Kidney failure is not defined by a creatinine number. It is defined by an estimated filtration rate below 15, and that was a deliberate choice by the people who wrote the definition, because the same creatinine means wildly different things in different bodies. Most people whose kidneys have failed do have a creatinine well above 5 mg/dL, often far higher. But the number alone never makes the diagnosis. This page explains what does.
There is no creatinine level that means kidney failure. Kidney failure, meaning stage 5 chronic kidney disease or end-stage kidney disease, is defined as an estimated glomerular filtration rate below 15 mL/min/1.73m², or being on dialysis. In practice most adults who reach that point have a serum creatinine somewhere above 5 mg/dL, roughly 440 µmol/L, and figures of 8, 10 or 15 mg/dL are common. But a frail 82-year-old woman with very little muscle can be genuinely in failure at a creatinine of 3.6, while a 27-year-old powerlifter can carry 4.0 with filtration that is reduced but nowhere near failure. If you are here because you have seen a number and you are frightened, the honest answer is that the number by itself cannot tell you.
What follows is written for two kinds of reader. One has just been told their kidneys are failing, or has read a result and worked it out before anyone explained it. The other is a son, daughter, partner or friend trying to understand what is happening to someone they love. Both deserve the same thing: accurate information delivered calmly, without softening the hard parts and without making them sound worse than they are. Kidney failure is serious. It is also, for most people who reach it, a condition managed over years rather than an immediate catastrophe. If your own numbers are not yet at this point, creatinine in stage 4 kidney disease and creatinine in stage 3 cover the earlier ground.
A creatinine number on its own is close to meaningless without age, sex and body size. The Waldev creatinine clearance calculator does that conversion, and what creatinine actually is explains where the molecule comes from in the first place.
On this page
What kidney failure actually means
The word failure sounds absolute. In kidney medicine it is not. It is a threshold on a continuous scale, chosen because it is roughly the point at which kidneys can no longer hold the body’s internal chemistry within survivable limits without help, or are approaching it.
Healthy kidneys filter blood at somewhere around 100 to 120 mL per minute, adjusted to a standard body surface area of 1.73 square metres. Kidney failure is defined as that rate falling below 15. A person in failure has therefore lost roughly seven-eighths of their filtering capacity. Two things follow from that definition, and both matter.
The first is that the definition is built on filtration, not on creatinine. Creatinine is the raw ingredient that goes into the estimate. It is fed into an equation along with age and sex, and the equation produces an eGFR. That is the number that stages the disease. Your laboratory almost certainly prints both, and if it prints an eGFR below 15, that is what the kidney team will act on.
The second is that failure is a category, not a cliff edge. Someone with an eGFR of 14 and someone with an eGFR of 6 are both in stage 5, but they are in very different situations. The first may feel entirely well and keep working for years. The second is usually symptomatic and close to needing replacement therapy. Lumping them together under one label is useful for administration and misleading for prognosis.
The formal definition. Kidney failure, in the international staging system, means an eGFR below 15 mL/min/1.73m² or treatment by dialysis. Someone whose kidneys have been replaced by a working transplant is classified separately, usually staged by the transplant’s own filtration rate. The older term “end-stage renal disease” describes the same territory but is used less now, partly because “end-stage” frightens people out of all proportion to what it actually means.
Why the definition avoids creatinine on purpose
Creatinine comes from muscle. Muscles convert a small, fairly constant fraction of their creatine stores into creatinine every day, and kidneys clear it. If muscle is the source, then the amount produced varies with how much muscle a person has. A body-builder produces far more creatinine per day than a small, sedentary, elderly woman, and so sits at a higher blood level with identical kidney function.
That is the whole problem in one sentence. Set a creatinine threshold for failure and you would over-diagnose the muscular and under-diagnose the frail. Under-diagnosing the frail is the more dangerous error, and it happens constantly: an 80-year-old with a creatinine of 2.4 mg/dL may be told her results are “only slightly up” when her actual filtration is under 20. The estimating equations exist precisely to correct for this. They do it by adjusting for age and sex, which are crude proxies for muscle mass but far better than nothing.
None of this makes creatinine useless. It is the most practical marker of kidney function available, it is cheap, it is measured everywhere, and tracked over time in one person it is genuinely informative. What it cannot do is carry a fixed diagnostic threshold across different bodies. Creatinine clearance and how clearance differs from GFR go further into the mechanics of that conversion.
The creatinine figures usually seen in kidney failure
People want numbers, and refusing to give any is unhelpful. So here they are, with the caveat attached rather than buried: these are typical, not diagnostic. Your own relationship between creatinine and filtration depends on your body.
| Situation | Creatinine (mg/dL) | Creatinine (µmol/L) | Notes |
|---|---|---|---|
| Average adult man, normal kidneys | 0.7 – 1.3 | 62 – 115 | eGFR usually above 90 |
| Average adult woman, normal kidneys | 0.6 – 1.1 | 53 – 97 | Lower because of lower muscle mass |
| Stage 4 (eGFR 15–29) | Roughly 2.5 – 5 | Roughly 220 – 440 | Wide overlap with stage 5 |
| Stage 5 / kidney failure (eGFR under 15) | Usually above 5 | Usually above 440 | Frequently 6 – 12; can be far higher |
| Symptomatic uraemia, often near dialysis | Commonly 8 – 15 | Commonly 700 – 1300 | Symptoms track the number poorly |
| Established on haemodialysis, pre-session | Often 6 – 14 | Often 530 – 1240 | Stays high permanently; this is expected |
| Frail, low-muscle older adult in failure | Can be 3 – 4.5 | Can be 265 – 400 | Low production masks the severity |
| Very muscular young adult, moderate impairment | Can be 2.5 – 4 | Can be 220 – 350 | High production overstates the severity |
Read the bottom two rows together and the point becomes obvious. A creatinine of 4.0 mg/dL could belong to a woman of 82 whose filtration is around 10, squarely in failure, or to a 30-year-old rugby forward whose filtration is closer to 25, which is serious but is stage 4. Same number. Different stage. A completely different conversation.
To convert between units, multiply mg/dL by 88.4 to get µmol/L, or divide µmol/L by 88.4 to go the other way. American and some Asian laboratories report in mg/dL. The UK, most of Europe, Canada, Australia and New Zealand report in µmol/L. A great deal of unnecessary panic comes from reading an American forum post in mg/dL while holding a British result in µmol/L, or the reverse.
µmol/L = mg/dL × 88.4
mg/dL = µmol/L ÷ 88.4
Example: 5.0 mg/dL × 88.4 = 442 µmol/L
Example: 700 µmol/L ÷ 88.4 = 7.9 mg/dL
Why the same creatinine means different things in different bodies
This section exists because it is the most common source of misunderstanding in the whole subject, and because getting it right changes how you read your own result.
Creatinine production is roughly proportional to muscle mass. Skeletal muscle holds around 95 percent of the body’s creatine, and something like 1 to 2 percent of that pool converts to creatinine each day through a spontaneous chemical reaction that no drug and no diet meaningfully alters. A person with 35 kg of muscle produces perhaps twice the daily creatinine of a person with 17 kg. If both have identical kidneys, the first sits at roughly twice the blood level of the second, permanently, and neither has a kidney problem.
Now run that in reverse. Take two people who both measure 4.0 mg/dL.
Margaret, 82
48 kg, uses a frame indoors, has lost a lot of muscle over five years of reduced mobility. She produces very little creatinine each day. For her level to reach 4.0, her kidneys must be clearing almost nothing. Her estimated filtration comes out around 10. She is in kidney failure, and the number understates it.
Daniel, 30
102 kg, trains five days a week, carries a great deal of muscle. He generates creatinine fast every single day. His kidneys have to work hard just to hold him at 4.0. His estimated filtration comes out around 22 to 25. Serious, needing specialist care, but stage 4 rather than failure.
Neither of those estimates is exact, and that is worth saying plainly. The equations were built on population averages, so they are least accurate at the extremes of body composition, which is exactly where Margaret and Daniel sit. For someone like Daniel a nephrologist may add a cystatin C measurement, a different filtration marker produced by nearly all cells rather than by muscle, and therefore far less sensitive to how big his arms are. Where the two markers disagree substantially, the truth usually lies between them.
Amputation, prolonged illness, spinal cord injury, malnutrition, advanced liver disease and long-term steroid use all shift the relationship the way Margaret’s has shifted. Heavy resistance training, creatine supplementation and a high meat intake shift it Daniel’s way. If any of those apply to you, tell whoever interprets your result, because it changes what the number is worth. The full list of things that raise creatinine covers the non-kidney causes, and it answers the question people in Daniel’s position ask most often, which is whether creatine supplements are to blame.
What this means for you specifically
If you are elderly, small, or have lost weight and muscle through illness, take your creatinine more seriously than its raw value suggests, and make sure someone has calculated an eGFR rather than eyeballing the creatinine. If you are young and heavily built, do not assume the worst from a single raised value, but do not wave it away either. Get a urine albumin test, get a repeat, and if the picture stays unclear ask whether cystatin C is available locally.
In both cases, watch the trend rather than the point. A creatinine that has climbed from 1.4 to 2.8 to 4.6 over three years tells a clear story. One reading of 4.6 with no history behind it tells almost none.
Kidney disease and kidney failure are not the same thing
People use these phrases interchangeably and they should not. The confusion causes real distress, because someone told they have chronic kidney disease often hears that their kidneys are failing, which in most cases is nowhere near true.
Chronic kidney disease means kidney damage or reduced filtration that has been present for at least three months. It covers an enormous range. Stage 1 and 2 involve normal or near-normal filtration with some marker of damage, usually protein in the urine. Stage 3 means filtration between 30 and 59. Stage 4 means 15 to 29. Only stage 5, below 15, is kidney failure.
The proportions matter. The large majority of people with chronic kidney disease are in stages 1 to 3, and most of them will never reach failure. Many will die of something else entirely, usually cardiovascular disease, without their kidneys ever becoming the main problem. That is not a grim observation. It is a genuinely reassuring one for the person who has just been told they have stage 3 and assumed it meant dialysis was coming.
| Phrase you may hear | What it actually means |
|---|---|
| Chronic kidney disease | Any lasting kidney damage or reduced filtration, stages 1 to 5. Says nothing about severity on its own. |
| Reduced kidney function | Vague and usually means stage 3 or 4. Ask for the eGFR number. |
| Kidney impairment | Same vagueness. Again, ask for the number. |
| Kidney failure / stage 5 CKD | eGFR below 15, or on dialysis. The specific thing this page is about. |
| End-stage renal disease (ESRD) | Older term for the same territory, still used in the United States and in insurance and registry documents. |
| Established renal failure | UK phrasing for the same thing, generally preferred over “end-stage”. |
| Acute kidney injury (AKI) | A sudden drop in function over hours to days. Often reversible. A different condition entirely. |
| Renal replacement therapy | Dialysis or transplantation. The treatments that replace what the kidneys did. |
One more distinction is worth having. Kidney failure describes a level of function. It does not describe a cause. Diabetes, high blood pressure, glomerulonephritis, polycystic kidney disease, obstruction, repeated injury and a long list of rarer conditions all end in the same place if they go far enough. The cause still matters, because some are treatable and some recur in a transplanted kidney, but by the time filtration is under 15 the management is broadly similar whatever the origin.
Acute kidney failure versus end-stage kidney disease
This is the distinction that changes everything, and it is invisible on a single blood result. Two people can walk into the same emergency department with a creatinine of 9 mg/dL. One will be back to normal in a fortnight. The other will never be. Telling them apart is the first job.
Acute kidney failure, more usually called acute kidney injury, is a rapid fall in function over hours or days. The kidney tissue is often structurally intact and being starved, poisoned, obstructed or inflamed. Remove the cause quickly enough and function frequently returns, sometimes completely. Common triggers are severe dehydration, blood loss, sepsis, a blocked urinary tract, contrast dye given during a scan, and certain drug combinations.
End-stage kidney disease is the destination of a slow process that has usually taken years. Filtering units have scarred and dropped out of service permanently. Nephrons do not regenerate. Nothing brings them back, and no treatment currently available reverses established scarring.
| Feature | Acute kidney failure | End-stage kidney disease |
|---|---|---|
| Time course | Hours to days | Months to decades |
| Previous results | Normal, often very recently | Progressively rising over years |
| Kidney size on ultrasound | Usually normal or enlarged | Usually small and scarred |
| Anaemia | Often absent early | Usually present, often for years |
| Bone chemistry | Usually near normal at first | Calcium, phosphate and PTH commonly disturbed |
| Symptoms | May be dramatic and sudden | Insidious, often tolerated for months |
| Reversibility | Often substantially reversible | Not reversible |
| Dialysis, if needed | Frequently temporary | Long term unless transplanted |
The single most valuable piece of information in that table is the second row. An old blood test from two years ago is worth more than almost any test that can be done on the day. If your creatinine was 0.9 last spring and is 6.5 now, that is acute, and the search is on for a cause that can be removed. If it was 2.1 three years ago and 3.8 eighteen months ago, the trajectory speaks for itself.
Anaemia and bone chemistry are the next best clues. Kidneys produce erythropoietin, the hormone that tells bone marrow to make red cells, and they activate vitamin D. Both of those functions decline slowly. So a person who has been in failure for a long time is usually anaemic, with a disturbed calcium, phosphate and parathyroid hormone picture, while someone whose kidneys stopped last Tuesday typically has a normal haemoglobin and near-normal bone chemistry. The blood tells you how long this has been going on.
There is a third possibility worth naming, because it catches people out. Acute-on-chronic failure is a sudden deterioration in someone whose kidneys were already damaged. It is extremely common. A person sitting stably at an eGFR of 22 gets a vomiting illness, becomes dehydrated, keeps taking a blood pressure tablet and an anti-inflammatory, and arrives in hospital with an eGFR of 8. Part of that drop is recoverable and part of it is not, and which part is which only becomes clear over the following weeks. Do not assume the worst on day one.
What kidney failure feels like
The symptoms of failure come mostly from uraemia, the accumulation of waste products that healthy kidneys would have removed, together with fluid overload, acid build-up and disturbed salts. What surprises people is how unspecific they are, and how late they arrive.
Plenty of people function at an eGFR of 12 and describe themselves as basically fine. That is not denial. Adaptation happens gradually enough that the loss of energy over two years gets attributed to age, work, or poor sleep. Only in retrospect, often after dialysis starts and they feel better, do people realise how unwell they had become.
Fatigue that sleep does not fix. Usually the earliest and most universal symptom. Driven partly by anaemia from lost erythropoietin production and partly by uraemia itself. People describe it as a flatness rather than sleepiness.
Loss of appetite and a metallic taste. Food goes off people in a specific pattern, with meat usually first. Weight loss follows quietly. This is one of the symptoms clinicians weigh most heavily when deciding about starting dialysis.
Nausea, especially in the morning. Common and often mistaken for a stomach problem for months before anyone connects it to kidney function.
Itching. Generalised, worse at night, often without any rash. Related to phosphate and to retained waste products. It can be genuinely miserable and it is treatable, so it is worth reporting rather than enduring.
Swelling of ankles, legs, hands or face. Fluid the kidneys cannot excrete. Shoes stop fitting. Rings tighten. Ankle swelling that leaves a dent when pressed is the classic sign.
Breathlessness. From fluid in the lungs, from anaemia, or both. Breathlessness lying flat, or waking at night gasping, is a more urgent version of this symptom.
Restless legs and cramps. An unbearable urge to move the legs at night, and cramping in the calves. Extremely common in advanced failure and often the symptom people complain about most.
Poor concentration and muddled thinking. Sometimes described by relatives before the person notices it themselves. Loss of sharpness, repeating questions, losing the thread.
Changes in urine. Less urine, or frothy urine from protein, or waking repeatedly at night to pass water. Not everyone in failure passes less urine, which is a point that confuses many people.
Easy bruising. Uraemia interferes with platelet function, so people bruise more readily and bleed for longer from small cuts.
The one that catches people out most is urine output. It is intuitive to assume failing kidneys stop producing urine, and in advanced failure output does often fall. But many people in stage 5 pass entirely normal volumes, because the kidneys are still making urine, just poorly filtered urine. Normal urine output is not reassurance that filtration is adequate. Conversely, a sudden and dramatic drop in output is a genuine warning sign and belongs in the next section.
Red flags: symptoms that need medical attention today
Most of what is described above develops slowly and is dealt with in clinic. The following are different. If any of these appear, get assessed the same day, and for the first four, go to an emergency department rather than waiting for a call back.
Go to an emergency department for these four.
Breathlessness at rest, or being unable to lie flat. Suggests fluid on the lungs. Sitting upright to breathe, or waking suddenly at night fighting for air, means the fluid has reached a dangerous level. This is treatable, and treated quickly it is usually straightforward.
Confusion, drowsiness, or a seizure. Severe uraemia can affect the brain. Confusion that is new, sleepiness that cannot be shaken off, twitching, or any seizure needs immediate assessment. Relatives usually spot this before the person does.
Palpitations with muscle weakness. Can indicate a dangerously high potassium level, which failing kidneys cannot excrete. High potassium disturbs heart rhythm and can do so without warning. It is invisible without a blood test, so symptoms are the only clue you get.
Passing very little or no urine. A sudden fall to a few hundred millilitres a day, or nothing at all, needs same-day assessment. It may mean an obstruction, which is often reversible if relieved promptly, or a sharp deterioration in function.
Get seen the same day for these four.
Vomiting that will not stop. Persistent vomiting both signals advanced uraemia and worsens it by causing dehydration. It is one of the more common reasons dialysis is started sooner than planned.
Chest pain, particularly worse lying down. Uraemia can inflame the lining around the heart. Pain that eases on sitting forward and worsens lying flat is the characteristic pattern, and it is an accepted reason to start dialysis urgently.
Rapid new swelling with weight gain. Several kilograms gained in a few days is fluid, not fat. It often precedes breathlessness by a day or two, which makes it a useful early warning if someone is weighing themselves regularly.
Any fever or infection in someone with advanced failure. Immunity is reduced in kidney failure and infections escalate faster. A temperature that would simply be watched at home in a healthy adult deserves a lower threshold for getting checked here.
Writing that list out is uncomfortable, and it is the part of this page most likely to alarm. So hold two things at once. These symptoms are the reason kidney failure is monitored so closely, and they are also the reason monitoring works. People under regular nephrology follow-up are usually treated well before any of this appears. The list is here so you recognise something if it happens, not because it is expected.
Where stage 5 sits in the staging system
Seeing the whole scale at once helps, because it shows how much distance sits between a first abnormal result and failure, and how uneven the stages are in width.
Normal filtration with evidence of damage, usually protein in the urine. Kidney function itself is intact.
Mildly reduced filtration plus a damage marker. Without that marker, this range is not classed as disease at all.
Moderate reduction, split into 3a (45–59) and 3b (30–44). The commonest stage by a wide margin, and most people here stay here.
Severe reduction. Specialist care, planning conversations, and preparation for possible replacement therapy begin in this band.
Filtration below 15, or on dialysis. This is the band this page is about. It can last months or many years.
Two features of that scale are worth noticing. Stage 3 is by far the widest band, covering a 30-point range, which is why it holds most people with kidney disease and why it is such a poor predictor on its own. Stage 5 is unbounded at the bottom, covering everything from 14 down to 3, which is why the single label tells you so little about how someone actually is.
Modern staging adds a second axis that many people never get told about: albuminuria, the amount of albumin leaking into the urine. Two people with the same eGFR but very different urine albumin levels have very different outlooks, and the one with heavy protein leak is at considerably higher risk of progressing. If you have been given an eGFR but never an albumin-to-creatinine ratio, that is a reasonable thing to ask for. How the albumin-to-creatinine ratio is calculated explains that measurement and the numbers that go with it.
There is also a validated tool called the kidney failure risk equation, which combines age, sex, eGFR and urine albumin to estimate the probability of needing dialysis or a transplant within two and five years. Many kidney clinics use it. It is far more informative than a stage label, and if your team has not mentioned it, it is worth asking whether your risk has been calculated. Predicting the future in medicine is always uncertain, but this particular estimate has held up well across different populations.
The three paths from here: dialysis, transplant, conservative care
When kidneys can no longer sustain life on their own, there are three options. All three are legitimate. The right one depends on age, other conditions, priorities and what a person actually wants their remaining years to look like, and a good kidney team will present all three rather than assuming.
Haemodialysis
Blood is pumped out of the body, through a filter that removes waste and excess fluid, and returned. In a hospital or satellite unit this is typically three sessions a week of about four hours each, plus travel. At home it can be done more often and for longer, which many people find suits them better physically, though it demands more of them and of the household.
Access is needed. The preferred route is a fistula, a surgically created join between an artery and a vein in the arm, made some months in advance so it has time to mature. The alternatives are a graft or a tunnelled neck line, both of which carry higher infection risk. This is the main reason nephrologists want to have the conversation early: a fistula created in good time is considerably better than a line inserted in a crisis.
What it is like varies. Some people feel washed out for hours after a session, particularly if a lot of fluid has been removed. Others read, work on a laptop or sleep through it and get on with their week. Fluid and dietary restrictions are usually part of the deal, often limiting potassium, phosphate, salt and fluid volume.
Peritoneal dialysis
The lining of the abdomen is used as the filter instead of a machine. Fluid is run into the abdominal cavity through a permanent soft catheter, left to draw waste across the membrane, then drained out. It is done at home, either by hand four or five times a day, or overnight by a machine while sleeping.
It suits people who value independence and flexibility, and it is often gentler on the circulation because the process is continuous rather than three sharp corrections a week. It also tends to preserve remaining natural kidney function for longer, which matters more than most people realise. The main drawbacks are peritonitis, an infection of the abdominal lining that needs prompt treatment, and the fact that the membrane’s performance declines over years, so many people eventually move to haemodialysis.
Transplantation
A working kidney from a living or deceased donor, placed usually in the lower abdomen with the original kidneys left in place. Where it is possible, it generally offers better survival and much better quality of life than long-term dialysis, and it is the only option that comes close to restoring what was lost.
It is not a cure, and describing it as one sets people up for disappointment. A transplant is a treatment that requires immunosuppressant medication for as long as it functions, which increases infection and some cancer risks, and it has a finite lifespan. Many transplants work well for well over a decade; some fail much sooner. Some people receive a second or third over a lifetime.
Not everyone is a candidate. Significant heart disease, active cancer, uncontrolled infection or frailty may make the surgery more dangerous than the alternative. Assessment is thorough and takes time. A transplant can also be done before dialysis ever starts, usually when filtration falls below about 20, which is called pre-emptive transplantation and tends to produce the best outcomes of all. That possibility alone is a strong argument for early referral.
Conservative kidney management
The third path is active treatment of everything except the filtration itself. No dialysis, no transplant. Anaemia is treated, blood pressure managed, itching and nausea and restless legs addressed, fluid controlled with diuretics where they still work, and the focus is on symptoms and quality of life rather than on the numbers.
It deserves to be described as a choice rather than as giving up, because that framing is both kinder and more accurate. For an older person with significant heart disease, frailty, or dementia, dialysis may add relatively little time while taking a substantial share of the time that remains: three days a week in a chair, plus travel, plus the recovery afterwards. Observational studies in this group have found survival differences smaller than most people assume, and better preserved independence in those who chose conservative care. That evidence is not from randomised trials and the comparison is difficult to make fairly, so it should be held loosely. But the choice is real, it is offered routinely in good kidney services, and nobody should feel they have failed by making it.
| Haemodialysis | Peritoneal dialysis | Transplant | Conservative care | |
|---|---|---|---|---|
| Where | Unit or home | Home | Hospital surgery, then clinic | Home and clinic |
| Typical commitment | 3 sessions weekly, ~4 hours | Daily, several exchanges or overnight | Daily tablets, regular clinics | Regular clinic review |
| Access needed | Fistula, graft or line | Abdominal catheter | Major surgery | None |
| Main risks | Line infection, blood pressure swings | Peritonitis, membrane failure over time | Rejection, infection, drug side effects | Symptoms of advancing uraemia |
| Diet and fluid limits | Usually significant | Usually less restrictive | Much freer | Individualised |
| Best suits | Those wanting supervised care | Those wanting flexibility and independence | Those medically fit for surgery | Those prioritising time at home |
Why people live in stage 5 for years before starting treatment
Reaching an eGFR of 14 does not mean dialysis next week. This is one of the most useful facts on this page, and one of the least widely known.
Dialysis in stable chronic kidney failure is started on symptoms, not on a threshold. Someone can sit at an eGFR of 12 for two or three years, feeling reasonably well, working, travelling, with nothing changing but the frequency of blood tests. Others deteriorate quickly and start within months. The variation is enormous and depends on the underlying cause, blood pressure control, protein leak, and a large amount of individual biology nobody can fully account for.
This approach is evidence-based rather than a matter of preference. A large randomised trial published in 2010, known as IDEAL, assigned people to start dialysis either early, at an eGFR around 10 to 14, or later, at around 5 to 7. Starting early produced no survival benefit. Most of the people assigned to the late group ended up starting sooner than planned because symptoms developed, which is itself informative: symptoms, not the number, drive the decision. That finding changed practice internationally and is why no responsible unit starts dialysis purely because a figure crossed a line.
Nausea and vomiting that will not settle. One of the most common triggers, because it makes eating and drinking difficult and undermines nutrition.
Fluid overload that diuretics can no longer control. When swelling and breathlessness persist despite maximum tolerated doses, dialysis becomes the only way to remove the fluid.
Potassium that keeps climbing dangerously. Recurrent high potassium despite dietary measures and medication is a hard indication, because the risk to the heart is immediate.
Acid build-up the body cannot compensate for. Persistent metabolic acidosis causes breathlessness, weakness and bone loss over time.
Malnutrition and unintended weight loss. Appetite failing and weight dropping is taken seriously, because nutritional state before starting dialysis strongly influences how well people do afterwards.
Confusion, twitching or pericarditis. The uraemic complications affecting brain and heart lining are urgent indications and are not waited out.
The practical implication is that the months or years spent in stage 5 before treatment are preparation time, not dead time. That period is used to create dialysis access, to complete transplant assessment and possibly find a living donor, to learn about the options properly, and to get affairs into whatever order a person wants them in. People who use that window well tend to have a considerably smoother transition than those who arrive at dialysis unexpectedly through an emergency admission. When dialysis is started and what creatinine level is involved goes into the decision in more depth.
It is also the window in which slowing progression still matters. Blood pressure control, treating the underlying cause, avoiding anti-inflammatory painkillers, staying well hydrated within whatever limits have been set, and attending appointments all continue to be worth doing at an eGFR of 13. The goal shifts from preventing failure to buying time, and time here is measured in months and years of ordinary life. Whether high creatinine can be cured takes an honest look at what can and cannot be reversed at each stage.
What creatinine does once someone is on dialysis
This confuses almost everyone, including relatives who have been tracking the number carefully for years and expect it to normalise once treatment starts. It does not.
Creatinine stays high on dialysis, permanently. A person on three-times-weekly haemodialysis typically has a pre-session creatinine somewhere between 6 and 14 mg/dL, roughly 530 to 1240 µmol/L. It falls substantially during a session, perhaps by half, then climbs back over the following two days. That sawtooth pattern is normal and expected. Nobody on dialysis has a normal creatinine, and nobody is trying to give them one.
The reason is straightforward. Dialysis three times a week for four hours provides roughly the equivalent of an eGFR in the range of 10 to 15, spread unevenly. Natural kidneys work continuously, twenty-four hours a day, at 100 or more. Intermittent treatment cannot reproduce that, and the aim was never to normalise blood chemistry but to keep it within a survivable and reasonably comfortable range.
Here is the part that genuinely surprises people. Among people established on dialysis, a higher creatinine is generally associated with better outcomes, not worse. It is a marker of muscle mass and good nutrition. Someone with a pre-dialysis creatinine of 11 is usually better nourished and stronger than someone at 4, and tends to do better. A falling creatinine in a dialysis patient often signals muscle loss and declining nutrition, which is a warning rather than an improvement. Relatives who cheer a dropping number have understandably read it backwards.
What is actually monitored instead. Dialysis adequacy is judged by measures such as Kt/V or the urea reduction ratio, which quantify how much clearance a session delivered. Alongside those, teams watch haemoglobin, potassium, phosphate, calcium, parathyroid hormone, albumin, fluid gains between sessions and blood pressure. Creatinine is recorded but is not the target. If you are following someone’s results, those are the numbers worth understanding.
Residual kidney function is another thing worth knowing about. Many people starting dialysis still make urine and still have some natural filtration, and preserving it is valuable: it makes fluid management easier, allows a more liberal diet, and is associated with better outcomes. It is one reason peritoneal dialysis is often favoured early on, and one reason nephrologists remain careful about anti-inflammatory drugs and contrast scans even after dialysis has begun.
Recovering from acute kidney failure, and what predicts it
If the failure is acute, the question everyone asks is whether the kidneys will come back. Often they do. Sometimes partially. Occasionally not at all, and it takes weeks to know which.
The rough shape of it: a large proportion of people who develop acute kidney injury severe enough to need dialysis in hospital recover enough function to come off it, usually within days to a few weeks. A minority do not, and go on to long-term dialysis. Precise figures vary a great deal by cause, by setting and by how sick people were to begin with, so any single percentage quoted at you should be treated with suspicion.
| Factor | Better recovery likely | Poorer recovery likely |
|---|---|---|
| Kidney function beforehand | Normal baseline | Pre-existing chronic kidney disease |
| Cause | Dehydration, obstruction relieved quickly, a single drug | Prolonged sepsis, severe inflammation, multiple insults |
| Speed of treatment | Cause removed within hours to a day or two | Days of untreated low blood flow or obstruction |
| Urine output | Maintained throughout | Little or no urine for a prolonged period |
| Age and general health | Younger, fewer other conditions | Older, frail, heart or liver disease |
| Duration of dialysis need | Days to a couple of weeks | Beyond about six to eight weeks |
| Number of episodes | First event | Repeated episodes over time |
Recovery is rarely a clean line. Creatinine tends to plateau, wobble, then fall over days to weeks, and urine output often returns before the blood tests improve. There can be a phase where large volumes of urine are passed while filtration is still poor, which needs careful fluid and salt replacement. Patience is genuinely required, and it is normal for the team to decline to give a firm answer in the first fortnight, because they do not have one.
One thing deserves saying that often goes unsaid. Even when someone recovers fully from acute kidney injury, the episode raises their long-term risk of chronic kidney disease. Function may return to what looks like baseline while reserve has been permanently reduced. That is why follow-up blood tests months after an admission are arranged, and why they are worth attending even when everything feels fine. Protecting kidney function over the long term covers what actually helps, and the NIDDK guidance on kidney disease testing explains the tests used to follow it up.
Supporting someone facing kidney failure
A large share of the people reading this are not the patient. They are the daughter who looked up the number after her father’s appointment, or the partner sitting in the car park working out what to say. This section is for you.
People retain very little of what they are told in a consultation that frightens them. A second set of ears and a notebook is one of the most useful things you can offer, and it is entirely normal to ask the clinician to slow down or repeat something.
Some people want survival statistics and the mechanism of every treatment. Others want to know when the next appointment is and nothing else. Both are legitimate ways of coping. Do not force information on someone who has not asked for it.
Kidney failure comes with appointments, transport, prescriptions, dietary limits and fatigue. Cooking a meal that fits the restrictions, or driving to a dialysis session, is worth more than most conversations about feelings.
Knowing that new breathlessness or confusion needs same-day attention, rather than waiting until Monday, is genuinely protective. Relatives notice confusion long before the person does.
If someone chooses conservative care, or declines a transplant assessment, that is theirs to decide. Ask what led them there, make sure they have accurate information, then support the decision. Pressure damages relationships at exactly the point they matter most.
If you are considering offering a kidney, the assessment process is designed to protect donors and includes independent counselling. There is no obligation, and no shame in deciding against it after looking into it.
Carer exhaustion is common and it creeps up. Most kidney units have social workers and counsellors who support families, not just patients. Using them is not an imposition; it is what they are there for.
One thing to avoid, gently: sending articles about cures. There is a large industry of supplements, herbal preparations and diets promising to reverse kidney failure, and none of them do. Some are actively harmful in advanced disease, particularly anything high in potassium or phosphate, or any herbal product with unknown content. Someone with an eGFR of 9 who spends their remaining energy on an internet protocol instead of preparing for dialysis has been badly served. What genuinely lowers creatinine, and what does not sets out the difference honestly.
What to ask the kidney team
Consultations are short and it is hard to think clearly in them. These are the questions that tend to produce the most useful answers, and taking a written list in is normal.
About where things stand
What is my eGFR, and what was it a year ago? Is this acute, chronic, or both? Do you know the cause? How much albumin is in my urine? Has my kidney failure risk been calculated?
About what happens next
How fast is this progressing? Is anything still likely to slow it? Roughly when might I need replacement therapy? What would make you bring that forward?
About treatment options
Am I a candidate for a transplant, and can it happen before dialysis? Which type of dialysis suits my circumstances? Can it be done at home? Is conservative management something we should discuss?
About day-to-day life
Which of my medicines need changing or stopping? What should I do if I get a vomiting illness? Are there dietary limits now? Can I still work, drive, travel, exercise?
About symptoms
Which symptoms should bring me in urgently? Can anything be done about the itching, the restless legs, the fatigue? Is my anaemia being treated?
About support
Is there a specialist nurse I can contact between appointments? Is there a social worker or counsellor? Are there patient education sessions, and can my family attend?
Two questions are worth asking even if they feel awkward. The first is what the realistic outlook is, if you want to know, and it is fine to say you would rather not. The second is whether anything about your case is uncertain, because clinicians are often more candid about uncertainty when directly invited to be. The National Kidney Foundation’s guide to eGFR is a useful thing to read before an appointment so the vocabulary is familiar.
Enter age, sex, weight and creatinine into the Waldev creatinine clearance calculator to see the estimated filtration behind your result. It sits alongside the rest of the Waldev health calculators.
Ways this number gets misread
Some of these come up so often that naming them is probably the most practically useful part of this page.
Assuming a specific creatinine equals failure. There is no such number. Someone quotes 5 mg/dL, or 10, and it spreads. Filtration defines failure, and filtration depends on the body the creatinine came from.
Comparing units without noticing. A result of 450 in µmol/L is a little over 5 mg/dL. Someone reading American figures with a British result, or the reverse, can frighten themselves badly over nothing, or reassure themselves wrongly.
Reading one result without the history. A single value is a snapshot. Whether it arrived over three days or three years changes the diagnosis, the treatment and the outlook entirely.
Treating normal urine output as reassurance. Plenty of people in stage 5 pass entirely normal volumes. Making urine and filtering it properly are different things.
Expecting creatinine to normalise on dialysis. It stays high, and among dialysis patients a higher value usually reflects better muscle mass and better nutrition.
Confusing stage 5 with imminent dialysis. Many people spend years in stage 5 before starting, and some never start by choice. The threshold is a category boundary, not an appointment date.
Mistaking a temporary rise for failure. Dehydration, a new medication, contrast dye or a bout of illness can push creatinine dramatically high and then let it fall back. A repeat test after rehydration resolves a surprising number of frightening results. Dehydration’s effect on creatinine covers this.
Ignoring a small rise in an elderly person. The opposite error, and the more dangerous one. A move from 1.1 to 1.9 in an 84-year-old can represent a large loss of filtration, because so little muscle is producing the creatinine in the first place.
Believing failure means immediate death. It does not. It means the kidneys can no longer sustain the body unaided, and there are established treatments for exactly that situation, used by hundreds of thousands of people who are getting on with their lives.
If you want to work through the wider interpretation of a raised result, what high creatinine means and when a creatinine level is genuinely worth worrying about take it from the beginning, and the normal creatinine range gives the reference figures. For the related blood measure often quoted alongside it, see the BUN-to-creatinine ratio. If you are trying to interpret a full report, creatinine on a blood test and the normal creatinine clearance range explain the surrounding numbers.
What creatinine level indicates kidney failure: frequently asked questions
What creatinine level indicates kidney failure?
No single level does. Kidney failure is defined as an estimated filtration rate below 15 mL/min/1.73m², or being on dialysis, rather than by a creatinine threshold. Most adults in failure have a creatinine above 5 mg/dL, around 440 µmol/L, and values of 8 to 12 are common. But muscle mass shifts the relationship enormously: a frail elderly woman can be in failure at 3.5, while a heavily muscled young man may reach 4.0 with filtration around 25, which is stage 4. The eGFR, not the creatinine, makes the diagnosis.
Is a creatinine of 5 kidney failure?
Often, but not always. In an average-sized older adult, a creatinine of 5 mg/dL usually corresponds to an eGFR below 15, which meets the definition of kidney failure. In a young, muscular man it may correspond to an eGFR in the high teens or low twenties, which is stage 4. The only way to know is to calculate the eGFR using age and sex, and to look at previous results, because a level of 5 that appeared in a week means something entirely different from one reached gradually over years.
Can kidney failure be reversed?
It depends which kind. Acute kidney failure, developing over hours or days, is frequently reversible when the cause is removed quickly, and many people recover enough function to come off temporary dialysis. End-stage kidney disease, the result of years of gradual scarring, is not reversible with any treatment currently available, because filtering units do not regenerate. What can be done is slowing further decline, treating the consequences, and replacing the lost function with dialysis or a transplant. Be sceptical of anything promising reversal of established damage.
How long can you live with kidney failure without dialysis?
Longer than most people expect, and it varies enormously. Many people remain in stage 5 for months or years before needing treatment, feeling reasonably well while their eGFR sits in single figures. When someone chooses conservative management rather than dialysis, survival is often measured in months to a few years, depending strongly on age, other conditions and how much function remains. Nobody can give an accurate individual figure, and a nephrologist who knows the whole picture is the only person worth asking.
What are the first signs that kidneys are failing?
Usually fatigue that sleep does not fix, then loss of appetite, nausea, a metallic taste, and disturbed sleep. Itching, restless legs, cramps and swelling of the ankles follow. The difficulty is that all of these are unspecific and develop slowly enough to be blamed on age or stress, so kidney failure is often well advanced before anyone suspects it. Breathlessness, confusion, very reduced urine output, persistent vomiting or palpitations with weakness are different: they need same-day medical assessment.
Does kidney failure always mean dialysis?
No. Reaching stage 5 does not commit anyone to dialysis, and it does not mean starting immediately. Some people receive a transplant before dialysis ever begins, which is called pre-emptive transplantation and generally produces the best outcomes. Others choose conservative kidney management, treating symptoms and quality of life without replacement therapy, which is a legitimate choice often made by older people with several other conditions. And many who will eventually need dialysis spend a year or more in stage 5 before symptoms make it necessary.
Why is creatinine still high on dialysis?
Because dialysis replaces only a fraction of what kidneys do. Three sessions a week deliver roughly the equivalent of an eGFR of 10 to 15, spread unevenly, against the 100 or more that healthy kidneys provide continuously. Creatinine falls sharply during a session and climbs back between them, so pre-session values of 6 to 14 mg/dL are entirely normal. Among dialysis patients a higher creatinine usually reflects better muscle mass and nutrition and is associated with better outcomes, so a falling value can be a warning rather than progress.
What is the difference between kidney disease and kidney failure?
Kidney disease covers all five stages, from normal filtration with a damage marker down to failure. Kidney failure is only stage 5, an eGFR below 15. Most people with chronic kidney disease are in stages 1 to 3 and will never progress to failure; many die of unrelated causes with their kidneys still doing their job. Being told you have chronic kidney disease is not being told your kidneys are failing. If nobody has given you a stage and an eGFR number, ask for both.
What creatinine level requires hospital admission?
Symptoms and the rate of change decide this, not the level. Someone stable at 7 mg/dL under regular clinic review may need nothing urgent, while someone whose creatinine has jumped from 1.0 to 4.0 in three days needs admission regardless of the absolute figure. Breathlessness at rest, confusion, drowsiness, chest pain worse lying flat, persistent vomiting, palpitations with muscle weakness, or a sudden drop in urine output all warrant emergency assessment whatever the number happens to be.
Can you feel fine with kidney failure?
Yes, and many people do, which is one reason kidney disease is diagnosed so late. Adaptation happens gradually, so declining energy over two years gets attributed to ageing or work. People at an eGFR of 12 often continue working and describe themselves as well. Feeling fine is genuinely good news, and it is also not evidence that treatment can be delayed indefinitely, because complications such as high potassium and fluid overload can develop without warning. Regular monitoring is what bridges that gap.
The short version
Kidney failure means an estimated filtration rate below 15, or being on dialysis. It is not defined by a creatinine number, because creatinine comes from muscle and the same value means different things in different bodies. Most adults in failure sit above 5 mg/dL, roughly 440 µmol/L, often much higher, but a frail older person can be in failure well below that and a very muscular young person can be above it without being there. Acute failure, developing over days, is often reversible; end-stage disease, built over years, is not.
From stage 5 there are three honest paths: dialysis, transplantation, and conservative care focused on symptoms and quality of life. Many people spend years in stage 5 before starting any of them, because the decision is led by symptoms rather than by a threshold. Get red flags checked the same day: breathlessness at rest, confusion, very low urine output, persistent vomiting, palpitations with weakness. Put your own value in context with the CrCl calculator, and read more across the creatinine blog category, the wider health blog, and the full tool library at waldev.com.
Medical disclaimer: This article is general educational information about kidney function tests and cannot tell you whether you or anyone else has kidney failure. It is not medical advice, and it must not be used to decide whether to seek care, delay care, start or stop dialysis, or change any treatment or medication. Reference ranges differ between laboratories, and results only mean something alongside your history, medications, symptoms and other tests. Always discuss your own results with a doctor or qualified healthcare professional, and seek urgent medical attention for breathlessness at rest, confusion or drowsiness, chest pain, persistent vomiting, palpitations with muscle weakness, or a sudden fall in urine output.
The National Kidney Foundation on eGFR, what the ranges mean, and how the five stages are defined. Estimated GFR explained →
NIDDK on the blood and urine tests used to assess kidney function and follow it over time. CKD tests & diagnosis →
MedlinePlus explains what the creatinine test measures, why it is ordered and how results are read. Creatinine test →