A kidney works 168 hours a week. Dialysis gets about twelve, and that one number explains nearly every rule a patient is given

Dialysis is usually introduced to families as an artificial kidney. That phrase does a lot of damage, because it sets an expectation the machine cannot meet and makes every restriction that follows sound arbitrary — why so strict about water, why the potassium warnings, why tablets with every meal if the machine is cleaning the blood anyway.

A more accurate description is that dialysis is a very good filter, run intermittently, that replaces two of the kidney's jobs and none of the others. Almost everything a dialysis patient is told to do follows from that sentence, and from one number: a working kidney runs 168 hours a week, and a standard dialysis schedule is about twelve.

What the machine is actually doing

Blood is drawn from the patient, pumped through a dialyser — a cylinder packed with thousands of hollow fibres made of a semipermeable membrane — and returned. Around the outside of those fibres flows dialysate, a precisely mixed fluid, running in the opposite direction to the blood so that a concentration difference is maintained along the whole length.

Two separate physical processes then do the work, and they are worth keeping apart because they fail in different ways.

Diffusion removes solutes. Urea, creatinine and potassium are far more concentrated in the blood than in the dialysate, so they cross the membrane down their gradient. Nothing pushes them; they simply spread. The dialysate deliberately contains substances the patient should keep — a set concentration of sodium, bicarbonate to correct acidity, sometimes calcium — so those do not wash out. The membrane's pore size decides what can cross at all: small waste molecules pass easily, and useful large proteins such as albumin are retained.

Ultrafiltration removes water. Diffusion does not shift much fluid, so the machine applies a pressure difference across the membrane and pushes plasma water through it. This is the part that decides how a patient feels afterwards. Between sessions the body accumulates every glass of water drunk and there is no urine to remove it, so a fixed volume must be taken off in four hours. Pull it too fast and blood pressure crashes, with cramps, nausea and the exhaustion that many patients describe for the rest of the day.

This is the real reason for fluid restriction. It is not fussiness about thirst. Every extra litre gained between sessions has to be removed faster within the same four hours, and the speed of removal is what makes the session punishing.

Why some things clear and others do not

Not all wastes behave alike, and the exceptions explain the prescriptions.

Urea and creatinine are small and distribute in body water. They clear well, and they are what dialysis adequacy is measured on — which is convenient rather than complete, since they are markers rather than the toxins that matter most.

Potassium clears well too, and quickly. But it is largely inside cells, so the blood level falls during the session and then rebounds afterwards as potassium moves back out of the cells. This is why a dietary indiscretion between sessions is genuinely dangerous: a high potassium level disturbs heart rhythm, and the machine is not there for the intervening two days.

Phosphate is the instructive one. It is a small molecule and ought to clear easily, and it does not, because the great majority of it sits inside cells and in bone and moves into the blood far too slowly to be caught in a four-hour window. As soon as dialysis lowers the blood level, more diffuses out of the tissues and the level climbs back. Longer or more frequent sessions help; four hours three times a week does not. This is precisely why patients are prescribed phosphate binders with meals — tablets that trap phosphate in the gut so it is never absorbed. Stopping them because "the machine cleans it" is a common and consequential mistake, because sustained high phosphate drives bone disease and vascular calcification.

Middle molecules, such as β2-microglobulin, are larger and cross conventional membranes poorly. Their accumulation over years causes dialysis-related amyloidosis, with joint and tendon problems. High-flux membranes and haemodiafiltration — which adds convective transport, dragging larger solutes along with a flow of fluid rather than waiting for them to diffuse — clear them better, and this is an active area of membrane engineering rather than a solved problem.

Dialysis is not a small kidney running continuously. It is a large clearance delivered in bursts, and a body that was designed for continuous regulation has to absorb the difference in between.

What dialysis does not replace at all

The kidney is not only a filter. It is also an endocrine organ, and dialysis does nothing for that side of the job.

It makes erythropoietin, the hormone that tells bone marrow to produce red cells. Without it patients become anaemic, which is why erythropoiesis-stimulating injections and iron are part of standard care. It performs the final activation step of vitamin D, without which calcium absorption and bone metabolism go wrong — the other half of the bone disease that phosphate drives. It participates in blood pressure regulation, and it fine-tunes acid–base balance continuously in a way a thrice-weekly bicarbonate correction only approximates.

So a dialysis prescription is never just dialysis. The injections, the binders, the vitamin D analogues and the blood pressure medicines are not add-ons; they are the parts of kidney function the membrane cannot do.

The weak point is the access

The most common reason a dialysis patient ends up in hospital has nothing to do with the membrane. Getting several hundred millilitres of blood per minute out of a person and back, three times a week for years, requires a durable access point, and this is the fragile link in the whole system.

An arteriovenous fistula — a surgically created connection between an artery and a vein in the arm, which over weeks makes the vein thick-walled and high-flow — is the best option, with the lowest infection rate and the longest life. It has to be created months in advance and needs time to mature, which means the single most useful thing in a patient's dialysis career often happens before dialysis starts. A central venous catheter can be used immediately, and is the usual fallback when someone arrives already in crisis, but it carries a substantially higher risk of bloodstream infection and tends to fail sooner. A great deal of avoidable harm traces back to late referral, where there was no time to plan an access.

The other membrane: the patient's own

Haemodialysis is not the only route. Peritoneal dialysis uses the peritoneum, the body's own membrane lining the abdominal cavity, as the exchange surface. Dialysate is run into the abdomen through a permanent catheter, left to dwell while wastes diffuse across into it, and drained out — several times a day, or overnight by machine.

The trade-offs are real and cut both ways. It is done at home, needs no dialysis centre and no needles, and its gentler, more continuous rhythm means less of the crash-and-rebound pattern, better preservation of any remaining kidney function, and fewer dietary restrictions. Against that, it demands scrupulous sterile technique because peritonitis is the main complication, it requires storage space and reliable supply delivery, and the peritoneum's transport characteristics change over years so it is not always a permanent solution.

For a country where a large share of patients live hours from the nearest dialysis centre and travel three times a week to reach it, the case for peritoneal dialysis is more than clinical. Uptake in India remains low relative to that logic, which is itself worth studying.

Why it matters for students and researchers

India carries an enormous chronic kidney disease burden, with estimates of new patients reaching end-stage disease each year running to roughly two lakh, and the national dialysis programme launched in 2016 has expanded access substantially in district hospitals. That expansion makes the practical, unglamorous questions more important, not less.

Several sit squarely in membrane science and engineering. Membrane materials that clear middle molecules without losing albumin are an open design problem. So is biocompatibility — a membrane is a large foreign surface in contact with blood, and the inflammatory response it provokes has long-term consequences. Water is another: a single haemodialysis session consumes on the order of a hundred litres or more of highly purified water, which makes reverse-osmosis plant design, monitoring for endotoxin contamination and water reuse genuine engineering and public-health issues wherever dialysis is scaled up. And the health-systems questions — why patients discontinue treatment, what distance and cost do to adherence, why peritoneal dialysis is under-used — are answerable with careful local data and largely unanswered.

That span, from membrane materials and characterisation to membrane processes and applications, is the scope of the International Journal of Membranes (ISSN 3049-4427), a peer-reviewed journal launched in 2024. For medical, biomedical engineering and materials students, dialysis is a good demonstration that a membrane is not passive equipment: its pore distribution, its surface chemistry and its interaction with blood decide what a patient's next ten years look like.

Frequently asked questions

How does a dialysis machine clean blood?

Blood flows through thousands of hollow semipermeable fibres while dialysate flows around them in the opposite direction. Wastes such as urea and potassium cross the membrane by diffusion down their concentration gradient, while excess water is removed separately by applying pressure across the membrane.

Why are dialysis patients told to restrict fluids?

Because there is no urine to remove what is drunk between sessions, so all of it must be taken off during the next four-hour session. The more fluid gained, the faster it must be removed, and rapid removal causes low blood pressure, cramps and prolonged exhaustion.

If dialysis cleans the blood, why are phosphate binders needed?

Because most phosphate is inside cells and bone and moves into the blood too slowly to be removed in a four-hour session — the level simply rebounds afterwards. Binders taken with meals trap phosphate in the gut so it is never absorbed.

What does dialysis not do?

It does not replace the kidney's hormonal functions. It cannot produce erythropoietin, so anaemia must be treated separately, and it cannot activate vitamin D, so bone and mineral metabolism need separate management, alongside blood pressure control.

What is the difference between a fistula and a catheter?

A fistula is a surgically created artery-to-vein connection in the arm that takes weeks to mature but lasts far longer and carries much less infection risk. A catheter can be used immediately but has a substantially higher rate of bloodstream infection and shorter useful life.

What is peritoneal dialysis?

A form of dialysis that uses the patient's own peritoneal membrane instead of a machine's. Fluid is instilled into the abdomen through a catheter, wastes diffuse into it, and it is drained and replaced several times a day at home. Its main complication is peritonitis, so sterile technique is essential.