What Percentage Kidney Function Before Dialysis?

Most people are told dialysis begins when kidney function drops to about 10 to 15 percent, but modern evidence and clinical guidelines tell a more nuanced story. There is no single percentage that triggers dialysis for everyone. The decision depends heavily on whether you have symptoms, how quickly your kidneys are declining, and your overall health. In many cases, asymptomatic patients can safely delay dialysis until kidney function falls well below 10 percent, sometimes to around 5 to 7 percent, with careful monitoring.

Why There Is No Universal Cutoff

For years, the assumption in nephrology was that starting dialysis earlier, at a higher level of remaining kidney function, would give patients a head start and better outcomes. That assumption was tested head-on in a landmark randomized trial known as IDEAL (Initiating Dialysis Early and Late), which enrolled over 800 patients in Australia and New Zealand. The early-start group began dialysis when their estimated kidney function (measured as eGFR) was between roughly 10 and 14, while the late-start group waited until it dropped to 5 to 7. After more than three and a half years of follow-up, about 38 percent of the early group and 37 percent of the late group had died. The difference was statistically meaningless.¹ Rates of cardiovascular events, infections, and dialysis complications were also similar between the two groups.

A meta-analysis pooling multiple studies went further, finding that starting dialysis at a higher eGFR was actually associated with a roughly one-third higher risk of death compared to starting later.² That finding likely reflects confounding: sicker patients with more complications tend to get referred to dialysis sooner, and their poor outcomes get attributed to the timing rather than their underlying disease burden. But even after adjustments, no clear benefit of early initiation emerged.

A large nationwide cohort study of over 10,000 patients with advanced kidney disease found a modestly lower mortality risk when dialysis was started at an eGFR of 15 to 16 compared to 6 to 7, but the absolute difference was small, about a five percent lower five-year mortality risk, corresponding to a mean postponement of death by roughly 1.6 months. The trade-off was that dialysis would need to be started about four years earlier to achieve that modest gain.³ When the researchers mimicked the design of the IDEAL trial, the results were consistent with what IDEAL had found: no meaningful survival advantage to starting earlier.

What Current Guidelines Actually Recommend

Given this evidence, clinical practice guidelines have shifted away from eGFR-based thresholds and toward symptom-driven decision-making. A systematic review of high-quality guidelines found that all of them agreed on one principle: start dialysis when a patient develops symptoms or signs related to kidney failure.⁴ Those symptoms include persistent nausea, uncontrollable fluid overload, worsening fatigue, confusion, dangerous electrolyte imbalances, and protein-energy wasting that does not respond to dietary interventions.

For patients who remain asymptomatic, only two of the six guidelines reviewed recommended a specific eGFR threshold, placing it at 5 to 7. That number represents roughly 5 to 7 percent of normal kidney function. The strength of even those recommendations varied, reflecting genuine uncertainty about whether any particular number should be a hard trigger. Two guidelines also emphasized that the decision should be made jointly between the patient and the care team after careful discussion.⁴

The Australian guidelines are particularly explicit about this: in asymptomatic patients with stage 5 kidney disease, dialysis can be safely delayed until eGFR is at least as low as 5 to 7, provided there is careful clinical follow-up and the patient understands their condition.⁵ The same source notes that creatinine-based eGFR estimates become unreliable in patients with end-stage kidney disease, which is another reason the decision should not hinge solely on a number.

Why the eGFR Number Can Mislead You

Most eGFR calculations are based on blood creatinine levels. Creatinine is a waste product of muscle metabolism, so anything that changes your muscle mass or nutritional status changes how much creatinine your body produces, which distorts the estimate. As kidney disease progresses, patients often lose muscle, eat less, and become less physically active. All of these reduce creatinine production, which can make kidney function look stable or even slightly improved when it is actually still declining.⁶

Cystatin C is an alternative blood marker that is not affected by muscle mass in the same way. When the two estimates diverge, with the creatinine-based number reading higher than the cystatin C-based number, it tends to signal declining overall health and a higher risk of poor outcomes. This is one reason nephrologists treat eGFR as a guide rather than a gospel: the number on your lab report may not accurately reflect how much filtering your kidneys are actually doing, especially as you approach end-stage disease.

Preparing for Dialysis Well Before You Start

Even though the actual start of dialysis can often be delayed, preparation needs to begin much earlier. If hemodialysis is the planned modality, you need a functioning vascular access, ideally an arteriovenous fistula created from your own blood vessels. Guidelines recommend placing that access when eGFR drops below 20, which corresponds to roughly 20 percent kidney function, and ideally at least six months before dialysis is expected to begin.⁷ A fistula needs time to mature before it can handle the blood flow rates required during dialysis sessions. Synthetic grafts require less lead time, around three to six weeks, but are generally considered a second-choice option.

This creates a practical tension. If you wait until eGFR hits 5 to 7 to start dialysis but only get your access placed at eGFR 10, you may not have a mature fistula when you need it and could end up starting on a temporary catheter, which carries a higher infection risk. The planning conversation with your nephrologist typically starts around eGFR 15 to 20, even though you may have years of dialysis-free life ahead.

How Dialysis Timing Differs Around the World

Where you live affects when dialysis begins. A multinational study comparing kidney replacement therapy initiation across countries found substantial variation. The median eGFR at which patients started dialysis was about 14 in Brazil, 11 in the United States, 10 in Germany, and under 10 in France.⁸ These differences held across subgroups defined by sex, ethnicity, and heart failure history, suggesting they reflect systemic differences in clinical practice and health system structures rather than patient characteristics alone.

In the U.S., the relatively higher starting eGFR has been attributed partly to insurance incentives. Medicare coverage for dialysis patients kicks in regardless of age, which may encourage earlier referral. Countries with more conservative traditions, like France, tend to monitor patients closely and initiate later. Neither approach has been shown to produce clearly better outcomes at the population level, which reinforces the idea that individual clinical judgment matters more than any target number.

The Danger of Waiting Too Long

While the evidence clearly does not support early initiation, delaying dialysis too long carries its own risks. A study tracked patients who initially refused dialysis when their kidney function fell below 10 percent. Over half of those patients eventually developed a uremic emergency, a dangerous buildup of toxins that causes severe symptoms and can be life-threatening. Among the patients who initially refused and then later agreed to dialysis, first-year death rates were substantially higher than among those who started electively, with all-cause mortality roughly triple and cardiovascular mortality nearly four times higher.⁹

The distinction matters: safely deferring dialysis under medical supervision is not the same as refusing dialysis and hoping for the best. Patients who delay within a structured care plan, with regular blood tests, symptom monitoring, and a prepared access, do well. Patients who disengage from care and show up in crisis do not.

Incremental Dialysis as a Middle Path

One approach that is gaining traction is incremental dialysis, where patients start with fewer or shorter sessions per week and gradually increase as their remaining kidney function declines. Standard hemodialysis is prescribed three times per week, but patients who still produce a meaningful amount of urine and maintain reasonable fluid and electrolyte balance may do well with just two sessions initially.

A systematic review and meta-analysis found that incremental dialysis preserved remaining kidney function for roughly a year longer than full-dose dialysis on both hemodialysis and peritoneal dialysis, without increasing mortality risk.¹⁰ In peritoneal dialysis specifically, an incremental approach was associated with a significantly lower risk of losing all remaining urine output compared to starting at full dose.¹¹

The appeal of this approach is practical. Fewer sessions per week means less disruption to daily life, fewer hours in a dialysis chair, and better preservation of vascular access. But it requires regular monitoring of residual function, because the kidneys will continue to decline and the dialysis prescription needs to keep pace. Candidates for incremental dialysis tend to be patients who still have substantial urine output, good nutritional status, controlled fluid levels, and a low burden of other chronic diseases.¹²

Why Residual Kidney Function Matters After Dialysis Starts

Whatever kidney function you still have when you start dialysis does not just disappear. Residual kidney function, meaning the filtering and urine production your own kidneys still provide, continues to contribute meaningfully even after you begin treatment. In a study of over 6,500 hemodialysis patients, those who retained more residual function after their first year on dialysis lived significantly longer. The relationship was graded: the faster residual function declined, the higher the mortality risk, with the fastest decliners facing roughly double the death rate compared to the reference group.¹³

Residual kidney function does more than just add a little extra filtering capacity. It helps control fluid balance between dialysis sessions, removes middle-sized toxins that dialysis handles poorly, and contributes to hormone production. Preserving it has become a recognized goal in nephrology. Peritoneal dialysis patients tend to retain residual function longer than hemodialysis patients, in part because hemodialysis involves large fluid shifts and episodes of low blood pressure that can damage already-vulnerable kidneys.¹⁴

Quality of Life Around the Transition to Dialysis

Starting dialysis is often imagined as a sharp dividing line between feeling terrible and feeling better, but the reality is more gradual. A study of older patients found that both mental and physical quality of life declined substantially in the year before dialysis started, then stabilized after initiation. The mental health decline was roughly 13 points and the physical decline about 11 points in the year leading up to dialysis, but neither worsened further in the year after.¹⁵ In other words, dialysis stopped the slide, but did not reverse it.

Another longitudinal study found that patients reported improvements in general quality of life and physical wellbeing within the first 6 to 12 weeks after starting dialysis.¹⁶ However, not all studies agree. An earlier study found that fatigue and lack of energy actually worsened after dialysis started, even though disease-specific symptoms and anxiety stayed about the same.¹⁷ The mixed findings probably reflect the enormous variability among patients: someone who starts dialysis with severe symptoms and fluid overload is more likely to feel better afterward than someone who was managing reasonably well on medical therapy alone.

The Cost of Starting Too Early

Beyond health outcomes, timing has financial consequences. The economic analysis from the IDEAL trial found that the early-start group accumulated significantly higher direct dialysis costs, averaging roughly $10,800 more per patient, simply because they spent more months on dialysis with no survival benefit in return.¹⁸ Total costs including hospitalization and management of complications were even higher in the early group, though the difference did not reach statistical significance. On a population level, these unnecessary months of dialysis represent a substantial burden on health systems.

Older Adults and the Conservative Alternative

The decision to start dialysis is different for older patients, particularly those who are frail or have multiple serious health conditions. Mortality is high in elderly dialysis patients, functional decline is common, and quality of life tends to be lower than in younger patients.¹⁹ This has led to growing interest in conservative kidney management, an approach where patients receive full medical care to manage symptoms, control complications, and maintain quality of life, but choose not to pursue dialysis.

A systematic review found that patients over 80 and elderly patients with significant comorbidities appear to lose the survival advantage that dialysis provides to younger, healthier patients.²⁰ In terms of symptom burden, hospitalizations, and dying at home rather than in a hospital, conservative management had potential advantages. A study comparing the two approaches found that adjusted median survival from recruitment was about 13 months shorter for patients choosing conservative management, but those patients maintained stable quality of life, whereas life satisfaction actually declined significantly after dialysis initiation in the dialysis group.²¹

A recent Cochrane review assessed the evidence comparing conservative management to dialysis in older people and rated the certainty of all major outcomes as very low.²² The honest state of the science is that we do not have high-quality randomized data to definitively say which approach is better for any given elderly patient. The conversation needs to center on the patient’s goals: maximizing lifespan at any cost, or preserving day-to-day quality with the understanding that survival may be shorter.

How Patients Misunderstand the Numbers

One underappreciated problem is how patients and families interpret the eGFR numbers they are given. A qualitative study of older patients with advanced kidney disease found that numerical and graphical depictions of kidney function were central to how people understood their disease, but they often created unintended misunderstandings. The concept of a “threshold” for dialysis led many patients to perceive a binary choice: hit the number, and you either start dialysis or die.²³ That framing obscures the reality that the transition is gradual, the timing is flexible, and alternatives like conservative management exist.

If your nephrologist tells you that your eGFR is 12 and you may need dialysis “soon,” that does not mean you have weeks to live without it. For most patients, it means you are entering a phase where closer monitoring matters, where access planning should be underway, and where you and your care team should be having detailed conversations about your symptoms, your preferences, and the approach that fits your life. The percentage on the lab report is a starting point for that conversation, not the whole answer.