Generations of medical students learn early and repeatedly in their training to begin the description of a clinical case with demographic information. As in, “AH is a 52 year old divorced, white female who presents with the chief complaint____.” We learn early that this info is important because each of those data points can help in arriving at a diagnosis by building a list of an appropriate differential diagnosis. ‘Differential diagnosis’ is a list of likely diagnoses based on information available so far. Thinking about differential diagnosis is an important safeguard against jumping to conclusions, a way to minimize diagnostic mistakes.
Race, age, gender, marital status, employment status, etc. all can give relevant clues to the listener or reader of the case who doesn’t have the patient in front of her – for example, the teaching physician who goes over the cases with the trainee during a grand rounds or clinical case conferences. Certain illnesses tend to cluster in certain races and certain age ranges, while marital and employment status can be clues to the burden imposed by the illness on the patient and also be predictors of prognosis.
Illnesses sometimes cluster by race due to genetics – for example sickle cell disease, a blood disorder, is more prevalent in blacks. But on other occasions illnesses cluster due to environment – for example, people living in a neighborhood or town with poor water or air quality are more likely to have illnesses that have an environmental trigger (e.g., breathing or skin problems)
But difference in likelihood of certain illnesses by demographics is not the only reason why on most measures of health status, blacks do worse than whites. Some of it has to with access to healthcare services and how people get treated, some of it the result of explicit prejudice and/or implicit bias. It appears that, for once, this prejudice may have given blacks some protection, rather than hurting them, in the current opioid epidemic.
Whites account for approximately 80% of opioid overdose deaths nationally, though that number is as low as 10% in Washington, D.C. and as high as 97% in Vermont. Dr. Kolodny, Co-Director of Opioid Policy Research at the Heller School for Social Policy and Management at Brandeis University, shares a theory about this in this interview:
Something that we do know is that doctors prescribe narcotics more cautiously to their non-white patients. It would seem that if the patient is black, the doctor is more concerned about the patient becoming addicted, or maybe they’re more concerned about the patient selling their pills, or maybe they are less concerned about pain in that population. But the black patient is less likely to be prescribed narcotics, and therefore less likely to wind up becoming addicted to the medication. So what I believe is happening is that racial stereotyping is having a protective effect on non-white populations.
On one hand it’s great that for once bias has worked in favor of black communities as far as the opioid crisis is concerned. However, Dr. Kolodny’s explanation also implies that black patients’ complaints of pain are not taken as seriously.
We should also note that opioid use is increasing in black communities, and that when we compare rates, rather than absolute numbers of opioid overdose deaths, the difference between whites and blacks is not as stark as the absolute numbers suggest:
Rates are a better way of making such comparisons because they allow us to compare the frequency of a particular condition in different groups. For example, over 33,000 whites and over 4,000 blacks died from opioid overdoses in 2016, suggesting overdose deaths in whites are about 800% higher than in blacks. But when we compare the proportion of overdose deaths in each racial demographic to the total population of that respective demographic, the difference – though still significant at 17.5 vs. 10.3 per 100,000 – is much smaller at about 70%.
We physicians need to reflect and learn that while taking race (and other demographic data) when thinking about diagnosis might be helpful, it also affects how we treat people. I don’t recall being taught in medical school on how to manage the potential impact of my biases – we all have them – on my work. My psychiatry residency did help, but it was not part of the formal teaching curriculum. I am not sure how it is today, but managing our biases must be part of the formal training of all who aspire to work in healthcare.
