eGFR (CKD-EPI)

Standard eGFR

CKD-EPI 2021 Race-Free

Serum Creatinine

Creatinine (mg/dL)

Patient Age (Years)

Biological Sex

CKD-EPI 2021 · Race-free · Accepts mg/dL or µmol/L

Renal Intelligence

Enter serum creatinine (mg/dL or µmol/L) and age to compute filtration rate.

Guidelines & Evidence

Verified

Last Review: 2026-07-17

When to Use

Primary tool for estimating GFR in stable adult patients with steady-state serum creatinine.

Staging Chronic Kidney Disease (CKD) per KDIGO 2024 guidelines and tracking longitudinal progression.

Screening at-risk populations: Type 2 Diabetes, Hypertension, Obesity, Cardiovascular disease, Family history of CKD, Recurrent nephrolithiasis, Autoimmune disease.

Determining eligibility and dosing thresholds for renally-cleared drugs including DOACs, Metformin, SGLT2 inhibitors, and NSAIDs.

Pre-procedural risk assessment prior to iodinated contrast administration or nephrotoxic agent exposure.

Transplant evaluation: donor and recipient workup for living kidney donation.

Monitoring disease trajectory in conjunction with UACR to guide referral and intensification of treatment.

Inclusion Criteria

Adults aged 18 years and older with a stable serum creatinine reflecting steady-state GFR. Both SI units (µmol/L) and conventional units (mg/dL) are fully supported. The calculator auto-converts between units in real time — toggling units will carry over the entered value. Always confirm which unit your laboratory report uses before entering a value.

Absolute Contraindications

Do not apply in Acute Kidney Injury (AKI) — creatinine is still rising and does not reflect true filtration. Do not use in patients on dialysis, post-nephrectomy with single kidney in non-steady state, or in acute tubular necrosis. Do not use for pediatric patients (use Bedside Schwartz or CKiD U25 equation instead).

Relative Limitations

Extremes of muscle mass (sarcopenia, bodybuilders, amputees, malnutrition) distort the creatinine-GFR relationship.

Pregnancy: GFR naturally rises 40–65% by second trimester; CKD-EPI underestimates true GFR.

Liver cirrhosis: reduced hepatic creatine synthesis lowers serum creatinine independently of GFR.

High meat intake acutely raises serum creatinine (cooked meat releases creatine → creatinine), transiently lowering the eGFR estimate.

Medications that competitively inhibit tubular creatinine secretion (Trimethoprim, Cimetidine, Cobicistat) raise serum creatinine without affecting true GFR.

How it Works

The 2021 Race-Free Equation

The 2021 CKD-EPI creatinine equation calculates eGFR using serum creatinine, age, and biological sex only. It eliminates the race coefficient present in the 2009 version after evidence showed race is a social construct with no consistent biological basis for inclusion in clinical equations.

Scoring Formula

eGFR = 142 × min(SCr/κ, 1)ᵅ × max(SCr/κ, 1)⁻¹·²⁰⁰ × 0.9938ᴬᵍᵉ × [1.012 if Female] κ (kappa) = 0.7 for females, 0.9 for males α (alpha) = −0.241 for females, −0.302 for males SCr = serum creatinine in mg/dL (the formula requires mg/dL; 1 mg/dL = 88.42 µmol/L) Unit conversion: SCr (µmol/L) ÷ 88.42 = SCr (mg/dL) Age in years (valid range 18–120 for adults) Result expressed as mL/min/1.73 m² (normalized to standard body surface area of 1.73 m²)

Understanding the Formula Components

The kappa (κ) term normalizes creatinine relative to sex-specific expected muscle mass thresholds.

The alpha (α) exponent governs the slope of the GFR-creatinine relationship below the kappa threshold — steeper for males due to higher average muscle mass.

The max(SCr/κ, 1)⁻¹·²⁰⁰ term captures the steeper decline in GFR at high creatinine concentrations.

The 0.9938ᴬᵍᵉ term reflects the well-established age-related decline in GFR (~1 mL/min/1.73m²/year after age 40).

The female sex multiplier (×1.012) corrects for systematically lower creatinine production in females at equivalent GFR.

Physiological Rationale

Creatinine is generated from creatine phosphate metabolism in skeletal muscle at a relatively constant daily rate. It is freely filtered at the glomerulus and undergoes modest tubular secretion (~10–15% of excreted creatinine). Because it is not significantly reabsorbed, serum creatinine concentration inversely correlates with GFR in steady state — doubling of serum creatinine approximates halving of GFR. However, this relationship is non-linear and confounded by muscle mass, hence the need for age and sex adjustment.

CKD-EPI 2021 vs CKD-EPI 2009

2021 equation performs comparably in accuracy to the 2009 race-inclusive equation when validated across diverse cohorts.

Eliminates systematic overestimation of GFR in Black patients under the 2009 equation, which delayed CKD diagnosis and referral.

Endorsed by NKF-ASN Joint Task Force (2021), KDIGO, and the American College of Physicians for immediate clinical adoption.

Lab systems and EHRs have been directed to transition all GFR reporting to the 2021 equation.

CKD-EPI Creatinine-Cystatin C Equation

A combined CKD-EPI 2021 equation incorporating both serum creatinine and Cystatin C is available and provides superior accuracy across all GFR ranges. Cystatin C is not influenced by muscle mass, making the combined equation preferred in sarcopenic patients, liver disease, and when a confirmatory GFR estimate is needed. Cystatin C alone can also be used when creatinine-based estimates are unreliable.

Clinical Pearls

Conditions Where eGFR Overestimates True GFR

Sarcopenia and cachexia: Low muscle mass → low creatinine production → falsely low serum creatinine → eGFR higher than actual GFR.

Amputees: Reduced total muscle mass leads to systematically lower serum creatinine.

Liver cirrhosis: Reduced hepatic creatine synthesis + fluid dilution lowers serum creatinine independent of kidney function.

Prolonged immobility or muscle wasting (e.g., ICU patients, advanced malignancy).

Strict vegan or low-protein diets: Lower dietary creatine intake reduces creatinine generation.

Conditions Where eGFR Underestimates True GFR

High skeletal muscle mass (e.g., athletes, bodybuilders): Higher creatinine production inflates serum creatinine relative to GFR.

Creatine or creatinine supplementation: Direct dietary creatinine load raises serum levels transiently.

High red meat intake acutely: Cooked meat releases free creatinine during digestion, raising serum creatinine within hours.

Medications inhibiting tubular creatinine secretion: Trimethoprim, Cimetidine, Cobicistat, Dolutegravir — raise serum creatinine without affecting GFR.

Superiority over MDRD

CKD-EPI is significantly more accurate than MDRD at GFR > 60 mL/min/1.73m², where MDRD systematically underestimates GFR.

MDRD was derived in a CKD population; CKD-EPI was developed across a broader range including healthy individuals.

CKD-EPI reduces false-positive CKD diagnoses in the general population — a major concern with routine MDRD reporting.

MDRD should no longer be used as the primary GFR estimation equation per KDIGO 2024.

GFR Decline Rate as a Clinical Signal

Normal age-related GFR decline: ~0.75–1 mL/min/1.73m²/year after age 40.

Decline > 5 mL/min/1.73m² over 1 year or > 10 mL/min/1.73m² over 5 years is clinically significant regardless of absolute GFR.

Rapid decline warrants investigation for superimposed AKI, obstruction, renovascular disease, or medication nephrotoxicity.

A patient with eGFR 55 → 45 over 12 months has a far higher risk trajectory than one stable at eGFR 35 for 5 years.

eGFR and the UACR Interaction

eGFR alone is insufficient for CKD staging. KDIGO mandates the addition of Urine Albumin-to-Creatinine Ratio (UACR) for complete risk stratification. Two patients with identical eGFR 45 can have vastly different prognoses based on whether UACR is < 30 (low risk) vs > 300 mg/g (very high risk, high cardiovascular mortality). Always interpret eGFR within the CGA framework: Cause, GFR category, Albuminuria category.

When to Order Cystatin C

eGFR 45–59 (G3a) without albuminuria or other CKD markers — Cystatin C can reclassify patients and avoid overdiagnosis.

Extremes of body habitus where creatinine-based estimates are unreliable.

Pre-chemotherapy GFR assessment where accurate filtration rate determines drug dosing (e.g., Carboplatin AUC dosing).

Assessment of GFR in potential living kidney donors.

Suspected discordance between clinical picture and creatinine-based eGFR.

eGFR in the ICU — a Special Caution

CKD-EPI should not be used in critically ill patients. Serum creatinine in ICU settings is confounded by: decreased muscle production (catabolism), large volume resuscitation causing dilution, altered tubular secretion, and continuous flux during AKI recovery. Use measured 8- or 24-hour creatinine clearance or Cystatin C trends for drug dosing in this population.

Next Steps

KDIGO GFR Categories & Clinical Action

≥ 90 (G1)	Normal or high. Diagnose CKD only if structural/functional damage markers are present (albuminuria, haematuria, imaging abnormality) for > 3 months.
60–89 (G2)	Mildly decreased. Assess UACR. Optimise BP and glycaemic control. Monitor annually if stable.
45–59 (G3a)	Mild-Moderate decrease. Review all nephrotoxic medications. Check phosphate, PTH, bicarbonate. Screen for anaemia of CKD. Consider Cystatin C for confirmation.
30–44 (G3b)	Moderate-Severe. Intensify RAAS blockade with close potassium monitoring. Restrict dietary sodium and phosphate. Refer Nephrology if progressive. Prepare for KFRE risk scoring.
15–29 (G4)	Severely decreased. Nephrology referral mandatory. Initiate CKD-MBD management (calcium, phosphate, Vitamin D). Discuss renal replacement therapy options. Evaluate for AVF creation.
< 15 (G5)	Kidney Failure. Initiate renal replacement therapy (haemodialysis, peritoneal dialysis) or expedited transplant evaluation. Palliative pathway if appropriate.

Medication Safety by GFR Threshold

Metformin: Do not initiate if eGFR < 45. Discontinue if eGFR falls < 30 due to risk of lactic acidosis from drug accumulation.

SGLT2 Inhibitors: Glycaemic efficacy diminishes below eGFR 45, but cardiorenal protective benefit (Empagliflozin, Dapagliflozin, Canagliflozin) now extends down to eGFR ~20 per landmark trials. Do not initiate if eGFR < 20.

Dabigatran (DOAC): Contraindicated if eGFR < 30. Dose-reduce (110 mg BD) at eGFR 30–50 in elderly or high bleeding risk. Monitor renal function every 3–6 months.

Rivaroxaban: Dose reduce to 15 mg OD (AF indication) if eGFR 15–49. Avoid if eGFR < 15.

Apixaban: More renal-safe than Dabigatran; can be used with caution down to eGFR 25 with dose adjustment using the 2-of-3 criteria.

NSAIDs: Avoid if eGFR < 60 in patients with CKD, heart failure, or volume depletion. Acutely reduce GFR by causing afferent arteriolar constriction via prostaglandin inhibition.

ACE Inhibitors / ARBs: First-line for CKD with proteinuria. An acute eGFR dip of up to 30% upon initiation is acceptable and expected due to reduced glomerular hyperfiltration — do not stop unless > 30% decline or hyperkalemia.

Contrast media (iodinated): eGFR < 30 warrants pre-hydration with IV isotonic saline. Stop nephrotoxins 24–48 hours before. Reassess GFR 48–72 hours post-procedure. Gadolinium MRI contrast: avoid if eGFR < 30 due to risk of Nephrogenic Systemic Fibrosis (NSF).

Aminoglycosides: Adjust dosing interval, not dose. Once-daily dosing preferred. Serial creatinine monitoring mandatory. Avoid if alternatives exist in eGFR < 30.

Allopurinol: Start at 50 mg/day in CKD; titrate slowly. Risk of allopurinol hypersensitivity syndrome is higher in CKD.

Digoxin: Renally cleared. Significant toxicity risk in CKD. Monitor levels closely; consider dose reduction at eGFR < 30.

Low Molecular Weight Heparins (LMWH): Accumulate in renal impairment. Use UFH or anti-Xa monitoring if eGFR < 30 and therapeutic anticoagulation required.

Cardiovascular Risk Integration

CKD is an independent cardiovascular risk factor equivalent to diabetes in many risk models.

Each 10 mL/min/1.73m² decrement in eGFR below 75 is associated with a stepwise increase in all-cause and cardiovascular mortality.

Patients with eGFR < 60 and UACR > 30 should be considered for statin therapy regardless of calculated ASCVD risk, per ACC/AHA guidance.

Anaemia of CKD (target Hb 10–11.5 g/dL with ESA therapy) increases cardiac workload and worsens heart failure — screen from G3b onwards.

CKD-MBD (elevated phosphate, low 1,25-VitD, secondary hyperparathyroidism) begins subclinically in G3a and contributes to vascular calcification.

KFRE — Quantifying Progression Risk

In patients with eGFR < 60, apply the Kidney Failure Risk Equation (KFRE) using eGFR, age, sex, and UACR to estimate 2-year and 5-year probability of kidney failure. A KFRE 5-year risk > 3–5% generally warrants nephrology referral and initiation of renal replacement planning. KFRE outperforms clinical gestalt and is validated across diverse populations including sub-Saharan Africa.

Monitoring Frequency by Stage

G1–G2, A1	Annual eGFR and UACR if stable risk factors present.
G3a–G3b	Every 6 months. Add FBC, electrolytes, bicarbonate, phosphate, PTH, 25-OHD.
G4	Every 3 months. Nephrology co-management. AV fistula assessment.
G5 (not on dialysis)	Every 1–3 months. Active RRT planning.

The Evidence

Foundational Evidence

New Creatinine- and Cystatin C–Based Equations to Estimate GFR without Race.

Inker LA et al. • New England Journal of Medicine. 2021;Development and validation of the 2021 race-free CKD-EPI equations using data from 10 development datasets (n = 10,528) and 18 external validation datasets (n = 13,225). The race-free creatinine equation showed comparable bias and precision to the 2009 race-inclusive equation across all subgroups, with meaningfully improved equity outcomes for Black patients.

A New Equation to Estimate Glomerular Filtration Rate.

Levey AS et al. • Annals of Internal Medicine. 2009;Original CKD-EPI 2009 validation study across 8,254 participants with measured GFR. Demonstrated that CKD-EPI had less bias and greater precision than MDRD, particularly above eGFR 60. This landmark paper established CKD-EPI as the new clinical standard for GFR estimation.

Kidney Disease, Race, and GFR Estimation.

Levey AS et al. • Clinical Journal of the American Society of Nephrology. 2020;Detailed analysis of the scientific and ethical arguments surrounding race in GFR equations. Concluded that race-based adjustment lacks a consistent mechanistic basis and that its removal is both clinically justifiable and necessary for equitable care.

KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease.

KDIGO CKD Work Group. • Kidney International. 2024;The current authoritative global guideline. Endorses CKD-EPI 2021 as the preferred creatinine-based GFR estimation equation. Introduces updated albuminuria categories, updated CKD-MBD targets, and expanded guidance on SGLT2 inhibitor use across GFR categories. Recommends KFRE integration for referral decisions.

A Unifying Approach for GFR Estimation: Recommendations of the NKF-ASN Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Disease.

Delgado C et al. • Journal of the American Society of Nephrology. 2021;Joint NKF-ASN task force report recommending immediate transition to the 2021 CKD-EPI equation across all U.S. clinical laboratories and health systems. Provided practical guidance on implementation, EHR transitions, and communication strategies.

A Predictive Model for Progression of Chronic Kidney Disease to Kidney Failure.

Tangri N et al. • JAMA. 2011;Original derivation and validation of the Kidney Failure Risk Equation (KFRE) using eGFR and UACR as primary inputs. Demonstrated excellent discrimination (C-statistic > 0.90) for 2- and 5-year kidney failure risk in CKD G3–G5 patients.

Origins & History

The CKD-EPI Collaboration

The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) is a research consortium coordinated through the Division of Nephrology at Tufts Medical Center, Boston, led by Dr. Andrew Levey and Dr. Lesley Inker. Founded to address the recognized limitations of the MDRD equation — particularly its systematic underestimation of GFR in healthier individuals — CKD-EPI pooled individual-participant data from multiple research cohorts with reference GFR measurements to build more accurate prediction equations.

Historical Timeline

1999 — MDRD Study equation published. Derived in patients with established CKD. Widely adopted but later shown to underestimate GFR in healthy and mildly impaired populations.

2009 — CKD-EPI creatinine equation published in Annals of Internal Medicine. Race and sex incorporated as covariates based on observed creatinine differences across groups.

2012 — CKD-EPI creatinine-cystatin C combined equation published, offering superior accuracy particularly in the normal-to-mildly reduced GFR range.

2020–2021 — Growing scientific and advocacy consensus that race is a social construct, not a biological one, and that race adjustment in clinical algorithms reinforces systemic disparities by delaying CKD diagnosis in Black patients.

2021 — NKF-ASN Joint Task Force recommends immediate adoption of the new race-free CKD-EPI 2021 equations. New England Journal of Medicine publishes the validation data.

2022 onward — Widespread transition in U.S. and international laboratory systems. KDIGO and major nephrology societies align guidance to CKD-EPI 2021.

The Race Coefficient — What Was Wrong

The 2009 CKD-EPI equation applied a multiplier of 1.159 for patients identified as Black. This was derived from observed differences in mean serum creatinine across racial groups in the development dataset — attributed at the time to differences in muscle mass or tubular secretion. However, this assumption conflated ancestry with race, ignored enormous within-group variation, and meant that a Black patient with identical creatinine, age, and sex to a non-Black patient received a higher eGFR result — leading to systematic underdetection of CKD, delayed referral to nephrology, delayed transplant listing, and inequitable drug dosing decisions. The 2021 equation corrects this by removing race entirely and recalibrating coefficients to minimize aggregate bias.

Global Applicability

CKD-EPI 2021 has been validated in diverse non-Western populations including sub-Saharan African, South Asian, East Asian, and Latin American cohorts. Some variability in performance exists across populations due to differences in mean muscle mass and diet, but it consistently outperforms MDRD. For populations where systematic bias is suspected, the combined creatinine-cystatin C equation is recommended as the more robust alternative.

Last Comprehensive Review: 2026-07-17

In Recent Clinical News

Scanning Medical Journals

No new significant updates or guidelines matching this topic were found today. We will check again soon.

≥ 90 (G1)

Normal or high. Diagnose CKD only if structural/functional damage markers are present (albuminuria, haematuria, imaging abnormality) for > 3 months.

60–89 (G2)

Mildly decreased. Assess UACR. Optimise BP and glycaemic control. Monitor annually if stable.

45–59 (G3a)

Mild-Moderate decrease. Review all nephrotoxic medications. Check phosphate, PTH, bicarbonate. Screen for anaemia of CKD. Consider Cystatin C for confirmation.

30–44 (G3b)

Moderate-Severe. Intensify RAAS blockade with close potassium monitoring. Restrict dietary sodium and phosphate. Refer Nephrology if progressive. Prepare for KFRE risk scoring.

15–29 (G4)

Severely decreased. Nephrology referral mandatory. Initiate CKD-MBD management (calcium, phosphate, Vitamin D). Discuss renal replacement therapy options. Evaluate for AVF creation.

< 15 (G5)

Kidney Failure. Initiate renal replacement therapy (haemodialysis, peritoneal dialysis) or expedited transplant evaluation. Palliative pathway if appropriate.

G1–G2, A1

Annual eGFR and UACR if stable risk factors present.

G3a–G3b

Every 6 months. Add FBC, electrolytes, bicarbonate, phosphate, PTH, 25-OHD.

Every 3 months. Nephrology co-management. AV fistula assessment.

G5 (not on dialysis)

Every 1–3 months. Active RRT planning.