eGFR vs Cockcroft-Gault — Which to Use for Drug Dosing
There are at least three formulas in routine clinical use for estimating kidney function. They give different numbers for the same patient. The choice of formula matters most for two things: drug dosing of renally-cleared medications, and CKD staging for diagnosis and prognosis. Different goals, different formulas.
The three formulas
Cockcroft-Gault (1976)
CrCl = ((140 − age) × weight_kg) / (72 × Scr_mg/dL)
× 0.85 if female
Reports in mL/min — NOT body-size normalised. The original "estimate kidney function from a serum creatinine" formula. Still widely used for drug dosing because most renal-dose recommendations were validated with it, and because absolute clearance (mL/min) maps better to drug elimination than body-surface-normalised clearance for many agents.
MDRD (1999)
eGFR = 175 × Scr^-1.154 × age^-0.203 × (0.742 if female) × (1.212 if Black, in older versions)
Reports in mL/min/1.73 m² — body-size normalised. The first widely-adopted eGFR formula. Less accurate at high GFR (>60). Being phased out.
CKD-EPI 2021 (race-free)
GFR = 142 × min(Scr/k, 1)^α × max(Scr/k, 1)^-1.200 × 0.9938^age × (1.012 if female)
where k = 0.7 (female) / 0.9 (male), α = -0.241 (female) / -0.302 (male)
The current KDIGO-recommended standard. The 2021 update removed the race coefficient that earlier CKD-EPI and MDRD versions included — a change driven by years of equity advocacy after the original race adjustment was shown to lead to under-treatment of Black patients. This is what most modern lab reports give you.
Why they give different numbers
For a 60-year-old female, 70 kg, serum creatinine 1.0 mg/dL:
- Cockcroft-Gault: ((140-60) × 70) / (72 × 1.0) × 0.85 = 66 mL/min
- MDRD: 175 × 1.0^-1.154 × 60^-0.203 × 0.742 = 62 mL/min/1.73m²
- CKD-EPI 2021: ~66 mL/min/1.73m²
The numbers are close in this case but can diverge significantly at extremes — particularly in obesity, very low or very high muscle mass, and in elderly patients. In some clinical scenarios, the same patient can be CKD stage G2 by one formula and G3a by another.
Which to use for what
Drug dosing — Cockcroft-Gault
Most renally-dosed drugs (vancomycin, gentamicin, NOACs / DOACs, many chemotherapeutics, antibiotics, gabapentin, allopurinol) have dose adjustments that were originally validated with Cockcroft-Gault and are still specified that way in package inserts and clinical guidelines.
The reason isn't tradition alone: absolute clearance in mL/min maps directly to how much drug is being filtered, while body-surface-normalised values (mL/min/1.73 m²) abstract that away. For a drug that's filtered without further metabolism, the absolute clearance is what determines dose.
This matters in extremes: a 50 kg patient and a 110 kg patient with the same eGFR (mL/min/1.73 m²) have very different absolute clearances — and probably need different doses.
CKD staging — CKD-EPI 2021
For diagnosing and staging chronic kidney disease, KDIGO guidelines specifically recommend CKD-EPI 2021. The stage cutoffs (G1 ≥90, G2 60-89, G3a 45-59, G3b 30-44, G4 15-29, G5 <15) are validated against CKD-EPI eGFR. MDRD was the previous standard but is being replaced.
Population research / lab reports — CKD-EPI 2021
Most labs in the US now report CKD-EPI 2021 eGFR. You'll see it on the basic metabolic panel as "eGFR" with a value in mL/min/1.73 m² and a footnote about the formula.
The 2021 race-free update
Earlier versions of MDRD and CKD-EPI included a coefficient that adjusted eGFR upward for Black race. This was originally based on observational data showing higher mean creatinine in Black populations attributed to muscle mass differences. Subsequent analysis raised concerns: the adjustment was leading to systematically higher eGFR (and therefore later CKD diagnosis, delayed referral to nephrology, and reduced eligibility for transplant lists) in Black patients.
The National Kidney Foundation and ASN convened a task force; in 2021 they recommended removing the race coefficient. The CKD-EPI 2021 formula is the result. Most major labs in the US adopted it through 2022-2023.
The cystatin C alternative
For patients where creatinine-based eGFR is unreliable (extremes of muscle mass, certain medications, amputation, severe obesity), cystatin C is an alternative biomarker. CKD-EPI also has a cystatin C formula (CKD-EPI Cr-Cys 2021) that combines both for the most accurate estimate. Cystatin C testing is more expensive and less universally available; it's typically used when there's reason to doubt the creatinine-based estimate.
Skip the math
The Creatinine Clearance + eGFR Calculator computes all three formulas (CKD-EPI 2021, MDRD, Cockcroft-Gault) in one tool with explicit "which to use when" guidance. CKD stage classification (G1–G5) for the eGFR results. mg/dL OR µmol/L unit toggle. On-device.
The short version
For drug dosing, use Cockcroft-Gault — most package inserts and renal-dosing guidelines still specify it, and absolute clearance (mL/min) maps better to drug elimination than body-surface-normalised values. For CKD staging and diagnosis, use CKD-EPI 2021 — it's the current KDIGO standard and the race-free update is real progress on diagnostic equity. MDRD is the older formula being phased out. When the choice would matter clinically (extremes of body size, sharp dose changes), check what your institution's protocol specifies.