There is no single GFR number that automatically triggers dialysis. Most clinical guidelines now recommend starting dialysis based on symptoms and clinical judgment rather than hitting a specific threshold, and the landmark trial that tested this question found no survival benefit to starting at a higher GFR versus a lower one. In practice, most people begin dialysis somewhere between a GFR of 5 and 15, but the wide range reflects how much individual circumstances matter. The story behind that range, and what actually does prompt the decision, is more nuanced than a simple cutoff.
Why Guidelines Moved Away From a Fixed Number
For years, there was a trend toward starting dialysis earlier, at higher GFR levels. The thinking seemed logical: intervene before things get too bad. That changed in 2010 with the publication of a major randomized trial called IDEAL, which assigned patients with advanced kidney disease to either an early-start group (planned dialysis at a GFR of 10 to 14) or a late-start group (planned dialysis at a GFR of 5 to 7). The result was clear: starting earlier did not improve survival or any other clinical outcome.1PubMed. A randomized, controlled trial of early versus late initiation of dialysis This finding held up across follow-up analyses, with no meaningful difference in patient-centered outcomes between the two groups.2PubMed Central. Timing of Dialysis Initiation: What Has Changed Since IDEAL?
The IDEAL trial shifted how doctors think about dialysis timing. After its publication, real-world practice changed measurably, with the trend toward starting at higher GFR levels reversing in countries like Canada.3JAMA Internal Medicine. Association Between the Publication of the Initiating Dialysis Early and Late Trial and the Timing of Dialysis Initiation in Canada Economic analyses also showed that the earlier-start trend had been costly, and reversing it could yield meaningful savings.4PubMed Central. Timing of initiation of dialysis: time for a new direction?
A systematic review of high-quality clinical practice guidelines found that half of the major guidelines addressed the timing question directly. All six that did so agreed that dialysis should start when symptoms or signs of advanced kidney disease are present. Only two of the six recommended a specific GFR starting point for people without symptoms, and even those set the range at 5 to 7.5PLoS ONE. Recommendations of high-quality clinical practice guidelines related to the process of starting dialysis: A systematic review The consensus today is that there is no recommended GFR threshold for starting dialysis, and the decision should come from shared discussion between you and your care team.6JAMA. Dialysis for Chronic Kidney Failure: A Review
What Actually Triggers the Decision
If not a number, then what? The answer is symptoms. As your kidneys lose more and more function, waste products build up in the blood. This causes a cluster of problems collectively known as uremia. Persistent nausea, severe fatigue, loss of appetite, difficulty concentrating, and an unpleasant metallic taste are common. Fluid overload can show up as shortness of breath, swelling in the legs and feet, and difficulty sleeping flat. Metabolic problems like worsening acidosis and dangerously high potassium also factor into the decision.6JAMA. Dialysis for Chronic Kidney Failure: A Review
These symptoms tend to appear at different GFR levels for different people. Someone with diabetes, heart disease, and poor nutrition might become symptomatic at a GFR of 12 or 13. Another person with fewer comorbidities might tolerate a GFR of 6 with surprisingly few complaints. This is exactly why a universal cutoff does not work well. Your nephrologist will track the trajectory of your GFR alongside how you feel, your lab trends, and your nutritional status. When those factors converge, it becomes time to start.
Emergency Situations That Force the Issue
Sometimes the decision is not gradual. Certain life-threatening complications demand urgent dialysis regardless of what your GFR reads. Clinicians often remember these with the mnemonic AEIOU:
- Acidosis: When blood pH drops to dangerously low levels that cannot be corrected with medications alone, dialysis can clear the acid.
- Electrolyte imbalances: Severely elevated potassium is the most common culprit and can cause fatal heart rhythm problems.
- Ingestion: Certain drug overdoses or toxic ingestions can be treated by dialyzing the substance out of the blood.
- Overload: Fluid buildup that causes or worsens heart failure, particularly when diuretics are no longer effective.
- Uremia: Severe uremic symptoms such as encephalopathy, pericarditis, or uncontrollable bleeding from platelet dysfunction.
These emergencies can arise at any GFR level. Someone who has been declining slowly might have a sudden potassium spike or develop fluid overload from a dietary indiscretion or a missed medication. In these cases, dialysis begins urgently, and the GFR number on the lab report is secondary to keeping the person alive.
The Wide Variation in When People Actually Start
Even though guidelines have moved away from a fixed number, the GFR at which people actually begin dialysis varies enormously. A large study of U.S. dialysis facilities found that about a third of patients started dialysis with a GFR at or above 10.5, but the proportion ranged from roughly 20% to 57% depending on the geographic region and from 10% to 67% depending on the specific facility.7PubMed Central. Variation in the level of eGFR at dialysis initiation across dialysis facilities and geographic regions
What explains that spread? Mostly the patients themselves. After accounting for individual characteristics and facility factors, patient case mix explained about 97% of the variability. The facility itself accounted for about 3%, and geographic region essentially zero.7PubMed Central. Variation in the level of eGFR at dialysis initiation across dialysis facilities and geographic regions In other words, patients are different from one another in ways that legitimately affect when dialysis becomes necessary. One person’s GFR of 8 may be another person’s GFR of 12, depending on their overall health, body composition, and how their kidneys are failing.
Why the GFR Number Can Be Misleading
The standard estimated GFR is calculated from blood creatinine levels. Creatinine is a waste product of muscle metabolism, and therein lies a problem: people with more muscle mass naturally produce more creatinine, while people who are older, malnourished, or have lost significant muscle mass produce less. A frail 80-year-old with very little muscle could have a GFR that looks higher than it truly is, because their low creatinine production masks how poorly the kidneys are actually working.
Cystatin C is an alternative blood marker that is not tied to muscle mass the same way creatinine is. Research has confirmed that higher creatinine levels are associated with higher muscle mass, while cystatin C shows either a modest inverse relationship with muscle or none at all. When both markers are measured together, the positive link between creatinine and muscle and the inverse link between cystatin C and muscle both become more apparent.8PubMed Central. Creatinine, cystatin C, muscle mass, and mortality: Findings from a primary and replication population‐based cohort For patients whose muscle mass is unusually low or high, a GFR estimate based on cystatin C or a combination of both markers can give a more accurate picture and help avoid misjudging the timing of dialysis.
Planning Vascular Access Before You Need It
Even if the actual dialysis start is symptom-driven, there is one aspect of timing that does depend on GFR thresholds: creating your vascular access. If you are heading toward hemodialysis, you need a functioning access point in your arm, ideally an arteriovenous fistula that has had time to mature. This requires minor surgery, and the fistula typically needs several months before it is ready to use.
A decision analysis found that referral for fistula creation about 12 months before the estimated time to dialysis, or when GFR drops below roughly 15 to 20, produced the best outcomes.9American Journal of Kidney Diseases. Optimal Timing of Arteriovenous Fistula Creation in Patients With Chronic Kidney Disease: A Decision Analysis In practice, only about 40% of patients have their fistula created within the recommended 3- to 12-month window before starting dialysis.10PubMed Central. Likelihood of starting dialysis after incident fistula creation The rest end up either too early (the fistula sits unused for years) or too late (they need a temporary catheter).
How late is too late? One study found that patients who had access surgery at a GFR below 10 waited a median of only 69 days before starting hemodialysis, compared with 156 days for those who had it at a GFR of 10 to 14 and over a year for those who had it at a GFR of 15 or above.11PubMed Central. Predialysis Vascular Access Placement and Catheter Use at Hemodialysis Initiation At 69 days, many fistulas have not had time to mature. That means those patients often start dialysis through a temporary catheter, which carries higher infection and complication risks.
The Danger of Crash Starts
The flip side of not starting too early is the risk of waiting too long or losing track of kidney function altogether. A “crash start” is when someone begins dialysis urgently, without adequate preparation, planning, or a functioning access site. This can happen when someone with known kidney disease does not follow up with their nephrologist, or when kidney failure is diagnosed for the first time in an emergency room. Crash starts are associated with higher rates of illness and death, as well as lower quality of life. The reliance on temporary catheter access during crash starts is a major contributor to worse outcomes.12PubMed Central. Risk factors for unplanned and crash dialysis starts: a protocol for a systematic review and meta-analysis
Certain patient groups are more vulnerable to early or unplanned starts. A study of patients with diabetic kidney disease found that older age was significantly associated with starting dialysis earlier, with an odds ratio of about 4.6 for elderly compared with younger patients.13PubMed Central. Risk factors for an early dialysis start in patients with diabetic nephropathy end-stage renal disease Older patients with diabetes were also more likely to have had pneumonia before starting dialysis, suggesting that their overall health had deteriorated to the point where delaying was no longer safe.
Incremental Dialysis as a Middle Path
One emerging approach tries to ease the transition rather than making dialysis an all-or-nothing jump. Incremental dialysis means starting with fewer sessions per week, typically two instead of the conventional three, and adjusting upward as your remaining kidney function declines. The rationale is straightforward: if your kidneys are still doing some of the work, you do not need the full dialysis dose right away.
A large cohort study of over 23,000 patients found that about half still had substantial residual kidney function at the time they began dialysis. Among those who started with an incremental schedule, residual kidney clearance and urine output were roughly 15-16% better preserved at three months compared with patients on conventional schedules, and that preservation persisted over time.14PubMed Central. Incremental Hemodialysis, Residual Kidney Function, and Mortality Risk in Incident Dialysis Patients: A Cohort Study Incremental hemodialysis has also been linked to extending the useful life of fistulas and grafts, which matters when you consider that these access sites can wear out over years of use.15PubMed Central. Incremental dialysis for preserving residual kidney function-Does one size fit all when initiating dialysis?
Incremental dialysis is not yet standard practice everywhere, and it requires careful monitoring to ensure that the reduced schedule is providing adequate clearance. But for the right patient, it can make the early period of dialysis less disruptive while protecting what kidney function remains.
When Dialysis May Not Be the Right Choice at All
For some older adults, particularly those with multiple serious health problems, starting dialysis does not always make sense. Conservative kidney management is a recognized alternative that focuses on symptom control, quality of life, and comfort without dialysis. This is not about giving up; it is about making a clear-eyed assessment of whether dialysis will actually help.
A systematic review found mixed evidence on whether dialysis prolongs survival in elderly patients compared with conservative management, and that any survival advantage shrinks as comorbidities pile up, especially heart disease.16PubMed Central. Conservative Management of End-Stage Renal Disease without Dialysis: A Systematic Review A single-center observational study sharpened the picture further: among patients aged 80 and older, there was no statistically significant survival advantage for those who chose dialysis over conservative management. For patients aged 70 and older with high comorbidity scores, the survival advantage of dialysis was substantially reduced, particularly when cardiovascular disease was present.17PubMed Central. Comparative Survival among Older Adults with Advanced Kidney Disease Managed Conservatively Versus with Dialysis
These findings do not mean dialysis is futile for all older adults. Many people in their 70s and 80s do well on dialysis and live for years. But for the subset of elderly patients who are already quite sick, the burdens of dialysis (travel, time on the machine, recovery days, catheter infections, blood pressure drops during treatment) may outweigh a modest or uncertain survival gain. This is a conversation that needs to happen well before GFR reaches emergency levels.
What Happens to Quality of Life Around the Transition
One of the most anxiety-producing aspects of approaching dialysis is the fear that life will become unbearable once it starts. The reality is somewhat more reassuring, at least in relative terms. A study of older patients tracked quality-of-life scores in the year before and after starting dialysis. During the year leading up to dialysis, both mental and physical health scores dropped substantially, with mental health declining by an estimated 13 points and physical health by about 11 points on standard scales. After dialysis began, those scores essentially stabilized: mental health ticked up slightly, and physical health held roughly steady.18PubMed Central. Quality of Life before and after the Start of Dialysis in Older Patients
In other words, the worst quality-of-life period tends to be the months before dialysis, not the months after. The steep decline is driven by worsening uremic symptoms and fluid overload that your failing kidneys can no longer handle. Starting dialysis does not magically restore everything to normal, but it often halts the slide. For many people, that stabilization itself is meaningful.
Frailty and Cognitive Function at the Start of Dialysis
Starting dialysis while already frail can set the stage for additional problems. A longitudinal study of over 300 new dialysis patients found that those who were physically frail at the time they started dialysis scored lower on cognitive testing both at the start and after a year on dialysis. The relationship showed a dose-dependent pattern: the more frail someone was, the worse their cognitive scores tended to be.19PubMed Central. The impact of frailty on outcomes in dialysis
This does not mean dialysis causes cognitive decline. It means that someone who arrives at dialysis already weakened and frail tends to do worse afterward than someone who arrives in relatively better shape. The practical takeaway is that maintaining physical function and nutrition in the months and years of advanced kidney disease, before dialysis, matters. It is not just about whether to start but about what condition you are in when you do. Prehabilitation, including exercise, dietary optimization, and managing anemia and bone health, can influence outcomes in ways that the GFR number alone does not capture.