When To Start Dialysis Creatinine Level Explained

Dialysis is started because of how a person is, not because of where a single blood result lands. Two people can walk into the same clinic on the same morning with an identical creatinine of 6.0 mg/dL. One is a 34-year-old with a lot of muscle who feels well, eats normally, sleeps through the night and has stable blood tests. The other is a 79-year-old who has been vomiting for a fortnight, cannot lie flat because of fluid on the chest, and is confused by mid-afternoon. The first may not need dialysis for another two years. The second may need it this week. The creatinine is the same. Everything that matters is different.

That is not a way of dodging the question. It is the actual position of kidney medicine, and it has been for well over a decade. Nephrologists do watch creatinine, and they do watch the filtration estimate that comes with it, but those numbers set the pace of the conversation rather than triggering the treatment. What triggers the treatment is symptoms, fluid, blood chemistry that will not behave, and nutrition. If you want to understand where a given value sits on the wider map first, what creatinine level indicates kidney failure covers the territory just before this one, and when to worry about creatinine levels covers the ground long before that.

Why there is no creatinine level that starts dialysis

The expectation of a threshold is completely reasonable. Most of medicine has them. Blood pressure has targets. Blood glucose has diagnostic cut-offs. Cholesterol has treatment lines. So when someone is told their kidneys are failing, they assume there is a creatinine value written down somewhere that means the machine starts. There is not, and the reason is specific rather than vague.

Creatinine is a waste product from muscle. Your blood level is a balance between how much your muscles produce and how fast your kidneys remove it. That means the number carries information about two things at once, and you cannot separate them by looking at the number. A person with a great deal of muscle produces more creatinine every day, so they sit higher at every level of kidney function. A person who has lost muscle through illness, age or immobility produces less, so they sit lower at every level of kidney function, including very poor kidney function.

This creates a specific and dangerous asymmetry. The frail, thin, elderly patient with genuinely awful kidney function may have a creatinine that looks almost reassuring, while the muscular younger patient with moderately reduced function may have one that looks alarming. If a treatment decision as consequential as starting lifelong dialysis were pinned to that number, it would be made too late in exactly the group who can least afford it and too early in the group who would gain the least. The article on what a normal creatinine level is covers how much the reference range itself shifts with body size and sex.

There is a second reason, which is about what dialysis is for. Dialysis does not repair kidneys. It substitutes for two of their functions: clearing waste and removing fluid. It does that imperfectly, at considerable cost to a person’s time, freedom and often their sense of themselves. So the question a kidney team is asking is not “has the number crossed a line” but “is this person now worse off with their own kidneys than they would be on treatment”. That is a comparison of lived states, not of laboratory values.

Why the same creatinine means different things in different people

It is worth going through this properly, because it is the single most common source of confusion and distress in kidney clinics. Someone reads that their friend started dialysis at a creatinine of 7 and panics because theirs is 7.4. Someone else reads that a relative managed for years at 5 and assumes they have years too. Neither comparison holds.

Each of those five people has a number that, taken alone, would be interpreted wrongly. The kidney team never takes it alone. They convert it into an estimated filtration rate, which adjusts for age and sex and gets much closer to the truth, and then they interpret that estimate in the context of the whole person. The mechanics of that conversion are set out in how to calculate GFR from creatinine, and the difference between the two main filtration measures in creatinine clearance versus GFR.

There is also a third variable most people never hear about: how much the number moves on its own. Creatinine is not a stable quantity. Hydration, a heavy meat meal, a bout of vomiting, a new medicine, an infection, a hot week, a hard gym session — all shift it. A single reading is a snapshot of a moving thing. That is why nobody sensible makes a permanent decision on one result, and why the article on whether creatinine levels fluctuate matters more than it sounds like it should.

One more point, and it is the one that most often gets missed by people reading their own results. Creatinine and filtration do not move in a straight line together. The relationship is a curve. In the early stages, kidney function can fall a long way while creatinine barely twitches, because the remaining filtering units compensate. By the late stages, the curve has turned almost vertical, so a small further loss of function produces a large jump in the number. That is why a rise from 1.0 to 1.4 can represent more lost function than a rise from 6.0 to 7.0, and it is why people in advanced kidney disease often see frightening-looking jumps that their team treats calmly.

What actually drives the decision to start

Here is the honest list. These are the things that get discussed in a clinic room when a start date is being considered, roughly in order of how much weight they carry.

Look at where creatinine sits on that list. It is at the bottom, and it is there because by the time it is high enough to be part of the conversation, everything above it is telling a much clearer story. Creatinine is how the person got referred, monitored and prepared. It is not how the start date gets set.

The clinical shorthand for the top three is that dialysis is started for symptoms or for a complication that cannot be managed any other way. Everything else is preparation. A kidney team that is doing its job well will have spent the previous year or two making sure that when one of those three arrives, the patient is ready for it — access created, decisions made, transplant options explored. That preparation is covered further down this page, and it is arguably more important than the start date itself.

Uremic symptoms: what they are and why they matter most

Uremia is what happens when waste products the kidneys normally remove build up in the blood to the point where they make you ill. It is not one substance. Urea is the one it is named after, but the syndrome involves dozens of retained compounds, and the symptoms are systemic rather than confined to the kidneys.

These are the ones that carry weight in the decision.

What makes this list awkward in practice is that almost every item on it is non-specific. Tiredness, poor appetite, itching and poor sleep have a hundred causes. The pattern is what counts: several of them appearing together, gradually, in someone with known advanced kidney disease, and getting worse month on month. A kidney team will ask about them at every appointment for exactly that reason. If you are supporting someone through this, being the person who notices and reports the change is genuinely useful, because the patient’s own baseline shifts so slowly that they adapt without registering it.

A practical note for anyone reading this on behalf of someone else. Symptoms that need same-day medical attention rather than a routine appointment include vomiting that prevents keeping fluids down, breathlessness at rest or when lying flat, new chest pain, marked confusion or unusual drowsiness, and a sharp drop in how much urine is being passed. Those are not things to wait out until the next clinic date.

Fluid overload that stops responding to treatment

Healthy kidneys manage your fluid balance without you ever thinking about it. Drink more, pass more. Drink less, pass less. As kidney function falls, that regulation gets progressively less reliable, and eventually fluid starts accumulating regardless of what a person does.

It shows up in a predictable order. Ankles and lower legs swell first, worse by evening and better after a night lying flat. Then the swelling climbs. Then fluid starts collecting in the lungs, and the person becomes breathless — first on exertion, then on lying flat, then at rest. Waking at night gasping, or needing three or four pillows to sleep, is a specific and serious version of this. Weight climbs quickly, sometimes several kilograms in a week, which is fluid rather than anything else.

Kidney teams manage this for a long time with fluid restriction, salt restriction and diuretic medicines that push the failing kidney to produce more urine. That works, until it does not. The point at which fluid overload can no longer be controlled by those means is one of the clearest indications for starting dialysis, because dialysis removes fluid directly and mechanically rather than relying on the kidney to cooperate.

Repeated hospital admissions for fluid overload are particularly significant. Each one is unpleasant, each carries risk, and a pattern of them is usually taken by a kidney team as evidence that the current approach has run out. It is one of the more common routes into dialysis for people who also have heart failure, since the two conditions worsen each other and fluid is the shared battleground.

Do not attempt to manage this yourself by changing how much of any medicine you take. Diuretic doses in advanced kidney disease are a balancing act between fluid removal and pushing the kidney further into trouble by drying it out, and getting it wrong in either direction causes real harm. It is a conversation for the kidney team, urgently if the breathlessness is new or worsening.

Potassium and acid that can no longer be controlled

Two blood chemistry problems can force the timing regardless of how someone feels.

Potassium

The kidneys are the main route by which potassium leaves the body. As function falls, potassium can accumulate, and high potassium is dangerous in a way that is quite different from most kidney complications: it interferes with the electrical activity of the heart and can cause a dangerous rhythm disturbance with very little warning. People often have no symptoms at all until something serious happens, which is why it is monitored closely rather than waited for.

Teams manage raised potassium with dietary restriction, reviewing medicines that raise it, treating the acidosis that contributes to it, and in some cases prescribing binding agents that remove potassium through the gut. When potassium repeatedly runs high despite all of that, dialysis becomes the answer, because a dialysis session removes potassium directly and reliably.

Potassium restriction in advanced kidney disease is genuinely difficult, because many of the foods people are told to eat for general health are high in it. Bananas, potatoes, tomatoes, oranges, avocado, dried fruit and nuts are all in that category. This is one of the many places where kidney diet advice contradicts standard healthy-eating advice, and where a renal dietitian is worth more than any article. If you are trying to understand the food side of kidney disease more broadly, how to lower creatinine levels covers the dietary levers and their limits.

Acidosis

Healthy kidneys remove acid produced by normal metabolism and regenerate bicarbonate to keep blood pH in a narrow range. When they cannot, acid accumulates. Metabolic acidosis is less immediately dramatic than high potassium but it does steady damage: it breaks down muscle, weakens bone, worsens the accumulation of potassium, and appears to accelerate the loss of remaining kidney function.

It is usually treated with oral bicarbonate, and that works for a long stretch. Acidosis that cannot be corrected by those means, particularly when it is combined with high potassium and falling muscle mass, is a recognized reason to start dialysis. The NIDDK guidance on kidney disease testing sets out the blood chemistry that gets tracked alongside creatinine for exactly these reasons.

Appetite, weight and malnutrition

This is the indication that gets the least attention from patients and a great deal of attention from nephrologists, and it deserves explaining because it is often the thing that tips a decision.

As uremia develops, appetite falls. Protein foods become unappealing first. Meanwhile the body in advanced kidney disease is in a low-grade inflammatory, catabolic state that breaks muscle down faster than normal. Add the dietary restrictions that come with advanced kidney disease — less protein, less potassium, less phosphate, less salt, less fluid — and it becomes genuinely hard to eat enough of anything.

The result is a gradual loss of muscle and body weight that is sometimes called protein-energy wasting. It matters enormously, because how well someone does on dialysis depends heavily on what physical condition they are in when they start. Someone who begins dialysis already depleted, weak and underweight has a harder time than someone who begins in reasonable shape. So a kidney team watching weight drift downward over successive clinic visits will often move the conversation forward rather than waiting for a dramatic symptom.

There is a cruel irony hidden in the creatinine number here. As muscle is lost, creatinine production falls. So a person who is becoming malnourished may show a creatinine that plateaus or even drops, at precisely the moment they are getting worse. Anyone reading that as improvement is reading it exactly backward. It is one of the strongest single arguments against using creatinine as a trigger, and it is covered from the other direction in what low creatinine means.

Where eGFR fits, and what “single digits” means

If any number is watched more closely than creatinine in this situation, it is the estimated glomerular filtration rate. eGFR takes your creatinine and adjusts it for age and sex using a validated equation, producing an estimate of how many milliliters of blood your kidneys filter per minute per 1.73 square meters of body surface area. A healthy young adult is around 100 or above. The five stages of chronic kidney disease are defined by it.

The phrase you will hear from a kidney team is that dialysis is generally considered when eGFR falls into the single digits — broadly below about 10, and often nearer 5 to 7 in someone who remains well. But that range is a rough guide to when the conversation usually becomes concrete, not a switch. Plenty of people are dialysed at an eGFR of 12 because they are unwell, and plenty manage at 7 because they are not. The stage-by-stage picture is set out in creatinine levels in stage 4 kidney disease and, earlier on, creatinine levels in stage 3 kidney disease. The National Kidney Foundation’s eGFR guide explains how the estimate is derived and where it is least reliable.

Two things about eGFR are worth understanding if you are tracking your own.

First, the trend matters far more than the value. An eGFR of 14 that has been 14 for three years is a different situation from an eGFR of 14 that was 22 nine months ago. The first suggests stability and possibly years of runway. The second suggests a trajectory that will reach the point of decision fairly soon, and it is the one that prompts a team to get access created and choices made. Kidney teams plot these values over time for precisely this reason, and a single reading in isolation tells them very little.

Second, eGFR becomes less reliable at the extremes, and the extremes are exactly where you are when this question matters. The equations were built and validated across broad populations, and they perform less well in people with very low muscle mass, very high muscle mass, amputations, severe liver disease, or unusual body composition. At very low filtration rates the estimate also loses precision. This is why cystatin C — a different marker that is much less affected by muscle — is increasingly used alongside creatinine when the picture does not add up. If you are the muscular 30-year-old whose creatinine looks alarming, cystatin C is worth asking about.

The specific numbers people search for, answered directly

People arrive at this page having typed a specific value into a search box, usually the one on their own report. Here are direct answers to the ones asked most often. Every one of them comes with the same caveat: the number alone does not decide anything.

Is a creatinine of 2.8 a level that needs dialysis?

Almost certainly not. A creatinine of 2.8 mg/dL is meaningfully raised and does indicate reduced kidney function in most people, but it is nowhere near the territory where dialysis is usually discussed. Depending on age, sex and build, 2.8 typically corresponds to an eGFR somewhere in the twenties to low thirties — stage 3b or stage 4 kidney disease. That is the zone where you would expect specialist referral, close monitoring, aggressive management of blood pressure and diabetes, medication review, and eventually education about future options. It is not the zone where dialysis is started.

What a 2.8 should prompt is a serious conversation about slowing progression, because what happens over the following years is genuinely influenceable. Blood pressure control, glucose control if you are diabetic, avoiding anti-inflammatory painkillers, staying hydrated, stopping smoking, and modern kidney-protective medicines have all changed the outlook for people at this stage. Many people at 2.8 never reach dialysis at all.

Is a creatinine of 3.5 a level that needs dialysis?

Generally no, though it is a more advanced position than 2.8. A creatinine of 3.5 mg/dL usually corresponds to an eGFR in the high teens to low twenties, depending on the person — stage 4 for most. This is the stage at which formal preparation typically begins: education sessions about the different treatment options, discussion of transplantation including whether a living donor might be available, and planning for dialysis access if that is the likely route.

That preparation is not the same as starting. Someone at 3.5 who feels well, eats normally, has controllable fluid and stable potassium may have a considerable stretch ahead of them. But it is the point where drifting is no longer sensible, because the single worst outcome in this whole area is arriving at the point of need with nothing prepared.

Is a creatinine of 4.5 the level needed for dialysis?

No. There is no “level needed for dialysis”, and 4.5 in particular is not one. This value appears in search queries often enough that it has clearly circulated somewhere as a threshold, and it is not one. A creatinine of 4.5 mg/dL is advanced — for most adults it corresponds to an eGFR somewhere in the low-to-mid teens — but plenty of people sit at 4.5 while feeling reasonably well, and they are managed rather than dialysed.

What is true is that 4.5 is usually deep into the preparation phase. If you or someone you care for is at that level and no one has yet discussed access, transplant assessment or treatment options, that is a reasonable and important thing to raise at the next appointment.

What is the highest creatinine level before dialysis?

There is no ceiling, and this is the question that surprises people most. Values above 10 mg/dL are not rare in people who have not yet started, and considerably higher figures are seen, particularly in younger people with a lot of muscle whose kidneys have failed slowly enough for the body to adapt. What determines whether someone at a very high level is dialysed is, again, how they are: symptoms, fluid, potassium, acid, nutrition.

It runs the other way too. People with very little muscle sometimes need dialysis at creatinine levels that would look almost unremarkable on a report, because the low number reflects low production rather than good clearance. The frail 85-year-old with a creatinine of 4 and an eGFR of 8 who is vomiting and confused needs treatment more urgently than the 30-year-old at 9 who is at work.

What creatinine level requires dialysis, then?

None, in isolation. The best short answer anyone can give you is this: dialysis is required when the consequences of kidney failure can no longer be managed any other way. That state usually occurs when eGFR is in the single digits, which usually means a high creatinine, but the creatinine is a correlate of the situation rather than its cause. A team looking at a person at 4.5, 7 or 11 asks the same questions each time — how are they eating, sleeping, breathing, thinking; what is the potassium doing; is the fluid controllable — and the answers, not the number, set the date.

Why starting early on numbers alone has not been shown to help

It is worth knowing that the current approach is not simply tradition or resource rationing. It came from evidence, and from a genuine change of mind within the specialty.

For a period, the prevailing assumption was intuitive: if kidney failure makes people ill, starting the replacement treatment earlier should make them less ill. Average filtration rates at the point of starting dialysis drifted upward in several countries as a result. It seemed obviously right.

It was then tested properly. A large randomised trial conducted across Australia and New Zealand, generally referred to as the IDEAL study and published in 2010, assigned people with advanced kidney disease either to start dialysis at a higher estimated filtration rate or to wait and start at a lower one, unless symptoms demanded otherwise. The early-start group did not live longer. Rates of cardiovascular events, infections and complications were not improved by starting sooner. What the early-start group did get was more months of their life spent on dialysis. Notably, a substantial proportion of those assigned to the later-start group ended up starting earlier than planned anyway, because they developed symptoms — which is itself the point. Symptoms, not the number, brought them to treatment.

Observational data has broadly pointed the same way, and international guidance has followed. The practical consequence is the position described throughout this article: an intent to start based on symptoms and complications rather than on reaching a filtration figure, with the figure used to make sure everything is ready in time.

There is an important nuance here that is easy to misread. “Later is not worse” is not the same as “later is better”, and it is emphatically not an argument for delaying once someone is genuinely symptomatic. Starting too late — after weeks of vomiting, after malnutrition has set in, after an emergency admission — is clearly worse. The evidence supports starting when the person needs it, not starting as late as possible. Those are different instructions, and the second one causes harm.

Planned starts versus urgent starts, and why the difference is large

If there is one practical message in this article worth acting on, it is this one. How dialysis begins matters a great deal, and the difference between a planned start and an emergency one is one of the few things in this whole area that is substantially within reach.

The difference is not just about comfort, although the comfort difference is real. Urgent starts more often involve a central venous catheter, which carries a meaningfully higher risk of bloodstream infection than a surgically created fistula. They frequently remove the option of peritoneal dialysis or home hemodialysis before the person has even heard about them. And they begin the whole experience with an emergency, which shapes how someone feels about the treatment for a long time afterward.

A significant proportion of people worldwide still begin dialysis this way, often because kidney disease was diagnosed late, sometimes because progression was faster than anticipated, and sometimes because preparation was deferred. Late referral to a kidney specialist is the single biggest contributor. It is the strongest practical reason to take a raised creatinine seriously long before dialysis is on the horizon — the whole point of identifying what is causing high creatinine early is to buy the years in which something can be done about it.

If you are reading this on behalf of someone at stage 4 or 5 who has no access, no education appointments booked and no transplant discussion under way, that gap is the thing worth raising. It is a far more useful conversation than any about a specific creatinine value.

The preparation that starts long before dialysis does

Good kidney care in the two or three years before dialysis looks like a series of unglamorous, practical steps. None of them commits anyone to anything. All of them preserve options.

Vein preservation deserves a second mention because it is so easily lost. It costs nothing, it is entirely preventable, and losing the option of a good fistula because of repeated cannulation in the wrong arm is a genuinely common and avoidable problem. If someone you care for has advanced kidney disease, it is a reasonable thing to mention to any nurse approaching them with a needle.

Hemodialysis and peritoneal dialysis, at overview level

There are two broad ways of doing dialysis, and the choice affects daily life far more than most people expect before they are faced with it. This is an overview, not a decision guide — the choice belongs in a conversation with a kidney unit that knows the person.

A few things are worth saying plainly. Neither is clearly better in terms of survival for most people; the differences are mostly about fit with a person’s life, home circumstances and preferences. Peritoneal dialysis tends to preserve remaining natural kidney function somewhat longer, which is one reason it is often favored as a first treatment. Home hemodialysis, where someone is trained to run their own machine, allows longer and more frequent sessions that many people find leaves them feeling considerably better, though it demands a lot of the household. And the choice is not permanent — people move between the two over the years as circumstances and their bodies change.

The one route that removes choice is arriving in an emergency. That is the thread running through this whole page.

Choosing not to have dialysis, which is a legitimate decision

Dialysis is a treatment, and like every treatment it can be declined. For some people, particularly those who are very elderly or living with several serious conditions alongside kidney failure, choosing not to dialyse is a reasonable, informed and entirely respectable decision. It has a name in kidney medicine — conservative kidney management, sometimes called supportive care — and it is a defined pathway rather than an absence of one.

Conservative management means treating everything that can be treated without dialysis. Anemia is managed. Fluid is managed as far as possible. Itching, nausea, restless legs and breathlessness are treated actively as symptoms in their own right. Blood pressure is controlled. Nutrition is supported. The person keeps their time, avoids the travel and the needling, and is looked after by a kidney team alongside palliative care specialists. It is not giving up, and describing it that way to someone who has chosen it does them a real disservice.

The reason it is a genuine choice rather than a lesser option is that the benefit of dialysis is not uniform. In fit people it generally extends life substantially. In people who are very frail, or who have advanced heart disease, significant dementia or serious limitations in daily function, the survival advantage narrows — in some groups it is modest — while the burden does not. Dialysis three times a week involves travel, several hours attached to a machine, recovery time afterward, and often a decline in independence. Some people, given that information clearly, decide the trade is not one they want. Others in similar circumstances decide the opposite. Both are defensible.

What makes the difference is the quality of the conversation. A decision made after a frank discussion of what each path actually involves is a good decision either way. A decision made by default, or by assumption, or because nobody raised the alternative, is not. If you are supporting someone through this and conservative management has never been mentioned, it is a fair thing to ask about — as is the option of trying dialysis and stopping later if it is not what they wanted, which is also permitted and more common than people realize.

What to ask the nephrologist

Appointments are short and it is hard to think clearly in them, particularly when frightened. These questions are worth writing down beforehand. They are ordered so that the most useful ones come first, in case time runs out.

Take someone with you if you can. Two people remember far more of an appointment than one, and in a conversation like this the second person is often the one who asks the question that mattered.

What happens to creatinine once dialysis starts

This surprises almost everyone, and it is a fitting place to finish, because it shows how completely the meaning of the number changes.

Once dialysis begins, creatinine stops being a measure of your kidneys and becomes a measure of your dialysis. It falls during each session as the machine clears it, then climbs again between sessions as your muscles keep producing it and your kidneys still cannot remove it. Someone on three-times-weekly hemodialysis has a sawtooth pattern: highest just before a session, lowest immediately after, and highest of all after the longer weekend gap.

Because of that, the timing of the blood test determines the result, and comparing a pre-dialysis value with a post-dialysis one tells you nothing useful. Units standardize when they take samples for exactly this reason.

Three further points follow from this, and they are the ones that catch people out.

A creatinine that is not especially high in someone on dialysis is not necessarily good news. It may reflect low muscle mass, which in this population usually means poor nutrition and is associated with worse outcomes. Kidney teams are often more concerned by a dialysis patient whose creatinine is unexpectedly low than by one whose creatinine is high, which is the reverse of how the number is read in everyone else.

Adequacy of dialysis is not judged by creatinine at all. It is judged mainly by urea clearance, expressed as a measure called Kt/V, along with fluid removal, blood pressure control and how the person actually feels. The way urea and creatinine relate to one another is covered in what the BUN creatinine ratio is, and the wider idea of measuring clearance rather than concentration in what creatinine clearance means.

And residual kidney function still counts. Many people continue to pass urine after starting dialysis, and that remaining function is genuinely valuable — it helps with fluid management, contributes to clearance, and is associated with better outcomes. Protecting it is one reason kidney teams remain careful about anti-inflammatory painkillers, contrast scans and episodes of dehydration even after dialysis has started. The underlying principles are the same ones described in what high creatinine means and in the routine interpretation of creatinine in a blood test.

After a transplant, the picture changes again. Creatinine typically falls substantially within days of a successful transplant and then settles at a new baseline that becomes the reference point for that person. From then on it is monitored closely, because a rise can be the earliest sign of rejection or of a problem with the new kidney. The number becomes important once more, but it is measuring something different: the health of a transplanted organ rather than the failure of the original ones. The ranges used to judge it are the ordinary ones described in the normal creatinine clearance range.

When to start dialysis and creatinine levels: frequently asked questions