Estimating glomerular filtration rate (eGFR) is central to the diagnosis, staging, and management of chronic kidney disease (CKD). The three most widely used equations are:
Clinicians often encounter different eGFR values for the same patient depending on the formula applied. Discrepancies can influence:
Understanding the degree of agreementor lack thereofhelps decide which equation is most appropriate for a given clinical scenario.
| Feature | MDRD | CKDEPI | CockcroftGault |
|---|---|---|---|
| Population Used for Development | Patients with CKD, mainly eGFR<60mL/min/1.73m | Broad range of kidney function, healthy volunteers and CKD patients | Pharmacokinetic study of drug clearance (1976) |
| Variables Required | Serum creatinine, age, sex, race (Black vs. nonBlack) | Serum creatinine, age, sex, race (Black vs. nonBlack) | Serum creatinine, age, weight, sex |
| Units of eGFR | mL/min/1.73m | mL/min/1.73m | mL/min (absolute) |
| Performance at High GFR | Underestimates >60mL/min/1.73m | More accurate >60mL/min/1.73m | Variable; depends on weight and body composition |
| Adjustment for Body Size | Standardized to 1.73m | Standardized to 1.73m | Uses actual body weight (or adjusted body weight) |
Large cohort studies (e.g., NHANES, CKDEPI validation) report Pearson correlation coefficients of 0.850.95 between MDRD and CKDEPI, indicating strong linear agreement. The CG equation shows a slightly lower correlation (r0.80) with the two eGFR formulas, largely because it does not standardize for body surface area.
BlandAltman analyses consistently reveal a systematic bias:
These differences are clinically relevant near staging thresholds (e.g., 45, 60mL/min/1.73m).
Age In patients >75years, CG often overestimates GFR because serum creatinine decreases with sarcopenia, while MDRD and CKDEPI remain closer.
Body Mass Index (BMI) In obesity (BMI>30kg/m), CG using actual weight may overestimate GFR; using adjusted body weight improves agreement with CKDEPI.
Race The race coefficient in MDRD and CKDEPI inflates eGFR for selfidentified Black patients. Recent guidelines suggest removing the coefficient, which reduces the discrepancy between MDRD/CKDEPI and CG in this group.
When a patients eGFR hovers around a staging cutoff, using CKDEPI is recommended because of its better performance at higher GFR values. If a decision about nephrology referral depends on a specific threshold, report the value from both MDRD and CKDEPI and note any discordance.
Many drugdose adjustment guidelines are based on CGderived creatinine clearance (CrCl). For medications with narrow therapeutic windows (e.g., aminoglycosides, vancomycin, direct oral anticoagulants), calculate CG in addition to eGFR and, when possible, verify with measured CrCl.
Registries that span many years often have MDRD as the historic standard. Transitioning to CKDEPI improves longitudinal consistency but requires a conversion algorithm or parallel reporting during the transition period.
Emerging biomarkers (cystatinC, trace protein) and combined equations (creatinine+cystatinC) further reduce bias, especially in extremes of muscle mass. Until these become routine, clinicians should remain aware of the systematic differences among MDRD, CKDEPI and CockcroftGault and choose the formula that aligns with the clinical question.
For a quick bedside tool, many electronic health records now display CKDEPI automatically, with an optional CG calculator accessible under medication dosing modules.
**Key Takeaway:** CKDEPI offers the best overall concordance with the other equations while maintaining accuracy at higher GFRs. CockcroftGault remains indispensable for drug dosing, but it should be interpreted alongside CKDEPI, especially in patients with atypical body habitus or advanced age.
For further reading, see:
