CKD-EPI 2021 eGFR Calculator

What the CKD-EPI 2021 equations are

CKD-EPI is the Chronic Kidney Disease Epidemiology Collaboration.

The 2021 creatinine equation (eGFRcr) and the 2021 creatinine-cystatin C equation (eGFRcr-cys) estimate adult GFR without a race coefficient. Enter cystatin C and this page reports both so you can see whether they agree near a cutoff. Cystatin C comes from nucleated cells and is filtered by glomeruli. It tracks muscle mass less than creatinine. Combined equations usually sit closer to measured GFR. Cystatin C is still missing from many routine panels, so labs still lead with eGFRcr.

NKF told labs to switch MDRD and 2009 creatinine reporting to 2021 eGFRcr, and to use eGFRcr-cys when cystatin C is on the report. The 2012 cystatin-only equation remains for unusable creatinine. This page runs the two 2021 equations.

[1][2][3]

Formulas used on this page

eGFRcr = 142 × min(Scr/κ, 1)^α × max(Scr/κ, 1)^-1.200 × 0.9938^age × 1.012 if female. κ 0.7/0.9 and α -0.241/-0.302 (female/male). Creatinine is mg/dL.

eGFRcr-cys = 135 × min(Scr/κ, 1)^α × max(Scr/κ, 1)^-0.544 × min(Scys/0.8, 1)^-0.323 × max(Scys/0.8, 1)^-0.778 × 0.9961^age × 0.963 if female. Here α is -0.219 (female) or -0.144 (male). Cystatin C is mg/L. Leave it at 0 if you do not have it and the combined row stays blank.

Both equations need age 18 or older and a stated sex. NKF lets labs print male and female estimates when sex is missing. This tool asks you to pick one.

[2][3][4]

Worked comparison

A 60-year-old woman, creatinine 1.0 mg/dL: eGFRcr 65. Cystatin C 1.1 mg/L: combined 66. Near the G2/G3a line, that match is useful.

A muscular person can have a lower creatinine-based estimate than the combined one. Untreated hyperthyroidism, high-dose glucocorticoids, and some inflammatory states can push cystatin C out of line with creatinine. The page shows the difference. It does not name the true GFR.

Miller and colleagues told labs not to trend 2021 values with 2009 output on the same flowsheet. Do not average the two eras.

[2][3][4]

How to use the CKD-EPI calculator

Enter adult age, sex, and standardized creatinine. Add cystatin C in mg/L from the same window if you have it.

A cystatin drawn days later in an acute illness does not pair with last week’s creatinine. NIDDK prefers the combined equation near a drug, donation, or transplant cutoff. Take a large gap to a clinician instead of keeping the friendlier number.

Use G categories as language, then add ACR before you say “stage.” This page has no cause or albuminuria, so it does not print a CGA string.

[1][5]

When cystatin C changes the interpretation

Creatinine misleads when muscle is far from average: amputation, sarcopenia, bodybuilding, neuromuscular disease, some eating disorders.

Cystatin C misleads in some thyroid disease, very high-dose steroids, and some inflammatory states. Neither is a gold standard. Cystatin C can rise faster than creatinine in acute hyperthyroidism and fall in untreated hypothyroidism. Glucocorticoids can raise it. Smoking and adiposity associate with higher cystatin C in epidemiologic work. The combined equation is a better average estimator. It is still not a biopsy.

NIDDK notes that 2021 creatinine-only eGFR can shrink apparent racial differences in kidney-failure risk versus combined or cystatin-only equations. That is a population finding. For one person, the question is whether the estimate is close enough to measured GFR for this decision. If the two estimates sit on opposite sides of 30, 45, or 60, pause metformin, SGLT2, or chemotherapy changes. Recheck labs, get ACR, and measure GFR when the stakes are high. Repeat both markers a week later on a stable outpatient day rather than averaging two same-day disagreements in a spreadsheet. NIDDK especially wants cystatin C near donation, chemotherapy, or anticoagulation thresholds.

[1][3][5]

Why 2021 replaced 2009

The 2009 equation used a 1.159 multiplier for people labeled Black. Race is a social label, not a filtration-marker term.

The NKF-ASN task force backed race-free 2021 equations after looking at accuracy and at who gets drugs, transplant, and referral. 2021 was refit on the same development data without race. Group performance stayed acceptable. It is not more accurate than 2009 in every subgroup. The choice was to drop race as an input and keep a creatinine report every lab can run.

Two old numbers labeled Black and non-Black came from 2009 or MDRD. Switch the chart to a 2021 value. Do not average the pair.

[1][3][4]

A discordant-marker case

A 28-year-old lifter, creatinine 1.4 mg/dL: eGFRcr in the high 60s, cystatin C 0.8 mg/L, combined near 90, ACR 8 mg/g. The creatinine equation is seeing muscle. Record both estimates. Do not label him G3 on creatinine alone.

A 78-year-old woman, BMI 19, creatinine 0.7, cystatin C 1.4: eGFRcr looks G2, combined falls to G3a. Get ACR, review drugs, and lower the bar for nephrology. The creatinine-only number is the reassuring one and the weaker one here.

[1][5][7]

Laboratory reporting details

LOINC 98979-8 is 2021 eGFRcr. 98980-6 is 2021 eGFRcr-cys. Age has to be known. Unknown sex is a reporting problem, not a reason to drop the female factor.

CKD staging uses indexed eGFR. For some drugs, NKF wants unindexed mL/min (indexed × BSA/1.73). This page skips that BSA step. Take height and weight to a clinician or pharmacist if the label wants mL/min.

Assays must be standardized. If the report does not say IDMS creatinine, ask the lab before a dosing talk.

[3][4][6]

Limitations

The equations were built in people with stable function. Skip AKI, pregnancy after the first trimester, and dialysis.

Combined estimates still scatter around measured GFR. An 8 to 10 mL/min/1.73 m² gap can be noise. This page does not run 2012 cystatin-only, CKiD, or EKFC. Trust the equation name on the lab printout when records disagree.

If eGFRcr is 44 and eGFRcr-cys is 62, do not keep 62 to dodge referral. Recollect both markers, measure ACR, and ask about muscle, thyroid, and steroids. Chemotherapy or donor work still wants measured GFR. Race-free reporting does not fix access to nephrology, transplant, or cheap drugs. Blood pressure, glucose, albuminuria, and follow-up still do the clinical work. Do not start, stop, or split prescriptions from this comparison alone. Obstruction and glomerulonephritis need a workup no equation can give.

[1][5][6]

How it works

Creatinine-only eGFRcr uses the 2021 CKD-EPI creatinine equation. If cystatin C is entered, eGFRcr-cys = 135 × min(Scr/κ,1)^α × max(Scr/κ,1)^-0.544 × min(Scys/0.8,1)^-0.323 × max(Scys/0.8,1)^-0.778 × 0.9961^age × 0.963 if female.

Frequently asked questions

• When should cystatin C be added?NIDDK notes that the combined equation is more accurate than creatinine alone, especially near a drug-dosing or transplant threshold, or when creatinine may misrepresent muscle mass.
• Can I trend 2021 results with 2009 results?No. Coefficients changed. NKF laboratory guidance says 2021 values should not be allowed to trend with older equations.
• Does this stage CKD by itself?Staging uses cause, GFR category, and albuminuria category together (CGA). This page reports GFR category only.

Related calculators

References

1. NIDDK. Estimated Glomerular Filtration Rate Calculators
2. National Kidney Foundation. CKD-EPI Creatinine Equation (2021)
3. Miller et al.. Laboratory engagement recommendations for CKD-EPI 2021
4. NIDDK. eGFR Equations for Adults
5. KDIGO. KDIGO 2024 CKD Guideline
6. National Kidney Foundation. NKF eGFR calculator notes on BSA adjustment
7. Inker et al., 2021. New Creatinine- and Cystatin C-Based Equations to Estimate GFR without Race