Howard Straker, PA-C, EdD, Susan LeLacheur, DrPH, PA-C, faculty members in the department of physician assistant studies at Washington, D.C.-based George Washington University, and Kara Caruthers, PA-C, associate professor at Nashville, Tenn.-based Meharry Medical College, recently sat down with Becker's to discuss how race is used in clinical algorithms and how that can impact patient care for people of color, and to offer some insight on how to remove racial elements from clinical medicine.
Editor's note: Responses have been lightly edited for clarity and length.
Question: In your experiences, how has race come up in clinical algorithms, and how is it affecting people of color?
Kara Caruthers: We have to acknowledge that racism in medicine exists. It truly is entrenched, baked in and really constructed in everything we do — even how we learn about things.
[One example is] our kidneys and kidney function. There is a particular calculation that all clinicians do, that allows us to know if someone's kidney is functioning at an optimal level. Based on that calculation it kind of depends on what medications we give, the strength and concentration of the medication, how much we give and if we need to refer to a specialist to do some additional treatment.
What is unfortunate — what has really been built into medical practice — is for that calculation, there's the non-Black calculation, which is everybody else, white, Asian, Hispanic, etc. Then there's the Black calculation. There's actually a difference in how the calculation factors, which put Black people specifically — not other people of color, but Black people specifically — in harm's way, meaning we don't treat them as fast when they have kidney disease or they're not eligible to be on the kidney transplant list [as an example.]
We used this race-based calculation for decades until late 2021 when the National Kidney Foundation and other organizations called for the use of a race-neutral calculation. The race-based calculation caused us to miss a group of patients whose kidneys were not functioning at optimal levels. And since there is no gene for race, why on earth would we have a medical calculation for kidney function that harms, delays treatment for Black people and their kidneys?
Dr. Susan LeLacheur: With lung function, there was until last summer a calculation that is built into most spirometers that we still use today that requires you to put in the patient's race when you're doing the test.
At a lower functional level, a Black person will be called normal, whereas a white person will not. Now think about the effect of that during COVID.
[Additionally], there's a test we use all the time called a pulse oximeter that tests your oxygen level. Next time you're in a public restroom, watch Black people and white people go to the sink with those automatic faucets and try and turn them on. How does that system work? It's shining a light, it's testing the light. It picks up [white] skin just fine.
Same thing happens with the pulse oximeter. It is calculated to work for people with white skin, not for people with darker skin. So, just like that automatic faucet, it's giving an abnormal or wrong reading and it gets increasingly wrong as the level of lung function decreases.
Dr. Howard Straker: Race is a social concept. If it's a social function, then how do we have biological parameters tied into this? How do they connect? I began realizing that all these objective measures where we put race into them are playing a social role, that we're adding into something that's supposed to be a biological objective function, as we see with the kidneys or with the lungs.
Those objective measures as we looked back through history are based on biases and opinions that people had. Any kind of algorithm is really based on some kind of functioning, and a lot of our algorithms are built on bias and built into the system.
Historically … people were starting to use the word structural racism. For us in medicine, one of the ways of structural racism is these biases are built into these objective measures. If you ask anybody, they will say, "Well, I get their creatinine level, and I get their [estimated glomerular filtration rate.] That's objective, I can't mess with that."
Now, we realize it does make a difference. People have made decisions about who gets a hospital bed, based on systems. They come in with acute heart failure, and there's a system: Do they go to the ICU? Do they go to a cardiac unit? Do they go to the regular floor based on this calculation? And then the calculation asks, 'Are they Black?' No. And they put that in there. How does that social [function] suddenly become biological?
Now, the truth is that one of the major social functions of race is discrimination, and one of the reasons we actually pay attention to race in this country is because of a historical system of discrimination. It's every day. So the federal government at one point really started classifying people by race so they could see if people were getting the same treatment, whether it was an education, or whatever else. That allows us now to see how we don't get the same treatment.
Q: What should healthcare organizations consider in respect to clinical guidelines?
HS: At the end of last year, the Biden administration put forth, through the Department of Health and Human Services, this intention to combat algorithms that have discrimination baked into them. It hasn't happened. They've put it out for public review and it hasn't happened yet. But one piece of it was going to put a level of responsibility on clinicians.
KC: This is bigger than each of us individually. But each of us individually [can be] more intentional and mindful about how we use these clinical shortcuts, because that's what they are.
[Clinicians] can ask: If I use this race modification, is it going to harm my patient? Is it delaying care, not giving care? If people were to think, "Well, why do we do this?" and "Wait a minute, that's not right, because I just had a patient that was white yesterday that had the same thing, and I admitted them to the hospital or to the ICU, because they need more care." Instead of, "This patient, because they're Black, they're just going to go to the floor to a nursing staff that has six to eight patients, versus going to ICU where they may have one or two."
So, how do we get people to use our critical thinking skills and to think and pause before you subscribe it as absolute truth? I think what some of us forget, in both medicine and science in general, it's a lot of theories, which means as we get more information we also have to be willing to adapt.
What we're doing is we're offering additional information, to encourage our colleagues to adapt, to petition, to advocate for patients. I tell students and potential students all the time PA stands for patient advocate. And if you cannot advocate for your patient, I don't want you to be a PA.
SL: One of the things that we want to push [organizations] a little [more] on is how do you fix the damage already done. It's not just not using race anymore in the algorithm. What about all those people who came last year before we change this? What about all those people who are on the transplant list, but way down, because their kidney function was miscalculated? What about all those people whose lung function requires a different level of care at this point? How do we fix the damage already done?
Q: What are the three questions providers should ask themselves when considering race-based algorithms?
HS: The three questions that we have, actually, come from a New England Journal of Medicine article. One is, if there's a need for race-based correction, that's what we call our race-based modification, is it based on robust evidence? We are taught in evidence-based medicine to look at the internal/external validity, bias and what kind of confounders are present. One of the reasons this is a difficult thing is that it forces you to have to go read medical literature and go back into that stuff. So, it's correct, but it's difficult.
The second question is, is there a mechanism that will make sense of this? If we're saying this is a social function, and we're applying it biologically, what's the biological mechanism that makes this work? And clinicians can start thinking that through.
Now, if you believe that Black people have more muscle mass than white people, then you're going to accept that idea for EGFR, right? So we want you to then also double check your own biases. There are people who believe that Black people have thicker skin and don't feel pain, right? And if you have that bias already, then the mechanism is already there for you to believe that you shouldn't give more pain medicine.
The third one, which is the easiest one to kind of look at is, if you implement race-based modification, does it exacerbate problems? Does it make them worse? And in what way?
I want to emphasize what Susan was talking about, which is, in essence, we have three official questions that are based in the literature, but we actually have four questions. The fourth being, how can we take care of the harm that we've already done?