Diesen Beitrag gibt es auch auf Deutsch.

An 83-year-old man with an eGFR of 12, heart failure and diabetes sits in clinic. The question is familiar: create a fistula and prepare for dialysis, or continue conservative care focused on symptoms and quality of life? Until now, this decision rested on observational data. A 2025 Cochrane review found 23 such studies and rated the certainty of evidence as very low throughout. The groups differed too much for a fair comparison.

Prepare for Kidney Care, published in The Lancet, is the first randomised trial on this question.

Who was included

The trial ran in 31 kidney centres across England, Wales, Scotland and Northern Ireland. Patients had an eGFR below 15 ml/min/1.73 m², confirmed by a second value within the previous year. They had to be 80 or older, or aged 65 to 79 with frailty (WHO performance status 3 or higher) or multimorbidity (Davies comorbidity score 2 or higher). Patients judged unfit for dialysis, expected to need dialysis within four weeks, or planning transplantation were excluded.

445 patients were randomised. Median age was 82 years, median eGFR 13. Three quarters were 80 or older, and almost half had diabetes.

Two strategies, no obligation to dialyse

The trial compared two care pathways, not two treatments. One arm prepared for dialysis: information on modalities, choice of modality, access creation, and dialysis start once kidney function and symptoms suggested it. The other arm followed a protocolised conservative pathway called responsive management. It included a comprehensive assessment with advance care planning, usually at home, monthly phone calls, annual home visits and extra appointments as needed. Both arms ended in palliative care according to local routine.

The primary outcome was quality-adjusted life-years (QALYs), combining survival with the EQ-5D-5L, measured every four months. A patient co-applicant shaped this choice, because length and quality of life matter equally to people with kidney failure.

Results

After a mean follow-up of 4.5 years, QALYs were 1.354 with conservative care and 1.472 with dialysis preparation. The adjusted difference was −0.147 (95% CI −0.350 to 0.056, p = 0.16), equivalent to about 78 days at a quality of life of 0.69. The prespecified minimal clinically important difference was 0.345 QALYs. Neither strategy was superior.

Mortality did not differ significantly (adjusted hazard ratio 1.10, 95% CI 0.87 to 1.40). 69% died in the conservative arm and 63% in the dialysis arm. Hospital-free days, place of death, symptom burden, treatment burden and physical function were similar.

EQ-5D-5L scores were slightly lower in the conservative arm (difference −0.039, 95% CI −0.074 to −0.005). The authors consider this unlikely to be clinically important.

Dialysis access complications were less frequent with conservative care: 3.74 versus 11.33 events per 100 patient-years (adjusted rate ratio 0.41, 95% CI 0.19 to 0.88). This outcome was defined as a harm only during peer review and rests on 74 events in 36 patients.

The number to remember

In the dialysis arm, only 60 of 217 patients (28%) started dialysis, at a median of 394 days. In the conservative arm, 42 of 228 (18%) did. Annual mortality was 25 to 30%. In this population, the risk of death far exceeded the risk of needing dialysis.

In practice, many access procedures in this group are never used. Kidney function stabilises, or another illness determines the course.

Limitations

The trial tested superiority. It does not prove equivalence or non-inferiority. It was open-label, and outcomes were not assessed blind. About one in four patients switched pathways (28% and 25%), which dilutes differences between arms. The per-protocol analysis showed a slightly larger difference but supported the main result.

Of 1,261 eligible patients approached, 445 agreed to randomisation. People who accept random allocation might differ from those who decline. 91% of participants were White, more than among new dialysis patients of similar age in the UK. Men had lower QALYs with conservative care than women, but this analysis was not prespecified and lacked power. Part of the trial fell into the COVID-19 pandemic. The results reflect UK care, where conservative kidney management is established.

The trial was funded by the UK National Institute for Health and Care Research. Some authors report travel support, speaker fees or employment with companies in pharma and data analytics, unrelated to this work.

What this means for practice

Dialysis preparation as the automatic default for people over 80 and frail older patients is hard to justify on this evidence. Conservative care is an equal option and deserves to be offered actively, not only when a patient declines dialysis.

1. Raise the conservative option early and explicitly, with concrete numbers instead of general prognoses.
2. Weigh access surgery against the likelihood of ever using it.
3. Revisit decisions regularly. One in four patients changed their mind during the trial.
4. Organise conservative care as active care, with regular contact, symptom control and advance care planning.

I am happy to discuss how these conversations work in your centre on LinkedIn.

Source: Caskey FJ, Worthington J, Liu Y et al., Lane JA. Preparing for responsive management versus renal dialysis in older people living with kidney failure and frailty or multimorbidity in the UK (Prepare for Kidney Care): a two-arm, superiority, parallel-group, open-label, pragmatic, phase 3, individually randomised controlled trial. Lancet, online 4 Nov 2026. doi:10.1016/S0140-6736(26)01818-0. ISRCTN17133653.

Source: DOI: 10.1016/S0140-6736(26)01818-0 · Original article