A peer-reviewed study of injuries sustained by participants and bystanders in last spring’s protests following George Floyd’s killing by Minneapolis police has drawn widespread media attention. The study was published in the New England Journal of Medicine.
Conducted by an interdisciplinary team from the University of Minnesota, the study concludes that “less-lethal” projectile weapons commonly used to disperse participants during cases of civil unrest “are not appropriate for crowd control.”
To come to this conclusion, the research team scanned thousands of medical records dated between May 26 and June 25 for patients seen in clinics and emergency rooms in two Twin Cities medical systems. The study’s authors discovered 89 records containing the words “riot,” “tear gas,” or other crowd-control methods that they separated and closely examined. The data collected from these records ultimately led the researchers to their conclusion.
I wanted to learn more about why this team of researchers decided to conduct their study and how they came to their conclusions, so I spoke with three individuals who worked closely on the project, Erika Kaske, a student at the University of Minnesota Medical School; David Satin, assistant professor of family medicine and community health at the University of Minnesota Medical Center and a family medicine physician at M Health Fairview Clinic – Smiley’s; and Joel Wu, adjunct professor of clinical ethics at the University of Minnesota Center for Bioethics.
In an open and informative conversation, Kaske, Satin and Wu spoke about their methodology, their motivation and the key role that data can play in sparking social change.
MinnPost: What inspired you to conduct this study?
Erika Kaske: This study was inspired by the injuries that we were seeing in the Department of Neurosurgery. [During the George Floyd protests], we started noticing a lot of head injuries come in. Specifically, there was a case of an adolescent with a depressed skull fracture from a rubber bullet. He wasn’t even participating in the protests: He was a bystander. This was concerning to us. We wanted to do research to see if there was a general pattern in the types of injuries.
MP: Did you enter into the study with a hypothesis?
Joel Wu: It’s not necessarily that we had a hypothesis that we wanted to test. It’s more that we wanted to get an accurate sense of what had happened. If you don’t measure, you don’t know if something is wrong in a community.
If there is a certain burden acquired by a certain group of people, you won’t know that unless you measure. You won’t see it if something happens to nonwhite people more frequently than to white people. We measured it to show the numbers. Without that you won’t be able to understand the problem and you won’t be able to intervene.
MP: It makes sense that you didn’t enter into this project with a specific hypothesis. Did you have a set of questions you hoped your work would answer?
EK: There are guidelines on how these weapons should be used. One main question we had was, “Were these weapons being used appropriately?” United Nations guidelines say they should not be aimed at the head, face and neck or near the heart or kidneys. We wanted, in as unbiased a way as possible, to measure these injuries so we could inform the community.
MP: How did you get involved with this project, David?
David Satin: I teach ethics, law and policy for the medical school. Erika contacted me as an ethics professor, thinking there is an ethics angle to this. In their design of the study and analysis it would be good to have me on board for guidance. I quickly assembled a team realizing that there are public health, race and racism and sociological aspects to this.
What surprised me was that there was a clear quality improvement aspect to this work, which is another course I teach. We teach med students in their first semester about the concept of quality improvement and patient safety. When you apply a quality improvement lens to this data, we see that we can’t know how to do any of this better until we measure and find out what is happening.
MP: Have other researchers looked into injuries from so-called less-lethal weapons?
EK: These weapons have been used in the U.S. since the 1970s, but there are very few reports in the scientific literature [about injuries], particularly from Black Lives Matter protests. These protests have been going on for about 10 years, along with the use of these crowd-control weapons. We know they were used in protests from media reports but there are not documents from the scientific literature. This is one of the first reports that documents the use of these weapons and injuries that occurred because of them.
MP: Do you have any theories about why there have been so few studies of the impact of these weapons before now?
JW: Erika’s observation that there is almost a complete vacuum is significant. It is not just that we care about injuries to the individual: It’s the data that matters. This is happening in a societal context that has public health and justice implications. In addition to thinking about how we can reduce injuries and harms to individuals, we are working to be able to substantiate the idea that people are not being treated similarly in similar situations.
Is one group being treated more harshly than another? From my perspective I would say that question is extremely important because our society holds up the idea that similarly positioned people should be treated similarly. If groups of people are participating in civil unrest, shouldn’t it be the case that law enforcement treats those people equally? Why is one group being treated differently than another?
That question is part of the reason why it is important to gather accurate, objective evidence.
MP: It sounds like you’re referring to the domestic terrorist attack that happened at the U.S. Capitol last month.
DS: Without even having to make any comparisons to the Capitol incident, it is very clear that we’re flying blind from an evidence-based policy perspective. If I’m going to be analyzing the burdens of appropriate use of force, I have to know what hangs in the balance on both sides.
In terms of the larger ramifications, the harm these kinds of injuries cause long term to someone who loses an eye or gets a brain injury, we’re flying blind. We don’t have enough data. Without this data, policymakers can’t weigh the benefits and burdens.
Clearly something went better in event one than it did in event two, or worse in two than in one. What can we learn from that? That’s the science of quality improvement and evidence-based policymaking. Policymakers need data to make good decisions. They need to learn the benefits and burdens. If they don’t have the data on the burdens, it is very difficult to decide when and how much force is appropriate.
MP: Can you give me an example of how data can inform decisions about use of force?
EK: So far we haven’t talked about the accuracy of these weapons. We don’t know the motivations or how these weapons were used, but from our research we know that they can cause serious harm to the people they are used against.
Even when officers are trying to use these weapons appropriately they can hit an unintended target or person or body region. There is a high risk of hitting and harming unintended individuals. We can use data to decide as a community, “Are these weapons appropriate for crowd control?”
DS: The analogy I would draw is to early studies on pharmaceutical drugs. It’s one thing to say, “Under perfect conditions with perfect subjects, this is how the drug behaves.” When you release that drug into the market, you get very different data. Pharmaceutical companies will say, “We didn’t study that because we studied the drug in perfect conditions in the lab.” That’s what we’re seeing with these weapons when they are used in uncontrolled, real-world situations.
MP: So when these weapons are used in the real world, they don’t work like they did when they were developed?
JW: When the manufacturer of the rubber bullets did a study, it was in a controlled environment. So nobody was ever hit in the face and nobody was ever blinded. They sell their weapons under the claim, “We did these studies. When used appropriately, no one was ever blinded.” What we’re saying is, “The evidence found that’s not true.”
EK: Rubber bullets are very large. They are designed to create a blunt, not a penetrating, injury — a bruise instead of a gunshot wound. Because they are bigger than regular bullets, they move slower. When you think in terms of physics, because these bullets are larger and slower, they have the potential to not hit exactly where they are being aimed. If you are shooting a rubber bullet from a closer distance than recommended, that can cause a penetrating injury. If you shoot from further away, your shot may be inaccurate.
MP: Did your study find examples of serious injuries from other less-lethal weapons?
EK: Tear gas is something a lot of people are talking about. What we noticed in our study was that, while the injuries from tear-gas exposure were generally quite mild, many people were hit with the tear gas canisters. These metal canisters can be launched in a similar way to rubber bullets.
If a police officer doesn’t want to put a canister on the ground, they can use a gun to launch it into a crowd. What’s been happening is those canisters are hitting people in the head. United Nations guidelines say that tear gas canisters shouldn’t hit anyone at all, especially not in the head or face.
In our study, someone lost their sight after being hit in the eye with a tear gas canister. There were eight people in our study that were hit with tear gas canisters. And these were the people who chose to seek medical evaluation.
DS: Before we did this study, my knowledge of these weapons came from movies and the TV news. I’m afraid most policymakers are in the same position. When I’ve seen tear gas canisters in movies, they are lying on the ground. It hadn’t occurred to me that a tear gas canister could be launched hundreds of feet into the air into a crowd and that the canister could come down and hit someone’s head or face.
We don’t know the evidence-based impact of this. Hopefully our study will prompt other researchers to take our methodology and measure what happens in these situations.
MP: Until recently, I didn’t realize that someone could be seriously hurt by a tear gas canister.
EK: Something that is interesting is how we talk about these weapons. They’re called “rubber bullets” or “chemical irritants.” How we name these weapons and talk about them diminishes their effect. It makes them sound less serious. How come it is called a “chemical irritant” when it used on a protester, but a “chemical weapon” when it is used in warfare?
Bean bag rounds are also used in crowd control. When you think of a beanbag, you think of a comfy chair. But these are sacks filled with lead pellets. Getting hit with them hurts. We should think about a better way to reframe this, especially in media reports about lawful protests.
MP: Language really is important.
DS: Our team had a debate about whether we call these “less-lethal weapons” in our report, or use the more common vernacular “non-lethal” weapons. We’ve all heard of non-lethal weapons, but less-lethal weapons seems more accurate. We finally decided to go with what we felt was the correct term. It is absurd to call these weapons non-lethal.
MP: Since your study was published and has been the focus of media attention, has your team been in contact with representatives from law enforcement or government?
EK: We were recently contacted by a lieutenant from the Minneapolis Police Department. He wanted to talk with us about our findings. We haven’t been approached by Mayor Frey or Governor Walz. I think it’s worth discussing, and I hope we can come together to talk about what we’ve found.
JW: None of this is happening a societal vacuum. It is important to realize that this this is an exercise of state power, an exercise of force by a government that has the power and obligation to protect and promote the general welfare. Under this obligation there are two important societal and ethical considerations about the appropriateness of the use of force.
The first is to the individual. You should only use the least-harmful means to achieve a goal. If a kid shoplifts a pack of gum, you shouldn’t shoot him.
The second ethical consideration is to society at large. An excessive use of force is a societal and moral problem. It is a population-level concern if one group is clearly being treated differently from another, if one group is not being exposed to this harm but another is. The only way to prove that one group is being treated differently from another is to have the data.
MP: Is this research a way of using science as a form of social activism?
DS: I want to be very clear about the link between this science and social activism. This is science. We took great pains to bring objective data to light. People can do with that objective data what they will. Our hope is that it will be used for good.
Priority No. 1 is to get accurate information. We took great pains to be sure that we followed the most rigorous protocols to get the most accurate data. This is not cherry picking. This is accurate data.
EK: One key thing I’ve learned from my mentors is the importance of finding the truth in my research and doing this in the most unbiased way possible. We went through over 6,000 medical records to find the 89 in our study. That way we knew we wouldn’t miss any potential subjects and we wouldn’t have a biased data set.
I think the results of our research have potential for advocacy. That’s the next step, using what we found and sharing it with the appropriate people, so we can create policy change for our patients. We came into this research with the goal of letting people know what was happening. We want people to be safe. We want them to be able to express their voice in a safe way. This is not happening right now.
MP: Do you think your research has the potential to influence major policy change?
DS: Policy is not the only lever. Like Erika said about the importance of accurate language, if our cultural norms start to reimagine what these things really are, then we will have a better grip on reality. A better grip on reality allows for better decisions by individuals, policymakers, society. There are multiple known and unknown levers that may be triggered here.
The starting point for all that is accurate information.
Do I hope that our research results in fewer people ending up in the hospital with these horrible injuries? Absolutely. But exactly how to get there is not the major thrust of our paper. First and foremost, it is the data. We didn’t come into this saying, “We need policy reform regarding these weapons.” That may ultimately be what happens. I don’t know. What I do know is that we had a lot of people get badly hurt.