What really killed George Floyd: Part 1 — some basic forensic pathology concepts

The trial of the police involved in the death of George Floyd was front page news a couple of years ago. Now that some time has passed, I thought I’d put in my two cents.   This is the first of a long series of posts about the case.  The rest will follow in the upcoming weeks.  Update: There are 13 parts to this series.

The controversy associated with this case was not limited to the news. There was a significant difference of opinion among the forensic pathologists involved in the case (including myself).  A lot of folk don’t understand why.  The thoughtless but easy approach is to attribute it to motivation. Dr.  Roger Mitchell, a former Medical Examiner in Washington, DC and now Chief of Pathology at Howard University, charged David Fowler, the forensic pathologist who testified for the defense, with ethics violation for his testimony.  It’s a cheap shot, of course, and is nothing more than another attempt at cancelling someone who departs from the imposed narrative, but it’s important to understand how we forensic pathologists look at a case.

As a matter of disclosure, I was also hired as a consultant in this case by the defense, as part of a team of around 14 experts in a number of fields.  Dr. Fowler’s testimony was not his own personal bias.  It reflected the considered opinion of a number of experts who signed on to his report.  As will always happen, when you have a committee of 14 people, there will be some variations on the final theme.  I’m going to give you my variation, but not because I believe it’s the absolute truth. — I’ll explain why that’s a bit of a reach for any conclusion in cases like this after this introduction.  But I believe it’s a good approximation of the truth.

It’s important to understand that there are two different levels on which this case was being evaluated, and two different definitions of truth.  One definition of truth is that the narrative is truth, and any facts that detract from that truth should be discarded.  There is no such thing as “objective truth.”  In such an instance, the “truth” must be crafted within the context of a narrative.  In an article regarding the problems with forensic pathologic diagnosis, one expert notes:

Although the county autopsy ultimately squared with what a private investigation commissioned by Floyd’s family found – that his death was a homicide – the examiner’s report was extrapolated from in the charging document for the former officer Derek Chauvin to assert that “underlying health conditions and any potential intoxicants in his system likely contributed to his death”.

Even if the practice of including medical information that isn’t directly relevant is common practice, Tsai said, “there’s character assassination in this process in describing the medical facts.”

Note the key here — the facts should be ignored because they are “character assassination.”

Dr. Mitchell concurred:

But first, let’s talk a little about some basic concepts in cause of death, mechanism of death, manner of death, and what causes death.

Cause of death is the underlying issue that lead to death.  It’s the gunshot wound, the influenza, the blow to the back of the head.  It may be immediate, or may be delayed.  For instance, a gunshot wound to the spine may result in paralysis.  The paralysis results in the inability to move or to void urine.  The development of pressure ulcers and urinary tract infections leads to recurrent pneumonia.  Recurrent pneumonia results in progressive loss of lung function.  Finally, 20 years after the injury the decedent dies of respiratory failure.  The cause of death is “gunshot wound of the spine,” because it is the underlying event that eventually lead to death.

Mechanism of death is the actual mechanical pathway that resulted in death.  How can a gunshot wound to the chest kill you?  Lots of ways, in fact.  You can get a hole in the lung, collapse the lung, and die of respiratory failure.  You can get a hole in the lung, develop a one-way tissue valve that builds up air pressure and puts tension of on the heart and die of a tension pneumothorax.  You can get a hole in the heart and bleed to death internally.  You can damage the heart’s conduction system and die of an arrhythmia.  And there are a number of other ways as well.  These are “mechanisms” of death through which the “cause” of death (gunshot wound) results in the demise of the decedent.   Sometimes you add a bit of the mechanism in the death certificate, and sometimes you don’t. I tend to add mechanistic details to the death certificate if I think it’s needed for a lay reader to understand why the cause is there.  In fact, for the previous example, most folk would likely create a multiple line death certificate of “Pneumonia Due to: Paraplegia Due to: Gunshot wound of spine,” but you get the idea.

Manner of death is almost universally misunderstood.  It is a classification of the death within a cultural context for the purpose of public health statistics.  The reason we do manner of death determination is so that people can go to the CDC website and see how many homicides or accidental deaths there were in 2020.  It is used in public health planning, policy making, etc. at the local, state, and national levels.  It is not, however, a legal determination in the sense of “murder” or anything like that.  In the vast majority of cases, manner of death is trivial — Uncle Joe died of cancer and it’s a natural death.  In a small number of cases, it is ambiguous.  Unfortunately, it is that small set of ambiguous cases that more often end up in court.  It is a statement of the success of Medical Examiners in the determination of manner that some courts use it in their deliberations.  However, it is not built for that, and it is fundamentally a misuse of manner for it to be determinative in some way in court.  Again, unfortunately, this misuse of manner in court is a profound headache for many Medical Examiners.  We are criticized for making manner determinations the way it’s designed rather than in a way that’s amenable for its misuse in court.

As one historical review noted(1):

By 1851, seven states had enacted death registration laws and a method was devised as an outgrowth of the Shattuck Report for the 1850 Federal Census to count deaths. However, registration worked well in only a few cities and two states (9). By 1880, resolutions of the American Medical Association and efforts of the American Public Health Association, National Board of Health, and Superintendent of the Census resulted in a registration area concept that was supported by a resolution of Congress. [As an aside, it was about this same time that the first medical examiner systems emerged in Boston in 1870 and in Baltimore in 1890]. Prior to 1900, the United States lagged behind other Western countries in developing a centralized death registration system, but by 1900 had drafted a model vital statistics law that could lead to a centralized system of death registration.

Note the designers: American Medical Association, American Public Health Association, National Board of Health, and Superintendent of the Census. Not Judges. Not District Attorneys. Not detectives. Not Lawyers. Not Juries,

Death is often a combination of insult and vulnerability. 

The body can respond to injury.  The better it responds, the better the chances of survival.  The greater the vulnerability, the less insult it takes to kill.  There are some very common examples of this.  For instance, there is an old rule of thumb about mortality in burn victims.  The old saw is that you add the age of the patient and the percent of body surface burned, and that’s your expected mortality.  Thus, a 20-year-old with a 20% body surface burn would have a 40% chance of dying, while an 60-year-old with a 20% body surface burn would have an 80% chance of dying.  It’s not a particularly good rule of thumb, but it illustrates the idea.  Another common example is falls in the elderly.  A common final pathway for the old and infirm is a fall resulting in head trauma or a broken leg.  Both of these are usually suvivable injuries in young people.  A broken leg in a young adult rarely kills.  In the elderly, however, it can be a death sentence.  While falling from standing height can result in lethal head trauma in younger people, it is very rare.  In the elderly, it commonly results in fatal hemorrhage in the brain or surrounding tissues.    A severe diabetic with peripheral vascular disease is a great risk from what would be a trivial injury to the foot in a person without diabetes; merely cutting one’s toenails can be a threat.

The claim that these vulnerabilities are irrelevant to the cause and manner of death because they obstruct a narrative or are “character assassination” is simply wrong.

This can lead to ambiguity in manner determination.

This interaction between vulnerability and insult plays a role in manner determination. Manner determination is based on the underlying event — what caused the fire that lead to the burn, what caused the fall, etc.  But how bad of an injury does it have to be?  At some point, people get so fragile that normal daily activities are life-threatening.

An often unrecognized fact is that even in cases of natural death, there is usually some physical event that precipitates the death.  For instance, a person with terminal cancer with growths in the bones may still be able to walk, but bumps against a  piece of furniture.  The bump hits the leg just at a place in the bone where tumor is growing and the bone collapses.  The decedent is hospitalized and develops a blood clot in the leg that travels to the lungs and the decedent dies of a pulmonary thromboembolism.  Most (but not all) people would classify this as a “natural” death because the disease was at a point where normal daily events were life-threatening. For many forensic pathologists, there’s a line where the insult is so small that it just no longer counts because the vulnerability is overwhelming.  Where that line happens to be in any individual case is a matter of individual medical judgment.

The other thing to remember is that the categories of manner are themselves cultural constructs.  There is no “scientific” definition of “homicide.”  Manner determination is a cultural and policy classification to put an event in a cultural context using cultural definitions and criteria.  While cause of death determination is a “scientific” effort, manner determination is a cultural and policy determination.  In the medical examiner world, many of us consider the difference between cause of death determination and manner of death determination a formal separation of the “science”  and the “art” of medicine as we practice it.

This means that manner of death determination necessarily involves some cultural judgment.  In many cases, the classification will be ambiguous, because it could fit into multiple categories.  For instance, “suicide by cop” where a person intentionally places himself or herself into a situation where the police will use lethal force could be considered either a suicide or a homicide.  In the medical examiner community the convention is to classify it as “Homicide,” but not all people do that.

It their guidelines for manner determination, the National Association of Medical Examiners is explicit(2):

It must be realized that when differing opinions occur regarding manner-of-death classification, there is often no “right” or “wrong” answer or specific classification that is better than its alternatives. When promulgating guidelines, however, one of the available options needs to be selected as the one recommended for use. Thus, the recommendations herein are ones selected to foster a consistent approach amongst certifiers, not because the recommended approach is the “right” or the “better” one.

The “arguments,” principles, and foundations used to support certain recommendations in this Guide cannot be applied uniformly to every conceivable death scenario because issues sometimes vary with the manner of death being discussed. As a result, there will be obvious, apparent “inconsistencies” in the rationale discussed for making some of the recommendations in this Guide. This problem is unavoidable because of the nature of the subject at hand. Thus, in some cases, one simply must select an available manner-of- death classification as the preferred one for use in a given scenario while recognizing that the logic used to select that option may not be applicable or directly transferable to other situations (and, in fact, may seem inconsistent with the logic employed in other scenarios). In short, it is sometimes necessary to simply select an approach and use it for the purpose of consistency, recognizing that other approaches may be “just as good.”

The misconception that manner determination is a “scientific” or “legal” determination that has only one answer in ambiguous cases is very destructive.

In the next post, I’ll talk a little more about vulnerability and insults and how it relates to manner of death.