Disponible en ligne sur

www.sciencedirect.com

Transfusion Clinique et Biologique 20 (2013) 416–421

Original article

Lessons from the response to the threat of transfusion-transmitted

vCJD in Ireland
Leçons tirées de la gestion du risque de transmission-tranfusionnelle du vMCJ en Irelande
W.G. Murphy a,b,∗
aIrish Blood Transfusion Service, James’s Street, Dublin 8, Ireland
b School of Medicine & Medical Science, University College, Dublin, Ireland
Available online 31 August 2013

Abstract
By the time vCJD was first described in 1996, it was already far too late to offset further disaster from transmission of the disease by blood
transfusion: almost all the humans who would be infected and infectious were already diseased. Nothing done by the blood transfusion services
around that time, with the exception of excluding transfusion recipients as blood donors, would have made any useful contribution to containing
the extent of the epidemic. The ability to spread emerging diseases before the problem is manifest or understood is a fixed and unavoidable
feature of blood transfusion as it is practiced today. A second fixed property of blood transfusion is that the root cause of disaster is not within
the control of the blood transfusion universe. Strategies that have emerged to cope with similar threat in other enterprises that also contain these
properties comprise the components of robust design: surveillance, preparedness for action, engagement, herding together, evasion or avoidance,
early adoption of potentially useful measures, engineered resilience, defence in depth, damage limitation including modularity and removal of
feedback loops, and contingency, redundancy and failure management, and ultimately, individual escape. Early adoption of leucodepletion based
on the possibility that it might work rather than any hard evidence was a good example of threat management. Exclusion of previously transfused
donors is a robust mechanism for containing any future infection; optimal blood use structures that provide a national transfusion rate as low as
possible also constitute an effective threat management strategy.
© 2013 Elsevier Masson SAS. All rights reserved.

Keywords: Prion; vCJD; Blood transfusion; Risk management

Résumé
Lorsque la maladie fut décrite pour la première fois en 1996, il était déjà trop tard pour mettre en place des actions préventives contre ce nouveau
désastre potentiel de transmission transfusionnelle de la maladie : en effet, tous les sujets contaminés et probablement infectieux étaient déjà
décédés. Aucune des mesures déjà mises en place à cette époque, à l’exception de l’exclusion des donneurs ayant été transfusés, auraient permis
de lutter efficacement contre l’épidémie. La capacité qu’ont les maladies émergentes de se propager avant même que le problème soit identifié
et compris est un caractéristique classique et courante en transfusion sanguine dans sa pratique actuelle. Une seconde caractéristique est que la
racine du mal n’est pas sous contrôle en la transfusion sanguine. Les stratégies préventives vis-à-vis de risques émergents similaires rencontrés dans
d’autres entreprises forment un plan d’actions robustes comprenant : la veille, la préparation à l’action, l’engagement, l’esprit d’équipe, l’évitement,
la mise en place de mesures préventives précoces, la gestion de l’urgence, etc. La mise en place anticipée de la déleucocytation fondée sur le pari
que cela pourrait avoir une action bénéfique est un bon exemple de gestion de risque. L’utilisation des produits sanguins à bon escient constitue
également une stratégie de management efficace pour diminuer le risque potentiel de transmission par transfusion.
© 2013 Elsevier Masson SAS. Tous droits réservés.

Mots clés : Prion ; vMCJ ; Transfusion sanguine ; Gestion du risque

In June 1996 the Lancet published an account by Robert

Will and colleagues of 6 patients who had developed what
∗ Irish Blood Transfusion Service, James’s Street, Dublin 8, Ireland. is now called vCJD [1]. They conjectured, almost certainly
E-mail address: nmd@indigo.ie correctly, that their patients had been infected with a lethal

1246-7820/$ – see front matter © 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.tracli.2013.06.002
 W.G. Murphy / Transfusion Clinique et Biologique 20 (2013) 416–421
neurodegenerative agent through eating material produced from
cattle infected with bovine spongiform encephalopathy (BSE).
At that time, there was no clear idea how extensively the
disease had been spread through the British population, or
indeed through the rest of the world. Nevertheless it was widely
accepted that there was a possibility that many thousands could
have been infected, and that a huge problem could be unfolding
[2].
Within the blood transfusion community a set of responses
to this problem began to form. National transfusion services and
plasma product manufacturing agencies explored the nature of
the disease, and the feasibility and utility of possible strategies
to counter it. Almost all chose a wait and see approach, at least at
first–waiting for more concrete evidence that the infectious agent
could be transmitted by transfusion, waiting to see if methods to
test for infectivity in donors or to inactivate infection in blood or
plasma would be developed, waiting to see if the problem really
was a problem.
But it was already far too late, by 10 years or so, to offset
disaster, if disaster there were to be. By the time Will and his
colleagues described the first cases of human disease, almost
all the humans who would be infected were already diseased,
and had been for several years. Worse, they were already poten-
tially highly infectious as blood donors, and had been for years.
Had the disease been more effective at infecting humans and
had tens of thousands been infected from eating BSE infected
material, as was plausibly estimated in the early years of the
human epidemic [2], then thousands of patients receiving blood
transfusions would have been lethally infected by then as well as
in the following 10 or so years. A far greater number of patients
would have died from vCJD-contaminated blood than had died
from HIV or HCV. That scale of disaster did not happen, but
only because the dietary-derived human epidemic was so small.
Nothing done by the blood transfusion services around that time,
with one notable exception in one country — France — would
have made any useful contribution to containing the extent of
the epidemic. Even the French initiative to contain any infection
(by breaking the chain of onward human to human transmis-
sion through stopping transfusion recipients from subsequent
blood donation) would have been ineffective in preventing or
attenuating the first, more lethal wave of infection.
The first human-to-human transmission that we know of
took place through blood transfusion in September 1996, a few
months after the first human cases were reported. The fact of
that transmission was not apparent for a further 7½ years [3].
By now, in April 2004, the dietary-derived cases were petering
slowly out, and the damage was done. A few more transfusion-
transmitted cases might occur, and it was worthwhile seeking to
prevent them; commercial and regulatory interest in prevention
was enhanced, but it was far, far too late — almost 20 years too
late — to have prevented disaster. Only chance, that the disease
was so refractory to crossing into humans from cattle via inges-
tion of food, prevented massive loss of life from contaminated
blood transfusions.
What could have been done differently? What methods could
have been applied that would, or at least could, have reduced
the impact of the emergence of a new, previously undescribed,
417
highly transfusion-transmissible and invariably lethal disease.
There was (and there remains) no test that could be applied
to donors or blood products; there was (and there remains) no
known method to remove infectivity from donated or processed
blood or plasma. Within a community there was no possible
effective strategy to exclude high-risk individuals — even veg-
etarians could be infected. Quarantining nations from donating
blood was possible outside of the British Isles; indeed quarantin-
ing Europe was possible outside of it. But the classical approach
of collection of data and designing and applying remedies to the
problem was always going to be monumentally useless as an
effective defence against the early waves of infection, no matter
how essential it would be later.
This was not the first time this had happened — it had hap-
pened on different scales with hepatitis B, hepatitis C and HIV:
diseases of unknown biology or unknown impact had been
widely spread by transfusion before the nature, or even the
existence, of the disease had become apparent. This property
to spread emerging diseases in this way before the problem is
manifest or understood is a fixed and unavoidable feature of
blood transfusion as it is practiced today. There is nothing that
can be done to get rid of it — it is an essential emergent part of
the practice. That is not to say that there is nothing can be done
to deal with it, but the first step is to recognise it.
There is a second fixed property of blood transfusion practice
that is essential to understand if meaningful lessons can be learnt
from the vCJD story: the root cause of disaster is not within the
control of the blood transfusion universe, and never can be. HIV
probably crossed into humans through bushmeat hunting and
butchering in the West African rainforest. From there it found its
way out of the forest, and into the world outside, killing millions,
devastating lives, devastating nations, changing history. A sim-
ilar remote, trivial event began the BSE/vCJD story, and could
have had similar consequences had the molecular properties of
the mutant bovine prion been subtly different. No conceivable
strategy by blood transfusion services could have prevented the
emergence or global spreading of these diseases in the several
years that their storms were slowly building. Blood transfusion
services might, however, have limited the impact on blood trans-
fusion recipients had they understood that as well as being prone
to early, silent, and widespread infection in emerging epidemics,
blood transfusion is at the mercy of seemingly trivial, and very
remote effects–exactly as weather is in the famous butterfly anal-
ogy. Perhaps closer to that example is the impact of a single
smuggled bird or passenger mosquito on a plane landing in New
York in 1999, bringing West Nile Virus to a virgin land. This
chaotic, or edge-of-chaos, nature of blood transfusion makes its
future safety unpredictable at best, and challenges, but does not
prevent, a scientific approach to preventing a repeat of HIV, or
hepatitis C, or of the near-miss of vCJD: “science does not pre-
dict the future...scientific advice (may be) nothing more than a
sharper set of questions to guide us through the fog” (Lord May
of Oxford, who first described the complexity inherent in the
biology of populations [4]).
By recognising that these two properties — fixed inherent
risk and edge-of-chaos — are integral in blood transfusion, we
can draw valid inferences from the set of human and other
418 W.G. Murphy / Transfusion Clinique et Biologique 20 (2013) 416–421
Table 1
Elements of robust design applied to blood transfusion.
Element of robust design
Assuming the presence of threat: preparedness for action; assessing the
threat: surveillance for existence, proximity, & magnitude of threat
Engagement of the threat
Herding together
Evasion/Avoidance
Early adoption of potentially useful measures
Engineered resilience
Defence in depth: Layered defences–potentially, though not necessarily
proven, (and clearly-not-completely-) effective actions are identified,
implemented and functional
Damage limitation: Detection of failure of defences.
Containment/Modularity & removal of feedback loops; repair of
defences, backup/redundancy
Contingency: system-wide critical failure
Individual escape
biological enterprises that also contain these properties, explore
what strategies in those systems attenuate the impact of disas-
ter, and consider whether and how they might be applied to the
enterprise of blood transfusion. Such systems include those of
ecology, evolution and biology, engineering, seismology, social
sciences, & military endeavours. These are all environments
where either the nature or the quantum, or both, of a threat can-
not be defined a priori, but its existence can be taken as given.
Strategies that have evolved or emerged to cope with threat
in these systems comprise the components of Robust Design
[5–9]: surveillance, preparedness for action, engagement, herd-
ing together, evasion or avoidance, early adoption of potentially
useful measures, engineered resilience, defence in depth, dam-
age limitation including modularity and removal of feedback
loops, and contingency, redundancy & failure management, and
ultimately, individual escape.
Extrapolating these elements of robust design to managing
the threat of emerging infectious agents for blood transfusion
yields the components in Table 1.
These strategies and tactics costs money, in the same way as
insurance, fire alarms, and the myriad ways we manage or avoid
Actions
Surveillance adopted as formal and functioning activity within a designated
division of the organisation; information gathering and processing, information
sharing, training, equipping, ensuring full range of skills in place. There must
be a specific budget, cost centre and governance structure for this function in
the organisation
Specialized function within the organisation to fully engage with the threat at
an early point–There must be absolute prioritization for this function once
mobilised
Collegiality–may have utility for information sharing, resource pooling and
legal defences in dealing with new threats
Alternative supply or avoiding blood use. Identify the available tactics and
strategies. Assess the feasibility and costs of alternatives; ensure that a viable
route to alternative supply can be activated. Move in good time–act as one
would for a threat within own domestic or personal environment
If it might be useful, it should be implemented. It is inappropriate to wait for
everyone else, or for absolute proof of effectiveness or cost effectiveness–one
should act as one would for a threat within own domestic or personal
environment
Comprises all the levels of complexity within an organisation, from the
elements in this table to the materials in every key component in the entire
process. Includes GMP and quality assurance and control as elements of threat
management
Identify the available tactics and strategies: Testing, donor exclusions,
filtration, leucoreduction, pathogen inactivation
Implement capability, test capability & maintain functionality; build in
redundancy where possible. Do not allow transfusion recipients to donate
Systematic and early sharing of information on all CJD cases as they appear
Developing plans with other blood services, and exercising them. Alternatives
e.g. cancellation of elective surgery. Stockpiling
Limit donor exposures, avoid non-essential transfusions. As threat escalates,
defer elective procedures, send children abroad for essential surgery, promote
bloodless surgery
threats and risks in everyday life cost money. We do not try to
budget for insurance when the house is on fire, for example, or
when the car has been stolen. And once any serious menace or
threat appears at a personal or domestic level, or even at a physio-
logical level, we focus every available resource to its resolution.
Managing threats in the complexity of blood transfusion differs
only, essentially, in scale. It is important that senior management
and the governance function in a transfusion service understand
this function of coping with an underlying, unavoidable chaos,
and is prepared to pay real money for it. It incorporates risk
management and GMP, quality control and quality assurance,
but it places these elements in a larger framework of threat
management in an unavoidably complex environment, capable
of generating unanticipated crises on a vast and unknowable
scale.
From early 1997 to date, the Irish Blood Transfusion Service
prioritised the threat of vCJD to the safety of the blood trans-
fused in Ireland. From an early stage it was apparent that the
level of threat was high — there is a very large amount of travel
between Ireland and the United Kingdom, where the epicentre
of infection lies; Ireland shares a large and open land border with

Table 2
Approaches taken by the Irish Blood Transfusion Service to address the risk of transfusion-transmitted vCJD.
Element of robust design
Assuming the presence of
threat: preparedness for
action; assessing the threat:
surveillance for existence,
proximity, & magnitude of
threat
Engagement of the threat
Herding together
Evasion/Avoidance
Early adoption of potentially
useful measures
W.G. Murphy / Transfusion Clinique et Biologique 20 (2013) 416–421
Actions
Surveillance adopted as formal and
functioning activity within a
designated division of the
organization; information gathering
and processing, information sharing,
training, equipping, ensuring full
range of skills in place. There must
be a specific budget, cost centre and
governance structure for this function
in the organisation
Specialized function within the
organisation to fully engage with the
threat at an early point–There must
be absolute prioritization for this
function once mobilised
Collegiality–may have utility for
information sharing, resource
pooling and legal defences in dealing
with new threats
Alternative supply or avoiding blood
use. Identify the available tactics and
strategies. Assess the feasibility and
costs of alternatives; ensure that a
viable route to alternative supply can
be activated. Move in good time–act
as one would for a threat within own
domestic or personal environment
If it might be useful, it should be
implemented. It is inappropriate to
wait for everyone else, or for
absolute proof of effectiveness or
cost effectiveness–one should act as
one would for a threat within own
domestic or personal environment
419
Actions taken at the IBTS after 1996 Actions that would have been
beneficial from mid 1980s had the
epidemic developed
Included local conferences with Unlikely to have reduced infections
invited experts, commissioned prior to 1996 had dietary vCJD been
reports, attendance by one or two more prevalent
designated individuals at all or
almost all significant scientific
meetings in Europe and the USA;
constant survey of the scientific
literature; early engagement with
manufacturers of emerging detection
and removal technologies. Formal
reviews took place twice monthly
(1997), but not formally captured in
the budgetary process
This was top priority for senior Unlikely to have reduced infections
medical staff (1997) prior to 1996 had dietary vCJD been
more prevalent
As above; formal links with the UK Unlikely to have reduced infections
in particular were established for prior to 1996 had dietary vCJD been
information sharing and more prevalent;
technological assessments. Measures there is also a risk that this tactic
under consideration or implemented could reinforce inertia rather than
in other countries were very seriously encourage early action
assessed for adoption on a
comparable timescale: early adoption
of strategies used by others was seen
as a vital function at individual,
managerial, governance and
Government levels. (From 1997;
until 2008, budgetary considerations
were given relatively little weight)
Alternatives to transfusion of blood Geographical exclusions as a general
components from Irish, UK or other measure are unlikely to be a
European risk area donors consistently effective approach to
Recombinant factor VIII and Factor risk reduction in the absence of
IX used exclusively from 1997; epidemiological knowledge of the
United States as preferred source for disease being confronted
plasma for non-recombinant factor Replacement of blood-derived
concentrates sourced by the Blood medicines with ones of similar
Transfusion Service from 1998 efficacy and pharmacological safety
Replacement of almost all plasma for will always be an effective threat
clinical use with US-sourced SD management strategy
plasma from 2002
Replacement of almost all
cryoprecipitate for clinical use with
fibrinogen concentrate from
US-sourced plasma from 2009
Adoption of universal leucodepletion Could have been very effective if
with full implementation by implemented earlier–however there
November 1999 based on tenuous was no reason to have guessed this
evidence of protective possibility. No effect. Nevertheless, given that
cost benefit assessment was possible leucodepletion reduces some virus
Phase 1 clinical trials of prion filtered risks it stands as a robust threat
red cells 2005 prevention strategy in its own right
(It is very probable that a test would
have been implemented as soon as
available–plans to do so were well
developed and the prior necessities
were in place)
420 W.G. Murphy / Transfusion Clinique et Biologique 20 (2013) 416–421

Table 2 (Continued)
Element of robust design Actions Actions taken at the IBTS after 1996 Actions that would have been
beneficial from mid 1980s had the
epidemic developed

Engineered resilience Comprises all the levels of
complexity within an organisation,
from the elements in this table to the
materials in every key component in
the entire process. Includes GMP and
quality assurance and control as
elements of threat management
Defence in depth: Layered Identify the available tactics and
defences–potentially, strategies: Testing, donor exclusions,
though not necessarily filtration, leucoreduction, pathogen
proven, (and inactivation
clearly-not-completely-)
effective actions are
identified, implemented
and functional
Damage limitation: Detection Implement capability, test capability
of failure of defences. and maintain functionality; build in
Containment/Modularity & redundancy where possible. Do not
removal of feedback loops; allow transfusion recipients to donate
repair of defences, Systematic and early sharing of
backup/redundancy information on all CJD cases as they
appear
Contingency: system-wide Developing plans with other blood
critical failure services, and exercising them.
Alternatives e.g. cancellation of
elective surgery. Stockpiling
Individual escape Limit donor exposures, avoid
non-essential transfusions. As threat
escalates, defer elective procedures,
send children abroad for essential
surgery, promote bloodless surgery
No effect from 1986 on vCJD spread,
other than as a systematic response to
the threat as it emerged. A high level
of engineered resilience would
already have provided for aggressive
reduction of donor exposure, for
example
Exclusion of donors with increased No effect, other than a possible
risk through UK residence from 2001 chance effect of leucodepletion had it
Leucodepletion of all fresh blood been adopted earlier as a general
components from 1999 threat management strategy
Central surveillance and notification
system for all cases of CJD in place
Exclusion of previously transfused Exclusion of previously transfused
persons from donation (2004) donors as a general threat
management strategy would have
been very effective in preventing
secondary transmissions
A number of contingencies were Not applicable as a general measure
seriously considered, including
sending children abroad for elective
surgery and collecting blood for Irish
use in purpose specific collection
programmes abroad in collaboration
with other national blood
services–e.g. organising collections
in geographical areas not fully used
by self-sufficient blood services in
countries with a large Irish diaspora
Use of apheresis platelets instead of Always going to be effective in part
pools, optimal blood use. From as a threat management approach
1999–partial

Northern Ireland, politically and epidemiologically part of the
United Kingdom, and a large percentage of Irish blood donors
spend or have spent significant amounts of time in the United
Kingdom. Although by 1997 it was, as outlined above, already
too late to do very much about the first and by far the largest
wave of transfusion-transmitted infections, a number of initia-
tives were taken, that would, or at least could, have attenuated
the transmission of disease after 1997. Some of these would
have been effective, at least in part, in 1986 too, when some-
thing meaningful might have been possible to counter the first
wave. (Table 2).
In conclusion, it has become apparent that blood transfu-
sion has two fundamental, existential traits that require to be
systematically addressed:
• emerging human infections may be spread widely in blood
transfusion recipients before the existence or the nature of the
infection is understood;
• serious risks to blood transfusion recipients can be generated
by events remote in time and place from the blood transfusion
services and from the recipients of blood transfusions, and
prevention of these events will always be impossible.
In retrospect:
• the early adoption of leucodepletion based on the possibility
that it might work rather than any hard evidence (although it
was known to be safe in general use) was a good example of
threat management;
• had a worse situation emerged, the proposal to collect
at least some blood for neonatal use in countries out-
side Europe would have been seen to have been under-
developed.
From knowledge gained, and once it is appreciated that threat
management strategies can provide only one element in a suite
 W.G. Murphy / Transfusion Clinique et Biologique 20 (2013) 416–421
of strategies and tactics to reduce the potential for and impact of
harm at a national, local and individual level:
• exclusion of the previously transfused is a robust mech-
anism for containing any future infection, and should be
implemented everywhere as a sound threat management
strategy;
• optimal blood use structures that provide a national transfu-
sion rate as low as possible–around 30 per 1000 of population,
or perhaps lower, for a typical western age distribution–also
constitute a universal and effective threat management strat-
egy.
Disclosure of interest
The author declares that he has no conflicts of interest
concerning this article.
421
References
[1] Will RG, Ironside JW, Zeidler M, et al. A new variant of Creutzfeldt-Jakob
disease in the UK. Lancet 1996;347:921–5.
[2] Ghani AC, Ferguson NM, Donnelly CA, Anderson RM. Predicted vCJD
mortality in Great Britain. Nature 2000;406:583–4.
[3] Llewelyn CA, Hewitt PE, Knight RS, et al. Possible transmission of variant
Creutzfeldt-Jakob disease by blood transfusion. Lancet 2004;363:417–21.
[4] May R. Risk and uncertainty. Nature 2001;41:891.
[5] Murphy W. Managing threats rather than risks in blood transfusion: robust
design for a complex system. Transfusion 2006;46:2011–3.
[6] Jen E. Stable or robust? What’s the difference? In: Jen E, editor. Robust
Design. Oxford: Oxford University Press; 2005.
[7] de Visser JA, Hermisson J, Wagner GP, et al. Perspective: evolution and
detection of genetic robustness. Evol Int J Org Evol 2003;57:1959–72.
[8] Voight CA, Mayo SL, Wang ZG, Arnold F. Directing the evolvable: utilizing
robustness in evolution. In: Jen E, editor. Robust Design. Oxford: Oxford
University Press; 2005.
[9] Robertson DS, Feedback theory, Darwinian evolution. J Theor Biol
1991;152:469–84.