Industries / Health / Healthcare / Dialysis
Machines and water
This is the work of building dialysis machines and the filters they use, called dialyzers, and of turning tap water into the fluid that draws waste out of a patient's blood.
7
Milestones achieved, from 1968 to 2024.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1968
Blood through hollow fibers
Early dialyzers ran blood between flat plates or through coils of cellulose membrane. The hollow-fiber dialyzer runs it through a bundle of hair-thin hollow fibers instead, with cleaning fluid flowing around them. By 1968 Richard Stewart's team was reporting treatments with this "capillary kidney". The Kidney Project at the University of California, San Francisco, dates Stewart's design and Dow Chemical's first cartridges to 1964 to 1966, and says the format is still in use.
pharm.ucsf.edupmc.ncbi.nlm.nih.govebi.ac.ukcollections.nlm.nih.gov
Credit
Richard Stewart, a professor at the Medical College of Wisconsin from 1966, who held a patent on the hollow-fiber artificial kidney; Dow Chemical
1976
Aluminum named as poison
Doctors in Denver described a fatal brain disorder in 12 long-term hemodialysis patients; it was then the leading cause of death in one dialysis unit there. A. C. Alfrey and colleagues measured 25 ppm of aluminum in the brain gray matter of patients who died of it, against 6.5 ppm in dialysis patients who died of other causes and 2.2 ppm in control subjects. Later work by D. N. Kerr's group found two routes: heavily contaminated dialysis fluid poisoned patients within weeks or months, and aluminum gels taken by mouth to control blood phosphate did so over years.
LessonA patient on standard hemodialysis may meet more than 360 liters of dialysis fluid a week, by Lucena and colleagues' count, so traces that are harmless in drinking water can build up in the body.
Credit
A. C. Alfrey, G. R. LeGendre and W. D. Kaehny; J. S. Burks and colleagues; D. N. Kerr and colleagues
1996
Algae toxins at Caruaru
In February 1996, 116 of 130 patients at one hemodialysis center in Caruaru, Brazil, fell ill after treatment. Both of the town's centers drew water from the same reservoir, but only the other center's supply was treated, filtered and chlorinated. Investigators found microcystins, liver toxins made by cyanobacteria (blue-green algae), in the reservoir, in the center's water and in patients' blood and liver. Sources differ on the toll: the 1998 report in the New England Journal of Medicine counts 101 cases and 50 deaths, and a 2002 review attributes 52 deaths by December 1996. Brazil added cyanobacteria and their toxins to its drinking-water rules in 2000.
LessonIn the 1998 authors' words, water used for hemodialysis "can contain toxic materials, and its quality should therefore be carefully monitored.".
Credit
E. M. Jochimsen, W. W. Carmichael, S. M. Azevedo and colleagues
2001
Test fluid left in filters
In late 2001 clusters of deaths after hemodialysis were traced to Baxter's A and AF series dialyzers, made at a plant in Ronneby, Sweden, that Baxter had bought with Althin Medical in 2000. Baxter first linked at least 33 deaths in Spain, Croatia and Texas; a week later HospiMedica reported 53 deaths under FDA review in seven countries, 21 of them in Croatia. A perfluorocarbon liquid used to test the fibers for leaks during manufacture had stayed in some filters and entered patients' blood. Baxter recalled the filters and stopped making them. In 2005 a team injected the liquid into rabbits and reproduced the gas bubbles in the heart that had been found in the patients.
LessonThe largest cluster of dialysis-related deaths came from new, single-use filters, not reused ones, as Z. J. Twardowski noted in 2006, so a filter must be checked for what its own manufacture leaves in it.
Credit
Baxter International (recall); B. Canaud and colleagues (animal confirmation)
2008
Japan sets fluid purity standard
Japan routinely uses central systems, unique to Japan, that mix dialysis fluid in one place and pipe it to every bedside machine. In 2008 the Japanese Society for Dialysis Therapy (JSDT) set a standard for the microbial quality of dialysis fluid, written for those systems. JSDT authors report that fluid quality then "markedly improved" across the country and that the number of patients on online hemodiafiltration, a treatment that infuses fluid made by the machine, "dramatically increased".
LessonThe 2008 standard covered microbes only, not toxic chemicals, and no standard yet treated water treatment equipment as a medical device, the authors wrote in 2017.
Credit
Japanese Society for Dialysis Therapy; Japan's clinical engineers, a profession created in 1987, who run the central systems
2016
Wearable kidney trial stopped
A wearable artificial kidney is a small dialysis machine worn on the body. Instead of using fresh water, it cleans a small volume of fluid over and over with sorbents, materials that bind waste. An FDA-approved trial planned to test it on up to 10 people for 24 hours each; seven enrolled and five finished, with no serious adverse events and an average urea clearance of 17 ml/min. The trial was stopped after the seventh person because of device problems, among them excess carbon dioxide bubbles in the fluid circuit and unsteady blood and fluid flows.
LessonA machine that reuses its fluid must handle on its own the gas and flow problems that a clinic machine never meets.
Credit
Victor Gura and colleagues, with a University of Washington Kidney Research Institute team that included Jonathan Himmelfarb; funded by the Wearable Artificial Kidney Foundation and Blood Purification Technologies
2024
US clears an HDF machine
Hemodiafiltration (HDF; see Dialysis in clinics) needs large volumes of sterile replacement fluid, which the machine makes from clinic water. On 2 February 2024 the FDA cleared Fresenius Medical Care's 5008X system with HDF among its treatment options, for patients over 40 kg. A filter in the machine makes ultrapure dialysis fluid and replacement fluid "as defined in ISO 23500-5", part of the international standard series for dialysis fluids published in 2019.
Credit
Fresenius Medical Care; US Food and Drug Administration
Watch for
A 13 kg sorbent machine
The Neokidney is a portable hemodialysis machine of about 13 kg that cleans 4.5 liters of fluid in a closed loop with sorbents and needs no water supply or drain. In the first two phases of the RENEW-HD trial at four European centers, 29 people completed 474 two-hour sessions by 31 August 2026, with average urea removal above the trial's target and no serious adverse device effects. The results so far are a conference presentation, and the trial is funded by the maker.
Achieved whenthe full results are published in a peer-reviewed journal and a regulator clears the machine for home use.
Claimed by
Karin Gerritsen, University Medical Center Utrecht, and Nextkidney, October 2026
Open
01
Chlorine tests side by side
An independent bench test of the tests that clinics use to check that their carbon filters have removed chlorine and chloramine, a chlorine compound many cities add to tap water. It would cover test strips, color kits based on DPD (a dye that turns pink with chlorine), handheld photometers and continuous monitors, all read on water held at known levels between 0 and 0.3 mg/L, around the 0.1 mg/L total chlorine line in the US Medicare rule.
The briefCredit · unclaimed02
Dialysis water rules compared
A public table of the legal or official limits for dialysis water in at least 30 countries: which standard each cites and which edition, limits for chemicals such as aluminum, chloramine and fluoride, microbial and endotoxin limits, how often each must be tested, and whether the rule is law or advice, with every cell linked to the document it came from.
The briefCredit · unclaimed
Open 01 · the brief
Chlorine tests side by side
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An independent bench test of the tests that clinics use to check that their carbon filters have removed chlorine and chloramine, a chlorine compound many cities add to tap water. It would cover test strips, color kits based on DPD (a dye that turns pink with chlorine), handheld photometers and continuous monitors, all read on water held at known levels between 0 and 0.3 mg/L, around the 0.1 mg/L total chlorine line in the US Medicare rule.
Why it is still open
The Medicare rule requires two carbon tanks in a row and more testing once water leaving the first tank reaches 0.1 mg/L total chlorine. The FDA's 510(k) database lists 12 cleared chlorine test strips, kits and meters under dialysis product codes. One maker's leaflet says its rinse-water strips reliably detect chlorine only down to about 0.25 mg/L and that strip readings depend on color perception and lighting. A Beijing hospital study compared just two instruments, and searches found no independent head-to-head test of strips against photometers at the 0.1 mg/L line. Lapses still harm patients: in 2023 doctors in Guadalajara, Mexico, reported eight patients with chlorine poisoning on a day their unit's water held 0.78 mg/L.
law.cornell.eduapi.fda.govhach.compmc.ncbi.nlm.nih.goveutils.ncbi.nlm.nih.gov
Built with what exists
chloraminated city tap water diluted with distilled water to set levels; a laboratory DPD photometer as the reference; test strips, kits and meters bought from dialysis suppliers; a light box with three lighting settings; volunteer readers who do not know the true level.
First step
Buy three strip brands and one photometer, and compare their readings on diluted tap water at 0.05, 0.1 and 0.2 mg/L.
Done looks like
at least 10 products tested at six or more set levels, each read by at least five people under three lighting conditions, with each product's share of wrong calls at the 0.1 mg/L line and all raw readings published.
Open 02 · the brief
Dialysis water rules compared
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of the legal or official limits for dialysis water in at least 30 countries: which standard each cites and which edition, limits for chemicals such as aluminum, chloramine and fluoride, microbial and endotoxin limits, how often each must be tested, and whether the rule is law or advice, with every cell linked to the document it came from.
Why it is still open
Standards differ by country and by edition. US Medicare still cites AAMI RD52 from 2004, while Fresenius-linked authors wrote in 2026 that international standards "have largely converged" and US references lag. Japan's 2008 standard covered microbes but not toxic chemicals. A PubMed search found reviews of water saving and of single countries' water quality, but no cross-country table of the rules themselves.
law.cornell.edueutils.ncbi.nlm.nih.goveutils.ncbi.nlm.nih.gov
Built with what exists
national health ministry and regulator websites; the ISO 23500 series and AAMI documents as reference columns; machine translation; a language model such as Claude to draft table entries from each rule for a person to check against the source.
First step
Fill the table for five countries (the US, Japan, Brazil, Germany and India) and fix the list of columns.
Done looks like
at least 30 countries, including the ten with the most dialysis patients by a cited registry count; every limit linked to a page of its source; a second person re-checking a random 20% of cells and agreeing on at least 95%.