At What eGFR Score Is Dialysis Needed?

There is no single eGFR number at which dialysis becomes mandatory. Current guidelines have moved away from fixed thresholds, instead recommending that the decision be driven by symptoms and shared decision-making between patient and clinician. A 2024 review in JAMA put it plainly: there is no recommended eGFR threshold for initiating dialysis.¹ That said, when guidelines do mention a number, it tends to fall around 5 to 7 mL/min/1.73 m² for people without symptoms, and the story of how the field arrived at this position involves a surprising reversal in thinking.

What the Guidelines Actually Say

The most recent clinical practice guidelines recommend that dialysis be considered when eGFR drops to around 6 mL/min/1.73 m² or lower, but only as a backstop for patients who have not yet developed symptoms.² A systematic review of high-quality clinical practice guidelines found broad agreement on one point: dialysis should start when symptoms or signs of kidney failure appear. Only two of the six guidelines reviewed even mentioned a specific eGFR range for asymptomatic patients, and those two recommended starting at 5 to 7 mL/min/1.73 m².³ All six emphasized that the decision should be made jointly between the patient and the care team, not triggered automatically by a lab value.

This is a dramatic change from how things worked in the early 2000s, when clinical practice drifted toward starting dialysis earlier, at higher eGFR values. Many patients were beginning dialysis at eGFR levels of 10 to 15 mL/min or even higher, driven partly by concern that waiting too long would lead to worse outcomes. That trend has since reversed, and the reversal is grounded in solid evidence.

Why Starting Earlier Did Not Help

The pivotal study that changed the field was the IDEAL trial, published in 2010 in the New England Journal of Medicine. It randomly assigned patients with stage V chronic kidney disease to begin dialysis either early (at an eGFR of 10 to 14 mL/min) or late (at an eGFR of 5 to 7 mL/min). After a median follow-up of about three and a half years, roughly the same proportion of patients in both groups had died, and there was no meaningful difference in cardiovascular events, infections, or dialysis complications.⁴ Starting earlier simply did not improve survival or clinical outcomes.

A separate meta-analysis looking at critically ill patients with acute kidney injury reached a similar conclusion: early versus late initiation of dialysis made no difference in mortality. Worse, the early-start group experienced significantly more episodes of low blood pressure and infections.⁵ These findings reinforced the idea that starting dialysis prematurely exposes patients to real harms without offsetting benefits.

Since the IDEAL trial, clinical practice guidelines have evolved considerably, moving away from fixed eGFR targets and toward symptom-based initiation.⁶ The earlier trend toward starting at higher eGFR values also carried a substantial financial cost. Analyses based on the IDEAL trial and large US registry data suggested that significant savings could be achieved by reversing the early initiation trend.⁷

Symptoms That Actually Trigger Dialysis

If it is not a number on a lab report, what actually pushes people to start? In practice, the most common reason is the onset of uremic symptoms: nausea, persistent fatigue, loss of appetite, and mental fogginess caused by the buildup of waste products in the blood. A large retrospective study found that about half of patients began dialysis primarily because of uremic symptoms, roughly a fifth because of fluid overload or uncontrolled blood pressure, and about a fifth because of worsening lab values without clear symptoms.⁸ Smaller fractions started because of dangerously high potassium or severe metabolic acidosis.

The JAMA review summarized the clinical picture neatly: persistent signs and symptoms of uremia such as nausea and fatigue, volume overload such as shortness of breath and swelling in the legs, worsening eGFR, metabolic acidosis, and high potassium all inform the decision.¹ No single lab result is sufficient on its own. A patient with an eGFR of 8 who feels well and has stable labs is in a different situation from someone with an eGFR of 12 who is vomiting, swelling, and unable to keep food down. The second patient needs dialysis sooner.

Why eGFR Can Be Misleading Near the End of Kidney Function

Part of the reason the field has backed away from fixed eGFR thresholds is that the number itself becomes less reliable as kidney disease advances. Creatinine-based eGFR formulas work by measuring creatinine in the blood and using it to estimate how well the kidneys filter. But as kidney function worsens, the body compensates by clearing creatinine through other routes, especially the gut. This extra-renal clearance means the formulas overestimate how well the kidneys are actually working, and the overestimation gets worse as kidney function declines.⁹ A patient whose eGFR reads 8 may in reality be filtering closer to 5 or 6.

To address some of these shortcomings, kidney disease guidelines have recommended using cystatin C, a different blood marker, to confirm the diagnosis of CKD and to estimate kidney function when precision matters for clinical decisions.¹⁰ Newer eGFR equations that combine creatinine and cystatin C, developed without race as a variable, have shown improved accuracy and smaller differences across racial groups.¹¹ These improvements matter because the eGFR number directly affects both treatment decisions and transplant eligibility.

Planning Ahead Before You Reach That Point

Even though the actual start of dialysis is symptom-driven, preparation needs to begin well before symptoms appear. If you are choosing hemodialysis, you typically need a surgically created access point in your arm called an arteriovenous fistula, and those take months to mature before they can be used. A decision analysis found that referral for fistula creation performs best when done at an eGFR of roughly 15 to 20 mL/min or within about twelve months of the estimated time to dialysis, whichever comes first.¹² Referring too early wastes resources, since the fistula may never be needed, while referring too late can force a patient into emergency dialysis through a temporary catheter, which carries higher infection risk.

The timing of referral should also account for how quickly your kidney function is declining. Someone whose eGFR drops by 5 points per year needs earlier preparation than someone stable at the same level. Older patients whose kidney function is declining slowly may benefit from later referral to avoid creating access that goes unused.¹²

For patients interested in kidney transplantation, a different eGFR threshold comes into play. Under current US policy, a patient becomes eligible for transplant waitlist registration once their GFR falls to 20 mL/min or below.¹³ Since transplant waiting lists can stretch for years, getting listed early matters. A preemptive transplant, one that happens before dialysis is ever needed, generally yields better outcomes than transplanting after dialysis has already begun.

What Happens When Dialysis Starts Without a Plan

A substantial number of patients end up starting dialysis on an emergency or unplanned basis, which consistently leads to worse outcomes. A systematic review found that the proportion of patients with an unplanned dialysis start ranged widely across studies, but most reported that 40 to 60 percent of patients fell into that category.¹⁴ The strongest risk factors for an unplanned start were late referral to a nephrologist, lack of predialysis education, older age, and more severe comorbidities. Patients who had not been followed in a predialysis clinic were particularly vulnerable.

An unplanned start typically means a temporary catheter instead of a mature fistula, a hospitalization to begin treatment, and poorer adjustment to the demands of dialysis. This finding underscores why the conversation about dialysis timing should start long before a crisis, ideally when eGFR is still in the 20 to 30 range, even though the actual start may not come for years.

Starting with Fewer Sessions Per Week

A growing body of evidence supports incremental dialysis, a strategy where patients start with two hemodialysis sessions per week instead of the standard three, then increase as their remaining kidney function declines. The rationale is simple: if your kidneys are still doing some of the work, you may not need a full dialysis schedule right away. Studies have linked incremental dialysis to better preservation of residual kidney function and improved quality of life.¹⁵

One study directly compared patients who started hemodialysis twice a week against those who started at the conventional three times per week. The twice-weekly group had significantly longer survival, with a median of about 5.6 years compared to 4.6 years in the three-times-weekly group. After adjusting for age, sex, comorbidities, and residual kidney function, twice-weekly hemodialysis was associated with a roughly 24 percent lower mortality risk.¹⁶ Peritoneal dialysis also tends to preserve residual kidney function better than conventional hemodialysis.¹⁷ Keeping whatever kidney function remains matters because even small amounts of natural filtration help control fluid balance, clear toxins that dialysis handles poorly, and improve survival.

When Patients with Heart Failure Start Earlier

Not everyone follows the typical trajectory. Patients with cardiorenal syndrome, where heart failure and kidney failure worsen each other in a vicious cycle, often need dialysis earlier than their eGFR alone would suggest. A study of peritoneal dialysis patients found that those with cardiorenal syndrome started dialysis at higher baseline eGFR values than patients without it, reflecting the fact that refractory fluid overload and heart failure forced their hand before the kidneys had deteriorated to the same degree.¹⁸ Paradoxically, starting at a higher eGFR in these patients was associated with higher mortality, likely because it was a marker for sicker patients, not because early dialysis itself caused harm.

This pattern illustrates a broader point: a high eGFR at the time of dialysis initiation does not necessarily mean the patient was started too early. It can mean the patient had other life-threatening problems that dialysis was needed to manage, problems that also carried their own mortality risk.

How Dialysis Initiation Affects Daily Life

The clinical trial data may show comparable mortality between early and late start, but what patients and families care about is often more immediate: will you feel better once dialysis begins? The honest answer is mixed. Many patients do experience relief from the worst uremic symptoms, but dialysis itself takes a toll, especially on older adults.

A landmark study of nursing home residents found that starting dialysis was associated with a sharp decline in the ability to perform daily activities like bathing, dressing, and eating. Three months after starting dialysis, functional status had been maintained in about 39 percent of residents. By twelve months, only 13 percent had maintained their pre-dialysis level of function, and 58 percent had died.¹⁹ A separate study of hemodialysis patients found that only about 15 percent managed to maintain good physical condition during follow-up, and among patients aged 75 and older, that figure dropped to under 4 percent.²⁰

These numbers do not mean dialysis is futile in older adults, but they do mean that the decision deserves serious thought about goals. For someone whose primary concern is staying active and independent, the trajectory of functional decline on dialysis is a critical piece of information.

Choosing Not to Start Dialysis

For older patients with significant frailty or multiple serious illnesses, conservative kidney management, sometimes called maximal conservative care, is a legitimate alternative. This approach focuses on managing symptoms, slowing progression, and maintaining quality of life without dialysis. A systematic review and meta-analysis found that dialysis roughly halved the risk of death compared with conservative care after adjusting for differences between groups.²¹ That is a meaningful survival benefit on paper. But when researchers looked more closely, the advantage diminished with increasing frailty and comorbidity, and conservative management provided comparable or superior quality of life along with fewer hospitalizations.²²

A longitudinal study comparing elderly patients who chose dialysis versus those who chose conservative management found that quality of life did not significantly improve in the dialysis group after six months, while the conservative group experienced a small decline.²³ Patients managed conservatively were also more likely to die at home, which for many people aligns with their end-of-life preferences.²² The takeaway is not that dialysis is wrong for older adults, but that it is not automatically right, and the conversation should be honest about the tradeoffs.

How the Question Differs for Children

Pediatric nephrologists face a different set of pressures. Growth, school attendance, and developmental milestones compete with the burden of treatment. A survey of Canadian pediatric nephrologists found wide variability in what they considered “early” versus “late” dialysis initiation, with thresholds ranging from above 10 to above 20 mL/min. In practice, most reported starting asymptomatic children at an eGFR of 10 to 14 mL/min, higher than the 5 to 7 range referenced in adult guidelines. Symptoms like fatigue, nausea, significant weight loss, and increasing missed school days were the factors that carried the most weight in the decision.²⁴

Despite the tendency to start children somewhat earlier, a large registry study of over 15,000 children found that those who began dialysis at a higher eGFR (median of about 13 mL/min) actually had a higher risk of death than those who started at a lower eGFR (median of about 7 mL/min). The risk was about 36 percent higher in the early-start group overall, and the effect was concentrated among children who started on hemodialysis rather than peritoneal dialysis.²⁵ As with adults, this likely reflects a mix of sicker children needing earlier intervention and genuine harms from unnecessary early exposure to dialysis. The children starting earlier were more likely to be underweight or obese and to have specific kidney diseases that themselves carry worse prognoses.

Racial and Socioeconomic Gaps in Dialysis Timing

Access to timely, well-planned dialysis initiation is not evenly distributed. Racial disparities affect both the timing of dialysis and access to transplantation. Under current US policy, patients become eligible for the transplant waitlist at an eGFR of 20 mL/min or below. Researchers have estimated that allowing Black patients to be listed at slightly higher eGFR levels, around 24 to 25 mL/min, could help reduce observed disparities in the amount of waiting time they accrue before transplant.¹³

The COVID-19 pandemic exposed additional cracks. A study of dialysis initiation during the first four months of the pandemic found that the mean eGFR at the time patients started dialysis declined slightly overall, but the decline was concentrated among non-Hispanic Black patients. In areas with the highest COVID-19 death rates, patients started dialysis at lower eGFR values than they had before the pandemic, suggesting delayed access to care.²⁶ A lower eGFR at initiation in this context does not reflect a thoughtful late-start strategy. It reflects patients who could not get to care in time, a very different clinical situation that carries its own risks.