Guess what! eGFR is racist!

[ValeRx]

Yep. You heard it here!

Medical students say the math doctors use to treat patients is racist

---

Members don't see this ad.

 

[confettiflyer]

Did they…just figure that out now? We’ve known this since…I don’t even remember.

This is why you treat patients, not numbers.

::eye roll::

 

[ValeRx]

Did they…just figure that out now? We’ve known this since…I don’t even remember.

This is why you treat patients, not numbers.

::eye roll::

Yeah. I like to ask patients how their kidneys are feeling.

 

[pion]

Well.. it is.

https://www.nejm.org/doi/full/10.1056/NEJMms2004740

 

[owlegrad]

What, medicine has a racist tradition that continues into the present day? Unthinkable!

 

[radio frequency]

Saw the NEJM article last summer.

Honestly though, we probably need a better kidney function metric for all. I have all these 40+ BMI obese patients living unhealthy lifestyles, and their kidney function looks fantastic. But they are so heavy I think it's skewing things.

 

[ValeRx]

Well.. it is.

https://www.nejm.org/doi/full/10.1056/NEJMms2004740

Is it? It is racist? Or potentially because data was limited when creating the calculation? Could it be possible that black patients did not have a large enough sample size? The article itself describing eGFR specifically addressed that elderly and ethnic minorities needed to be studied further. Or is it automatically racist and the reason they did that to black patients is to keep them from proper care because of ethnic superiority?

 

[ValeRx]

Saw the NEJM article last summer.

Honestly though, we probably need a better kidney function metric for all. I have all these 40+ BMI obese patients living unhealthy lifestyles, and their kidney function looks fantastic. But they are so heavy I think it's skewing things.

Similar to elderly. Is this ageism? Should we not assume SCr of 1 rather than what lab values we actually get back?

 

[owlegrad]

Similar to elderly. Is this ageism? Should we not assume SCr of 1 rather than what lab values we actually get back?

There is no biological basis of race, it’s a social construct. There is more genetic diversity within a racial group than there is between racial groups.

I mean did you read the article you posted? Do you have a counter argument?

 

[ValeRx]

There is no biological basis of race, it’s a social construct. There is more genetic diversity within a racial group than there is between racial groups.

I mean did you read the article you posted? Do you have a counter argument?

The article says this:

"The most recent eGFR equation, known as the CKD-EPI equation, was developed using data pooled from 26 studies, which included almost 3,000 patients who self-identified as Black. Researchers found the equation they were developing was more accurate for Black patients when it was adjusted by a factor of about 1.2. They didn't determine exactly what was causing the difference in Black patients, but their conclusion is supported by other research that links Black race and African ancestry with higher levels of creatinine, a waste product filtered by the kidneys.

Put simply: In the eGFR equation, researchers used race as a substitute for an unknown factor because they think that factor is more common in people of African descent."

So they used race as the variable, based off of other research showing higher levels of creatinine, which made the equation more accurate. I am failing to see how that makes eGFR RACIST. Now if the variable made the equation less accurate, I think there would be a basis for such claims.

 

[radio frequency]

Similar to elderly. Is this ageism? Should we not assume SCr of 1 rather than what lab values we actually get back?

I mean...maybe? Honestly though, I'm wondering why we aren't looking more closely at other types of labs to get a better idea of kidney function. This isn't my specialty, but cystatin C might be promising. I wish all the medication approval processes could use something other than Cockcroft Gault; there are just too many variables among individuals that may be missed with our current labs and equations.

 

[ValeRx]

We should also probably take out 0.85 as a factor for females when calculating Cockcroft Gault, since it's probably sexist.

 

[radio frequency]

The article says this:

"The most recent eGFR equation, known as the CKD-EPI equation, was developed using data pooled from 26 studies, which included almost 3,000 patients who self-identified as Black. Researchers found the equation they were developing was more accurate for Black patients when it was adjusted by a factor of about 1.2. They didn't determine exactly what was causing the difference in Black patients, but their conclusion is supported by other research that links Black race and African ancestry with higher levels of creatinine, a waste product filtered by the kidneys.

Put simply: In the eGFR equation, researchers used race as a substitute for an unknown factor because they think that factor is more common in people of African descent."

So they used race as the variable, based off of other research showing higher levels of creatinine, which made the equation more accurate. I am failing to see how that makes eGFR RACIST. Now if the variable made the equation less accurate, I think there would be a basis for such claims.

The concern is that it is inaccurate for individuals, so because it overestimates for larger muscles, it misses early cases of kidney disease (and also denies transplants to those who probably need them already but we presume they do not) primarily in those of African descent as we guesstimate they have larger muscle mass as a whole, meaning we guesstimate their kidneys work better than they may. So it may systematically discriminate against people of African descent.

Never mind that Africa is a very large continent, technically there isn't a biological basis to race, and there is a lot of variation in muscle mass from person to person in all backgrounds. Treat the individual. The currently used equations don't do that. Too bad these are the equations we use for dosing and medication approval processes, too.

 

[Pronounce Doxazosin]

My kidneys identify as females so you better multiply by .742 when you calculate my GFR

 

[ValeRx]

The concern is that it is inaccurate for individuals, so because it overestimates for larger muscles, it misses early cases of kidney disease (and also denies transplants to those who probably need them already but we presume they do not) primarily in those of African descent as we guesstimate they have larger muscle mass as a whole, meaning we guesstimate their kidneys work better than they may. So it may systematically discriminate against people of African descent.

Never mind that Africa is a very large continent, technically there isn't a biological basis to race, and there is a lot of variation in muscle mass from person to person in all backgrounds. Treat the individual. The currently used equations don't do that. Too bad these are the equations we use for dosing and medication approval processes, too.

Which I can go with that, along the premise that the limited information that they used was based in good faith, using data that was based on actual research to represent a patient population - you know, rather than purposeful marginalization and oppression based off a social hierarchy due to ancestry and ethnic group membership.

 

[owlegrad]

Which I can go with that, along the premise that the limited information that they used was based in good faith, using data that was based on actual research to represent a patient population - you know, rather than purposeful marginalization and oppression based off a social hierarchy due to ancestry and ethnic group membership.

I respectfully submit that you have a very narrow definition of racism.

 

[ValeRx]

I respectfully submit that you have a very narrow definition of racism.

: a belief that race is a fundamental determinant of human traits and capacities and that racial differences produce an inherent superiority of a particular race

: the systemic oppression of a racial group to the social, economic, and political advantage of another

 

[owlegrad]

: a belief that race is a fundamental determinant of human traits and capacities and that racial differences produce an inherent superiority of a particular race

: the systemic oppression of a racial group to the social, economic, and political advantage of another

Where in that definition does the oppression have to be purposeful to be racist as you stated above?

 

[radio frequency]

Which I can go with that, along the premise that the limited information that they used was based in good faith, using data that was based on actual research to represent a patient population - you know, rather than purposeful marginalization and oppression based off a social hierarchy due to ancestry and ethnic group membership.

Here's the thing though -- racism doesn't have to be purposeful or intentional. It often isn't. It's sometimes created by well-intentioned policies even. It just has to result in harm to a non-majority power group. I'm pretty certain these equations were meant to help and not harm when created. But their usage has caused harm.

 

[mentos]

All eGFRs matter.

 

[VA77]

Always makes me think of this scene.

 

[confettiflyer]

Makes me wonder if y’all have different PRN hypertension sequences in inpatient order sets for Black vs non-black patients.

 

[mentos]

How do Hispanic people feel when job applications ask if they're Hispanic or non Hispanic?

 

[Sine Cura]

What if that scene used, instead of the word Oreo.... check dat pun

 

[johnpharm01]

What if that scene used, instead of the word Oreo.... check dat pun

at least its nephrology, not phrenology

 

[zelman]

I was always confused by the MDRD using “African American” as a modifier. If somebody is visiting from Africa they can change their kidney function just by becoming a citizen? And Ethiopians with near-black skin and South Africans who look like Dave Matthews have the same kidney-affecting genes? It’s bananas.

 

[confettiflyer]

I was always confused by the MDRD using “African American” as a modifier. If somebody is visiting from Africa they can change their kidney function just by becoming a citizen? And Ethiopians with near-black skin and South Africans who look like Dave Matthews have the same kidney-affecting genes? It’s bananas.

I think you just have to translate African American to mean “descendant of slaves that were subject to a genetic selection event in favor of a CPY3A5 allele resulting in increased sodium retention.”

Cuz you know an Afro-Cuban is probably going to follow with the same issues. Terms are problematic, hence again why we all need to treat patients, and not numbers or labels.

 

[KidPharmD]

CKD-EPI (and MDRD) using race as a substitute for an unknown genetic effect isn't inherently racist. It does, however, make it a bad equation. What makes it a "racist" equation is that we use it to calculate an eGFR that we use for so many decisions. This correction (that may not actually be accurate in an individual) is used to determine who gets offered what drugs, who gets prepared for dialysis, who gets placed on the transplant list (and where). All of that has meant that Black kidney patients have had worse outcomes at "higher" eGFRs than their white counterparts.

The failure to create an eGFR equation that actually works shouldn't punish black patients.

 

[owlegrad]

Terms are problematic, hence again why we all need to treat patients, and not numbers or labels.

I have to confess I’ve never fully understood this term. What does "treat the patient" mean in the context of not treating numbers or labels? I mean numbers such as LDL or labels such as pregnant are the basis of evidence-based medicine. I am not trying to be flippant but I just don't know what it means to "treat the patient". Treating the patient means using their numbers and labels to figure out the best treatment plan using evidence-based medicine, no? To me, it sounds like a useless platitude. Functionally what is the difference between treating a number that belongs to a patient and treating a patient using their numbers?

I mean for example I take a statin. I would not like my provider telling me that he doesn't use labels like hyperlipidemia or treat numbers like LDL level but instead treats me as the patient. Uhhhh I don't need to be validated as a patient, I need you to treat my medical conditions using evidence-based medicine. What am I missing in the term "treat the patient not the number"? What would be an example where you would treat the patient not the condition or the number?

 

[lord999]

I have to confess I’ve never fully understood this term. What does "treat the patient" mean in the context of not treating numbers or labels? I mean numbers such as LDL or labels such as pregnant are the basis of evidence-based medicine. I am not trying to be flippant but I just don't know what it means to "treat the patient". Treating the patient means using their numbers and labels to figure out the best treatment plan using evidence-based medicine, no? To me, it sounds like a useless platitude. Functionally what is the difference between treating a number that belongs to a patient and treating a patient using their numbers?

I mean for example I take a statin. I would not like my provider telling me that he doesn't use labels like hyperlipidemia or treat numbers like LDL level but instead treats me as the patient. Uhhhh I don't need to be validated as a patient, I need you to treat my medical conditions using evidence-based medicine. What am I missing in the term "treat the patient not the number"? What would be an example where you would treat the patient not the condition or the number?

When you run into contradictory symptoms that the numbers don't work to paint a real picture, improvise.

For instance, what's the real ceiling dose for opioids? It's actually respiratory rate. If there's an oncology patient that is still breathing fine but is in obvious agony in the hospice with 1 g (1000 mg) of Dilaudid, you go to 2 g or higher (bear in mind that just 500 mg is part of a capital punishment-approved dose in most humans). If the patient is still breathing and still suffering, whatever it takes. And my personal max that I personally observed in practice was some terminal osteosarcoma sufferer who was lucid and functional at 8 g/d of Dilaudid and that is nowhere near the highest I've heard of.

The vast, vast majority of the time, the guidelines work. But it is our job to recognize when we are in a situation where we should go off-script. Those sorts of instincts aren't explicitly teachable, they come only with practice. Your own experience in nursing home and geriatrics gives you a feeling for those that no textbook records all of those judgment calls.

In your lipids example, someone who is absolutely treatment resistant to statins. That patient has other problems and that's when the health care team starts breaking out options like ileal surgery and other drastic plans. It is very rare, you'll probably encounter it two or three times in your entire career if you deal with hyperlipidemia all the time, but you do need to know when the guidelines are not sufficient.

But yeah, you and I are like everyone else, unless we're not. Hopefully, we don't win the rare case lottery as its usually bad.

The guidelines work. You shouldn't just deviate from them to be contrary, but when you do have an explicit reason (and usually that means treatment refractive with proof), then you treat the patient different. Guidelines do treat the patient, but you go beyond them if the guidelines don't get you the result you want out of your patient.

The philosophical difference in the balanced idea of "treat a patient" from the extremes of guideline automaton versus erratic iconoclast is the difference between artisanal versus artistic technique. We (the professoriate) train you on the artisanal values of craft: tradition, repetition, and worksmanship of pharmacy. But the practice demands you raise that craft to the level of art: emotion, expression, and creation beyond what is written to what you experience and what your judgment derives from your synthesis of both. I wish that was communicated better between the didactic and the practical portions of pharmacy training.

 

[confettiflyer]

I have to confess I’ve never fully understood this term. What does "treat the patient" mean in the context of not treating numbers or labels? I mean numbers such as LDL or labels such as pregnant are the basis of evidence-based medicine. I am not trying to be flippant but I just don't know what it means to "treat the patient". Treating the patient means using their numbers and labels to figure out the best treatment plan using evidence-based medicine, no? To me, it sounds like a useless platitude. Functionally what is the difference between treating a number that belongs to a patient and treating a patient using their numbers?

I mean for example I take a statin. I would not like my provider telling me that he doesn't use labels like hyperlipidemia or treat numbers like LDL level but instead treats me as the patient. Uhhhh I don't need to be validated as a patient, I need you to treat my medical conditions using evidence-based medicine. What am I missing in the term "treat the patient not the number"? What would be an example where you would treat the patient not the condition or the number?

What lord999 said, lol.

Unrelated examples here:

1) aggressive treatment of blood sugar in someone with limited life span. guidelines call for a1c goals, but if the benefits accrued are long term (microvascular complications, etc), you’re just making the patient miserable. Don’t do that.

2) pressing a statin on a child-bearing age diabetic…okay that’s not treating a number, but c’mon… I’ve had providers do this because it’s a metric they’re being judged against by some insurance company.

3) accepting an a1c that’s within goal, but was obtained by monster hypoglycemic events

Some of these I’m just describing bad clinician habits and failure to really dig in, but if all you do is stick to treating numbers, you’re missing the bigger picture.

Like, say if I came in with high LDL and a clinician just prescribed me a statin, but isn’t addressing a root cause of over eating — depression, or even lack of dietary information— that’s problematic.

 

[VA Hopeful Dr]

What lord999 said, lol.

Unrelated examples here:

1) aggressive treatment of blood sugar in someone with limited life span. guidelines call for a1c goals, but if the benefits accrued are long term (microvascular complications, etc), you’re just making the patient miserable. Don’t do that.

2) pressing a statin on a child-bearing age diabetic…okay that’s not treating a number, but c’mon… I’ve had providers do this because it’s a metric they’re being judged against by some insurance company.

3) accepting an a1c that’s within goal, but was obtained by monster hypoglycemic events

Some of these I’m just describing bad clinician habits and failure to really dig in, but if all you do is stick to treating numbers, you’re missing the bigger picture.

Like, say if I came in with high LDL and a clinician just prescribed me a statin, but isn’t addressing a root cause of over eating — depression, or even lack of dietary information— that’s problematic.

My experience has been 2 fold in your last paragraph there.

First, LDL doesn't seem all that effected by diet. It can make some difference but not usually enough to not need other treatment.

Second, patients almost never succeed in making significant lifestyle changes. I spend a fair bit of time on that topic, but usually doesn't happen.

 

[Sparda29]

I think you just have to translate African American to mean “descendant of slaves that were subject to a genetic selection event in favor of a CPY3A5 allele resulting in increased sodium retention.”

Cuz you know an Afro-Cuban is probably going to follow with the same issues. Terms are problematic, hence again why we all need to treat patients, and not numbers or labels.

I had an argument with one of the pharmacists here about Vancomycin dosing. I was saying that the docs here are very conservative in their Vancomycin dosing and rarely go beyond q12h frequency or 1500 mg per dose.

My argument is, "my vancomycin Pk calculator said that this was the appropriate initial dosing", her argument is "you've never seen the patient, all you know is the lab work and whatever height/weight the nurse claims the patient is, the doctor has seen the patient, maybe they are dehydrated, etc".

Isn't it better to get too high of a trough and then adjust the dose lower/less frequent instead of too low of a trough? Too high of a trough, you are taking care of the infection with a slight risk of nephrotoxicity. Too low of a trough and you are risking antibiotic resistance.

 

[BenJammin]

I have to confess I’ve never fully understood this term. What does "treat the patient" mean in the context of not treating numbers or labels? I mean numbers such as LDL or labels such as pregnant are the basis of evidence-based medicine. I am not trying to be flippant but I just don't know what it means to "treat the patient". Treating the patient means using their numbers and labels to figure out the best treatment plan using evidence-based medicine, no? To me, it sounds like a useless platitude. Functionally what is the difference between treating a number that belongs to a patient and treating a patient using their numbers?

I mean for example I take a statin. I would not like my provider telling me that he doesn't use labels like hyperlipidemia or treat numbers like LDL level but instead treats me as the patient. Uhhhh I don't need to be validated as a patient, I need you to treat my medical conditions using evidence-based medicine. What am I missing in the term "treat the patient not the number"? What would be an example where you would treat the patient not the condition or the number?

I'm honestly shocked that something like this is coming from you.