Medical Mistrust and the Legacy of Tuskegee
by Dr. Marcella Alsan
June 13, 2020
This interview with Dr. Marcella Alsan, a Professor of Public Policy at Harvard Kennedy School, was conducted and condensed by franknews.
Can you start with an overview of the Tuskegee Study?
"The Tuskegee Study of Untreated Syphilis in the Negro Male," was it's official title. It was run from 1932 to 1972 by the United States Public Health Service, which was a branch of the Centers for Disease Control and Prevention. It's pretty unheard of for a study to last for four decades, but it did. The core idea was that black men, both those with syphilis and a control group without, were going to be recruited in the name of observing the progression of the disease. The men were led to believe that they were getting treatment for what was euphemistically called "bad blood", but they weren't. And crucially, when penicillin was discovered in the 1940s as a highly effective, "magic bullet" for this disease, they were denied access to it. Syphilis is a disease that is transmitted sexually and also prenatally, so they were effectively allowed to spread it to their loved ones - their wives and to their children. And this was all done in the name of understanding the natural history of the infection, a lot of which was already known.
How did it end?
Well it lasted so incredibly long – beyond the 1960s, beyond the Martin Luther King Civil Rights Act. There was finally a whistleblower, a young man Peter Buxton, who came to work in Macon County and was completely alarmed by what he saw. He started writing to the CDC, exhorting them to end the study, saying that it was completely unethical. Actually online you can still see the notes from some of the internal meetings that his letters prompted. And it's scary to see that the response in the meeting from the CDC, from all these doctors is, "Well, it has become a political problem. What we need to do is not halt the study, not compensate the men, not treat them and their families, but find political cover. We need to get buy in from local chapters of the AMA and the county health department." And that is the response of the government at the time, to hide it. Eventually, Peter Buxtun tells someone with the Associated Press about the study, and the story gets published and becomes breaking news in 1972. So it is because of the press, not because of a moral consciousness, that there was cessation of the study. Eventually there was a settlement to the subjects themselves and their families, though it was comparatively meager.
And by looking at this event, what were you hoping to find?
Along with Marianne Wanamaker, who is an economist out of University of Tennessee Knoxville, we decided that we wanted to look at this event around the time the Ebola pandemic was starting. We started to hear about the idea of mistrust and about how people in West Africa weren't "complying" with the public health officials. For example, they still wanted to bury their dead, even though they were being told that it was high risk. There were all these questions around why they weren't trusting authority, and this issue of trust and medicine and history really came into consciousness.
In economics, we're going through a causal revolution where we try to be very precise about what we say is a correlation versus what we say is a causal relationship. We are developing all these new habits, including embracing randomized trials. But of course, in this case, you would never want to randomly assign some people to medical mistrust and some people not. So we really do need to kind of look back to what we call natural experiments, things that happened in history, to answer this question.
Which is how you used the Tuskegee Study.
Right. As economists, we leveraged the sharp timing of the 1972 disclosure. Our expertise is really taking observational data and trying to wrangle it into a causal analysis, using all of these differences- the time, the geography, the demography.
We hypothesized that for African American after 1972, when the news of Tuskegee broke, there would be a heightened degree of mistrust compared to before 1972. We also hypothesized that there would be a reduction in demand for health services. We also thought that the effects would be strongest for men who were close to the epicenter of Tuskegee, Alabama.
And what do you find out about how generations of men were affected by the study?
What we can confidently say is that we see about a one-third drop in demand for outpatient visits, lengthier hospitalization stays, and an increase in mortality up to 10 years after the disclosure.
Clearly not all mistrust and all of healthcare inequality can be explained by this one event, and that wasn’t our goal. What we are able to capture in this specific instance is how mistrust affects demand, why there are reasons for mistrust.
Mistrust doesn't just get helicopter dropped from the sky, it is directly related to historical events.
And it’s not just Tuskegee, it's the broader history going back to when this country enslaved African American people, and a repeated history of African American people being used against their will and knowledge in the name of science. Tuskegee is one example of that. We were trying to point to a specific instance and say, here we identify the historical episode, the deceit, and the spillover effects.
One of the findings of your study is that men who are most similar to the men in the Tuskegee study, in terms of geography, in race, in age, are most affected. Why is that such an important note?
Right. Our finding is not about the exact men that got experimented on, it's about other men who heard about it and felt like they were at risk too. Our results identify subjects who felt like, "this could be me." Which I think is important because it identified this spillover phenomenon that recognized that these are not one-off instances, that there are reverberations throughout entire networks of individuals.
It speaks to the idea that you are not just hurting that one person, you are hurting a whole community of people through intergenerational trauma.
Which brings up what I think is a really important, I think, a very interesting research question, and very interesting social question – if we apologize and we atone, can we avoid some of the spillover effect? It wasn't until Bill Clinton's presidency that there was a formal apology. That is decades later. There was only a smattering of the initial cohort involved in this study still alive when Bill Clinton formally apologized. Had the release of information in 1972 been different, and had it come with acknowledgement and atonement, would the effects have been different? We just can't know for sure.
Can you talk about the mechanisms of medical mistrust that lead to worse health outcomes?
I ran a trial in Oakland, California. This was a randomized trial now where it was published in the American Economic Review, where myself, along with other researchers from the Bay area, recruited African American men, from barber shops in the East Bay. We built a pop up clinic, and Uber donated ride sharing services to bring these men to our clinic. We randomly assigned the men to either have an African American male doctor or a white or Asian doctor. We focused the question on how that assignment affects the demand for preventative care.
Preventative care allows you to improve health outcomes dramatically, if you can catch cancer early on, take steps against diabetes, control blood pressure, et cetera. To have effective preventative care, you need partnership, and what we found was that specifically for invasive services, and by invasive I mean like a small prick of blood or a flu vaccine, it mattered incredibly to the African American men that we recruited, to have an African American doctor.
Our best look at the data said that this was really driven by the fact that they felt comfortable in communicating with doctors. We’ve done studies that have replicated that finding. It does really matter to have a diverse population of people who are helping bridge these historical inequities and helping to overcome a very valid, trust gap.
Diversity is not just a platitude, we need to more closely represent the population that we're trying to serve on the supply side, in order to produce more equitable outcomes.
Do you feel like there's an appropriate conversation around the role that mistrust is playing in the disparity of health outcomes in this particular pandemic?
In economics, we think about mistrust on the demand side of things - as lowering demand for healthcare. But if we think about it as more of a feedback loop, then we can understand it as the actions of the supply side by our providers and healthcares system that then affects trust and then demand. When we talk about alleviating mistrust, it has to include what we can do on the supply side to help overcome these historical traumas.
With COVID-19, and contact tracing, it is imperative to think about these dynamics. This pandemic is not coming out of the blue, it is being layered on top of this vast history of gross racial inequality and injustice. We can’t pretend that just because it's a crisis, that it can be all hands on deck. I think we still really need to be thoughtful about having a diverse set of communicators and people that we're employing to do this work in the communities that are being hardest hit. The conversation around mistrust is a fruitful conversation to have, as long as it didn't start and end with, 'There's mistrust."