March brought World Kidney Day and UK Kidney Week. But the kidney stories landing in primary care this month go beyond awareness campaigns.

The Kidney Failure Risk Equation: Who should we actually refer?

Before you dive in, if you want the one pager infographic to download and share with colleagues you can DOWNLOAD it below.

NICE added the kidney failure risk equation (KFRE) to CKD guidance in 2021. It helps find patients most at risk of needing renal replacement therapy. To find the five-year risk score, just use age, sex, eGFR, and urine albumin-creatinine ratio. Then, refer anyone with a score above 5% to nephrology.¹

A new study published in the BJGP this week suggests the picture is more complicated than that.²

Stewart and his team looked at CKD stages 3 to 5 patients. They used the Greater Manchester Care Record from 2018 to 2023. This is a large primary care dataset covering a diverse urban population. They calculated KFRE scores. They tracked the outcomes. Did patients start dialysis, receive a transplant, or die first?

The findings are striking.

For patients with a KFRE score below 5% (the group NICE says does not need referral), the probability of needing renal replacement therapy was just 0.1%. The probability of death was 11.9%. For those scoring 5 to 20%, the probability of renal replacement therapy was 2%, but the probability of death rose to 23%. Even in the highest risk group, with KFRE scores above 20%, death (29%) was more likely than renal replacement therapy (14%).

In other words, for most patients flagged by the KFRE, the competing risk of dying from something else outweighs the risk of ever reaching dialysis.

That does not mean the referral is wrong. Nephrology input can support conservative management, medication optimisation, and advance care planning. But it does change the nature of the conversation.

The most important finding was the equity dimension. Black and Asian patients with KFRE scores over 20% were more likely to need renal replacement therapy than to die. This is the opposite pattern to the overall cohort. Current NICE guidance doesn’t address ethnicity in referral decisions. It also doesn’t mention the risk of death - The authors argue it should.

Around 11% of patients with CKD stages 3 to 5 in this cohort had a KFRE score above 5%, which gives a rough sense of the workload implications. It also raises a practical question: if someone does not measure a uACR, they cannot calculate the KFRE. Including uACR in every CKD review is a prerequisite for making this tool useful at all.

What could this mean for primary care?

The KFRE is a useful tool, but it tells you about kidney failure risk in isolation. It doesn't consider that many CKD patients, especially elderly with other health issues, are much more likely to die than to need dialysis.

For these patients, a nephrology referral might focus on conservative management and advance care planning instead of preparing for renal replacement. For younger patients and those from Black and Asian backgrounds, the KFRE score is more important. This is because they have a lower risk of dying from other causes, but a higher chance of reaching kidney failure.

References:
1. NICE. Chronic kidney disease: assessment and management [NG203]. London: NICE; 2021 (updated 2023).
2. Stewart S, Kalra PA, Kontopantelis E, Blakeman T, Tilston G, Sinha S. Use of the kidney failure risk equation: a regional retrospective primary care cohort study in England. Br J Gen Pract. Published online 24 March 2026. doi:10.3399/BJGP.2025.0490

When a patient stops a medication, the assumption is usually that the benefits stop too. For most drugs, that is exactly what happens. The EMPA-KIDNEY post-trial follow-up, published in NEJM, checked if this idea is true for SGLT2 inhibitors in CKD.¹

In the original EMPA-KIDNEY trial, 6,609 patients with CKD at risk of progression were randomised to an SGLT2 inhibitor or placebo and followed for a median of two years. The trial showed a 28% reduction in kidney disease progression or cardiovascular death.²

So, what happens next?

In the post-trial phase, 4,891 patients (74% of the original cohort) were observed for a further two years. No trial drug was provided. Blinding was maintained. Local practitioners could prescribe open-label SGLT2 inhibitors if they chose to, and roughly equal proportions in both groups did (43% in the original treatment group, 40% in the placebo group).

Over the combined four years (two years on treatment and two years off), 26.2% of the treatment group had a primary outcome event. In contrast, 30.3% of the placebo group experienced the same event. The hazard ratio was 0.79, with a 95% confidence interval of 0.72 to 0.87. Kidney disease progression was lower (23.5% vs 27.1%), end-stage kidney disease was lower (9.0% vs 11.3%), and cardiovascular death was lower (3.8% vs 4.9%).

The American College of Cardiology described this as a “legacy effect.”

Most of the residual benefit is seen in the first 12 months after stopping the trial drug. After that, the curves begin to converge.

This aligns with the idea that the structural kidney benefits of SGLT2 inhibition last for a time but do not persist indefinitely once the drug is removed.3

A few important caveats. The trial included patients with advanced CKD. Their average eGFR was 37 mL/min/1.73m². They also had significant albuminuria, with a median uACR of 317 mg/g.

These are patients under nephrology care, not the broader primary care CKD population. The post-trial follow-up was observational. It did not control for open-label SGLT2 inhibitor use. However, usage was similar in both groups. The findings show what occurs after a set treatment time in a specific group. They don't say if shorter or longer treatment times would have the same carry-over. They also don’t mention if the same results would appear in earlier-stage CKD.

What could this mean for primary care?

This is a secondary care trial, but the principle it demonstrates is relevant to any clinician managing CKD. For patients with advanced CKD and high albumin levels, SGLT2 inhibitors helped cardiorenal health. These benefits lasted even after stopping the medication. Whether this translates to earlier-stage disease managed in primary care remains an open question. The data does not answer it, but it adds to the growing evidence base around these agents in kidney disease.

References

The NHS App is becoming the front door of the health service. The Government’s 10-year health plan, released in July 2025, aims for patients to access their health records, test results, and referral information. That sounds like progress. In many ways it is. But a report from Kidney Care UK suggests the reality is running ahead of the safeguards.¹

The report, titled Falling through the GApp, surveyed more than 1,300 people with kidney disease across the UK in October 2025. The headline finding: 11% of respondents first discovered they had chronic kidney disease by seeing it recorded in the NHS App. Not from a GP. Not from a letter. From a screen.

Almost 40% of those surveyed did not have the opportunity to discuss their diagnosis with a healthcare professional at all. Only a third (36%) found out they had CKD at the same time the diagnosis was first recorded. For the remainder, there was a gap, sometimes a long one. Kidney Care UK says about 10% of calls to its support line are from people who recently discovered a CKD diagnosis in their records. They had no prior discussion with their doctor.

NHS guidelines say patients shouldn't get serious diagnoses online. They need proper support and context. But the mechanism is passive: once a CKD code is added to a patient’s record, it becomes visible in the App. There is no flag, no accompanying explanation, no prompted follow-up. A patient looking at their blood results after a routine test might see “chronic kidney disease stage 3” for the first time. They may feel confused and alone, with no clinical context.

The downstream effect is predictable. Sixty-eight percent of people used the internet and social media to learn about their diagnosis. Kidney Care UK's report showed that 65% of people with diabetes and hypertension didn’t know about their kidney risk before getting diagnosed with CKD. This happened even though NICE guidelines suggest screening for these groups.

This is not an argument against digital access. It is an argument for getting the sequencing right. Transparent records are valuable. But transparency without a conversation is not the same as good care.

The conversation about the diagnosis needs to happen before or at the time of coding, not afterwards. Consider checking how your practice handles communication about blood results, diagnostic coding, and patient notifications. We should also think about whether patients in high-risk groups, like those with diabetes or hypertension, are informed about their kidney risk before a CKD code shows up.

References