So to do so, we created an ecologic study model using a range of national databases, notably all publicly accessible except for the AMA Physician Masterfile, which we use to analyze 3,141 counties. We then define novel metrics for radiation resource density and oncologic need.

Physician and utilization to person at risk were operationalized as the number of attending level radiation oncologists and the number of billed Medicare beneficiaries to a hundred thousand persons at risk respectively, with persons at risk defined for all stage prostate cancer as men greater than or equal to 35 years old. Now, counties with high PPR and UPR were classified as those exhibiting greater than or equal to 2 standard deviations above the mean PPR and UPR densities. For oncologic need, counties with greater than 2 standard deviations above mean prostate cancer specific incidents, and or death rate were classified as hot zones. So one of the first questions that we were interested in was the potential associations between select epidemiologic and sociodemographic factors and radiation resources and utilization densities. And as you can see in this figure at the left on univariable analysis, there were greater odds of high PPR and UPR densities in counties that were metro and had greater percent black population. And in the case of physician to person at risk density, lower percent poverty was also associated with high density.

Now, when we looked at these same relationships on multi-variable analysis, all of the associations remained except for percent black population with UPR density. We hypothesized that this may be due to affect modification by the rural urban continuum variables, specifically the observed clustering of above average percent black population in densely populated metro counties, which may have the most nominal access to healthcare resources as this association reemerged for metro counties when we stratified analysis by metro versus non metro counties. And importantly, these findings seem to suggest that proximity to healthcare resources does not necessarily equate to access and receipt of care.

Next, we wanted to not only characterize hot zone counties, but also investigate potential relationships attributable to increased risk of this designation. In our initial analysis, we found that hot zone counties for both prostate cancer specific incidents and death rate were predominantly rural with percent black population within the greater than 80th percentile for black constituents. On variable analysis highlighted at the table at left higher hot zone county risk for both incidents and death rate was seen for non metro counties with increasing percent black population and increasing percent poverty. Moreover, as noted in the last two rows, we found an inverse relationship between the odds of hot zone county status for death rate and high PPR/UPR density. Most of these significant associations on univariable analysis persisted on multi-variable.

Now after this analysis phase, we then ask how might we convey these results in a manner that would enable localization and visualization of counties exhibiting the greatest mismatch between radiotherapy resources and utilization versus oncologic need, regions that we termed radiotherapy deserts? And what you can see here on the left is a series of maps illustrating interstate variation in hot zones and in high PPR and UPR densities. And in comparing these, you can begin to appreciate even at a glance mismatch in these distributions. What our formal analysis showed was a disproportionate mismatch in resource and utilization densities and oncologic need in non metro counties with greater percent black population. And the existence of such radiotherapy deserts is particularly evident in the south and mid east for prostate cancer. Yet we didn't want to stop here in conveying our findings. We wanted to develop an innovative way of enabling translation of this observational data to targeted interventions.

So to this end, our team constructed an open access interactive web platform, which I've displayed a screenshot of here. And of note, we've completed the entire outline analysis for all stage breast cancer as well with the attention to expand to other cancer types. And as you can see, the platform enables several levels of filtering based on investigative variables at several geographic scales from county and statewide down to individual counties. Moreover, there is incorporation of epidemiologic and sociodemographic data, which populates upon user selection of a region of interest. And here I've shown the data for Baltimore City.

Now, in making the platform open access and interactive, our intent is to support interdisciplinary use from providers, policy makers, and advocacy groups alike, and a customizable experience that allows for narrowing to geographic regions and patient populations of interest. So future directions with the platform are really rooted in translational epidemiology. We feel that translation of such observational work in health equity really calls for acknowledgement of historic and persistent structural inequities that have and continue to impact communities of black, indigenous and people of color and those impacted by historic and persistent poverty. And if we return to the notion that I previously mentioned, that proximity to healthcare resources does not necessarily result in increased care access or receipt then defining and addressing barriers to care in not only radiotherapy deserts, but also hot zone counties with high resource density may be a particularly impactful route.

So to that end, we plan to conduct focus groups comprised of patients and caregivers residing in selected counties and the interdisciplinary providers involved in their treatment. Through these focus groups, we hope to characterize unique and shared barriers between different patient populations, and we envision that such characterization will be well positioned to inform development and implementation of targeted interventions. So with that, I'd like to thank our entire study team and co-authors as well as of course, the Prostate Cancer Foundation and Pfizer Global Medical Grants for the award that has continued to fund our work. Thank you very much. Look forward to taking any questions from you. Dr. Ryan.

Charles Ryan: Thank you very much. Really interesting data, really interesting perspective. I guess what I'm seeing is a couple of key points, one is density of radiotherapists or radiotherapy does not necessarily correlate to high radiotherapy use in certain populations. Really highlighting, I guess the inverse of that, which is minority populations in rural areas are those with the greatest disparity or deficiency, I guess, in this access to care. But one of the things that comes to mind is when we think about prostate cancer, radiation therapy, at least for localized diseases is in many cases a choice that somebody makes between surgery or radiation. Not always, of course, because it depends on risk and stage and other factors. But do you have any sense that there's inverse data, for example, with density of urologists or use of surgery in localized disease, or is there any possibility that could be what's going on here as a confounding piece of data?

Sara Alcorn: I think that's an interesting question. Our data shows primarily the strongest data is for the provider density, I would say. So we are looking mainly at that association on a cross-sectional level with other epidemiological factors. So I don't know that we can answer specifically utilization as well as we would like to. Our utilization data is from Medicare utilization for radiotherapy, which does only look at a small sector of the utilization for prostate cancer treatment in general, of course. So I don't know that our specific data can answer that question on how it's balanced with surgery or other interventions. Annie, would you add anything to that?

Annie LaVigne: I think I would add that if we remove just thinking about radiation from the mix, that really one of the most striking conclusions I think from our analysis is the fact that we're seeing persistence in oncologic need in regions that have both low and high density of radiation oncologist and resources. And I do like to think that whatever barriers are creating those similar patterns such that someone who is two blocks away from our hospital here is not able to access our resources might also translate to ability to access surgical resources or the healthcare system in general. So, well, Sarah mentioned, I'm not sure we can specifically comment on that, I do hypothesize that there are parallel trends and that work to reduce barriers to care from a radiation standpoint will also translate and benefit other disciplines in the oncologic field.

Charles Ryan: Yeah, I strongly suspect you're absolutely right there. And it's not that radiation is low where urology is high or that there's some sort of bouncing effect. Interestingly enough, we do have people who are studying neurologists density and are reporting somewhat similar data on outcomes and availability of urologists, not to mention medical oncology, which will be another situation altogether given how the disease is sort of parsed out amongst the specialties.

So you've taken on a big task, you've got the whole country here, you've got the whole country map, and it's really fascinating data. Tell us a little bit more about how you will know that your focus groups and that your next step is going to be a success. What are the metrics for success on a sort more granular level that you'll say, Okay, we're onto something and how's that going to be replicable?

Sara Alcorn: So as far as our metrics for success, so we will be using our focus groups, again, both patients, providers, caregivers, anyone in that sort of multidisciplinary setting, and that does include urology and medical oncology as well, and we will be ascertaining a number of questions in terms of what they determine to be barriers to care. And we expect that'll be a very interesting dialogue in terms of how the difference in perspectives maybe perceived between these different levels of patient to provider, caregiver, et cetera.

Now, in terms of the success of this project, I think that a lot will depend on what is gleaned from those conversations, but that is sort of one step of many as you might anticipate. I think that success in this project for us would be dispersal of this information in a way that not only benefits from this small focus group, but any next steps that are taken by ourselves and honestly probably much more than ourselves, the variety of stakeholders that could be interested in this. So as far as the focus group, I think that that would be a stepping stone to further projects, but watching as other groups, stakeholders including policy makers, epidemiologists, public health, and both prostate and then as you said, breast cancer, using these data as a way to narrow their own focus, that will be, I think, our true measure of success.

Annie LaVigne: I think I would just stemming from that add that there's really two components to this. Right. And this work needs to account for the heterogeneity between patient populations and regions in this country that might be facing unique barriers. But there's also a component to it that I think we're hoping to unearth shared barriers to care. And so I think that ultimately we're hoping to make an impact by identifying those in tandem. And I would envision success being measured as Sarah alluded to, of changes at a policy level. So impacting more broadly perhaps on the level of these shared barriers. And then also the empowering of community partnerships. The way that this data and our discoveries may help to leverage and support those because I do think that avenues for interventions on that front are really where the work lies and where we're actually going to start to make an impact. And so I think we're very hopeful that our work in an innovative fashion will be able to support those in developing across the country.

Charles Ryan: So perhaps you could educate me on something that makes radiation oncology unique, which is in radiation oncology you have the interface of a provider, a person, a clinician who sees a patient and makes a treatment plan, and the technology that exists in the particular area. To what extent is that a potential barrier, especially in the rural areas? I know that there are radiation oncologists who treat patients hundreds of miles from where they actually live because they can do that. They can travel to a place 200 miles away, see patients for a full day, execute a number of treatment plans, et cetera. But what is the interface between the physician density that you've described and technology, and to what extent is it something that could be done by redeploying radiation oncologist in a geographically different way?

Sara Alcorn: I think that is a great question, and it hearkens back a bit to the question you had asked before about if this could be used to think about urologist density or medical oncologist density. And thinking about that question a little bit further in an alignment with this... For the former question, I think that especially maybe not in completely straightforward cases, but no case is truly straightforward. And when you have multidisciplinary options, you should have multidisciplinary team members. And so radiation oncology with technology being our key feature may be the most rate limiting step, if you would call it that. And so from that perspective, that is a unique way to visualize this platform. It may just be more of a better proxy for multidisciplinary team density if we are truly the rate limiting step. We did do some analysis as well on the level of the type of machine that was being used, and we didn't include that in our final analysis because we did not feel that our data set was clean enough to truly give us that information.

Now, it's very, very true that where provider density is low it's probably a ratio of one provider or two providers to one machine versus where radiation oncology density is high. It may be 17 providers to one machine or something of that nature. We can't really comment on that specifically. But as far as the type of equipment goes, yes, we would expect that what is limiting in the rural setting, especially where density is very low, the machine is going to be probably the limiting feature. Where density is high, it's a little bit harder to comment on. And anything else you'd add to that, Annie?

Annie LaVigne:  No, I think it's a terrific question, and honestly, one of the questions that really drove and inspired this project, because that is really where the groundwork for a lot of this was being driven, was this question looking at geographic distribution of radiotherapy resources in terms of machines and providers. And I think Sarah and I, in seeing all of this developing data and results emerge where questioning, it can't be just as simple as putting providers and machines in rural areas. Yes, that's a component of it and proximity. We know their disparities from a standpoint of patients having to travel great distances for care. But I think through our work here in Baltimore, understanding that we were seeing patients who again, live a few blocks away, who're still not being able to access and receive care, that it wasn't just a question of proximity. And I think that is really at the crux of our conclusions and what we're hoping to convey to the community is that there really is such a nuanced layering of this problem.

Charles Ryan: Also, correct me if I'm wrong, but the technology of radiation has really advanced very rapidly in the last 10, 15 years. And so if there's a certain area that has not had an update in their machinery in the last 10, 15 years, those patients could be getting inadequate care based on the 2022, 2023 standards, whereas they would be getting standard of care based on the 2008 standards at the time the radiation machine was put in. I don't know. That's more of a question. Are there individuals out there who are receiving therapy on antiquated, if I can use that term, machines

Sara Alcorn: Not only antiquated machines, but also because technology and innovation, of course takes time to propagate throughout the population. We can now treat prostate cancer in a much shorter timeframe for certain people than we used to. And while that may not be better than the longer duration of treatment, it certainly may be better in terms of quality of life for patients. So that is another thing that we see time and time again in our particular area of study as well. In medicine in general, of course, when new information is out, it may take 10 or more years for it to be commonly in practice. So the more providers there are, maybe they're slightly protected against that and the more isolated that provider is perhaps less likely they may be to undertake more innovative strategies.

Charles Ryan: Yeah, it's a fascinating area and it does recapitulate I think what we're seeing in medical oncology and urology. There's nuances to the different specialties in terms of the disparities we're seeing. With oncology, what we're seeing I think, is that there are oncologists all throughout the country, but how much prostate cancer they see is a challenge. And so the urologist or the oncologist who sees relatively few cases of prostate cancer may be more likely to be less up to date on exactly the particular therapies of the therapeutic strategy that one would employ, that we would consider to be standard of care or state-of-the-art today.

Annie LaVigne: So as we've continued to talk about not only disparities and resources, but then also this educational component that you mentioned, Dr. Ryan, I would just add that in our subset analysis that we've also been working on, looking specifically at advanced stage prostate cancer, we've also detected similar trends and disparities in use of palliative radiation. And so I think this work of ours is also very important in making sure that patients not only have full access and proper receipt of definitive care, but also from a palliative standpoint.

Sara Alcorn: And I would add to that that another study we had attached to what Annie was just describing at the time we were analyzing the radiation oncology alternative payment model, which was being trialed across certain counties in the United States and areas in the United States, and it appeared that there was a mismatch between radiation oncology need for palliative radiotherapy for prostate cancer, and the areas selected as trial RO-APM, radiation oncology alternative payment model counties. So what that told us is that information glean from this analysis of the RO-APM would not well reflect the counties at highest need for palliative radiotherapy. And we thought that was in particular of interest since if that model was enacted, it would perhaps not well attend to the needs of these people. Now, in reality, the RO-APM model is undergoing additional analysis, but we do hope that that sort of data is kept in mind to prevent there from being disparities in the level of care that can be provided, especially at these really high risk areas.

Charles Ryan: Yeah, it's a really interesting area because I would think that there's a bit of a self-fulfilling prophecy when it comes to palliative radiation. Because if I have a radiation oncologist next door to me and I have a patient in front of me with pain, I'm probably more likely to send that patient to the radiation oncologist for pa... Amongst other choices for how I'm going to palliate that patient, I'm probably more likely to choose the radiation approach, which I do and I believe in, as opposed to if I don't have that person available, I may just choose to do other medical analgesics, et cetera, and not think about palliative radiation therapy. So again, part of the cycle that you're describing, which is lack of density, corresponds to lack of delivery of care, which probably potentiates the disparity a little bit.

Sara Alcorn: And it's particularly important in the palliative setting because comparative effectiveness studies have shown that radiotherapy is one of the best modalities for treating painful bone metastasis, which are very common in prostate cancer better than, or at least from a cost and a comparative effectiveness perspective than palliative chemotherapies or medication therapies and surgeries. So you're absolutely right. I think particularly if we're thinking of our folks with advanced stage disease who have symptomatic concerns from prostate cancer, that proximity question becomes incredibly important.

Charles Ryan: Yeah, interesting. So it highlights a need in the US in general for advancing education, advancing commitment of clinicians to the specialty of prostate cancer. And that's something that we're hoping that these Health Equity Challenge awards will help highlight. Your work, certainly, will help do that, and it will dovetail nicely with some of the educational and communications efforts that we're doing with the Prostate Cancer Foundation. So I suspect we'll be talking a little bit more as time goes on, about how you can help in that messaging as you not only complete your academic work, but you deploy what you find to helping physicians deliver better care across the country. So congratulations again, and thank you so much for spending this time with us, educating not only myself, but also all the listeners on this really important area.

Sara Alcorn: Thank you for having us.

Annie LaVigne: Thank you so much again.