Plasma donation is safe
And commercial plasma donation is not less safe than non-commercial donations
Source plasma donation (also called “plasmapheresis”) is inordinately safe (so is whole blood donation). And the best publicly-available donation safety data give us no reason to think that commercial plasma collection is less safe than non-commercial plasma collection.
That claim may be surprising in light of the recent heartbreaking deaths reported after plasma donations in Winnipeg. These tragedies have raised questions about the safety of plasma donation in general, with some critics suggesting that commercial plasma donation is inherently less safe than non-commercial plasma donation.
The evidence for the claim that plasmapheresis, including commercial plasmapheresis, is safe can be found in countries with the largest plasmapheresis programs, which publish annual reports on serious donor adverse events. Some of these countries have exclusively non-commercial plasma collection, while others have predominantly commercial systems.
European Union
The available data include Serious Adverse Reactions and Events (SARE) data across the European Union, published by the European Commission. These data include France and Italy which (along with Australia) have the largest non-commercial plasmapheresis programs in the world. The reports also include Germany, Austria, Hungary, and Czechia, all of which have long-running and well-established commercial plasmapheresis programs.
In its latest annual SARE report, covering 2023 data, the European Commission recorded 16,068,007 whole blood collections from 28 of 30 countries and 7,286,075 apheresis collections from 25 of 30 countries. The same report recorded 3,534 serious adverse reactions (SAR) in donors, with total donor SAR incidence rates of 16.0 per 100,000 whole blood (WB) collections and 13.1 per 100,000 apheresis collections.
Looking at plasmapheresis collections, the reported rates for serious adverse reactions (SAR) per 100,000, Major Cardiovascular Events (MCE), Fatalities, Apheresis donations, and Collection rates were as follows:
* This is inferred from the collection rate and estimates for Italy’s population in 2023
A scan of the European haemovigilance data suggests that both whole blood and apheresis donation are generally safe, with serious donor adverse reactions remaining rare overall. Because the report groups all apheresis procedures together, and does not separately identify plasmapheresis, the apheresis figures should be interpreted cautiously; still, in countries with large commercial plasma sectors such as Germany, Austria, and Czechia, the apheresis data most likely consists overwhelmingly of plasmapheresis donations. On that basis, this data does not show that countries with commercial plasma collection have worse SAR rates than countries without it. So the European data does not support the claim that commercial plasma collections are less safe than non-commercial plasma collections.
France
The most in-depth information comes from France.
Some background: France has a strictly public, non-commercial blood and plasma collection system, run by the Etablissement Français du Sang (EFS). The EFS has one of the best donor adverse reaction systems in the world. Reporting is automatic, resulting in immediate notification to the Agence nationale de sécurité du médicament et des produits de santé, or French National Agency for Medicines and Health Products Safety.
It is this robust safety system that allowed researchers to conduct what is, to my knowledge, the most comprehensive safety study on blood and plasma donations to date. Published in May 2025, the study, entitled “Cardiac serious adverse reactions in France 2010–2021,” looked at a 12-year period to assess the extent to which blood and plasma donation was associated with specifically cardiac serious adverse reactions (CSAR).
Over the 12-year period, there were a total of 35,055,266 donations, including 5,066,943 apheresis donations (combining thrombocytapheresis with plasmapheresis), with 125 reports of cardiac serious adverse reactions, including five deaths.
Two of the 125 reports were excluded because of “lack of data” (sometimes described as “non-assessable” in other hemovigilance reports), and 39 were excluded because the cardiac event occurred more than 48 hours after donation (and so is reasonably considered unrelated to the donation).
After all exclusions, 75 cases remained, including the five deaths, 58 coronary events (38 myocardial infarctions, 13 acute coronary syndromes, and seven cases of angina pectoris), and 17 rhythm disorders.
All of these were imputed to causes other than the donation itself (and so given a SAR imputability of “0” and “Excluded”) in the official data.
However, the researchers introduced a novel criterion of imputability, which they describe like this: “An additional criterion of imputability was added for the CSARD attributed to causes other than the donation (e.g., coronary atheroma) but aggravated or triggered by the donation: AT1 possibly (>24–48 h post-donation), AT2 probably (>12–24 h post-donation) or AT3 certainly (within 12 h or pre-donation start).”
According to this novel criterion, 20% of the cases were given an “aggravated” or “triggered” imputability score of 1 (“possibly”), 24% received a score of 2 (“probably”), and 56% received a score of 3 (“certainly”). Apheresis donations were more frequently associated with these cardiac events.
This gives us a CSAR incidence rate of 2.14 per million donations, or about 1 in 467,404 donations. For apheresis donations, the incidence was 5.33 per million, or about 1 in 187,665 donations.
The incidence of fatality, meanwhile, would be 0.143 per million, or 1 in 7.01 million donations. One of the fatalities followed a plasmapheresis donation. The study does not give us a breakdown of how many of the 5,066,943 apheresis donations were plasmapheresis donations, but it is safe to assume that the vast majority were. So the incidence of fatality for a plasma donation would be around 1 in 4 to 5 million.
An important note of caution: the authors stress, and it is worth repeating, that their study does not show a causal link between the cardiac events and donation. So these incidence rates should be regarded as the most conservative rates, not the most likely ones.
Over 12 years, there were five fatalities after non-commercial donation, and one fatality after a non-commercial plasma donation, in France. In Canada, Health Canada said that over the past decade (2016-2026) there have been three fatalities, all in Winnipeg, following plasma donations. They did not reveal details about the third incident.
To do a proper incidence rate comparison, we would need to know how many plasmapheresis donations took place at both commercial plasma centres as well as non-commercial plasma centres in Canada. I don’t have this data, and have not yet put together an estimate, but will do this soon (if you have estimates, please let me know).
Germany
Germany is the largest commercial plasma collector by total volume in the EU.
The Paul-Ehrlich-Institut provides robust data, similar in spirit to the French data. I pulled the most recent five years of hemovigilance data. Here are the results:
* Sudden cardiac event during blood donation. Investigation showed underlying heart rhythm condition, ruled “unlikely”
** Myocardial infarction, died one day after donation. Not enough data for imputation, although donor had undiagnosed coronary artery disease.
*** 27-year-old male suffered a cardiac arrest, but recovered.
**** The event occurred more than 48 hours after donation and so was not imputed to the donation
Germany is essentially unique among high-income countries in having both a commercial and a non-commercial whole blood collection system. In Germany, it is not uncommon to receive compensation for a whole blood donation intended for transfusion (unlike the U.S., for example, where this is basically non-existent; I am aware of one blood collector that compensates donors). Despite this, the non-commercial Deutsches Rotes Kreuz collects the vast majority of blood in Germany, about three-quarters. Meanwhile, plasmapheresis collections are almost entirely, more than 90%, conducted by the commercial sector.
A few things are worth noting. First, the apheresis donations with the highest donor adverse event rates were platelet apheresis donations, not source plasma donations, which had the lowest. Applied to the French data, we can infer, though not with certainty, that platelet apheresis rather than plasmapheresis probably explains why apheresis looks riskier in France.
Second, in Germany, commercial plasma collections had lower SAR rates than blood and platelet donations.
Meanwhile, the incidence of serious cardiovascular events (including cerebrovascular events) was 19 in 32.466 million, or 1 in 1.708 million, with little difference between (mostly non-commercial) blood donation and (overwhelmingly commercial) plasma donation. One of the German cases is relevant here in a narrower sense: in 2021, a 27-year-old donor suffered cardiac arrest while donating whole blood. That is not the same as the October 2025 Winnipeg case involving a 22-year-old plasma donor, but it does show that a donor’s young age, by itself, cannot tell us that donation caused the event.
If we assume the two fatalities were imputable to donation, which they were not, then the incidence of fatality across this five-year period was two in 32.466 million, or 1 in 16.233 million donations, and 0 in 13.9 million for commercial source plasma donations.
Germany had 2.78 to 3.5 times as many commercial plasma donations in just five years as France had non-commercial apheresis donations over 12 years. If we compare the fatality figures, we get roughly 1 in 4 to 5 million for non-commercial French plasma donations and 0 in 13.9 million for German commercial plasma donations.
The evidence from the European Union supports the claim that plasma donations are safe, and it provides no evidence that commercial plasma donations are less safe than non-commercial plasma donations.
United States
The best evidence we have for the safety of specifically commercial plasma donations comes from the United States.
Let me first give a sense of the scale here, because it is genuinely staggering. Across 12 years, France had just over 35 million total donations of both blood and plasma. The United States had more than twice that many commercial plasma donations in 2025 alone (around 75 million).
France and Italy, the two largest non-commercial plasma collectors in the EU, collected 922,605 litres and 919,700 kilograms of plasma in 2025. After converting kilograms to litres, using 1 litre of plasma as roughly 1.025 kilograms, they collected a combined total of about 1.82 million litres. The U.S. collected 62.5 million litres that same year, or more than 34 times as much plasma.
Put differently, it would take those two countries more than 34 years to collect as much plasma as the U.S. does in one year. Or, to make the point as starkly as possible: in 2025 the U.S. collected more plasma in less than 11 days than France and Italy together managed to collect in a whole year.
Recovered plasma makes up a large share of France’s and Italy’s total plasma collected, while it makes up a trivial share of America’s. Around 390,000 litres and around 400,000 litres, respectively, came from source plasma donations, for a total of about 790,000 litres. That implies roughly 1.2 million source plasma donations, assuming an average volume of 650 mL per donation. The U.S. had 75 million source plasma donations, or about 61.7 times as many. Or, in the starkest terms: it took the U.S. just under 6 days to carry out as many plasmapheresis procedures as France and Italy together carry out in a whole year.
The U.S. FDA publishes an annual report entitled “Fatalities Reported to the FDA Following Blood Collection and Transfusion.” I pulled data from 2015 through 2022 (the most recent available year):
The FDA does not always include an estimate of the total number of source plasma donations. If we exclude the years with missing estimates (2015, 2017, and 2020), we get a dataset with a total of 230.5 million individual donations. For those years, there was a total of 27 fatalities after a plasma donation with an imputation of either “probable” (5) or “possible” (22), and no cases imputed as “definite.”
If we take the most pessimistic view, as we did for France, and assume all 27 of those cases were definitely caused by the donation, which they were not, that results in a fatality rate of 0.0000117%, or roughly 1 in 8.54 million donations.
What the evidence shows
The deaths reported in Winnipeg after plasma donation are, like all deaths after donation, deeply tragic. We collect blood and plasma, and encourage people to donate, in order to save lives. That is the point and purpose of these systems.
But the data we have do not show that plasma donation is unsafe, and it does not show any inherent safety difference between commercial and non-commercial donations when it comes to serious donor adverse reactions and events.
None of this tells us, by itself, what happened in Winnipeg. But the broader hemovigilance evidence from millions of commercial plasma donations gives us good reason not to assume causation: when deaths occur after donation, they are often ultimately found to be unrelated to donation, doubtfully related, or better explained by underlying medical conditions.
We should wait for Health Canada to release its findings before jumping to extreme and unwarranted conclusions.
Corrections:
An earlier version reported that Health Canada said there were four fatalities in the span of a decade, three in Manitoba and one in Quebec. CBC corrected their story on March 16: “An earlier version of this story said Health Canada had received four reports of fatal adverse reactions after giving plasma, including one death in Quebec. Health Canada later said the information was given in error, and no fatal adverse reaction was reported in Quebec. Mar 16, 2026 5:57 PM EDT” I have therefore changed the text to reflect this. March 17, 9:38 a.m. (Thank you to Peter O’Leary for bringing this to my attention).