Getting serious about platelet contamination
Bacterial contamination of platelets—long recognized as the leading cause of
transfusion-related infection-has perennially hovered in the shadow
of HIV and HCV.
"The FDA felt the issue was important enough to convene three workshops
on it in the last seven years," says Roslyn Yomtovian, MD, director
of the blood bank-transfusion medicine service at University Hospitals
of Cleveland. "Yet nothing has been done to deal with the problem."
Now, however, bacterial screening of platelets is making the transition
from agenda item to action plan. Earlier this year, the FDA approved
the first two methods for detecting bacterial contamination, and
the manufacturers, Pall Corp. and BioMérieux, are actively marketing
them in the United States.
The American Association of Blood Banks’ standards committee is
considering a revised standard that would require blood centers
to screen platelets for bacteria starting late next year. And presentations
on the pros and cons of bacterial detection and reduction packed
in participants at the recent AABB meeting.
"There were two things people were talking about at AABB: West
Nile virus and bacterial contamination testing," says Mark E. Brecher,
MD, acting director of clinical laboratories at University of North
Drs. Brecher and Yomtovian, along with three colleagues in transfusion
medicine, helped heighten attention to bacterial contamination last
August when they signed a letter calling on the blood banking community
to step up adoption of screening technology now rather than waiting
until methods to inactivate or reduce bacteria are perfected.
"Because bacterial contamination of platelets is the largest infectious
disease risk patients currently face, we felt we could not wait
for pathogen reduction to be approved in order to address this problem,"
says James AuBuchon, MD, professor and chair of pathology, Dartmouth-Hitchcock
Medical Center, Lebanon, NH, and a co-signer of the letter.
Steady contamination rate
Solely based on the number of patients they kill, infectious agents
in blood product transfusions, such as HIV, hepatitis C virus, and
West Nile virus, pale in comparison with bacterial contamination
of platelets. The latter may be up to 250 times as likely to transmit
Of the 4 million platelet units transfused each year in the United
States, says Dr. Brecher, 1,000 to 4,000 are contaminated with bacteria,
and 167 to 1,000 cases of clinical sepsis result. Twenty to 40 percent
of patients with clinical symptoms die, he adds. By some estimates,
one patient per day in the United States develops life-threatening
sepsis because of bacterial contamination of platelets.
"Although HIV is definitely the grand leader in what scares people
about blood, that’s in part because most people getting platelets
are much sicker, by and large, than people getting red cells," Dr.
Yomtovian says. Platelets go to patients in oncology, trauma, and
transplant who have high mortality compared with recipients of red-cell
transfusions. Moreover, even though bacteria can kill, HIV is a
lingering infectious agent. "Once you get it," Dr. Yomtovian adds,
"you’ve got it forever, whereas bacteria doesn’t linger, and you’re
not going to pass it on to anyone else."
That doesn’t make it any less lethal, however. Jim MacPherson,
chief executive officer of America’s Blood Centers, says there has
been long-standing interest in preventive screening. "The number
of contamination-related transfusion reactions has been steady for
20 years. What is new is that deaths and infections from other types
of viruses are so low that bacteria now stands out. We know that
about 10 to 15 deaths from bacterial contamination are reported
to the FDA each year. We also know that is probably the tip of the
Although bacterial contamination is the cause of 17 percent of
transfusion-associated deaths, it wasn’t a hot issue in blood banking
until recently. MacPherson compares it to another anomaly of blood
product safety—that the leading cause of transfusion reaction
continues to be giving the wrong blood product to the patient, yet
available error prevention methods have scarcely been used.
The five signers of the letter decided to make their concern public
to prod the blood banking community to move forward. Dr. AuBuchon
is surprised it has taken blood bankers so long to wake up to the
threat. Whether you look at infection rates or fatalities, bacterial
contamination is "hugely more important than HIV," he says. "It’s
also interesting that when someone dies of a bacterially contaminated
platelet unit, they’re dead within a matter of hours or a day or
two, whereas someone with HIV is likely to live many, many years."
The seeds have been planted. Dartmouth-Hitchcock and University
of North Carolina are already routinely testing platelets, and Oklahoma
Blood Center, New York Blood Center, HemaQuebec, and Florida Blood
Services plan to do so in 2003.
Switching to apheresis
In addition to implementing bacterial detection, many of these
same institutions are slashing the risk of contamination by transfusing
apheresis platelets exclusively. Only about half of all transfusions
in the United States are apheresis platelets from single donors;
the remainder are pools of platelet concentrates from whole-blood
donors, or random concentrates. But studies in the past year confirm
that random-donor platelets carry about five times the risk of post-transfusion
sepsis and fatality that apheresis does.
"You’re increasing the risk because, number one, you have many
more donors involved in random concentrates," MacPherson says. "The
second issue is that apheresis platelets tend to be used first,
so they’re fresher and less likely to develop significant bacteria."
This means that if a blood center converts to all-apheresis, the
risk of contamination is lower but the advantage of preferential
use disappears. "There will be lots of units sitting around for
four or five days, and those are the ones, if there are bacteria,
that will represent a risk to the patient," MacPherson adds.
Other factors, too, complicate a move to all-apheresis, not the
least of which is cost. "We do not do single donor exclusively,"
Dr. Yomtovian says. "We would love to be there, but we’re not. We’re
now about two-thirds apheresis. If we wanted to be all-apheresis,
it would entail a considerable addition to our budget because we
would have to purchase the remaining apheresis units from an outside
The economics of using random donor versus apheresis platelets
are complex. An apheresis unit from a blood supplier might cost
$500, while individual units of platelets could run about $50 each,
so depending on the pool size—typically five or six units—apheresis
might cost almost twice as much, Dr. Yomtovian says.
But that cost factor may change over time. An apheresis unit has to be tested
only once for HIV and other pathogens, while each donor unit of random concentrates
must be tested on its own. On the other hand, donated blood is divided into
red cells, platelets, and plasma, so the cost of testing is typically divided
among those three parts. In spite of that, "for us to make single-donor apheresis
is cheaper than buying those non-apheresis" units from an outside supplier,
Dr. Yomtovian says. "The marketplace sets the price, and what a provider may
charge for a unit may not be a true reflection of what it costs them to make
Ahead of the curve
The experience of European countries demonstrates that random concentrates
don’t necessarily pose a higher risk if bacterial detection is used.
Countries that routinely screen platelets, such as the Netherlands
and Sweden, have maintained a low rate of bacterial contamination
while using random platelets almost exclusively.
Differences in regulatory environments have made it difficult to
compare contamination rates between the United States and Europe,
however. In Europe, prepooling random platelets, then testing them,
is allowed, while the FDA forbids prepooling on the basis that it
might lead to a greater chance of bacterial growth.
"The FDA’s rationale is that storing pooled plasma gives it a larger
volume in which bacteria could grow and could allow the innoculum
to reach a higher level," Dr. AuBuchon says. "The other concern
is that white cells from donors may interact with each other and
create an immunologic response in the bag."
The problem with the FDA policy, he adds, is that it means each
unit of whole blood-derived platelets has to be cultured individually,
which removes a substantial portion of the product to get the culture
and is more expensive.
A study conducted at Johns Hopkins Medical Institutions shows the
benefits of switching to single-donor platelets. During a 12-year
period, the use of single-donor platelets was increased from 51.7
percent of all platelet transfusions to 99.4 percent, bringing septic
platelet transfusion reactions down from three events annually to
"Everyone says random platelets will soon be extinct," MacPherson
says-in part because the chemicals used for pathogen inactivation
or reduction are aimed at apheresis platelets, not random concentrates.
On the other hand, if pathogen reduction is used in Europe, where
there’s little apheresis and little bacterial contamination, then
it will be used with pooled platelets.
Adds MacPherson: "There’s a movement in this country to say, Why
waste all these platelets [from whole blood donations]? If we can
ensure they’re safe, why not use them? Why incur hundreds of millions
in expense per year" by converting to all-apheresis? But it’s difficult
to tell where the competing cost-effectiveness arguments will lead,
The perishability factor
Any blood product can be contaminated with bacteria, but platelets
are much more susceptible because they cannot be refrigerated. To
maintain their functionality, they must be stored at room temperature,
"Most platelets don’t have any bacteria in them," says Paul M.
Ness, MD, director of transfusion medicine at Johns Hopkins Medical
Institutions and a co-signer of the August letter. "We think they
[bacteria] can be introduced in two different ways. One, even though
we try to sterilize the skin before venipuncture, there are sometimes
subcutaneous areas of bacteria." Two-thirds of bacteria-associated
reactions are due to this type of skin contamination, Dr. Ness estimates.
"Our studies show one-third are from a donor who seems to be healthy
but has a transient bacteremia."
Storage at room temperature also makes platelets especially perishable.
In the United States, the shelf life of platelets is limited to
five days, but until the mid-1980s it was seven days, and it is
still seven days in some European countries. Some transfusion medicine
experts argue that the seven-day standard makes economic and clinical
"There certainly is a difference," Dr. AuBuchon concedes. "A platelet
stored for seven days is not as good as one stored for five days.
There’s a continuous decline in function and viability. However,
seven-day platelets appear to function more than adequately clinically."
In fact, Dr. Ness notes, advances in plastics in the last 20 years
have made it easier to preserve the function and viability of seven-day
platelets. There’s always a shortage or a potential shortage of
platelets with a five-day shelf life, he says, because of the time
needed to test and process the units-and testing is taking up more
time rather than less. "It used to be possible to collect platelets
in the morning and use them later that day," he says. "But that’s
becoming increasingly difficult."
Can improved methods of detecting bacteria justify extending the
shelf life for platelets? Some say yes, arguing that an extended
shelf life offers a safe and reasonable way to finesse the extra
expense of detection methods as well as the loss of a certain amount
of each unit of platelets to the testing process.
According to the French company BioMérieux, through its BacT/Alert
system (formerly made by Organon Teknika), Sweden has saved an estimated
$300,000 by combining bacterial screening of platelets with a seven-day
shelf life, easily covering the cost of screening.
Similarly, the Netherlands has mandated platelet screening since
November 2001. Combined with an extended shelf life of seven days,
the screening has reduced the number of outdated platelet products
by as much as 15 percent.
A short shelf life makes it difficult to manage inventory without
wasting it, Dr. Yomtovian says. "The margin of error is very small
in terms of planning." The outdate rate at the University Hospitals
of Cleveland is in the five percent range for non-apheresis but
is extremely low—usually one percent—for apheresis.
"We make our own apheresis platelets, so we can, to some degree,
modulate our supply based on what we think the demand will be,"
The outdate rate for platelets at Dartmouth-Hitchcock is 15 percent,
Dr. AuBuchon says. "Back in the mid ’80s," he adds, "when shelf
life was seven days, it was around five percent. I can’t say it’s
necessarily going to drop all the way back to five percent. However,
if the cost of culturing an apheresis unit is $20 and the cost of
an apheresis unit is $500, you don’t have to avoid outdating of
many units before you’ve paid for the culturing.
"It’s not often in transfusion medicine that you get to spend less money and
improve safety," he notes.
More sensitive assays
In the past, methods to detect bacterial contamination typically
were insensitive, Dr. Yomtovian says. At her institution, the laboratory
began using Gram stain in 1991, before culturing was possible, because
of a cluster of contamination cases.
"We felt we had to do something," she recalls, "so for one year
we gram-stained all platelets before issuing them, and as time went
on, we tested only four- and five-day-old platelets. And everything
we gram-stained, we cultured. We’d get a lot of cases of gram-stain-negative
but culture-positive, and of course by then the platelets had been
"We’d then look at what happened to the patient," Dr. Yomtovian
adds. "In some cases the patient did have signs and symptoms that
were related to transfusion of a contaminated unit. Between the
summer of 1991 and spring of 2000 we had 36 bona fide cases of bacterially
contaminated platelets. Not all of them were transfused. We were
able to interdict seven cases out of those 36, so we were able to
do something with gram staining—it just isn’t very sensitive."
Since that time, more sensitive methods have been developed. Two
technologies have been approved by the FDA and are available. Although
FDA approval does not entitle a blood center to claim a unit is
sterile, and the systems may be used only for quality control, Dr.
Brecher says the approvals are significant. "It’s very important
that there is now choice in the market."
Pall Corp.’s BDS filter, recently introduced in Canada and the
United Kingdom, detects decreases in oxygen, which indicate bacterial
growth. "It’s really a surrogate culture," Dr. Yomtovian says. "You’re
incubating a sample removed from the platelets, then you’re probing
the air above the pouch you sampled to see what the oxygen content
is, so if bacteria are growing they will consume oxygen." The method
would not detect bacteria that grow anaerobically, but Dr. Yomtovian
says those are not a major cause of bacterial contamination of platelets.
BioMérieux’s BacT/Alert system detects the presence of bacteria by tracking
their production of carbon dioxide. It is already being used in a number of
countries for platelet screening.
While adopting detection systems and switching to apheresis are
the leading methods for reducing bacterial contamination, another
measure—diversion of the first part of the blood draw—has
received considerable attention since most contamination is from
the skin. This is not a total solution, Dr. Ness stresses, because
one-third of reactions come from donor bacteremia.
Most blood centers now recognize that the optimal arm scrub uses
alcohol and tincture of iodine, Dr. AuBuchon says. But systems that
divert the first 30 mL of the draw may further reduce contamination
Such diversion systems aren’t approved for use in this country—but
that’s coming, MacPherson says. "Again, there are no good data to
show this will reduce blood contamination, but it intuitively makes
sense, because if you look at the bacteria, most of what is cultured
in blood products is skin bacteria," he points out. "If you’re cleansing
the skin, then, it resides deep in the pores of the skin, so the
theory—although there’s not a lot of data to support it—is,
if instead of the tiny plug of skin going into the bag, it getsdiverted
into a pouch or a tube, you will have reduced the possibility of
This may already happen as a side effect of some blood centers’
current practices. "When we collect apheresis, we’re taking out
the first 30 to 40 ccs to do the required infectious disease testing
based on what the manufacturer tells us to do," Dr. Yomtovian says.
Manufacturers have been slow to move on diversion technology, Dr. AuBuchon
adds, because they think they might have to demonstrate its effectiveness via
a costly trial. But the FDA has indicated that manufacturers would not need
to make claims associated with safety to offer diversion technology.
It is difficult to evaluate some of the benefits of measures to
address bacterial contamination because so many deaths and infections
from this cause go unreported.
"Remember, the majority of blood transfusions are given either
intraoperatively, where there is a lot of blood, or in the emergency
room, where the patient may be bleeding out as much as you’re putting
in, so all that kind of trauma can sometimes mask reactions," MacPherson
says. "The patient may die, and you ask: What did they die of? Was
it bacterial sepsis, a reaction to the wrong blood, or did they
just die because they were on the verge of dying?"
For example, it took extensive medical sleuthing to determine in
a recent case of salmonella infection, reported in the Oct. 3 New
England Journal of Medicine (Jafari M, et al. 2002;347:1075-1078),
that the infection was linked to the platelet donor’s pet snake.
The contamination caused one death and one severe illness among
transfused patients, Dr. AuBuchon says. "Salmonella is not a common
organism found in platelets, and the donor had donated numerous
times before and appeared healthy and culture-negative for salmonella.
But salmonella is an organism that can establish a low-level chronic
infection without the donor’s awareness."
Dr. AuBuchon heard of another case in which an apheresis unit was
collected and split into two. Then the blood center received a report
from one hospital that a terrible thing had happened, while the
second recipient’s hospital said there was no problem. "But when
you go digging, you see the patient did get ill or even die—and
no one associated it with the transfusion," he says.
Only serendipity uncovers some cases. "In many cases I know of,"
Dr. Brecher adds, "it was almost an accident. They found out only
because the red cell unit was contaminated, and there may even have
been a catastrophic outcome. Then they asked what happened to the
platelets"—and it turned out the patient who received the
The number of platelet transfusions an institution performs and
how carefully it probes the causes of reactions are also factors,
Dr. Ness says. "If the volume of transfusions you administer is
not that great, you may be lucky not to experience any reactions.
One reason we see them frequently is that we have a high volume
of platelet transfusion. The other reason is, any time we have a
transfusion reaction to platelets that suggests there might be contamination,
we do cultures."
MacPherson suggests that part of the interest in bacterial contamination
detection stems from a desire to preempt bacterial inactivation,
or bacterial reduction as the FDA refers to it. He and others argue
that this would triple the cost of blood to eliminate unknown pathogens.
Many transfusion medicine specialists are concerned about the effects
of bacterial reduction products now in clinical trials. "Their processes
seem very capable of eradicating bacteria," Dr. Ness says. "But
as part of the chemical process used, there is damage to cells,
so the survival of platelets has some compromise."
The chemical used in the inactivation process is similar to drugs
used in cancer chemotherapy, Dr. Ness adds. "They try to remove
it from the process when they’re done, but there are questions about
whether it could be toxic to newborns or pregnant women."
The audience at a recent FDA Blood Products Advisory Committee
meeting was startled, MacPherson recalls, when an FDA official stated
bluntly that pathogen reduction methods are toxic and that no approval
of the technology can be expected for at least five years. "Some
of the chemicals you’re using in bacterial reduction are very toxic
and you have to get rid of them before you can give the product,"
he says. "The FDA has publicly expressed concern about this and
the benefit of doing it, so it’s really not clear what will happen.
The blood community will likely use bacterial reduction techniques
at some point, but right now the timing and technology are uncertain."
The FDA has again placed bacterial detection on the BPAC agenda
for December. The AABB is considering a draft standard that includes
a tincture of iodine requirement for arm preparation and a new standard
requiring either the blood center or the hospital to test platelet
units in some way. If adopted, it will go into effect in November
2003, Dr. AuBuchon says. (AABBmembers can view and comment on the
draft standard at www.aabb.org until Jan. 8.)
Many transfusion specialists agree with Dr. Brecher that U.S. hospitals
and blood centers need a push to catch up to bacterial contamination
screening done in other countries. "Practically speaking," he says,
"it would be very difficult to implement a requirement in the U.S.
unless there were a mandate from some higher power—AABB, CAP,
or the FDA. Someone has to take a stance. Otherwise hospital administrators
are probably not going to allow it to happen."
Anne Paxton is a writer in Seattle.