Early observations and pilot data that first suggested a new direction
For decades, an observational trend toward earlier dialysis initiation emerged, driven by the assumption that starting dialysis before severe uremic symptoms would improve outcomes. By the early 2000s, average GFR at dialysis initiation in many countries had risen above 10 mL/min. However, observational studies paradoxically showed that earlier initiation was associated with worse outcomes, likely confounded by indication (sicker patients starting earlier). The question of optimal timing demanded a randomized trial.
Landmark RCTs and pivotal trials that established the evidence base
The IDEAL trial was the definitive RCT addressing dialysis timing. Between 2000 and 2008, 828 patients in Australia and New Zealand with CKD stage V were randomized to planned early initiation (eGFR 10-14 mL/min) or late initiation (eGFR 5-7 mL/min). Over a median 3.59-year follow-up, there was no difference in all-cause mortality (37.6% early vs 36.6% late; HR 1.04; 95% CI 0.83-1.30; p=0.75). No differences were observed in cardiovascular events, infections, or dialysis complications. Notably, 76% of patients in the late-start group eventually started dialysis before reaching the planned eGFR threshold due to symptoms.
Follow-up studies, subgroup analyses, and real-world validation
Post-IDEAL analyses and subsequent observational studies reinforced that symptom-guided initiation was safe and preferred. The trend toward earlier initiation reversed in many countries following publication. Registry data from Canada, Europe, and the United States showed a shift toward lower eGFR at dialysis initiation after 2010. Health economic analyses confirmed that avoiding premature dialysis reduced costs substantially without compromising outcomes, supporting at least 6-12 months of avoided dialysis per patient.
Integration into clinical practice guidelines and recommendations
KDIGO and national guidelines shifted from GFR-based initiation thresholds to symptom-guided initiation. The recommendation is to prepare for dialysis when eGFR falls below 15 mL/min but initiate based on clinical assessment including uremic symptoms, fluid overload, nutritional status, and quality of life rather than a specific GFR number.
KDIGO
Initiate dialysis based on clinical assessment of symptoms, nutritional status, and quality of life rather than a specific eGFR threshold
ERBP
Symptom-guided dialysis initiation; preparation for dialysis when eGFR <15 mL/min
Now
Current standard of care and ongoing research directions
The IDEAL trial settled the early-versus-late dialysis debate in favor of symptom-guided initiation, reducing unnecessary early dialysis exposure. The focus has shifted to optimizing pre-dialysis care, timely vascular access creation, and integrating conservative management as a legitimate option for selected patients. With the expansion of CKD-slowing therapies (SGLT2 inhibitors, finerenone, GLP-1 RAs), more patients may delay or avoid dialysis entirely, making the timing question increasingly relevant to a smaller population.
The IDEAL trial (828 patients, NEJM 2010) showed no survival benefit of planned early dialysis initiation (eGFR 10-14 mL/min) versus late initiation (eGFR 5-7 mL/min), with mortality of 37.6% vs 36.6% (HR 1.04; p=0.75). Guidelines now recommend symptom-guided initiation rather than GFR-based thresholds.
When is dialysis preparation recommended?+
KDIGO recommends preparation for dialysis (education, modality selection, access creation) when eGFR falls below 15 mL/min, but actual initiation is guided by clinical assessment of uremic symptoms, volume overload, nutritional decline, and quality of life. A specific GFR threshold is not used as an indication to start.
