Alumni Perspective: Mark One or More Boxes
A young alumnus debates the utility of clustering patients into demographic categories.
The 80-year-old woman came to my ER with lower abdominal pain. I started thinking through my differential: diverticulitis, urinary tract infection, maybe appendicitis. Her labs came back showing nothing. During a second round of questions, she mentioned a new boyfriend at the nursing home. It turned out that my frisky octogenarian had a case of chlamydia.
We categorize people in medicine all the time. Young, old, black, white, female, male, this, that. Every demographic survey has a slew of boxes that tries to compartmentalize us as people. These boxes and I have a bipolar relationship. I find boxes complicated because they are frustratingly inexact and reductionist but still point out significant societal trends to address.
The analytical side of me contests the boxes. What should be an orderly and intuitive grouping generally ends up a messy hodgepodge of categories forming an impossible Venn diagram.
Consider the U.S. government’s five minimum categories for collecting data on race: American Indian or Alaska native, Asian, Black or African American, Native Hawaiian or other Pacific Islander, and White — the race boxes are a mix of skin tone, people group and geographic region. What about my Russian friends? Do they pick Asian to match geography or white to match skin color? Or friends from Algeria or Pakistan, both of whom the government unceremoniously dumps into the white category.
“THESE BOXES AND I HAVE A bipolar relationship. I find boxes complicated because they are frustratingly inexact and reductionist but still point out significant societal trends to address.”
The category of Pacific Islander includes less than 2 million people globally, while the category Asian includes 4.5 billion and would lump together the experiences of Chinese and Indian Americans. If someone orders Asian food, they would be surprised if they got dosas and tandoori chicken, which means that Grubhub somehow outpaces many electronic health records at differentiating people groups. Boxes reduce individuals to often ill-fitting categories that may not reflect their experience.
Boxes perform better in revealing trends of inequality in populations, where large sample sizes average out disparate individual experiences. The ignoble groupings have a brutal simplicity, yet still manage to reveal large disparities. What the roughhewn categories lack in specificity, they make up for in unearthing areas to research: White people are x times more likely to have health insurance than black people. The statistics shed light on aspects of American life in which certain groups have unequal outcomes. As physicians, we must both be cognizant of these patterns and seek to eliminate them.
As a physician, I use demographic information daily to risk stratify patients while taking their history. There are many determinants of health, and understanding the risk ratios associated with certain population groups helps to steer my workups. That being said, individuals are unique. I anchored on my 80-year-old’s likely diagnoses differently because we less frequently associate sexually transmitted infections with the elderly. It is crucial to remember that our patients may come from anywhere on the bell curve.
At a population level, knowledge of disparities between groups helps me advocate to eliminate those differences. I chose to pursue a master of public health degree during medical school so that I could better understand the disparate epidemiology of disease and address it through policy, especially among victims of gun violence and the residentially displaced. My hope is that through advocacy and hard work, the demographic boxes that our patients check will cease being risk factors for disease.