Core Revision
Revision aid, not clinical advice. This page is educational material written for exam preparation. It is not a substitute for the current guideline or for clinical judgement, and must not be relied on for the care of any individual. Always check up-to-date national guidance and verify independently before acting on anything here.
CKD has changed since we qualified — here’s what, why, and how to actually remember it
Chronic kidney disease used to be five numbers on a lab report. It’s now a grid, a risk equation, and a drug class that didn’t exist a few years ago. Understanding why it changed is the fastest way to remember it.
If you qualified more than a few years ago, the CKD you were taught is not quite the CKD tested now. The stages are the same five numbers, but they’re no longer the whole story — and the two biggest changes (the risk grid, and a whole new drug class) both happened for good clinical reasons. This piece is the history behind the numbers: not so you can recite it, but so the numbers stop being arbitrary and start being obvious.
Core Lens
What changed in CKD, in five points
- Two axes, not one. KDIGO 2012 added albuminuria, graded A1 to A3 by urine ACR, to the eGFR-based G stages, creating the grid now used for staging.
- NG203 consolidated. NICE NG203, published in August 2021, replaced CG182 and absorbed the separate anaemia (NG8) and hyperphosphataemia (CG157) guidelines.
- SGLT2 inhibitors. NICE now recommends an SGLT2 inhibitor for CKD in defined circumstances; the criteria have changed since 2022, so check the current technology appraisal rather than older notes.
- Referral by risk equation. NG203 (recommendation 1.5.5) uses a 5-year risk of needing kidney replacement therapy greater than 5%, on the 4-variable Kidney Failure Risk Equation (KFRE), as its referral threshold.
- For the exam. The pattern is that the changes moved from single cut-offs towards estimating the individual patient’s own risk.
Revision aid, not clinical advice — checked against NICE NG203 and KDIGO 2012 on 28 September 2026.
You are reading the Core Lens — the points that matter. Open the Deep Dive for the full article.
Sitting the AKT? Try twenty questions free.
Twenty original AKT questions in the exam's 80/10/10 split: sixteen clinical, two on evidence-based practice and two on organisation and management, with the reasoning after every one and a marked report. Free, no card and the first five are open to all.
Memory research is clear that facts anchored to a reason stick better than facts memorised in isolation. “ACR ≥22.6 triggers an SGLT2 inhibitor” is a coin-flip to recall under exam pressure. “SGLT2 inhibitors got added because a major trial showed they protect the kidney independently of blood sugar, and NICE fast-tracked them once albuminuria was high enough to matter” is a story — and stories outlast facts.
The old model: five stages, one number
For most of the 2000s, CKD was staged the way you probably first learned it: Stage 1 to Stage 5, based purely on eGFR. It was a genuine advance at the time — the 2002 K/DOQI guideline was the first to standardise “chronic kidney disease” as a diagnosis at all, and it’s why labs started reporting an eGFR alongside creatinine instead of creatinine alone.
But a single axis has a problem: an eGFR of 55 means very different things in different patients. Some people sit there for decades with a healthy kidney that’s simply smaller or older. Others are on a fast slide toward kidney failure. Same number, opposite prognosis — and staging by eGFR alone couldn’t tell them apart. It also over-diagnosed CKD in a lot of people (particularly older adults) whose kidneys were normal for their age.
2012: KDIGO adds a second axis — and the grid is born
The fix, published by KDIGO in 2012, was to stop asking one question (“how well does the kidney filter?”) and start asking two. Filtration still gives you the eGFR-based G stage — and G3 was split into G3a and G3b because that band was doing a lot of prognostic work and needed finer resolution. But alongside it, a second measurement was added: albuminuria, via the urine albumin:creatinine ratio (ACR), graded A1 to A3.
Albuminuria matters because it’s a direct marker of damage to the filtering membrane itself — not just a proxy for how much blood the kidney can clear, but evidence that the barrier is leaking. Two people with identical eGFR but different ACR are, it turns out, on very different trajectories: the one with heavier proteinuria has a substantially higher risk of both progressing to kidney failure and having a cardiovascular event. Cross the two axes and you get the “heat map” — the coloured grid this tool is built around — and it predicts outcomes far better than either number alone.
This is exactly why “G1 or G2 with a normal ACR (A1) is not CKD on its own” keeps appearing as a trap. It isn’t a technicality — it’s the whole point of the 2012 change. Reduced filtration with no damage marker isn’t disease; it’s often just age. The grid was built specifically to stop over-labelling those patients.
What NICE actually changed, and when
The UK guideline caught up to KDIGO’s model in NICE CG182 (2014), which adopted the G/A grid for primary care. Then in August 2021, NICE went further: it retired CG182 and folded it — along with the separate anaemia (NG8) and hyperphosphataemia (CG157) guidelines — into the single guideline this tool is built on, NG203, with a fresh evidence review behind the headline recommendations.
The single biggest change in years: SGLT2 inhibitors for the kidney itself
If you qualified before 2021, this is the one that will genuinely surprise you. SGLT2 inhibitors (dapagliflozin, empagliflozin and colleagues) were built and licensed as diabetes drugs. Then large outcome trials — most notably DAPA-CKD — showed something nobody had fully expected: they slow the progression of kidney disease and cut cardiovascular and heart-failure risk even in people without diabetes, simply because someone has significant albuminuria.
NICE responded quickly. NICE technology appraisal TA775 (March 2022) first recommended dapagliflozin for CKD in its own right; it was replaced by TA1075 in July 2025. TA1075 recommends dapagliflozin as an add-on to optimised ACE inhibitor or ARB treatment when eGFR is 20 to below 45, or when eGFR is 45–90 with a uACR of 22.6 mg/mmol or more (within A2) or with type 2 diabetes. That’s a genuinely new pillar of CKD management that simply did not exist in the guidance most of us first learned — which is exactly why it’s such a popular place for exam-writers to test whether your knowledge is current.
Referral changed too: from a cut-off to an equation
The old rule of thumb was blunt and memorable: eGFR under 30, refer to nephrology. It was easy to teach and easy to remember — and also wrong often enough to matter, because a stable eGFR of 28 in an 84-year-old is a very different referral decision to the same number in a 40-year-old with rising creatinine.
Current guidance replaces the cut-off with the Kidney Failure Risk Equation (KFRE): a validated, 4-variable calculation (age, sex, eGFR, ACR) that outputs an actual probability — refer if the 5-year risk of needing kidney replacement therapy is greater than 5%. The UK Kidney Association points to the UK-validated calculator — worth bookmarking, since the generic North American versions that turn up in a plain search are not calibrated the same way. It’s a small conceptual shift with a big practical one: referral became about an individual’s trajectory, not a population threshold.
Every one of these changes follows the same pattern: a single blunt number got replaced by something that predicts the individual’s outcome better. eGFR alone → eGFR + ACR. A flat referral cut-off → a personalised risk score. A diabetes-only drug → a kidney-protective drug used on its own merits. Once you see that pattern, you don’t need to memorise five unrelated facts — you need to remember one idea, applied four times.
Try reasoning it out, not looking it up
Reading this history is useful once. Actually placing patients on the grid, committing to a risk colour, and building the NG203 plan from scratch is what makes it stick for the exam — and for clinic. That’s the whole idea behind the CKD Reasoner: instead of handing you the answer, it makes you find the square, then checks your reasoning against the guideline, square by square, until the pattern above stops being something you read about and starts being something you just know.
Reason your way to the KDIGO square
Free to use — place a patient on the grid, commit a risk level, then build the NG203 plan yourself. The tool remembers the squares you find hard and brings them back until they stick.
Open the CKD Reasoner →It’s built for trainees revising for the AKT and MSRA, but the reasoning underneath — the grid, the risk equation, the SGLT2 threshold — is exactly what changed in real clinical practice too. Qualified GPs use it the same way: a two-minute refresher on what the current guideline says before a CKD review — and then checking the guideline itself. It is a revision aid, not a clinical tool.
This article summarises NICE NG203 and the KDIGO 2012/2024 CKD classification for revision purposes. Guidance is updated periodically — always check the current guideline before managing a real patient.
