Hunting for zebras

Avatar photo
Post by Kirsten Bibbins-Domingo, PhD, MD, MAS

My 21-year old niece died in 2017. I visited her at one of the nation’s premier medical centers, where she had received an emergency liver transplant soon after her diagnosis of Wilson’s disease. Although her body accepted the new liver, her brain function never returned and her family withdrew support soon after my visit.

My niece was smart and beautiful. She was an only daughter with three brothers, so she was the light of both her parents’ and her brothers’ lives. She was a very active college senior who, over many months, had returned home on several occasions because of extreme fatigue and unexplained abdominal pain.  She was noted to have mildly elevated liver tests, and her symptoms were variably attributed to anxiety, stress, depression, oral contraceptives, and non-alcoholic steatohepatitis linked to obesity. As her symptoms worsened, so did her liver tests. The diagnosis of Wilson’s disease came too late because her clinicians reasoned that the disease did not occur in Black patients; they only tested for the condition after her liver failed. 

Wilson’s disease is a rare (1 in 30,000) autosomal recessive genetic disorder caused by loss of function of the ATP7B copper-binding protein. It may present through hepatic, neurologic or psychiatric problems. It has been diagnosed throughout the world, without clear patterns by race, ethnicity or region (except for founder effects in Sardinia and a region in Costa Rica). 

The clinicians who commented that “We don’t really think of Wilson’s disease occurring in Black patients” are technically correct. It’s a rare disorder that we don’t necessarily think of in most patients and doesn’t occur at proportionately higher rates in Black patients. But they were completely wrong in opting not to evaluate for the possibility of this disease, particularly given my niece’s age, constellation of symptoms, and chronic, worsening liver tests. Their comments highlight the concept of “statistical discrimination,” a less-discussed contributor to unequal and inequitable patterns of care that result from our racialized approach to medical decision-making.

The landmark Institute of Medicine report “Unequal Treatment” identified three types of discrimination that occur in the health care setting – affective bias, cognitive stereotyping and statistical discrimination. The basic idea of statistical discrimination in health care is that uncertainty about the patient’s severity of illness can induce clinicians to behave differently with otherwise identical presentations for patients with different racial backgrounds. If the underlying prevalence of the illness is associated with race, the doctor might take race into account in deciding about the diagnosis and treatment of a particular patient. In statistical discrimination, physicians apply correct information about a population to reduce their clinical uncertainty about an individual patient. Statistical discrimination may be appropriate if based on reliable data on group differences in the absence of data about the individual, but ignoring information about the individual results in inequities ultimately rooted in the patient’s identity and perpetuated by our racialized shortcuts in medical decision-making.

Many of us have drawn attention to diseases and outcomes that pattern by race and ethnicity, important work because these patterns often mirror inequities in the social and structural determinants of health that contribute to these diseases and must be addressed if disparities are to be eliminated.  But as we teach and as we practice, that focus on patterning of disease risks over-emphasizing group effects, a practice that is often reinforced by our own racial biases, and ultimately interferes with good clinical decision-making for the individual patient.

Could an earlier diagnosis have averted the devastating outcome for my niece? It’s hard to know, but it is certainly tempting to speculate since it was neurologic decline that led to withdrawal of support – neurologic decline that was thought to be the result of accumulated exposure as the disease progressed and her liver failed. Wilson’s disease is treatable with medications or liver transplantation, the latter of which my niece was fortunate to receive, albeit too late in the disease course.

My niece loved math and economics and was planning on pursuing a degree in public health once she graduated. She wanted to combine her interest in numbers with her desire to help people. I search to find a lesson in her death that can also help others in medical decision-making overcome our historical and contemporary racial biases.

Whether looking for horses or hunting for zebras, as physicians we must always guard against blinders that lead us to look only in particular places for the zebras. And, as physicians who care for minoritized patients and who observe and comment on how diseases and outcomes pattern by race and ethnicity, reminding ourselves about how patterns themselves can reinforce our own biases, and ultimately affect good clinical decision-making, is critical so that we can provide all patients with the care they deserve.

Kirsten Bibbins-Domingo, PhD, MD, MAS is the Professor and Chair of the Department of Epidemiology and Biostatistics, and the Lee Goldman, MD Endowed Chair and Professor of Medicine at the University of California, San Francisco. Dr. Bibbins-Domingo is a general internist and cardiovascular epidemiologist whose scholarship includes observational epidemiology, pragmatic trials, and simulation modeling to examine clinical and public health approaches to prevention in the US and globally.