CKM Staging vs KFRE - Nephrology Lectures 2025

The single most common misuse of CKM staging is reading it as a kidney prognosis. It is not one. CKM staging is a cardiovascular risk framework in which the kidney appears as a risk axis. The Kidney Failure Risk Equation answers a different question, and the two disagree systematically — by design, not by defect.

⚖️ Different questions

	CKM grid (KDIGO heat map)	KFRE
Endpoint	Cardiovascular events; CKD enters as a risk equivalent	Kidney failure (kidney replacement therapy) at 2 and 5 years
Inputs	eGFR, UACR	eGFR, UACR, age, sex (4-variable); plus calcium, phosphate, bicarbonate, albumin (8-variable)
Output	Four ordinal categories	Continuous absolute risk
Validated range	All eGFR	G3–G5 only; not validated above eGFR 60

KDIGO action thresholds for KFRE

Risk	Action
5-year risk 3–5%	Nephrology referral
2-year risk above 10%	Multidisciplinary care
2-year risk above 40%	Modality education, access planning, transplant preparation

🔄 Three ways they disagree

CKM overcalls — G4 A1

eGFR 25 with a UACR of 12 is CKM stage 3 regardless of patient characteristics. An 80-year-old woman in that cell often carries a 5-year KFRE in the low single digits. She will die with her kidneys, not from them. The grid is correct about her cardiovascular risk and actively misleading if read as progression risk.

KFRE overcalls — G3a/G3b A3, younger man

Identical stage 3 label, but a 2-year KFRE that can exceed 40% — access-planning territory. Same stage, opposite management.

Shared blind spot — G1/G2 A3

eGFR 95 with a UACR of 1,200 mg/g is CKM stage 2, sitting alongside uncomplicated hypertension. KFRE will not run at all above eGFR 60. Neither tool captures the steepest slope in the clinic — only the UACR value itself does.

🧩 Why the divergence is structural

                Age enters the two models with opposite sign. In KFRE, older age
                lowers predicted kidney failure risk, through competing mortality — the patient dies
                before the kidney fails. In CKM staging, age is deliberately excluded, because it is
                carried separately by PREVENT. Any given G × A cell therefore fans out across a wide
                KFRE range by age and sex while the CKM stage stays fixed. This is a design consequence, not a flaw
                in either instrument — but it means the disagreement is systematic and predictable rather
                than random.
            

🎯 Practical reconciliation — run all three

Instrument	Answers	Drives
CKM stage	How much cardiovascular risk does this patient carry, and from which axis?	Treatment intensity and agent class
KFRE	How likely is this kidney to fail, and by when?	Referral, multidisciplinary care, access placement, transplant workup
PREVENT	What is the absolute 10- and 30-year cardiovascular risk?	The pharmacotherapy thresholds (7.5%, 20%, HF 5%)

UACR is the shared lever. It moves a patient from CKM stage 2 to stage 3, and it dominates KFRE within any eGFR band. That convergence is the quantitative case for the UACR thresholds that drive therapy — 30 mg/g for finerenone, 100 mg/g for semaglutide. One inexpensive urine test moves the stage, the kidney prognosis, and the drug choice simultaneously.

The triage heuristic. Run all three in stages 2–3. A stage 3 patient with a 5-year KFRE of 2% and a PREVENT-CVD of 28% is a cardiology problem. The reverse belongs to nephrology. The stage alone will not tell you which patient you have.

🔗 Where to go next

🧬 Back to staging and the kidney axis 🧪 The HbA1c problem in CKD

🔍 Source and verification

Primary source: the 2026 AHA/ACC/ADA/ASN guideline for the prevention, detection, evaluation, and management of cardiovascular-kidney-metabolic syndrome (Circulation. 2026;154(4):e50–e158, PMID 42263157), published Free Access and read in full for these pages. Glycemic triggers are from the ADA Standards of Care in Diabetes — 2026, Section 9 (PMID 41358900), read in full via PubMed Central. Kidney staging and glycemic targets in CKD are from KDIGO 2024 and KDIGO 2022 (PMID 36272764), the latter verified against its peer-reviewed synopsis (PMID 36623286).

Every numeric threshold on this page was checked against the source text. Figures that could not be confirmed in a primary source were removed rather than published with a caveat — including a set of pre-HF biomarker cut points that circulate in secondary summaries but appear nowhere in the guideline.

⚖️ CKM Staging vs KFRE