Dr Roger Mitchell, former DC chief medical examiner, on how homicides are mislabeled as ‘accidents’ and how ‘we must hold the system accountable’

The crisis of US police shootings has been increasingly well-documented by advocates and journalists, with data now suggesting officers fatally shoot an average of more than three people every day.

Since George Floyd’s murder four years ago, there has been growing scrutiny of a more hidden epidemic of police violence: deaths at the hands of officers who did not use guns. An Associated Press investigation in March found that more than 1,000 people died in US police custody from 2012 to 2021 after officers used “less lethal” tactics, including pinning victims face down and stunning them with Tasers. In hundreds of those cases, medical officials deemed the deaths “accidents” or “natural” despite officers’ use of force.

Dr Roger A Mitchell Jr, former chief medical examiner of Washington DC, has become a leading voice in the push to uncover the true scale of these fatalities. In his book Death in Custody published last year, he and co-author Jay Aronson, founder of Carnegie Mellon’s center for human rights science, lay out the nation’s systemic failure to track deaths caused by police and correctional officers. The book scrutinizes bias in death investigations, exploring how coroners and medical examiners have produced autopsies that minimized or erased the role of police, and in effect, blamed victims for their own deaths.

Mitchell has worked with victims’ families, testified in Congress and in high-profile criminal trials and reviewed “accidental” deaths that he determined should have been labeled police “homicides”. Now chief medical officer for ambulatory care at Howard University, Mitchell spoke with the Guardian about the lack of progress since Floyd’s killing and his proposed solutions.

This conversation has been edited and condensed for clarity.

How do you reflect on what’s happened in the four years since George Floyd’s murder?

There are so many more people who have died since then and few people know their names or even that they died. We don’t know how many George Floyds there have been. We look at these deaths as single travesties and terrible events instead of as a system. There is an epidemiology that goes on with deaths at the hands of law enforcement and in jails or prisons where the system is responsible for providing adequate healthcare and protections against suicide. So I’m reflecting on all these additional cases in the last four years, and how we don’t have a robust, complete, fully comprehensive data collection system for death in custody. The federal Deaths in Custody Reporting Act – meant to track deaths from the pre-arrest phase to arrest to transport, booking, jail and long-term incarceration – is not currently operational in the sense that we don’t actually gain accurate data.

Why is this lack of data such a problem? What is the impact of our failure to track these cases?

Mortality data, or how people die, has always been used to leverage policy and programs for prevention. We focus so much on cancer, because cancer causes so much death. It’s the same with heart disease and motor vehicle collisions. When it comes to deaths in custody, we don’t pay much attention or do much research or create many public health and law enforcement policies to decrease the burden of disease or injury of those coming into contact with law enforcement. But that is what is needed.

There’s such a stigma when it comes to the carceral system. Families of loved ones who died in custody often don’t know where to turn. And quite frankly, I don’t think people care. There have always been advocates and community members and families who have cared and now journalists are really taking on the mantle. But there hasn’t been any real movement to hold the system accountable for capturing the data. It’s a function of the fact that those who are marginalized in this country are suffering the most, and we don’t have that level of visibility so that we can create prevention.

How do medical examiners potentially obscure our understanding of the scale of these deaths?

The medico-legal death investigation system includes medical examiners, coroners [who generally aren’t required to be physicians] and sheriff-coroners [coroners who are part of the law enforcement agency they may be investigating]. They are critical, because they certify death, and the death certificate should be a ground truth as to the burden of disease or injury within society. Death certificates tell us about the health and wellness of our community. But this system is completely disjointed. Some are county-based systems, others are state-based. Funding varies between systems. There’s a shortage of forensic pathologists to provide necessary expertise for autopsies. There’s no federal oversight. So there’s a lack of uniformity and consistency in how cases are decided.

There are five manners of death: homicide, suicide, accident, natural and undetermined. Usually, it’s pretty clear. If someone is shot, that’s a homicide. If there’s a fight between two men on the street and one of them dies, that’s called a homicide. But there’s a philosophy shift when the altercation is between law enforcement and a civilian, with some calling those cases “accidents”, “undetermined” or even “natural”.

What are some of the most common types of police homicides that get mislabeled accidents?

Often we’ll see a struggle between an officer and a civilian, and that leads to a prone restraint, meaning the person is face down. There may be an officer’s knees or body on the back of the individual, and the individual may be expressing their inability to breathe. The actions of the officers precipitated the death – whether an asphyxial death, meaning a true inability to breathe, or a cardiac death, meaning the heart’s inability to pump and get oxygen. The individual may have also had drugs in their system, and in those circumstances, there are some forensic pathologists who would then call it an accident.

There is also a culture and philosophy in forensic medicine that for police-related altercation deaths, there needs to be “intent” to call it a homicide. But in most homicides, there doesn’t need to be intent. If a death is due to the actions of another person, it’s medically a homicide.

How do these misclassifications impact grieving families seeking answers and accountability?

Families are often left without clear communication about what was found in the autopsies, which leads to a lot of frustration and anxiety. When there’s an altercation with law enforcement, and without that fight, that individual would have walked away and been OK and the case is called an accident and not a homicide, it creates a distrust of the system. Keep in mind, homicide doesn’t mean murder. The legal system makes the decision on whether the force was warranted. But if a person died because of the level of force, to call it anything other than a homicide is to minimize the importance of the truth. It has a huge impact on not just the families, but on the larger community and society, when the cases from a medical standpoint aren’t called what they should be called.

California recently outlawed the use of “excited delirium” [a discredited cause of death that has long been cited to justify the sudden deaths of Black men restrained by police]. So what about the thousands of excited delirium cases that have happened since the 1990s? Cases where families were told, ‘Sorry, your loved one was high on drugs and all of the sudden just collapsed.’ Who is going to look at those cases, when we know excited delirium is not going to be used as a cause of death? We need to go back and correct the record and get closure for families.

How can we actually better track these deaths at the hands of police?

We have a real need for a public health approach. We can’t just rely on the department of justice. I’ve proposed that we have a checkbox on the US standard death certificate that says whether it was a death in custody. The death certificate tells us so much: age, gender, race, geography, employment. There’s so much we can learn from death certificates. If it says whether it was arrest-related or someone who was incarcerated, it won’t be the answer to prevention of these deaths. But it will be the beginning of us truly understanding the occurrences and the rate of these incidents throughout our nation. I really want people to think of this as a system of deaths that we need to understand, that we need to study and that we need to control for and prevent.