Those in our lives whom we consider to be most objective, who in many ways are paid to be objective, have repeatedly been shown to be anything but objective. Rather, they make decisions that are consistently influenced by unconscious (implicit) bias. White NBA referees call more fouls against black players than against white; MLB umpires are more likely to erroneously call a ball a strike if an all-star pitcher is on the mound; Israeli judges are more likely to grant parole at the beginning of the day than before lunch; doctors are less likely to provide adequate pain medication to patients of color and, because of bias, are prone to predictable errors in diagnostic reasoning.
Doctors’ cognitive errors and biases have been well studied and chronicled, most notably by Jerome Groopman in his book “How Doctors Think.” These include premature closure- latching on to a diagnosis too quickly; attributional bias- being unduly swayed toward a particular diagnosis because of a single patient attribute (age, race, gender) and ignoring evidence that points to alternative diagnoses; and confirmation bias- (often accompanying premature closure) accepting evidence that supports one’s diagnosis and discounting evidence that goes against it. This topic is, or at least should be, a critical component of every medical student’s education. I teach about it each year to medical students and a central theme that I emphasize is the need to recognize that unconscious biases are universal: that no one is immune; and that one needs to accept their existence, be aware of, rather than blind to, them. Deny them, ignore them, pretend they don’t exist and the risk of errors in diagnostic reasoning is certain to rise. As I followed the actions of the prosecuting attorney and the proceedings of the grand jury in the People of Missouri versus Darren Wilson case, I felt an eerie sense of familiarity with the risks, potential pitfalls, and resulting negative clinical outcomes that I teach about in class– I was watching a possible misdiagnosis unfold.
Soon after Michael Brown was killed, St. Louis County Prosecutor Robert McCulloch stated that he was not biased and if anything, he would be biased in favor of the victim. This, despite being the son of a police officer killed in the line of duty by an African American; this, despite more than 20 years working intimately with a mostly white police force; this, despite working during those same years to put mostly African American defendants behind bars. This despite having said about two unarmed black men killed in 2001 by DEA agents, “These guys were bums.”
McCulloch took a tack early on in the proceedings, however, that could shield him from charges of bias. He wouldn’t cherry-pick the evidence; he would show it all to the grand jury. How could that be biased? What McCulloch appears to have failed to recognize is that the way that evidence is put forward is susceptible to bias. One could argue that simply putting forward all of the witnesses (in this case 45) produced a situation in which greater inconsistencies between witnesses was virtually assured and the element of doubt, which would support a ruling of no true bill, would be raised.
McCulloch may have been swayed early in the case by attributional bias. Who was he to believe? Darren Wilson- a white police officer or Dorian Johnson- a dreadlocked African American who was with Michael Brown during a “strong-armed robbery” earlier that day, who had a criminal record and a history of lying to police? But McCulloch took another step to try to assure objectivity: he would place the foundation of his case on the “physical and scientific evidence.” In his announcement of the grand jury decision, McCulloch asserted that “Physical evidence does not change because of public pressure or personal agendas, physical evidence does not look away when events unfold, nor does it block out or add to memory. Physical evidence remains constant and as such is a solid foundation on which cases are made.” What McCulloch failed to recognize is that interpretation of physical evidence is not completely objective. Several interpretations of the autopsy advanced in the media are completely false. The autopsy did not prove that there was a struggle for the gun, rather it simply showed that Michael Brown’s hand was near the gun when he sustained the wound to his thumb. The autopsy did not prove, as some believe, that Brown could not have had his hands up with his palms facing Wilson. About the only thing that the autopsy could prove is that Wilson did not come up and shoot Brown in the back as he lay on the ground as a few witnesses asserted.
So how does it appear that McCulloch used his false interpretation of the evidence? In his announcement of the decision, he stated, “witnesses were confronted with the inconsistencies and conflicts between their statements and the physical evidence. Some witnesses admitted they didn’t see the shooting or only saw part of the shooting or only were repeating what they heard on the street. Some others adjusted parts of their statements to fit the facts. Others stood by their statements even though their statements were completely discredited by the physical evidence.” Sadly, the interpretation of key components of the physical evidence, the science that McCulloch stated was the foundation of his case was terribly flawed and was not actual fact.
Once McCulloch had the evidence he thought he needed to support Darren Wilson’s narrative, it appears that confirmatory bias came into play. Witness 10, who was one of the few witnesses to claim that Michael Brown charged the officer, was subject to virtually no cross-examination even though he initially stated that he was 100 yards away from the shooting. For that matter, Darren Wilson himself received gentle treatment from the prosecutors. Other witnesses whose narratives contradicted Wilson’s were often subject to aggressive questioning. Witness 14, an apparent senior citizen who was concerned about those in the neighborhood “embellishing” and who stood within feet of the incident, provided detailed and compelling testimony that Brown had his hands up at his shoulders as he turned to face Wilson; that he was staggering, not charging; and was falling to the ground, clearly disabled, as Wilson unleashed the final fatal series of shots. This witness was repeatedly grilled about the details of his recollection. A detective was even brought into question the conclusions that this witness had made.
McCulloch and his colleagues put together a case, a narrative, that appears to have been built on an illusory foundation. They appear to have been blind to the implicit bias that may have contributed to this. In medicine, this would have led to a misdiagnosis; in this case, it may have led to a miscarriage of justice.
Stuart Slavin, MD, MEd, is Associate Dean for Curriculum, Saint Louis University School of Medicine.