Industries / Health / Healthcare / Plasma and blood centers
Collection
Collection is the act of taking blood from a donor, either whole into a bag or through a machine that keeps one part, such as plasma or platelets, and gives the rest back.
7
Milestones achieved, from 1914 to 2021.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1914
Plasma out, cells back
J. J. Abel, L. G. Rowntree and B. B. Turner published "Plasma removal with return of corpuscles (plasmaphaeresis)", the method plasma centers still use: take blood, keep the liquid part, called plasma, and return the cells. The AABB, the US standards body for blood banks, dates the use of plasmapheresis to collect plasma for making medicines to 1964.
Credit
J. J. Abel, L. G. Rowntree and B. B. Turner
1950
Plastic bags replace glass
Carl Walter and W. P. Murphy Jr. introduced the plastic blood bag. It replaced breakable glass bottles and, by the AABB's account, made it possible to split one donation into several components, such as red cells, plasma and platelets.
Credit
Carl Walter and W. P. Murphy Jr
1965
A machine separates blood
IBM engineer George Judson and Emil Freireich of the US National Cancer Institute built a blood cell separator that spun a person's blood in a small centrifuge, kept one part and returned the rest in a continuous flow. Freireich needed large amounts of platelets, the cell fragments that stop bleeding, for leukaemia patients, and the National Inventors Hall of Fame notes that such separators are now used by blood banks, hospitals and laboratories worldwide.
Credit
George Judson (IBM) and Emil Freireich (NCI)
1990s
Pooled red cells spread HIV
In the early 1990s the Chinese province of Henan allowed paid plasma donation at commercial stations that took whole blood, removed the plasma and returned pooled red cells from donors of the same blood type, a practice the study authors call unsanitary. Commercial plasma donation was made illegal in 1996, and a 2010 study counted 37,084 people identified as infected with HIV through plasma donation, about 95% of them in parts of Henan and neighbouring provinces.
Lessonthe cells returned to a donor must be that donor's own, kept apart from everyone else's.
Credit
Zhihui Dou, Fujie Zhang and colleagues (the 2010 study); the Chinese government (the 1996 ban)
2008
Shared words for donor harm
The International Society of Blood Transfusion (ISBT) issued a standard for recording complications in blood donors, so that blood services could count and compare them the same way, and a revised version was endorsed by the ISBT, the AABB and the International Haemovigilance Network (IHN). The IHN's ISTARE database gathered 594,477 donor reactions from 25 countries between 2006 and 2016, but the IHN says the data were limited by inconsistent use of the definitions.
Lessona shared definition helps only if every country applies it the same way.
Credit
ISBT Haemovigilance Working Party (M. Goldman and colleagues); IHN
2021
Plasma volume set per donor
Since 1992 US plasma centers had used an FDA table, called a nomogram, that set one of three volumes, 625, 750 or 800 mL, by donor weight. In the IMPACT trial, with 3,443 donors and 23,137 donations at three US centers, a rule aiming for 28.5% of each donor's estimated plasma volume collected 841.7 mL per donation on average against 777.8 mL, 8.2% more, and was no worse for serious drops in blood pressure. The lead authors work for Haemonetics, which makes the machines, and in May 2024 the FDA also cleared Terumo BCT's Rika machine with its own per-donor nomogram.
Credit
Haemonetics (Jan Hartmann and colleagues) with Susan F. Leitman of the NIH Clinical Center; Terumo BCT
2021
Plasma donor harm counted
The Plasma Protein Therapeutics Association (PPTA), the plasma industry's association, and three member companies, CSL Plasma, Grifols and Takeda BioLife, pooled records of 12.18 million donations by 1.08 million donors at 513 US centers from May to August 2018, about 72% of US plasma donations in that period. They found 15.85 adverse events per 10,000 donations, 90% of them drops in blood pressure or needle-related, with young, lightweight donors at higher risk. The study used the industry's own definitions, which its authors say are not directly comparable with the AABB's.
Credit
George B. Schreiber, Toby Simon and colleagues at the PPTA
Watch for
A second per-donor nomogram
The FDA cleared Haemonetics' NexSys PCS plasma machine with Persona PLUS on 20 February 2026, on the basis of a randomised, controlled pivotal trial, by the FDA's summary. Haemonetics says the trial covered nearly 3,000 donors and more than 30,000 donations, and that Persona PLUS collects a mid-single-digit percentage more plasma per donation than its first Persona.
Achieved whenthe trial is published in a peer-reviewed journal and a group not paid by the maker reports volumes and donor reaction rates from routine use.
Claimed by
Haemonetics, February 2026
Open
01
An open check on nomograms
An open calculator that applies published plasma-volume rules, the FDA's 1992 three-volume nomogram and the percent-of-plasma rule described in the IMPACT paper, to synthetic donors built from public body-size statistics, and reports the volume each rule would collect per donation.
The briefCredit · unclaimed02
A footprint for a litre of plasma
An open life cycle assessment, a count of the greenhouse gas behind a product from raw materials to delivery, for one litre of plasma collected by machine: the single-use kit and bottle, saline and anticoagulant, machine and building power, freezing and frozen storage, and shipping to the drug maker's plant. Every input is cited, the model and its numbers are published, and the result is given as a range.
The briefCredit · unclaimed
Open 01 · the brief
An open check on nomograms
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open calculator that applies published plasma-volume rules, the FDA's 1992 three-volume nomogram and the percent-of-plasma rule described in the IMPACT paper, to synthetic donors built from public body-size statistics, and reports the volume each rule would collect per donation.
Why it is still open
The IMPACT paper gives the 28.5% target, the 1,000 mL cap and a blood-volume formula from Lemmens and colleagues, but not the units or how the machine applies them, so its method is only partly reproducible. The newer rules, Persona PLUS and Terumo's, are known only from their makers' trials, and a GitHub search for plasma nomogram code found nothing.
Built with what exists
Python or R; public height and weight statistics, such as the summary tables of the US CDC's NHANES survey; the IMPACT paper's published donor summary (mean body mass index 32, 65% men); the FDA's 1992 memo for the weight bands.
First step
Code the three FDA volumes and the 28.5% rule, and run both on 10,000 synthetic donors drawn to match the IMPACT summary.
Done looks like
Working only from the paper's aggregate donor summary, the model reproduces IMPACT's mean volumes of 777.8 mL and 841.7 mL within 2%, with code, inputs and a sensitivity analysis published.
Open 02 · the brief
A footprint for a litre of plasma
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open life cycle assessment, a count of the greenhouse gas behind a product from raw materials to delivery, for one litre of plasma collected by machine: the single-use kit and bottle, saline and anticoagulant, machine and building power, freezing and frozen storage, and shipping to the drug maker's plant. Every input is cited, the model and its numbers are published, and the result is given as a range.
Why it is still open
The first such study of transfusion, by Hibbs and colleagues in England (Transfusion, 2024), put a red-cell unit at 7.56 kg CO2 equivalent, with 17% from donation, mostly the plastic packs, and Australia's Lifeblood notes that it did not include plasma or platelets. A bag maker, Macopharma, has studied one whole-blood bag set, and a 2023 TU Wien thesis studied one plasma medicine's manufacture from a company's closed data. No open model of machine-collected plasma turned up in Europe PMC or on GitHub.
Built with what exists
openLCA or Brightway, free life cycle assessment software; public emissions factors for plastics, electricity and freight; the machines' published specifications and instructions for use; an AI model that pulls each figure from those documents with its page, every figure then checked by a person (Note: "Findings that cite notes").
First step
Copy the England study's seven stages into a spreadsheet and list, for a litre of machine-collected plasma, each material and energy input with a public source or a marked gap.
Done looks like
A published model for at least two collection systems and two countries' electricity, with every input cited, a stated range, and the same model run on the England study's red-cell inputs landing within 10% of its 7.56 kg.