publication date: May. 22, 2020
Conversation with The Cancer Letter
Erin Kobetz: How Sylvester’s cancer outreach is used to monitor COVID-19 in Miami-Dade
Erin N. Kobetz, PhD, MPH
Associate director for population science and cancer disparity,
Sylvester Comprehensive Cancer Center;
Chief of Population Health and Cancer Disparities,
UHealth Oncology Service Line;
Program co-leader, Cancer Control Research Program,
Incoming vice provost for research, University of Miami
This story is part of The Cancer Letter’s ongoing coverage of COVID-19’s impact on oncology. A full list of our coverage, as well as the latest meeting cancellations, is available here.
Programs designed to meet the NCI Community Outreach and Engagement requirements for cancer center designation have positioned the University of Miami Sylvester Comprehensive Cancer Center to monitor the prevalence of SARS-CoV-2 in South Florida.
“At NCI-designated cancer centers, we have the potential to be at the forefront of helping drive solutions in a pandemic. This is typically outside the scope of what we do, but the community relationships we’ve developed are deep. And they can serve a purpose beyond what we want to accomplish for our COE requirements,” Erin N. Kobetz, associate director for population science and cancer disparity at Sylvester and incoming vice provost for research at the University of Miami, said to The Cancer Letter.
“And I think that’s what we found here in Miami at Sylvester: Our relationship with our catchment area allowed us to be a resource in a time of unprecedented need.”
This conversation is part of an informal series of stories, interviews, and commentaries that track cancer institutions as they seek to reopen, reorganize, and reinvent in the wake of the COVID-19 pandemic:
Three months after the start of the COVID-19 pandemic, the Seattle Cancer Care Alliance is ramping up plans for a comeback of cancer services (The Cancer Letter, May 15, 2020).
Health systems and academic cancer centers are cutting expenses to make up for operational shortfalls resulting from the pandemic—laying off employees, furloughing staff, and cutting salaries and benefits (The Cancer Letter, May 8, 2020).
Community oncology practices are experiencing a significant decrease in patient volume, as weekly visits dropped by nearly 40%, while cancellations and no-shows have nearly doubled (The Cancer Letter, May 1, 2020).
The COE program Kobetz leads at Sylvester has been working with the Office of the Mayor of Miami-Dade County to ascertain the prevalence of infection within the county’s 2.75 million residents.
The program tested 2,500 residents for antibodies. The sampling was random, and it was weighted across the county’s municipal statistical areas.
The program is currently on pause, following an FDA guidance mandating the use of 12 serological tests that meet the agency’s requirements (The Cancer Letter, May 15, 2020). Sylvester is in the process of switching to one of the tests listed in the guidance.
During two weeks of testing, 6% of participants were found to be positive for COVID-19 antibodies, which can be extrapolated to equate to 165,000 Miami-Dade County residents. This figure directly contrasts with testing site data, which indicated that there were 10,000 positive cases in Miami-Dade, suggesting that the actual number of infections is potentially 16.5 times the number of those captured through testing sites and local hospitals. Using the 95% confidence interval of 4.4% to 7.9%, this would estimate that the number of people infected falls between 123,000 and 221,000.
More than half of individuals who tested positive for the antibodies were asymptomatic in the seven to fourteen days prior to screening.
Individuals from African American and Caribbean communities were twice as likely to be infected with COVID-19 than other racial groups.
At this writing, Florida has had 46,944 confirmed cases of COVID-19 and 2,052 deaths linked to the disease.
Sylvester attained its NCI Cancer Center designation last July (The Cancer Letter, July 29, 2019).
Kobetz’s COE program relies in part on Game Changer vehicles, which bring evidence-based interventions to underserved communities in the cancer center’s catchment area (The Cancer Letter, April 27, 2018). The center’s cancer control program also includes the Firefighter Cancer Initiative, a long-term study of exposures to carcinogens and ways to reduce and prevent cancer risks for Florida firefighters.
Kobetz spoke with Paul Goldberg, editor and publisher of The Cancer Letter.
You run outreach and engagement at Sylvester. You are, in effect, the NCI person focused on disparities in South Florida. Focusing on disparities, what do you see now, in the middle of the COVID crisis? What do you see in terms study opportunities? What are the scientific questions that can be asked because of COVID?
The COVID epidemic has brought into national media dialogue the observed difference in the burden of COVID infection between Hispanics, blacks, and other minorities. And so, I’m hoping that with the raised consciousness about the disproportionate burden of COVID in minority and underrepresented communities, that there is opportunity to leverage greater attention to the fact that those very same communities are often overrepresented in the statistics for most chronic conditions, including cancer.
Maybe the question is not just why do they have more COVID or why are they more susceptible to death from COVID?
The question is, why are they contributing to excess incidence and mortality for most health conditions overall? And where there might be an opportunity to attenuate the risk conditions underlying these disparities independent of COVID. Let’s have that conversation. Now.
What have you been doing since COVID struck Florida?
Steve Nimer, our cancer center director, is in a position of unique leadership within our health care system, and at the outset of the epidemic, he started to mobilize institutional testing resources and enhance testing capacity to support increased demand within our catchment area.
And as a complement to that, my team started to build the necessary infrastructure for contact tracing within our healthcare workforce, and then for other university faculty, staff and students, who were symptomatic for infection. We felt strongly that testing paired with contact tracing needed to be squarely in place to reduce risk of transmissibility.
Can we talk about NCI’s Community Outreach and Engagement—COE—requirements for designation of cancer centers?
There is something very powerful about the roles that NCI-designated cancer centers can play in this epidemic.
One, because from a community perspective, with increased attention to COE, many of us have established a pretty significant foothold in our local catchment areas.
As a result, we are able to work collaboratively with community stakeholders, particularly in communities that are disproportionately affected by the epidemic, and to generate solutions through research and intervention to try to attenuate that impact. I have been so impressed by how many of my Sylvester colleagues have stepped up, and out of their traditional work roles, to support our institutional response.
It is perhaps one of the silver linings of this entire COVID-19 insanity. In trying to navigate collective chaos, people have demonstrated the best versions of themselves, even in the worst of situations.
Steve sets that example. He harnessed cancer center resources to support the institutional response and encouraged cancer center members to use our skills, such as those that have a role in population science to mitigate disease risk. It was through this example that our contribution to contact tracing was born.
My team has been doing a lot of contact tracing since mid-March. To date, we’ve traced over 2,000 individuals with a 96% response rate, which is really promising.
Now, we’re working with the CIO of the university to develop an app that provides operational scalability for what we’ve been doing, which has really been very old-school gumshoe epidemiology contact tracing, where the primary goal is to identify close contacts over a known positive and then advise them to self-isolate for 14 days while monitoring symptoms.
I’ve also had the opportunity to work really, really closely with the mayor of Miami-Dade County, Carlos Gimenez, on a community surveillance effort, to try to estimate the prevalence of coronavirus infections through serologic antibody testing.
You have the Game Changer vehicles, you’re in the community. Have you been in the community, by the way, recently?
As part of this community surveillance effort—yes.
In terms of other research or outreach—no.
Our university has been encouraging anybody who can work from home or work remotely to do so. And then, also in terms of research, right now we’re still in a phase where only critical research has been approved.
Normal outreach activity is also on pause. But Sylvester’s outreach team is still in the field, staffing the community surveillance effort, and making sure that this work resonates with our understanding of local need, and, more importantly, that we effectively dissuade concerns about participating in an effort that involves giving blood “to the government!”
So, the Game Changer buses are going around?
The vehicles are supporting community surveillance, particularly in the larger sites, which tend to have less shade.
I mean, this is South Florida in May—it’s really hot!
Think about having outreach teams in the community in full PPE to support serologic antibody testing. It is brutal. So, for those sites, without appropriate shade, the Game Changers have been a much needed way to provide some reprieve for the teams.
I suspect that the focus of the Game Changers is going to change, probably sooner than I expected, to play a more active role in some of our university testing efforts. But right now, their deployment has really been, in terms of the community surveillance effort that we’re doing with the county.
A Game Changer vehicle pictured at the launch event, Feb. 2018.
What have you learned about the prevalence of SARS-CoV-2 in South Florida? What are some of the numbers around it?
In a random representative sample of Miami-Dade County residents, the prevalence of infection is significantly higher than what is reported from testing sites alone, which is not surprising, because testing availability has been very much limited to individuals who are symptomatic.
We found that about half of the individuals who had antibodies for coronavirus infection had no known symptoms in the seven to 14 days prior to participating in the screening initiative. We also saw about a two-times higher burden of antibodies in blacks or African-Americans.
How did your role come about?
The mayor’s team randomly contacted me early in March and said, “We need to figure out how many people in South Florida have been exposed to this infection.”
And so, together, we crafted a unique community surveillance effort using serologic antibody tests.
Our primary goal was to have a random sample. We wanted to be avoid overestimating the burden by simply recruiting the worried well or those had symptoms at some point.
We divided the county into its 25 minor statistical areas, and then randomly selected individuals from these areas, proportionally to population density, and with consideration for their racial/ethnic, age, and gender distribution.
We are really mindful about representing the multiculturalism that’s present in South Florida, understanding that there may be unique disparities that we would see here that haven’t been reported nationally. We’re still analyzing the data.
We did the surveillance for a series of about four weeks, and we were supposed to go back into the field this past week. But with the FDA’s new guidance on serologic antibody tests for serologic antibody testing, we’ve actually had to pause what we were doing and regroup, and think about how we could do this work with the new expectations in place.
So, you need to change the test you’ve been using?
We are definitely changing the test that we’ve been using, likely to the Roche serologic antibody test.
And we’re just figuring it out, because the community surveillance worked exceptionally well with the finger prick test by Biomedomics that we were using at the outset. The ease of test administration made if feasible to get 700 people screened in one day across the county, which, you know, is not small, geographically speaking.
I always think the beauty of good science is that it has to be nimble, particularly in an epidemic situation. I’m not at all surprised by the fact that we’re having to regroup and reconfigure, because as we gain more knowledge collectively, not only about COVID, but also about the technology to detect it, we have to be able to modify what we are doing in real time to accommodate new information.
And this is something that I was very clear with the mayor and his team about from the get-go, that our work together would have to be very, very fluid.
I anticipated that what we were doing in one moment may actually not be what we’d be doing two weeks later, given new technology or improved understanding of the disease that would influence surveillance aims or implementation strategy.
But basically, if you were to draw a preliminary conclusion based on the number of samples you have screened, what would be two or three of the most important things you’ve learned?
I’ll give you three.
I think number one, because this is The Cancer Letter, and you typically write about issues of importance to NCI-designated cancer centers and other community-based cancer centers…
I’m not sure that I could appreciate how critical the COE requirement is, in allowing cancer centers to be more to their catchment area than just a resource for cancer.
And I feel really proud that Sylvester had such established community infrastructure and resources like the Game Changer in place that could be deployed in a moment of national crisis, to help our catchment area navigate things a little easier.
I think that’s a really important lesson learned. Some of what we have been doing provided us necessary flexibility to be a partner in the truest sense of the word in a time when human connection and access to information was potentially more important than ever before.
And then I think the other lesson I’ve learned is building the plane and learning to fly at the exact same time requires a degree of intellectual flexibility—and tremendous patience. A lot of patience.
Some of my junior faculty that have been working with me on this have struggled with the flexibility part. As scientists, we are trained to be somewhat rigid in the way we approach study design and implementation.
In this unprecedented COVID situation, we have had to marry need with opportunity. Since our understanding about the disease is evolving very rapidly, in real time, we have to be able to modify our approach to accommodate the new knowledge and still uphold scientific rigor. It is somewhat of an uncomfortable position, but one that likely accelerates personal and professional growth.
And last, I’ve learned to look for the silver linings or to better reframe things as my mentor, Jo Anne Earp used to encourage me to do somewhat unsuccessfully at the time. I think my team’s willingness to push themselves outside of their comfort zone to take on new work and potentially new risk is amazing.
And, our new shared understanding that we must constantly critically appraise what we are doing and become more nimble will pay dividends. I’m certain of it.
Yes. You have to be able to change the test, for example.
You have to be able to adapt. I’m not sure that we always do that so well in science.
This situation is forcing us to do that, and I think forcing a lot of us outside of our comfort zones. Me—certainly. And when I’m outside of my comfort zone, it’s an opportunity for growth. Whether that’s growth as an individual researcher or growth in terms of the field, it’s the potential for progress, nonetheless.
Speaking of which, what are your thoughts on reopening?
As the person who is leading the Miami Dade County community surveillance effort, my job is to ensure that there’s necessary objectivity and integrity in data collection. And that our methods are sound, and that we’re attending to issues like randomness, and we’re flexible when we need to reconsider which tests to use, to be consistent with FDA guidelines.
If my job is to be the scientist and collect the data, I need to actually do that, and that alone. Once I start to make comments about reopening, then people will think that I’m politically motivated in my data collection efforts. And I want to stay above that.
What’s so interesting about the whole COVID pandemic is that science has been really politicized. Even the discussion of antibody tests has been really politicized. And not that I think you ever really take politics out of science, but this has been much more explicit, Paul. You know what I mean.
I have become very practiced at saying that I’m not in a position to draw any conclusions about reopening, because we’re in the process of data collection and I don’t want to unintentionally undermine my objectivity or credibility. We can talk about my thoughts about reopening once surveillance is done!
Can you say more about the prevalence SARS-CoV-2—and about disparities?
The only thing we’ve really commented on thus far in terms of disparities was that there seems to be two times the rate of antibodies in blacks or African-Americans who are participating in this work.
Also, we tend to see a higher burden of antibodies in the minor statistical areas that are predominately minority in composition. I believe that this finding requires further examination that the community surveillance effort alone is not well suited to contribute to.
When the mayor and his team ask me about observed disparities, I typically say that I think that we need to engage community leaders and have real, meaningful conversation about what they think or even know is driving the observed disparity.
Working with community leaders in those neighborhoods, and to potentially do more targeted testing, facilitating access, certainly, to RT-PCR for individuals who are symptomatic. And then thinking about how, through Sylvester’s outreach and engagement team and some other similar groups across the institution, how we may be able to bridge gaps in resource allocation and other social determinants that may be really, really important in why we’re seeing these disparities.
Something that struck me that I read in The New York Times, but isn’t surprising, is that in a number of these communities, there is a scarcity of access to primary care physicians. Individuals don’t have anybody to check in with to describe symptoms and to assess whether those symptoms are severe enough that they require further medical intervention.
And so, many people are delaying, and delaying, and delaying, until there’s nothing they can do except show up at the ER. And at that point there are not great clinical algorithms to prevent them from having to be intubated.
And so, we’re thinking about maybe bringing a group of our nursing students or some of our medical school students together to fill this gap. Also, we’re starting to think about education, trying to dispel common myths in many communities around the etiology of COVID, what happens if you have to go to the hospital etc. This is not anything that we would decide alone
We would have to do so in collaboration with key stakeholders, who, with the appreciation of what’s happening in their community, could ultimately inform the scope and delivery of an evidence-based multi-level intervention to attenuate risk factors and risk conditions that underlie the excess burden of COVID and other disparities, including cancer disparities within those areas.
If you can help me understand this issue with the communities of color having twice the prevalence of antibodies, are you able to also measure this among—I don’t know how best to say that—among rich white people?
We used random sampling to select representative subsets of our county.
When I say that we see two times a higher prevalence of antibodies in blacks, that’s compared to other racial ethnic groups that are in Miami-Dade County. And proportionally represented within our sample.
We did this for four weeks in a row, and then we paused, and we were going to a month-to-month basis when the FDA issued new guidelines.
I also believe with the surveillance work, given the reliance on serologic antibody testing, you must assess how the data varies over time, and also in relation to hospitalization and death data.
Together, these individual datasets tell an important story of what’s happening.
The hospitalization, and the testing data that’s being reported by our Department of Health are people who were symptomatic and met criteria for screening.
In my mind, these data represent the tail end of the viral distribution in the county. The community surveillance probably captures the rest of that distribution, because it reflects what is happening in the community, regardless of symptomology.
But if you really want to understand impact, you can’t just think about any of these data sources in a vacuum. We’re trying to figure all of this out. I don’t think it’s perfect, but it is arguably better than drawing conclusions in the absence of any data at all.
I think that there is, really, utility for community surveillance to fill that gap. And I think there’s a real opportunity for NCI-designated cancer centers to be at the forefront of that kind of work, because it’s so aligned with what we already do, as part of our COE efforts anyway.
You aren’t exactly building a plane while flying it, are you? You were working with viral issues, with the Game Changer. You were screening for HPV, you were working with HIV.
I mean, we’re doing HPV testing, HIV testing, testing for other sexually transmitted infections, including hepatitis B. And so, this is really not that different than our normal capacity. And whether COVID screening will become part of what the Game Changer routinely does, I think that’s open for conversation, certainly with Steve, who ultimately drives a lot of what we do in our COE space.
I think, because it’s COE, we were well poised to be responsive. And to help think about not only a healthcare system response, but a community response, using the power of data to ultimately inform operational decision-making and public health planning.
Maybe it’s too early to ask this: Are you finding what’s the prevalence versus the known rates of infection? Is it twentyfold? Fortyfold?
It’s 16 times higher, if I remember correctly, but recognizing that there are very wide confidence intervals surrounding that estimate, given our need to account for test sensitivity and specificity, which was highly variable in different publications.
I do think there’s something that’s really interesting, Paul, about how we communicate science in a way that makes sense to people who are outside the scientific community, but who are hearing new information that’s driven by the principles of research.
Figuring this out has been a fun exercise, because the work has involved a lot of interdisciplinary collaboration on the university side, and with each discipline came a different discourse for how to talk about the work and its findings . It’s been a fun group of people who are working together. There are public health experts, there are data scientists. And many, many students.
Arguably, the best part of this story is that, with the contact tracing and with the community surveillance, we were able to offer 150 medical school and public health students a way to fulfill their capstone and field experience requirements, because all of their clinical rotations were paused or summer internships were cancelled.
We were able to immerse the students in a public health learning opportunity that was very real, very timely, and helped fill gaps in their educational curriculum, given that the pandemic had imposed real constraints on what they could do to fulfill experiential requirements.
There’s a lot of nice stuff that’s come out of that, including that I think parts of the institution who haven’t traditionally worked together have started to do so, because we needed each other’s resources and capacity to build something that could fulfill the intent of what we were trying to accomplish, with both community surveillance and the contact tracing.
Is there anything we forgot, anything you’d like to add?
At NCI-designated cancer centers, we have the potential to be at the forefront of helping drive solutions in a pandemic. This is typically outside the scope of what we do, but the community relationships we’ve developed are deep. And they can serve a purpose beyond what we want to accomplish for our COE requirements.
And I think that’s what we found here in Miami at Sylvester: our relationship with our catchment area allowed us to be a resource in a time of unprecedented need.