Neighborhood Deprivation, Race and Ethnicity, and Prostate Cancer Outcomes Across California Health Care Systems - Isla Garraway

June 3, 2024

Isla Garraway discusses her team's study that explores how neighborhood deprivation, race, and ethnicity impact prostate cancer outcomes across California healthcare systems. Dr. Garraway highlights the significant disparities in prostate cancer incidence and outcomes, particularly among Black men and those in deprived neighborhoods. The study compares outcomes between veterans treated in the VA system, which offers open access to care, and those treated in community care systems. Results show that racial and socioeconomic disparities are less pronounced in the VA system compared to community care. Dr. Garraway emphasizes the importance of accessible healthcare and the need for further research to understand and reduce these disparities. She notes the potential benefits of standardized clinical pathways and additional support services in the VA system. The study aims to inform strategies to mitigate disparities and improve prostate cancer outcomes for all patients.


Isla Garraway, MD, PhD, Associate Professor and Director of Research, Department of Urology, UCLA School of Medicine, UCLA Health, Los Angeles, CA

Andrea K. Miyahira, PhD, Director of Global Research & Scientific Communications, The Prostate Cancer Foundation

Read the Full Video Transcript

Andrea Miyahira: Hi everyone. I'm Andrea Miyahira at the Prostate Cancer Foundation. With me is Dr. Isla Garraway, a professor at UCLA in the VA Greater Los Angeles Healthcare System. Dr. Garraway will discuss the recent paper from her team, "Neighborhood Deprivation, Race and Ethnicity, and Prostate Cancer Outcomes Across California Healthcare Systems," published in JAMA Network Open. Dr. Garraway, thank you for taking the time to share this study with us today.

Isla Garraway: Thank you, Andrea. I'm very happy to be here and thanks to the Prostate Cancer Foundation, who of course supports so much of the work that we'll be presenting today. Again, thank you for this opportunity to share the work of our latest paper that was just recently published last month in JAMA Network Open. It deals with neighborhood deprivation, race and ethnicity, and prostate cancer outcomes across California healthcare systems. So, similar to many other cancers, prostate cancer does not affect all individuals in the same way. There are critical differences according to different population groups. We call this oftentimes prostate cancer disparities when different groups are impacted in different ways related to their prostate cancer incidence. Basically, the rate at which they develop prostate cancer as well as their outcomes, which basically means once they have a diagnosis of prostate cancer, how do they fare?

How do they compare in terms of their survival, in terms of their progression, and their overall outcomes relative to other patients? So as we know, prostate cancer incidence is extremely associated with age. It occurs in older men. The average age is in the mid-sixties to develop prostate cancer and that risk increases with each decade of life. Another group that has a higher association of prostate cancer incidence is based on race and ethnicity. Black men have a much higher incidence of prostate cancer compared to other racial and ethnic groups. Finally, ancestry is also at play here. Men with a family history of prostate cancer and men of African ancestry have a higher incidence of prostate cancer than other groups. As far as outcomes go, Black men, again, tend to have worse outcomes than other racial and ethnic groups. Individuals that have lower socioeconomic status or reside in specific areas of the country or in specific neighborhoods that may be more deprived or rural also tend to have worse outcomes than other individuals or other groups.

So it's important to consider these disparities and how we can reduce them and intervene to make prostate cancer outcomes better. This data comes from the National Cancer Institute's SEER database, which accumulates data from cancer registries all across the United States. Basically, what we see here is this demonstration of year-over-year disparities based on race and ethnicity. You can see the clear outlier is the group in brown, and that represents Black men in the United States. They have a clearly higher incidence of prostate cancer. They also have a higher incidence of advanced disease at diagnosis. We call that metastatic prostate cancer at diagnosis. And then their outcomes also are worse. You can see that there's higher prostate mortality rates in this population compared to other groups. Basically, one in seven Black men are going to be diagnosed with prostate cancer in their lifetime versus one in nine white men.

Now, when it comes to measuring associations of socioeconomic status and prostate cancer outcomes, we can evaluate individual factors like income and wealth, and we can also evaluate factors associated with the neighborhoods where our patients reside. These neighborhood deprivation indices measure things like transportation access, the proximity of healthcare centers to where our patients live, income and employment levels in the community, as well as education and environmental exposures. They kind of composite all of these different factors into one measurement and then we can basically divide those rankings into groups and evaluate any associations of these rankings of socioeconomic disadvantage of a neighborhood and prostate cancer outcomes. There are two different ways that we have looked at these measurements. We've used the area deprivation indices to measure neighborhood deprivation among our prostate cancer patients, and we've also used neighborhood socioeconomic status indices. The difference is the kind of level of granularity that you have with these different measures.

The area deprivation index is the most granular measure because it measures at the block group level. That's when you literally just have a few blocks where patients live linked together, and it really is reflective of what that community is like. On the other hand, you can also use the census tract measurement, which is a little bit larger than the block group, but still a pretty small measure when you're thinking about these kinds of geographic regions compared to things like county level or zip code level or state level. It really does kind of get down to what a neighborhood might really be like. There have been some studies across different cancers looking at the association of the neighborhood deprivation level and cancer survival, including studies in prostate cancer. This is another paper that was published in JAMA Network Open a couple of years ago that looked at the association of neighborhood deprivation and prostate cancer survival.

What you can see is in neighborhoods that are least deprived, so that would be the first quintile, you see much better outcomes compared to the neighborhoods that are most deprived in terms of socioeconomic status measures. So that's the Q5 or the lowest quintile. And you can see this really big gap in survival in patients with prostate cancer in these cases. Now, one thing that's really important to consider when we're looking at prostate cancer disparities is the connection between some of these factors. Historically, there has been an association between race, ethnicity, and socioeconomic status. This is a nice figure that kind of really demonstrates this by looking at the median family wealth according to race and ethnicity over several decades. You can see that white families have a much higher degree of wealth than African American families or Hispanic families. This gap has been very prevalent for many, many years and it's important to consider this association between race and socioeconomic status.

So that led us to our key question when we're really trying to differentiate these factors that are associated with prostate cancer outcomes. Do racial, ethnic, and socioeconomic status disparities vary among individuals with prostate cancer that reside in the same neighborhoods but seek care in different healthcare systems? So in other words, what we want to kind of differentiate are disparities associated with race from those that are associated with socioeconomic status by examining prostate cancer patients that are treated in different types of healthcare systems. So if we think about healthcare systems, we can look at healthcare systems like the Veterans Affairs Healthcare System.

This is what we consider an open-access healthcare system because the healthcare is free to most veterans, so they don't have any costs out of pocket to access this healthcare system, as long as you're a veteran and you meet certain requirements. These healthcare systems are located throughout the country, which also improves access to this healthcare system. You can compare that to other types of healthcare systems in our communities, such as academic healthcare systems, state and county hospital systems, or other nonprofit hospital healthcare systems where essentially you are often required to have insurance or pay cash to access healthcare in that system.

And then, of course, you have the state or county hospitals, which are more of a safety net system, which also can be harder to access because of the number of patients and the high demand for those services. So we wanted to compare what happens in the federal system, the VA system, compared to some community care systems. So that's what we did in our recent publication that was published in JAMA Network Open. And so I'm going to walk you through a couple of the key findings of this study. First of all, I want to talk about the cohort that we put together. So again, what we decided to do is we wanted to compare the outcomes of patients who were treated in the VA versus those who were treated in non-VA care. But again, we wanted to match patients based on where they lived so we can compare the socioeconomic status of their different neighborhoods and the correlation with different outcomes that we've previously seen.

In the VA population, we focused on the VA where I serve, which is the Greater Los Angeles VA in Los Angeles, California. And so we had about 2,000 patients that we treated over a period of about 18 years that we were able to focus on. We had long-term outcomes for these patients, and we were able to match these patients with those patients who lived in the same block groups that they lived in but were receiving care in the community for prostate cancer. So to find these patients, we were using the California Cancer Registry Cohort, and there were nearly 50,000 of those patients that we could intersperse and compare their outcomes with the VA patients. So essentially, this is the map of California as you can clearly see, and this basically represents where these patients came from that we matched in VA care as well as non-VA care.

And of course, most of them are concentrated in Southern California where the Greater Los Angeles VA is. But there are also other regions in California where we're treating patients at the Los Angeles VA that we also matched according to their location from the California Cancer Registry. And then basically, these maps just show different kinds of attributes of the communities where they live, like the population density and different socioeconomic status measurements of these different areas that were being treated at the VA or in community care. So one of the most striking findings that we found in this paper, which you can see from the graph shown here, these are basically just basic Kaplan-Meier curves, which basically are just survival. It's looking just at simply how long do patients live with prostate cancer. And it's stratified by either their race, which are the top curves, or their neighborhood where they live in terms of the socioeconomic status or neighborhood deprivation index of that neighborhood.

So one thing that should be clear from these two graphs is that the lines on top, all the colored lines, are just totally interspersed with each other. They're all intertwined. They're not clearly separated from each other. And that's both in the top graph, which represents race, as well as the bottom graph, which represents the neighborhood deprivation level. Just no correlation between where you came from in terms of where you lived and survival or your race or ethnicity and survival. And these are coming from the VA population. This is in total contrast to what you see in patients who live in the same neighborhood but are receiving care in the community. Here you see what we classically see and what we've seen for decades, which are the racial and ethnic disparities. So now you can see clear separations of the blue, red, and green lines with the worst survival in our Black patients.

The same thing with socioeconomic status, which is measured through the Neighborhood Deprivation Index. You can see that patients with a higher socioeconomic status reflected by this index have the best survival and those with the lowest have the worst survival. This is very similar to the data that I showed you a couple of slides ago from a different group that was looking at prostate cancer survival and area Neighborhood Deprivation Index. So again, very different results when you're looking at patients who have an equal access open care system versus those who are accessing care through the community. Now, when we do a more sophisticated analysis, meaning that we control for other patient characteristics to try to make the patients even look more similar, such as the age when they were diagnosed with prostate cancer, the year they were diagnosed, and other factors, we basically see the same kind of result.

So on the top, you can see the results from the patient from the VA in panel A. You can see the overall or all-cause mortality rates. Basically, what you're looking at here is a graph. You want to see where these bars cross the red line. If the bars are crossing the red line, that means there's no significant difference based on race or based on where they lived in terms of the degree of neighborhood deprivation. So everything crosses the line in the VA population. In contrast, when we're looking at the California Cancer Registry population and these patients that live in the same areas as where veterans live, you see a lot of things are far away from this red line, meaning they're strongly significant. Including the overall survival, which is what we saw in that previous graph, which is much worse depending on how deprived the neighborhood of residence is, the prostate cancer-specific mortality, and also even the stage at which patients present.

So stage patients with metastatic disease at diagnosis, it's much more likely in patients who are living in more deprived neighborhoods in the California Cancer Registry, but not the case in the VA patients. So just to summarize the findings of these data that I have shown you, statistically significant associations were not observed between low socioeconomic status indices like the ADI or socioeconomic status index and adverse outcomes in the VA population in prostate cancer patients. In the California Cancer Registry, however, significant associations between neighborhood deprivation and adverse outcomes were observed in patients who were geographically matched to our VA patients but receiving community care, non-VA care.

So the racial disparities of Black, white disparities were attenuated in the VA population but were broadly observed in patients receiving community care. So what do these differences suggest when we're looking at patients who are receiving care in the VA versus patients who are receiving care in the community? Well, basically what we think is that interventions that might target access barriers that are experienced by patients with prostate cancer who have low socioeconomic status may mitigate these neighborhood deprivation and racial disparities. So really focusing on targeting access to care barriers may be important in reducing these disparities. And why is this important? The reason why this is so important is because there are so many factors that affect or are likely to affect prostate tumors. There are factors that can affect the actual tumor cells and the biology of the cells, and basically how the tumor is going to behave, whether it's aggressive or indolent.

And there are other factors that are probably equally important and sometimes more important that affect a patient's ability to access the healthcare that they need to effectively treat or manage their cancer. So there are some factors that are just kind of in one area, but some factors that kind of cross both lines that affect both biology as well as access to care. So we really need to understand the role and the weight of each of these factors so that we can target our interventions to improve the lives of all veterans and non-veterans who have a diagnosis of prostate cancer. So I really hope that everyone will feel free to go to JAMA Network Open and take a look at our paper. I'm very happy to discuss more results in detail. So thanks again for the opportunity to present.

Andrea Miyahira: Thank you so much, Dr. Garraway, for sharing this really important study. So are you able to tell at what stage in the prostate cancer diagnosis and treatment trajectory that disparities become apparent, whether it is initial access to care? So for instance, being screened and diagnosed versus access to equal care after diagnosis, which is causing the disparities that you're seeing?

Isla Garraway: Yes, this is a great question. So, yeah, the question is basically at what point do we start seeing these disparities? I think the answer is both. I think that one of the issues related to access to care, which is associated so strongly with socioeconomic status, is the ability to have access to PSA screening, for example, our early prostate cancer detection. So I think that's reflected in the results that we see in the community where you see patients who present with more advanced disease at diagnosis. It's possible that they don't have the same access to early screening and early detection in certain community situations. Not all but some. And so that definitely is a factor. And the same thing with outcome. If there's not the ability to access timely facilities and specialists that can really help treat your cancer, then it might translate into worse outcomes in terms of overall survival and also progression.

Andrea Miyahira: Okay, thank you. How can lessons from the VA equal access model be implemented in other healthcare systems to reduce health disparities?

Isla Garraway: Yeah, that's a really great question, and one of our next steps is to really understand that a little bit better because right now we don't know... I mean, we presume that the open access system definitely takes down one of those barriers in order to be able to have access to screening early detection and timely prostate cancer treatment. But there may be other factors at play here as well because the VA doesn't only offer access to care for the veterans, there are also many other services that could benefit the patient being able to get the care that they need. For example, there's transportation available, there's housing available, there are other kinds of support systems available through the VA. Another factor is that the VA has a lot of standardized clinical pathways that are accessible to providers so that the care is kind of standardized and that might also contribute to better outcomes overall.

Andrea Miyahira: Okay. And what are some of the key components that are needed to reduce racial and neighborhood socioeconomic disparities?

Isla Garraway: Yeah, so I think some of the key components are again, just really trying to make healthcare more accessible so that patients do have the opportunity to seek care early on and preventative and early detection approaches and plans. And I think that could be the key, but then we do really need to have more research to really understand other factors that might be at play here in terms of really reducing these disparities associated with socioeconomic status, neighborhood deprivation, and race ethnicity.

Andrea Miyahira: Thank you. Are you able to compare outcomes of patients at the VA with specific other types of healthcare institutions or networks and insurance types?

Isla Garraway: Not yet. And this is something we really would like to do and understand in more detail is exactly again, where these disparities are at play. We see that they're reduced in the VA, at least in terms of outcomes. We don't see that in terms of incidence. We still see the incidence of prostate cancer is much higher in veterans as a whole and also in Black veterans and other of the at-risk populations that we described. But it would be really interesting to see in the community exactly what hospital systems or healthcare systems are more impacted by disparities so that we can have a plan to mitigate those disparities.

Andrea Miyahira: Thank you. And what are the next steps in your studies?

Isla Garraway: Yeah, so I think as we just discussed, really taking a deeper dive into why we see these differences, both looking at the different factors that are making VA care so great. And I guess I'm a little bit biased since I'm a VA provider, but also looking at, again, across the different types of healthcare systems in California and seeing where, again, these disparities lie and how other places that have seen reduction disparities, what they are doing. The other thing to do is really look across the nation. We only focused the study in California looking specifically at one VA healthcare system. So we might've just been really lucky, and it might be different in different geographic regions and different states. So it really is important to understand the variation in these disparities so that we can understand why the variation is occurring, and for those places that have reduced disparities, what they're doing that might lead to this mitigation.

Andrea Miyahira: Okay. Well, thank you again, Dr. Garraway, for sharing the study with us.

Isla Garraway: Thank you.