Diagnosing kidney disease: Removing race from the equation

• Scientists generally estimate the glomerular filtration rate (GFR) — a measure of kidney function — by assessing the levels of creatinine or cystatin C in the body.
• Current equations for estimated GFR (eGFR) that use creatinine or cystatin C incorporate age, sex, and race to get this measurement.
• However, the inclusion of race in these calculations is coming under increasing scrutiny. The reason for this is that race is a social construct, not a biological one.
• Now, a recent study has evaluated the accuracy of GFR-estimating equations that use race, comparing them with new equations that do not factor in race.

Racial disparity in medical practice is not a new phenomenon.

However, conversations addressing disparities have only recently taken on a new urgency in the healthcare industry.

One example is an increase in the awareness of the consequences of using race as a factor in the diagnosis and treatment of kidney disease.

In a new study, scientists have shown that using race-free equations for calculating kidney function — while still assessing creatinine and cystatin C — leads to more accurate results and smaller differences between Black and non-Black people.

Medical News Today spoke with the study’s corresponding author, Dr. Neil Powe, MPH, MBA, leader of the University of California San Francisco Medicine Service at the Priscilla Chan and Mark Zuckerberg San Francisco General Hospital.

He said that the results of the study show how important it is for any clinical research that informs interventions in medicine to recruit a diverse group of participants.

“We found a path forward for estimating kidney function that was a victory for patients, a victory for valuing evidence-based medicine, and a victory for the trainees who were troubled by using race to estimate GFR.”

– Dr. Neil Powe

“We need to do more to discover the drivers of health disparities and do our best to devise interventions to address them,” he concluded.

The study findings appear in The New England Journal of Medicine.

Developing race-free eGFR equations

The study compared the current Chronic Kidney Disease Epidemiology Collaboration equations for estimating kidney function with two new equations that do not incorporate race.

There are three equations that medical professionals currently use to assess GFR. The first uses creatinine (eGFRcr), the second uses cystatin C (eGFRcys), and the third uses both creatinine and cystatin C (eGFRcr-cys).

The researchers developed the new race-free eGFR equations using data from two development datasets:

• 10 studies for serum creatinine, which included 8,254 participants, 31.5% of whom were Black
• 13 studies for both serum creatinine and cystatin C, which included 5,352 participants, 39.7% of whom were Black

All participants were 18 years or older, and in most studies, race was self-reported.

In the first set of new equations, the researchers used the same coefficients for age, sex, and creatinine levels as the current eGFRcr and eGFRcr-cys equations. However, they removed the Black race coefficient when computing eGFR and assigned eGFRcr and eGFRcr-cys values for non-Black individuals to Black individuals.

For the second set of equations, the team completely removed race as a variable for calculating eGFR.

Using a validation dataset of 12 studies — which included 4,050 participants, 14.3% of whom were Black — the researchers compared the performance of the current and new equations.

In addition, the researchers compared the previous equations with the new equations to estimate the prevalence of chronic kidney disease among a representative sample of adults in the United States.

These data came from 4,563 participants from the 1999–2000 and 2001–2002 cycles of the National Health and Nutrition Examination Survey.

Validating the study results

The authors note that in the validation dataset, the current eGFRcr equation overestimated measured GFR in Black participants by a median of 3.7 milliliters (ml) per minute per 1.73 meters squared (m²). It overestimated measured GFR in non-Black participants by a median of 0.5 ml per minute per 1.73 m².

In the first new equation, where the scientists assigned eGFRcr values for non-Black people to Black people, they noted that the measured GFR in Black participants was underestimated by a median of 7.1 ml per minute per 1.73 m².

The race-free eGFRcr equation also underestimated measured GFR in Black participants by a median of 3.6 ml per minute per 1.73 m², although it overestimated measured GFR in non-Black participants by a median of 3.9 ml per minute per 1.73 m².

Additionally, the current eGFRcr-cys equation overestimated measured GFR in Black participants by a median of 2.5 per minute per 1.73 m² and minimally overestimated measured GFR in non-Black participants by a median of 0.6 per minute per 1.73 m².

In an interesting twist, the two new eGFRcr-cys (non-Black coefficient and race-free) equations had a smaller bias in Black participants, a similar bias in non-Black participants, and an overall smaller differential bias than the corresponding new eGFRcr equations.

In comparison with the current eGFRcr equation, the new eGFRcr equations, but not the new eGFRcr-cys equations, increased population estimates of CKD prevalence among Black participants and yielded similar or lower prevalence among non-Black participants.

These findings led the authors to conclude:

“New eGFR equations that incorporate creatinine and cystatin C but omit race are more accurate and led to smaller differences between Black participants and non-Black participants than new equations without race with either creatinine [eGFRcr] or cystatin C alone [eGFRcys].”

Plainly put, the new eGFRcr-cys (non-Black coefficient and race-free) equations were more successful in minimizing the overdiagnosis and underdiagnosis of kidney disease.

What are the experts saying?

MNT reached out to experts for their opinions on the study.

Dr. Eniola Bada, a senior house officer at the Birmingham Women’s and Children’s National Health (NHS) Foundation Trust in England, expressed optimism over the study results. She told MNT that the results might lead to the earlier diagnosis of reduced renal function in Black people:

“[Another] clinical implication [of this study] is that [Black] people who have been on a level of care such as watchful waiting can have their management escalated to dialysis or transplant, thereby potentially improving their clinical outcomes.”

Closing thoughts

According to the authors, the strength of the study is its design. It uses separate large datasets that include both Black and non-Black participants for the development and validation of new race-free equations for estimating GFR.

However, the study is not without its limitations. Firstly, the study categorizes race into two groups — Black and non-Black — which does not adequately represent the diversity within and among ethnic groups. Also, some of the studies that the researchers used to develop the new equations were old, and none were in representative populations.

Furthermore, there were fewer Black participants than non-Black participants in the validation studies, and this may result in less precise estimates of accuracy in Black people. In addition, the representation of ethnic groups other than Black and white was insufficient.

Lastly, the researchers only used data from adult populations without serious coexisting conditions. Therefore, the results may not apply to individuals with more serious illnesses.

Regardless of the shortcomings, one thing is clear: The results from this study highlight racial disparity in medical care and also offer a solution that may, one day, have great clinical significance.

Dr. Bada’s final comments on the study provide pause for thought. She said, “the beauty of science is in its evolution, and as such, scientists, healthcare workers, and the public should be committed to evaluating our current clinical practices, especially the ones that seem ‘not right,’ and we just might find something that is better for everyone.”

Source: Read Full Article