Early Referral Matters: Why Nephrology Should See Kidney Disease Sooner

This article is written for clinicians and health professionals.

Late referral to nephrology is one of the few genuinely preventable harms in kidney medicine. It is associated with catheter starts instead of fistula or peritoneal dialysis starts, with unplanned "crash landing" initiation, with no time to discuss conservative care properly, and with no realistic chance of pre-emptive transplantation. The patient arrives at kidney failure with the options already closed. In the Indian Ocean, where the fallbacks that soften late referral elsewhere are largely absent, the cost is higher still.

KDIGO 2024 moved referral from stage to risk

The most important change in the 2024 KDIGO CKD guideline is the shift from staging by eGFR to risk-based decisions. Two patients can both be labelled "stage 3" and sit in completely different risk categories depending on albuminuria, which is why KDIGO classifies by Cause, GFR category and Albuminuria category together, and why it recommends an externally validated risk equation, in practice the Kidney Failure Risk Equation, to estimate absolute risk of kidney failure over 2 to 5 years in CKD G3 to G5.

Referral thresholds (KDIGO, 2024):

• A 5-year risk of kidney failure of about 3 to 5% on a validated risk tool, or
• eGFR below 30, or
• A sustained fall of 20 to 30% or more in GFR,
• Plus the classic triggers: A3 albuminuria, rapidly progressive CKD, refractory hypertension, recurrent stones, hereditary kidney disease, persistent haematuria.

Practical corollary: if you are checking eGFR without an albumin-to-creatinine ratio, you are not staging kidney disease, you are guessing at it.

The preparation threshold is a different decision from the starting threshold

This distinction is routinely collapsed in practice, and collapsing it is how patients end up with a tunnelled line.

When to prepare (KDIGO, 2024): consider planning for pre-emptive transplantation and/or dialysis access when GFR is below 15 to 20, or when the 2-year risk of kidney replacement therapy exceeds 40%. NICE NG107 (2018) puts it in time rather than numbers: start assessment for renal replacement therapy or conservative management at least one year before therapy is likely to be needed.

When to start, by contrast, is a symptom decision, not a number. The IDEAL trial randomised 828 adults to an early or a late start; the groups converged in practice, to a mean eGFR of about 12.0 versus 9.8, and there was no significant difference in mortality or in any clinically important secondary outcome (New England Journal of Medicine, 2010). Tellingly, about three quarters of the "late" arm started before reaching the planned window because they became symptomatic. KDIGO frames initiation as a composite assessment, typically in the eGFR range of 5 to 10; NICE advises considering dialysis when uraemic symptoms affect daily living, or at an eGFR of around 5 to 7 if asymptomatic. Both are guidance, not a rule, and the regional difference between them is worth knowing.

So: do not use eGFR to decide when to start. Do use risk to decide when to prepare. The two are separate decisions and only one of them can be safely deferred.

Preparation takes the year that NICE asks for

A fistula matures over months, a graft over weeks, and fistulas fail to mature reasonably often. The old "Rule of 6s" from KDOQI 2006 (by 6 weeks: flow of at least 600 mL/min, diameter at least 6 mm, depth 6 mm or less, a usable straight segment of at least 6 cm) is a guide, not a verdict: a 2022 Journal of Vascular Surgery analysis found many fistulas that fail the rules still support dialysis, and that flow volume and vein depth predict maturation about as well as all three criteria.

That fallibility is why the 2019 KDOQI Vascular Access guideline abandoned a rigid hierarchy in favour of the ESKD Life-Plan and its ViP ACCeSS framework: vessel preservation, access creation, contingency plan, succession plan. The principle is "the right access, in the right patient, at the right time", not a fistula for everyone. A fistula is the wrong answer for someone with a short life expectancy, poor vessels, or a high probability of failed maturation followed by months of catheter dependence anyway.

Two practical points follow from a year of lead time. Vessel preservation starts before the referral: protect the forearm veins of anyone with advanced CKD. And peritoneal dialysis needs planning too: NIDDK advises placing the catheter at least three weeks before the first exchange, since it works better with 10 to 20 days to heal.

Catheters remain the right answer for someone who needs dialysis this week, or who has exhausted the alternatives, or who is on a palliative trajectory. They are last choice, not no choice. The KDOQI 2019 facility goal is a catheter-related bloodstream infection rate below 1.5 per 1,000 catheter days.

Transplant is the conversation that has to happen earliest

Wolfe et al. (New England Journal of Medicine, 1999) reported annual death rates of 16.1 per 100 patient-years for all dialysis patients, 6.3 for those on the waiting list, and 3.8 for transplant recipients. The honest comparison is transplant versus waiting list, because the waiting-list population is fitter to begin with. The bet pays, but the payment comes first: relative risk of death in the first 2 weeks after transplantation was 2.8 times that of matched waiting-list patients, while long-term mortality was 48 to 82% lower.

Time on dialysis before transplantation is itself an adverse prognostic factor in registry data, and pre-emptive transplantation requires early referral, because the work-up takes months and finding a living donor takes longer. NICE NG107 recommends offering pre-emptive living donor transplantation, or pre-emptive listing, to eligible people. And keep transfusion policy restrictive in transplant candidates, to minimise allosensitisation (KDIGO Anemia, 2026).

Conservative care has to be offered, not merely implied

NICE NG107 requires that conservative management be offered as a genuine choice alongside kidney replacement therapy, and that conversation is impossible in an emergency. Present it honestly: the meta-analysis of 22 cohort studies (21,344 patients, Nephrology Dialysis Transplantation, 2022) found an adjusted hazard ratio for death of 0.47 favouring dialysis, but the conservative-care cohorts were a median of 7.0 years older with more comorbidity, and the authors concluded that the results "cannot be translated to an individual level". In those aged 80 and over the advantage was substantially reduced, and quality of life showed no distinct advantage either way. For a fit patient, dialysis very probably buys years. For a frail, very elderly patient it may buy little time and will certainly consume time. Revisit that decision, do not make it once.

Why this is sharper in the Indian Ocean

Seychelles reported 230 people on haemodialysis in 2023, in a population of roughly 100,000, and 3.6% of adults aged 18 to 74 had moderate or severe reduction in kidney function or kidney failure, rising to 9% of both women and men aged 55 to 74 (Ministry of Health National NCD Survey, 2023). The causal chain is not a detection failure: among adults aged 45 to 74 with hypertension or diabetes, 70 to 90% were aware and over 80% of those were on medication, but only about 40% of treated hypertension and about 30% of treated diabetes were controlled to target. Kidneys pay for that over decades. Polycystic kidney disease adds a locally distinctive burden, estimated at 57 cases per 100,000 (Yersin, Bovet et al., 1997).

The regional context changes what "prepare early" means. Mauritius had 1,578 patients on haemodialysis at the end of 2020 and one on peritoneal dialysis, with no transplant service in the country (Kidney360, 2021). The Maldives has no peritoneal dialysis programme at all, and transplantation is done abroad (Kidney360, 2023). Across the Indian Ocean islands, haemodialysis dominates, peritoneal dialysis is rare or absent, and domestic transplantation is largely unavailable. Where transplant means flying out and PD may not exist, the vascular access plan is not one option among several. It is the plan, and that makes the referral you make this month more consequential, not less.

What to do differently on Monday

• Order an albumin-to-creatinine ratio with the eGFR. Stage by CGA, not by eGFR alone.
• Run a validated risk equation in CKD G3 to G5, and refer at a 5-year kidney failure risk of about 3 to 5%.
• Start the modality and conservative-care conversation a year before therapy is likely to be needed, not when symptoms arrive.
• Preserve forearm veins in advanced CKD, and say so in the notes.
• Ask the transplant question early, and ask it out loud.
• Document a life-plan with a contingency, because fistulas fail to mature and people change their minds.

Sources

• KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. kdigo.org
• NICE NG107, Renal replacement therapy and conservative management, 2018. nice.org.uk
• KDOQI Clinical Practice Guideline for Vascular Access: 2019 Update, American Journal of Kidney Diseases.
• Cooper BA et al. (IDEAL trial). A randomized, controlled trial of early versus late initiation of dialysis. New England Journal of Medicine, 2010.
• Wolfe RA et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. New England Journal of Medicine, 1999.
• Rules of 6 criteria predict dialysis fistula maturation. Journal of Vascular Surgery, 2022.
• Survival of patients who opt for dialysis versus conservative care: a systematic review and meta-analysis. Nephrology Dialysis Transplantation, 2022.
• Seychelles National Survey of Noncommunicable Diseases 2023, Ministry of Health, published April 2024. health.gov.sc
• Global Dialysis Perspective: Mauritius. Kidney360, 2021. Global Dialysis Perspective: Maldives. Kidney360, 2023.

---

This article is general information, not medical advice. It does not diagnose, it gives no doses, and it must never be used to start, stop or change a treatment. Your dialysis prescription, your fluid allowance, your diet and your medicines depend on your own blood results, your own remaining kidney function and your own other conditions, and those decisions belong with you and your kidney care team. If you are unwell, contact your dialysis unit. In an emergency, call your local emergency number.