REVIEW

I Review I

Transfusion medicine: looking to the future

Lawrence T Goodnough, Atyeh Shander, Mark E Brecher

The evolution of transfusion medicine into a clinically oriented discipline emphasising patient care has beenaccompanied by challenges that need to be faced as specialists look to the future. Emerging issues that affect bloodsafety and blood supply, such as pathogen inactivation and more stringent dono r screening questions, bring newpressures on the availability of an affordable blood supply. Imminent alternatives for management of anaemia, such asoxygen carriers, hold great promise but, if available, will require close oversight. With current estimates of HIV orhepatitis C viral (HCV) transmission approaching one in 2000000 units transfused, keeping to a minimum bacterialcontamination of platelet products (one in 2000) and errors in transfusion, with its estimated one in 800 000 :mortality rate, assume great urgency. Finally, serious difficulties in blood safety and availability for poor, developingcountries require innovative strategies and commitment of resources.

Transfusion medicine has evolved from a mostly In an attempt to anticipate and manage such shortageslaboratory-centred service with a focus on the serological better, various organisations and govemment agenciesaspects of blood, into a clinically oriented discipline that have started to track production and use of such products.emphasises patient caTe. This evolution has taken place For example, in the UK, the Blood Stocks Managementover the past 20 years, mostly because of the recognition Scheme was established in 1997 as a collaborative venturethat HIV and hepatitis C virus (HCV) are transmissible by betWeen the National Blood Service and the hospitalblood. The resultant emphases on blood safety, sector to understand and improve management ofappropriate use of blood components, informed consent supplies.3 This system uses web-based submission of datafor blood transfusion, and altematives to blood have led to and instant graphic feedback and was implemented insubstantial advances in reduction of potential risks and April, 2001. As ofFebruary, 2002,167 (54%) ofhospitalscomplications associated with blood transfusion. served by the National Blood Service were tracking 1.44Nevertheless, this evolution has been accompanied by million red cell units/year (65% ofNational Blood Servicechallenges for which specialists in transfusion medicine issues).3 In the USA, the Department of Health andand their clinical colleagues must chart an uncertain Human Services began in August, 2001, to monitor thecourse into th~ future. Here, we overview selected issues, ability of the US blood supply to meet demand.' Thisand discuss how their resolution could be tumed into system monitors three sentinel community blood servicesopportunities for the future. and 26 sentinel hospital transfusion services, which

account for about 10% of the US inventory of red bloodBlood centres cells and platelets.Blood availability A comprehensive national survey of blood collectionsSporadic shortages of blood and blood products (eg, and blood transfusions in the USA has been done by thepacked red cells, platelet products, albumin, intravenous Center for Blood Research from 1982 to 1994, andimmunoglobulin, and clotting factor concentrates) are subsequently by the National Blood Data Resourcepotentially life-threatening occurrences. Such shortages Center. Blood transfusion and collection activities peakedhave been attributed to various causes, including in 1986, and then declined.' However, blood transfusionsdisruptions in production, increasingly strict criteria for and collections increased from 8.0% in 1997 to 10.2% indonor deferral, product recalls, increase in use (including 1999; Quikount surveys from the National Blood Dataoff-label), and possibly supply disruptions because of Resource Center of US blood centres (80.4% responsestockpiling or other market issues. Additionally, the ability rate) in 2000 estimated that 12.54 million allogeneicof blood centres to supply blood in response to acute blood units were transfused-a yearly increase of 4-5%crises has assumed particular importa1'u:e after the from 1999 (M Sullivan, National Blood Data ResourceSept 11,2001, terrorist attacks in the USA.,,2 Center, personal communication).

Search strategy and selection criteriaLancet 2003; 363.: 161-69 We searched the National Llbrary of Medlclne (NLM) database

by the NLM Gateway web site for the past 5 years, and fromDepartments of Medicine, and Pathology and Immunology, review of the authors titles abstract and source locationWashington University School of Medicine, St Louis, MO, USA articles in full were s~lected for furth~r examination. This '

(L T Goodnough MO); Department of Anesthesiology, Mount Sinai h d .th .T h L t ' g "d I. f rt " I . .searc was one WI In, e ance S UI e Ines or a IC es InSchool of Mediclne, New York, and Department of Anesthesiology, .t S . f t M di. h d .th O .

dI S emlnar orma. e Ine searc es were one WI VICrltlcal Care and Hyperbarlc Medlclne, Englewood Medlcal Center, T h I . V . 4 4 O h h h U .. f N hE gl d NJ (A Sh d ) D rt t f P th I g d ec no ogles, erslon ..,t roug t e nlverslty o ortn ewoo , an er MO; epa men o a o o y an .." .

Laboratory Medicine Universltyof North Carolina School of Carolina Health Sclences Llbrary. We searched the InternetMedicine, Chapel Hi;l, NC (M E Brecher MO) using the Google search engine. Specific searches were

.dane at the http://www.fda.gov/cber/index.html,~ Correspondence to: Or Lawrence T Goodnough, Washington htt .11 d g I d htt .11 h . t/h g IU "s h I f M d ". O "". f L b M " " p. www.c c. ov an p. www.w o.ln ome-pa e

nlverslty c 00 o e IClne, Ivlslon o a oratory edlclne,. .,Box 8118, 660 S Euclid Avenue, St Louis, MO 63110. USA ~Ites: Ref~rences were sele~ted accordln~ to the authors(e-mail: goodnough@labmed.wustI.edu) Identlficatlon of relevant tOplCS for the revlew.

THE LANCET. Vol361 .January 11, 2003' www.thelancet.com 161

REVIEWc ;

1/100

1/1000 1/2000co'u;.2~ 1/10000~+"'

'~ 8000 to 1/149000~ 1/100000Qjc.§ 1/872000 to 1/1.7xl06.., 1/1000000 t t t t~ t 1/1.4Xl06 to 1/2.4Xl06,S OD"O 5 >. OD[D cn 5 >. 5 c OD>-.so> -.-"O.StE -.-"8 -.-0> .ss:o c OD OD.8 c o+"' OD.o OD'~ 1;) "-

~ o>g c~ o>c~ c~ C'C 0»cn ~ .c: '" c ~ .O> ." c "t; '" +"' u;:;: uu 0>'P u':f.c: 0>'P 0>q-:g:I:~ cn", 0» cnc 0» I-N u~

5';: t;s:o>g t;u c. "'..2c 2?, cn (ô ~ cn :I: ,Qo o;: OD o O>

ou e-.- õ~ "" zcn

I I I 1 I I I I I I I I 1 I 1 I 1 I I

.<o~ # .<o~.<o~ .<o" .<oCO .<00) ~O ~~ ~'}, ~~ ~~ ~~ ~~ ~" ~co ~O) "p C~~~ ~O)' ~O)' t$ t$ ~O)' ~O)' .!;J ~~ .!;J ~~ .!;J .!;J ~~ .!;J ~~ ~~ '},O" '},O

Years

Figure 1: Rlsks of transfusion-related transmission of HIV, hepatitis B virus (HBV), and hepatitis C virus (HCV) in the USAEvery unit is exposure to one danar. Modified from reference 11. with permission of the publisher,

Allogeneic blood collections in 2000 were 13.37 mil1ion become regular donors! The challenge is to successfullyunits-a surplus of 6'2%. This fractional surplus has recruit and maintain an adequate blood supply before thediminished from the 7.4% ofuntransfused units in 19975 blood is needed, not after.because of increased demand for blood and loss ofdonors. Current estimates, modified by events of Sept 11, Blood safety200~ predict that US blood collections in 2001 surpassed Blood centres have now implemented nucleic acid testing15.0 million units. of minipools (16-24 donation samples/pool) from blood

Faced with future prospects of acute demand for blood donations to reduce HIV and HCV transmissions duringsupplies or transient decreases in donor collection, blood the infectious window period (before serologicalcentres and transfusion services have begun to reexamine conversion). Current estimates of the risk per unit oftheir approach to inventory reserves. In the USA, there is blood in the post-nucleic acid testing era are 1 in 1.4X 106much heterogeneity in how reserves are maintained. to 2'4X106 for HIV and 1 in 872 000 to l'7X106 forWithin the American Red Cross (which supplies about hepatitis C viruS.lO,ll Figure 1 shows the changes in risk ofhalf of ali blood products in the USA), a 3-4 day supply of transfusion-related transmission of HIV, hepatitis C, andred blood cell products is typical.6 Some independent hepatitis B in the USA as a result of the various strategiescentres allow for higher reserves. For example, the implemented. In the USA since 1999, over 25 millionOklahoma Blood Institute typically has a 14-17 day blood donations have been screened. To date, three casessupply of red cell product! Use of frozen red cells as a of an apparent HIV transmission by a unit negative byhedge against inventory shortages has generally not been both minipool nucleic acid testing and HIV serologypractical because the shelf-life of thawed units is only (antibodies against HIV and antigens to HIV p24) have24 h. However, an automated, functionally closed system taken place,'2 and one documented case of apparent(ACP 215, Haemonetics, Braintree, MA, USA) for the hepatitis C transmission by a unit negative by bothglycerolisation and deglycerolisation processes has now minipool nucleic acid testing and hepatitis C serology hasbecome available and allows for a 2-week post-thaw shelf been reported from Germany.13 These cases suggest thatlife." With the availability of such a system, the Oklahoma adoption of single donor nat testing will be likely.Blood Institute plans to increase its available red cell At present, the greatest risk of transfusion-transmittedinventory to 21-24 days, with a rotating frozen red cell disease is bacterial contamination of platelets. Unlikereserve of 2000 units.7 Similarly, the Yale-New Haven viroses, bacteria can proliferate from low concentrationsHospital transfusion service plans to freeze 200 units of « 1 colony-forming units/mL) at the time of collection tored cells with this new system to provide them with a very high concentrations (>lX10" CFU/mL) during the3-day supply ofblood products for emergencies.9 liquid storage period ofblood components. Since platelets

Because of the time needed for blood processing and are stored for up to 5 days at 20-24°C, they constitute antesting, it is the blood donated before a disaster that is excellent growth medium for bacteria. Culturetransfused for lives placed at risk. It is now apparent that surveillance suggests that bacterial contamination ofafter disasters (human or natural) in the short-term, on- platelet concentrates and apheresis platelets occurs inhand blood supplies are adequate and can be rapidly about 1 in every 2000 units.I4-16 Although the truemobilised over great distances.' The wave of donations prevalence of severe episodes of transfusion-associatedafter the Sept 11, 2001 attack overwhelmed the donor bacterial sepsis is not known, it is estimated to occur withsystem in the USA, and few of the new donors have a quarter to a sixth of contaminated transfusions.17-19 With

162 THE LANCET. VaI 361 .January 11, 2003. www.thelancet.cam

REVIEW

4 million (1 million apheresis platelets and 3 million Number of Seroprevalence Screening Potentialplatelet concentrates) platelet units transfused yearly in blood in blood donors coverage (%) casesthe USA alone, it would bf!: expected that 2000 to 4000 donationsbacterially contaminated units would be transfused and be Countryassociated with 333 to 1000 cases of severe and possibly Chile 220686 9.7% 79.8% 16fatal sepsis. Pooled platelet concentrates have a higher risk Colombi,a 422300 11-1% 99.9% 2

...Costa Rica 58436 25.7% 6.9% 487than apheresls platelets (a functl0n ofthe number ofurnts Ecuador 110619 1.3% 72.3% 9

pooled, reflecting the increased donor exposure and EI Salvador 34091 19.0% 100.0% Onumber of phlebotomies required to obtain the Honduras 27963 11.9% 99.0% 1products).20 Nicaragua 46539 3.9% 62.1% 21

Recent reports from Europe and North America have Panama 42342 NA NA NA.. d . 1 Paraguay 39904 37.7% 100.0% O

advocated use of automated 11quld me Ia cu ture systems Peru 203690 2.0% 60.0% 36

for testing of platelets to reduce the risk of bacterial Uruguay 115490 6.5% 100.0% Ocontamination.21-25 Culturing of platelets is mandatory in Venezuela 262295 7.8% 100.0% OBelgium (Flemish Red Cross) and the Netherlands. Data are from reference 37. NA=not available.Although not mand~tory, platelets are algo cultured in Estlmates of blood donation, seroprevalence, screenlngmost blood centres m Sweden, Norway, and Denmark coverage and number of potentlal cases for Trypanosoma cruz;and in selected sites in the UK, Germany, Canada, and in Latln "':merica (1997)the USA. Culturing is typically done after some storagetime has elapsed (eg, 1-2 days) to allow for bacterial hospital were judged to be at risk of T cruzi infection, andgrowth to optimise detection. In some cases, such of these, one in 500 were confirmed antibody positive.37culturing is used as a rationale to extend the shelf life of To date, seven cases of transfusion-transmitted and threeplatelets to 7 days.'1-27 cases of transplant-associated Chagas disease (from one

Another promising approach is pathogen inactivation organ donor) in the USA and Canada have been(eg, psoralens with ultraviolet irradiation), with the reported.38.39 Although donor screening for T cruzipotencial to eliminate viral and bacterial contaminants in antibodies is important in endemic conditions in Latinplatelets.28 However, these techniques have difficulty and South America (table), there is no direct evidence toinactivating spore-forming bacteria (eg, Bacillus Spp).'9 suggest that introduction of routine donor screening forAnother concern is that such processing can lead to antibodies to this organism in all donors woulddecreased platelet recovery and in-vivo survival (thereby measurably improve the safety of the USA blood supply.leading to the need for increased platelet transfusions).30 A pilot programme to screen and test blood donors forThis occurrence was seen in the US clinical tria13! with T cruzi infection is underway in Canada.apheresis platelets, with lower post-transfusion corrçcted Transfusion-associated babesioses is another examplecount increments and increased platelet transfusion of a transmissible zoonotic disease. Red cells and plateletsrequirements for maintenance of haemostasis, when prepared from asymptomatic donors have been implicatedcompareci with umreated platelets. Additionally, potential in more than 30 transfusion-transmitted cases. Asplenic,concerns about mutagenicity and teratogenicity of such elderly, or severely immunocompromised patients are atcompounds, along with costs and limited availability (for the greatest risk of developing haemolytic anaemia,example, the pivotal US phase-3 trial for platelet products coagulopathy, and renal failure. At present, no test iswas done with products derived from apheresis available for mass screening to detect asymptomatictechnology from one vendor) leave the eventual role of carriers ofbabesia species.4Othis technology uncertain. Donor referral criteria to deal with the potential

problems of Creutzfeldt-Jakob disease (CJD) and variantDanar screening CJD (vCJD) were implemented in 1987. Although CJDDonor-deferral questions remain important com- has been transmitted from human to human byplementary methods for increasing transfusion safety, transplantation of dura matter or cornea, injection ofespecially for diseases for which routine laboratory testing pituitary growth hormone, or reuse of electro-is not done. Deferral policies for selected diseases are encephalograph electrodes,41 no cases of transmission bybeing re-evaluated. The US Food and Drug blood transfusion have been reported. Results of newerAdministration (FDA) revised deferral criteria for malaria epidemiological studies42 confirm earlier studies that failedrisk in 1994,32 but another revision is being considered; to show a link between transfusion and transmission of103 cases oftransfusion-transmitted disease were reported CJD. Despite concerns about potencial transmission ofto the US Centers for Disease Control from 1958 through circulating prions, there is an emerging consensus that1998, with an estimated occurrence of 0'25 per million CJD is rarely, if ever, transmitted by blood transfusion.units transfused.33 Two-thirds of these diseases arose in Strategies implemented such as donor deferral schemes,donors who should have been excluded under current universal leucoreduction, etc, are therefore consideredscreening criteria, but the remaining third were from precautionary.donors whose last traveI exceeded the time limits (a Deferral of donors who have spent longer thanminimum of 3 years for immigrants or residents from 6 months in the UK from 1980 through 1996 wasendemic regions) in the FDA's guidelines.34Transfusion- implemented in the USA in April, 2000, in a pre-emptivetransmitted malaria is similarly a potential infectious effort to keep the potential risk of vCJD to a minimumthreat in Europe. while avoiding disruption of the blood supply. As of

Chagas disease (Trypanosoma cruZt) is endemic in many October, 2002, more than 137 patients, mostiy in the UKparts of Central and South America. Transfusion- (128), but algo six patients in France, and one in ltalytransmitted Chagas disease is a major concern because have been diagnosed with vCJD. As a furthermeasure toT cruzi establishes a chronic, asymptomatic carrier state in reduce the potential risk of vCJD by transfusion, newmost infected persons (table).35.36 Donor history screening donor deferrals were implemented in the USA in October,for risk factors associated with T cruzi infection have poor 2002: cumulative time spent from 1980 to present in thespecificity; in one study, 39'5% of donors at a 1.os Angeles UK longer than 3 months, in Europe longer than 5 years,

THE LANCET .Vol 361 .January li, 2003 .www.thelancet.com 163

REVIEW

or on a US military base in Europe longer than 6 months. less is known about the benefit of blood transfusion, orThese criterià are estimated to reduce by 90% the total conversely, the risks associated with (untreated)donor person-days of exposure to the putative agent of anaemia." The estimated risks of a blood transfusion arevC}D, with an estimated loss of about 5% of the US quantifiable and can be communicated to patients, butdonor pool.43 The effect of these enhanced donor- the risks of anaemia (ín the absence of a bloodscreeníng measures is expected to be most severe in the transfusion) are poorly understood and cannot bemetropolitan New York City region, which ín the past has accurately conveyed.derived about 25% of its blood ínventory via importation Anaemia has traditionally been regarded as anfrom the European Union. abnormal laboratory value, rather than as a serious

disorder associated with, or probably the cause Df, adverseTransfusion services clínical outcomes. By contrast with treatment algorithmsErrors in transfusion medicine or recommendations that have been developed for otherErrors are inevitable in any process, and mistransfusion disorders such as hypercholesterolemia or hypertension,resulting in death has been recognised for half a century.44 guidelines for management of anaemia have beenThe mistransfusion rate (blood transfused to other than developed for only a few patients. The relation betweenthe intended recipient) is about one in 14000 units ín the anaemia and morbidity'&-'9 and mortality'Q-62 has beenUSA4' and one in 18000 in the UK.46 About half the established best ín patients with chronic kidney disease,errors occur in the clínical arena (incorrect identification leading to development of guidelínes that these patientsof the recipient to the blood unit, phlebotomy errors, be maintaíned at haemoglobín concentrations betweenfailure to recognise a transfusion reaction), and about 110 g/L and 120 g/L.6330% in the laboratory.47 A similar error rate (one ín Information about the effect of anaemia on mortality is17 000) has been identified for autologous blood units ín becomíng íncreasingly apparent ín patients in otherCanada.48 One in 33 000 units are ABO-incompatible clinical settings. A growing body of evidence suggests thatbecause of error, half of those are associated with a anaemia affects the outlook ín patients who havetransfusion reaction, and about 10% are fatal.4' The congestive heart failure or who have ischaemic heartfrequency of death due to ABO-error is one per 800 000 disease. Two large observational studies64,6' noted anblood units,47 compared with about one per 2 000 000 association between haemogIobin concentrations oftransfusions for transmission ofHN or hepatitis C.IO 95-100 g/L and íncreased mortality ín patients with

The current data probably underestimate the cardiovascular disease, suggesting that such patients didmagnitude of non-infectious serious hazards of not tolerate anaemia as we11 as those without knowntransfusion, sínce they are derived from passive cardiovascular disease.haemovigilance reporting systems. In the USA, only fatal Investigators of a randomised trial66 ín patients withtransfusion reactions have been required to be reported moderate to severe congestive heart failure (New Yorkto the FDA since 1975. Nevertheless, the FDA reported Heart Association class 111 to IV) assessed whethertransfusion-related death rate was more than twice that treatment of anaemia affected clínical outcomes. Overdu&. to a11 infectious hazards combined, and the UK 8 months, haemoglobin concentrations in the treatment

survei11ance system reported an adverse event rate cohort íncreased from 103 g/L to 129 g/L, with improvedattributed to mistransfusion that was 10 times higher left ventricular ejection fractions, reduction ín diureticthan the rate attributed to ínfectious-disease therapy dose, and reduction in hospital days comparedtransmission.49 with controls. 25% of the control group died during the

A system of voluntary, confidential, non-punitive study interval, compared with none in the treatmentreporting is believed to be an important approach to group.improvíng transfusion safety, similar to a quality Results of a retrospective observational analysis67 ofimprovement system for safety existent ín the airline 78974 elderly patients admitted with acute myocardialindustry.'o A medical event reportíng system (MERS) for infarction ín the USA showed that ín those with admissiontransfusion medicine now beíng piloted in several packed-ce11 volumes of less than 33%, blood transfusionslocations." As we11 as haemovigi11ance schemes in the UK were associated with significantly lower 30-day mortality.and ín France, the MERS strategy implements an In the absence of prospective data or other data to theinternal tracking system to identify and investigate errors contrary, many lives could be saved when patients who(sentinel events, such as mistransfusions that result in present with acute myocardial infarction are maintained athaemolysis or death) and precursor events (near misses, packed-ce11 volumes of greater than 33%.68 For the firstsuch as phlebotomy sample errors) facilitating causal time, data are emerging that undertransfusion can lead toanalysis so that action can be taken to prevent recurrence. adverse outcomes.69,70Additiona11y, because 61 % of errors originate in clínical, Guidelines for management of anaemia need to bepatient-related settings,'2 processes to ensure developed in several clínical settings. The requ~st foridentification of patients with blood samples and blood applications by the US National Institutes of Health tounits need to be considered as an adjunct to existing establish a clinical trials network in transfusion medicine71

national voluntary, confidential, and non-punitivereporting system developed in partnership with Erythropoiesis stimulants Haemostasisregulatory agencies should be developed in para11el with Erythropoietin Recombinant factorlocal, operational systems such as medical event N$w erythropoietin stimulating Recombinant facto r VIIIreporting systems for transfusion medicine. factor Recombinant factor IX

.Artificial oxygen carriers AnticoagulantsUndertransfus/on ;HaemOgloblns"ojcutllOns """'i'Vc"' Antr"thlrrombt'nltl ,i"'"Alth h h . h b . d th . k d "',' ""'" !,i"i;:"1!r"",,,,"~ '" ,J ".I ",&Wi&&i"

oug muc attentlon as een paI to e ns s an 1?~1;f1 o 00 I)D~iiiii&iii;:ii;:;' &';:;"".~t1"'àtedihii;~""':iii&ii

complications associated with blood transfusion, much &W;:ii'~~i,~:&iiff~1;:i!1!J~&i~i~~.'&ii~~m:&1mf. '" i'ii~'iiii~ti';~;~K;:;:~!i~ii!;t':i&

164 THE LANCET. VaI 361 .January 1 1,2003. www.thelancet.com

REVIEW!

erythropoietin stimulating factor), which has a longer half-disappearance time. Recombinant factor VIla is now

oxygen carriers approved for patients with haemophilia who haveinhibitors,73 but there are also ongoing clinical trials of this

~dvantages haemostatic a.gent ~ pat.ients without pr~-existingStroma-free coagulopathy m penoperanve haemorrhage, m thoseNqt antigenic with thro~bo~openia after. pe~pheral stem ocellQ~ygen unloading devoid transl?lantanon, m ~ose undergol~g; ltver transplantanon,g~Q,3-diphOSPhoglycerate effect an~ m trauma pane?ts, The ablltty of factor VIla toPiIb about20 mm Hg(normal haemoglobin P$,,27mm Hg) acnvate factor :x .dlrec~~ on the oplatele~ mem~raneeqthogen inactivation surface and res~ct ItS acnV1ty to lo~altsed regto~s of t~ssue$!~e about 1 fLm fa~tor makes ~IS product potenna~ly useful ~n panentse~ssibility of unlímitedavaiJabiljty W1th substannal haemorrhage, W1th or W1thout the~ended shelf life >2 years presence ora coagulopathy,

H.(gh oncotic properties .o ..Ai.tered oxygen affinity and unloading characteristics (bovi Artificial o~ygen carners o, o! ' .Progress m developme~t of artIficIal oxygen carrlers hasLifllltatlons ,"" accelerated.74,75 Potennal advantages for cell-free?hort intravascular life1-2 h f'11 haemoglobin solutions and perfluorocarbon emulsionsP6ssible,renal toxic effects ~zz,',: (as synthetic oxygen carriers) are listed in panel 2.Negative inotropic effect :&ill& Possible disadvantages of such products includePQssible pulmonary and ""'W'8 interference with interpretation of some laboratory tests 76':i; t . h rt o I"" ',sys e.mlc. ype enS10n .',"A"ii",zz," and their short time in circulation (24-48 h)}7 ane,PO$Slble Immune suppresslon ;W:i::'.i"" potential difficulty associated with haemoglobin solutions

,'~~sslble anaphylaxls ~I~ is vasoconstriction, a consequence of their ability to bindLi~ited resuscitation ab11ity limited to haemoglobin 50-7t&", nitric oxide.78 The nitric oxide binding properties of

haemoglobin are thought to be respo~si?le f~r thegastrointestinal discomfort observed in clmlcal tnals ofsome products}5

is an opponunity for clinical studies to answer these and The vasoconstrictive effect could also be explained byother important questions. autoregulation, since small molecules, such as

haemoglobin solutions, overdeliver oxygen to vesselEmerging technologies walls,79 Results of one study"" assessing oxygen deliveryMore and more biotechnology products are becoming after administration of clinically relevant doses of aavailable as altematives to blood transfusion,72 some of haemoglobin solution to anesthetised surgical patientswhich are liste<ir in panel 1. Stimulants of red blood cell showed that the ability of the haemoglobin-based oxygenproduction include recombinant human erythropoietin carrier to increase oxygen delivery was limited by itsand an altered erythropoietin molecule (new vasoactivity. In another prospective, randomised trial,8!

Compound , AttributesPerflubron (perfluorooctyl. bromide), Oxygent,AIJiance 60% emulsion of perfluorocarbon has now compteted one phase 3"r:~armaceutica1s, San Diego, CA evaluation in Europe,83 USA, and CanadaoDuetoii';i,; unexplained results possibly related to clinical tríal design flaws, the.cardiac and non-cardia.c transfusion avoidance studies in the USA:,'Ei;i: were term1nated early In 2001;1~LHB-HemAssist, Baxter Healthcare,. Roundtree,. IL Negative outcomes in phase 2 and clínical trials caused research to be""" "terminated in 1999GlutaraldehYde poJymerised bovine haemoglobin, Hemopure has a long storage capabtl.ity at roam temperatureo Oxygen+

i,~~mopure,. Biopure Corp, Boston,. MA del1very is three times greater then allogeneic blood because of betterti::: oxygen afflnityo AppJications includetrauma and military settings along+,"';"'Ei surgical andacute heedsi~~lyoxyethylene glycol conjugated bovine haemoglobin,. Due to long half..life,. adm1nistration once a week makes tumours more'Enzon,.Píscataway,. NJ radiosensitive,Rqffinose crossltnked and polymerised human Controlled clinical trials in cardiac surgery have shown reduced exposu ,h~emoglobin HemoLink,. Hemosol,. Toronto, Canada allogeneic transfusion with few side.effects and no nega tive effect on

i;,I',: mortality~.ridoxylated and glutaraldehyde polymerised human Closest to human blood properties, Polyheme can be administered in6~emogl.Obin, PolyHeme, Northfield Laboratories,. quantíties (up to 5000 mL). This product is intended to be used in

'Chicago.,. IL patients. Clinical vasoconstriction has not been reported:,Recombinant human haemoglobin- Optro,. Somatogen, Early problems with oxygen yield and production were observed.~Qulder. CO, Baxter Healthcare, Roundtree, IL was algo a problem with Escherichia coli, but now is noto Further work+i E:' improving oxygen yield is underway with additional recombinants

+P~lynitroxylated polymerisedhaemoglobin,. PNH,. Haemoglobin-based oxygen carriers crosslinked with Dextran is associ",,$ynZyme Technologies, Irvine,. CA with a right shift of oxydissociation and nftroxylation resulting ín

-~;:,:"

THE LANCET. Vol361 .January 11, 2003. www.thelancet.com 165

REVIEW

more trauma patients died than controls, resulting in Blood centrediscontinuation of further product development from lhe Imanufacturer. +

Perfluorocarbon emulsions can dissolve large amounts Hospital transfusion committee

of any gas, including oxygen and carbon dioxide. These Structure Functionemulsions are effective for delivery of oxygen during Transfusion practices Use reviewhaemodilution in patients undergoing orthopaedic surgery BI d f ty T f ' .

t O 9 d t 1 8 g/k rfl b d 82 In 00 sa e rans uslon reactlons,

a .an a' g pe uorocar on oses. a I' t.1.. 1 d . d d .,83 f rfl b comp Ica lonsmu tmatlona ran omlse stu y o a pe uorocar on solution to augment açule normovolaemic haemodilution Blood Inventory Component avallablllty,

during orthopaedic surgery, perfluorocarbon combined wastagewith 100% oxygen was more effective than autologous Policies and procedures Accreditation

blood in reversal of physiological transfusion triggers. ~Because of lhe small sample size, efficacy of lhe H .t I t f ..

rfl b 1 .. 1... f 11 . bl d OSpl a rans uslon servlcepe uorocar on so utlon m e lInmatlon o a ogenelc 00exposure was not recorded. With their high affinity to Structure Functiondissolve gases, prevention of, and therapy for micro- Inventory safety To provide adequate andembolic bubbles from cardiopulmonary bypass or' safe blood products forpreservation of solid organs for transplantation are other health-care deliverypossible and desirable applications for whichperfluorocarbons seem to be ideally suited.84-86 Blood transfusion and To coordinate blood

The two principal applications for lhe artificial oxygen conservation transfusions and bloodcarriers under clinical investigation are in patients with conservation activitiestrauma8? and in those who are undergoing surgery, with orwithout açule normovolaemic haemodilution. The Quality improvement To oversee regulatory andrationale for use of artificial oxygen carriers with accreditation issueshaemodilution is three-fold: lhe cellular haemoglobin related to these activities

collected during haemodilution would be used to replace ~lhe haemoglobin solution or other synthetic oxygen carrier T f . d .. 1.. d f .fi . 1 . ld rans uslon me Iclne speclallstas lt IS e lmmate ; use o artl cla oxygen carners woupermit more aggressive haemodilution with lower targeted ~cellular haemoglobin concentrations than would otherwise Clinicianbe tolerated; and an artificial oxygen carrier could serve as Ia replacement fluid during blood 10ss.88 +

At present, artificial oxygen carrier products are in Patientvario~ stages of clinical development (panel 3). Ifapproved by lhe FDA, they would most likely be readily Figure 2: Flow of blood components from "veln to vein"applied in military casualties and in trauma patients, The physiology of the transfusion committee and the transfusion servicemassive surgical blood loss settings or in fulminant helps define the job description of the transfusion medicine specialist,h 1 ..89 ' .whose position between the transfusion service and the clinician

aemo ytlC anaemlas. ~e role ?f these subst~ces m emphasises the clinical role as consultant. Reprinted from reference 99,these and other arenas will most l1kely be determmed by with permission from the publisher.issues related to blood inventory and costs, rather than lhesafety ofthe blood supply. technologies (panel 1) such as pathogen inactivation and

Other beneficial molecules in human blood are being artificial oxygen carriers is not yet established. As bothidentified, isolated, and characterised, leading to lhe regulated and marketed pharmaceuticals, artificial oxygenavailability of blood-derived or ex-vivo synthesised carriers might be applied clinically as more traditionaltherapeutic agents. One likely candidate molecule is lhe von pharmaceutical therapeutics, similar to factorWillebrand factor-cleaving metalloprotease, which has been concentrates.associated with thrombotic thrombocytopenic purpura. It is important that oversight of these biotechnologyTreatment for this potentially fatal disorder requires daily products and other specialised blood products (such asplasma apheresis (typically for 1-2 weeks or longer) with solvent detergent plasma, leukoreduced blood products,plasma replacement.9o During a therapeutic course, patients irradiated blood components, and cytomegalovirusare generally exposed to over 200 blood donors.91 In negative blood components) be placed under lhe auspicesCanada, 39% of ali plasma apheresis procedures are for of standing hospital medical committees such as atreatment of thrombotic thrombocytopenic purpura.92 In transfusion medicine committee, a pharmacotherapeuticsmany patients with non-familial, idiopathic thrombotic committee, a surgical operating room committee, or athrombocytopenic purpura, activity of lhe von Willebrand critical care committee (figure 2).99 This process wouldfactor-cleaving metalloprotease is reduced (typically as a mean that longitudinal reviews could be done about use ofresult of an inhibitor), and lhe rarer familial recurrent these agents, not only to assist in proper selection ofthrombotic thrombocytopenic purpura is a result of gene patients, but algo to establish efficacy, track serious adversemutations yielding a dysfunctional metalloprotease.93-95 The events, and assess costs. The promotion of such productsactual protease has been isolated and partly sequenced.9&-98 by lhe commercial sector directly to consumers couldAvailability of a metalloprotease preparation might obviate undermine both institutional oversight and an evidence-lhe need for plasma apheresis and improve outcomes in this based rationale for use.serious disorder.

Global perspectiveOversight Worldwide, over 75 million units ofblood are estimated toWhether transfusion services will participate in be donated every year. In lhe USA, lhe yearly transfusionimplementation or distribution of emerging bio- of 12.5 million units (M Sullivan, National Blood Data

166 THE LANCET. Vol 361 .January 1 1,2003' www.thelancet.com

REVIEW

Resource Center, personal communication) corresponds and in vivo values in human RBCs frozen with 40-percent (wt/vol)to transfusion of one blood unit every 0,39 s. Only 43% of ~Iycerol and stored after deglyceroliz,atio,n for 15 days at 4 degrees CWHO's 191 member states test blood for HIV, and mAS-3: assessmentofRBCprocessmgmtheACP 215. Transfuswnh .. C d h .. B . A 1 13 mill" 2001; 41: 933.

epatltls an epatltls VIroses. t east 10n unltS 9 Snyder E Blood community perspective on assuring blood safety andof blood donated every year are not tested for these availability during a disaster. Presentation at the HHS Advisotytransmissible viroses. 80% of the world's population is Committee on Blood Safety and Availability. What lessons can beestimated to have access to only 20% of the worldwide leamed from the events of Sept 11, 2001, that would strengthen the

1 f afI bl d E ali tr fu' d safety and availability ofthe US blood supp1y? Transcript of 16th~u.pp '! o se. 00. ve~ year, uns e ans slon an meeting. Vol I. Jan 31, 2002. http://www.hhs.gov/bloodsafetym)eCtlOn practlces are estlmated to account for 8-16 /transcripts/20020131.html (accessedJan 5, 2002).million hepatitis B infections, 2-3-4'7 million hepatitis C 10 Kleinrnan SH, Glynn SA. Busch MP, Wright oJ, McMullen Q,infections, and 80 000-160 000 HIV infections.IOO,IOl Schreiber GB. Declining incidence raIes and risks of transfusion-

In the poorest countries, access to safe blood is transmitted viral infections in us blood donors. Vox Sang 2002; 83:

financially prohibitive, since testing costs between US$4011 A106b. h JP B. km JD B h MP S-& f th bl d I ' ...u uc on , Ir eyer ,usc .wety o e 00 supp y m

and $50 per blood donatlon. Testmg mlght algo not be the United States: opportunities and controversies. Ann Intern Medreliable, especially if done by inadequately trained staff or 1997; 127: 904-..09.with inadequate equipment. Transmissible diseases are not 12 CBS News. HIV-tainted b100d infects two in Florida. httP://WWW.the only issue, since many countries do not have an cbsnews.com/stories/2002/07/19/health/main515694.shtml. d fu . d d d bl d. .1 (accessed Dec 6, 2002).organlse trans Slon system an onate 00 IS slmp y 13 d S.

I C d M K K K b kB Th firs fHCV.a I va ar oso , Gemer , u ane .e t case ounavallable. ..seroconversion after 3 years ofHCV NAT screening in Baden-

Although safe and aval1able blood IS demanded and Wurttemberg. Transfusion 2000; 40: 1422-23.expected in developed countries, limitations in resources 14 Yomtovian R, Lazarus HM, Goodnough L:r , et ai. A prospectivemake this goal unattainable for poor developing countries. microbi?logic surveil.iance progr~m to detect and prevenI theThe British philosopher John Stuart Mill (1806-73) stated transfuslon ofbactenally contammated platelets. Transfuswn 1993; 33:th ". .gh th d th th d 902-09.

at actlons are n t to e egree at ey ten t~ 15 Dyskstra A, Jacobs J, Yomtovian R. Prospective microbiologicalpromote the greatest good for the greatest number. surveillance (PMS) ofrandom donor (RDP) and single donorAlthough our continued efforts to increase the safety of apheresis platelets (SDP). Transfusion 1998; 38 (suppl): S104.blood in developed countries are both impressive and 16 Leiby DA, Kerr KL, Campos JM, et ai. A retrospective analysis ofcommendable, more fundamental problems in developing micr?bial ~ontaminants in outdated random-donor platelets from

. d b dd d mulnple sltes. Transfuswn 1997; 37: 259---63.countnes nee to e a resse. .' .

17 Goldman M, BlaJchman MA. Blood product-assoclated bactenalsepsis. Transfus Med Rev 1991; 5: 73-83.

Conflict of interest statement 18 Association Bulletin No 96-6. Bacterial contamination ofbloodWithin the past 3 years, L:r Goodnough has been a consultant for Ortho components. AABB Faxnet, No 294, August, 1996. AmericanBiotech, Amgen, Novonordisk, and Hemosol; A Shander has been a Association ofBlood Banks.consultant ~or Ortho Biotech,.Amgen, Novonordisk, Abbott, Astra 19 Jacobs MR, Palavecino E, Yomtovian R. Don't bug me: the problemZeneca, Alhance Pharmaceuncals, and Hemosol; and M E Brecher has ofbacterial contamination ofblood components-challenges andbeen a consultant for BioMérieux (formerly Organon Teknika), Pall I t ' ." ...,,' 2001.41. 1331- 34B. d. I M . T hn I . d H I M E B h h so u lons. .l ransJ~wn ,. .I lome Ica, osalC ec o ogles, an emoso .rec er as 20 N PM B ' HG Kin K, I S. gl d I I d th. d h~. fr B. M " dH . d ess ,rame , g era. m e onorpateetsre uce erecelve researc -.pport om 10 eneux an aemonencs an , , 7- 1

I honoraria from BioMerieux, Pall Biomedical, and Mosaic Technologies. nsk ofsepnc transfuslon reacnons. Transfuslon 2001, 41. 85 6,

21 Ollgaard M, Albjerg I, Georgen J. Monitoring ofbacterial growth inplatelet concentrates-one year's experience with the BactAlertTMSystem. Vox Sang 1998; 74 (suppl): 1126,

References .22 Laan E, Tros C, Improved safety and extended shelf-hfe ofleuco-

1 Schmidt PJ. Blood and disaster: supply and demando N EnglJ Med depleted plate)et concentrares by automated bacterial screening.2002; 346: 617-20. Transfusion 1999; 39 (suppl): 5S.

2 Klein HG. Earthquake in America. Transfusion 2001; 41: 1179-80. 23 Vuetic D, Taseki J, Balint B, Mirovic V. The use ofBacTAlert system3 Chapman J .Monitoring the capacity of the blood supply to meet for bacterial screening in platelet concentrares, Vox Sang 2000;

demando Presentation at the HHS Advisory Committee on Blood 78 (suppl): P371.Safety and Availability. What lessons can be learned from the events 24 McDonald CP, Roy A, Lowe P, et aI. The first experience in theof Sept 11,2001, that would strengthen the safety and availability of United Kingdom ofthe bacteriological screening ofplatelets tothe US blood supply? Transcript of 16th meeting, Vol I, Jan 31, 2002, increase shelflife to 7 days. Vox Sang 2000; 78 (suppl): P375.http://www.hhs.gov/bloodsafety/transcripts/20020131.html (accessed 25 Brecher ME, Means N, Jere CS, et ai, Evaluation oftheJan 5, 2002). BacT/ALERT 3D microbial detection system for platelet bacterial

4 Nightingale S, The federal govemment role in assuring blood safety contamination: an analysis of 15 contaminating organisms. Transfusionand availability during a disaster. Presentation at the HHS Advisory 2001.41: 477-82,Committee on Blood Safety and Availability. What lessons can be 26 AuB~chon JP, Cooper LK, Leach MF, et aI. Bacterial culture ofleamed from ~e :,,:ents ofSept 11,2001, that would sr:engthen the platelet units and extension ofstorage to 7 days. Transfusion 2001;safety and avallablhty ofthe US blood supply? :rranscnpt of 16th 41 ( I). SImeeting. Vol I, Jan 31,2002. http://www.hhs.gov/bloodsafety supp. , .

I ' I wth./transcripts/20020131.html (accessedJan 5, 2002), 27 ~recher ME, Holland PV~ PI~eda A, et a. ~actena gro m

5 Go dn ghT'" B h ME Kan MH A b h JP moculated platelets: Imphcanons for bactenal detecnon and theo ou , rec er , ter, u uc on ., f I I ." ...,.' 2000. 40. 1308 12T fu ' di '. I NE IJM d )999.340.438.7 extenslon o p ate et storage. .lransJ~wn ,. -.

rans slon me cme. part. ng e ,. -'. , ' G I Ph h , ai ., . f6 O II G Th .d . fth W h. DC 28 Lin L, Cook DN, Wlesehalm P, et a. otoc emlC macnvanon o

ue ette .e expenence an acnons o e as mgton ...bl d ' d aft S b 11 P .' th' Vlruses and bactena m platelet concentrares by use of a novel psoralen00 communlty on an er eptem er .resentanon at e ." ..HHS Ad . C . BI d S-& d A il bil. Wh and long-wavelength ultraVlolet hght. Transfuswn 1997, 37. 423-35.

VlSOry ommlttee on 00 wety an va a Ity. at ,. ...lessonscan be learned from the events of Sept 11,200), that would 29 Knutson F, Alfon~o R, DupuIs K, 7t aI. Photochemlcal macnvanon ofstrengthen the safety and availability ofthe US blood supply? bacte.n:a and HIV m bu~-c~at-denved platel:t concentrares underTranscript of 16th meeting. Vol I, Jan 31,2002. condlnons that preserve m Vltro platelet funcnon. Vox Sang 2000; 78:http://www.hhs.gov/bloodsafety/transcripts/20020131.html (accessed 209-16.Jan 5, 2002). 30 Corash L, Behrman B, Rheinschmidt M, et ai. Post-transfusion

7 Gilcher R. Lessons from previous disasters. Presentation at the HHS viability and tolerability of photochemically treated plateletAdvisory Committee on Blood Safety and Availability. What lessons concentrares. Blood 1997; 90: 267a.can be learned from the events ofSept 11, 2001, that would 31 McCullough J, Vesole D, Benjamin RJ, et ai. Pathogen-inactivatedstrengthen the safety and availability ofthe US blood supply? platelets using Helinx technology are hemostatically effective in theTranscript of 16th meeting. Vol I. Jan 31, 2002. thrombocytopenia patient. Blood 2001; 98: 45a.http://www.hhs.gov/bloodsafety/transcripts/20020131.html 32 Nahlen BL, Lobel HO, Cannon SE, Campbell CC. Reassessment of(accessed Jan 5, 2002). blood donor selection criteria for United States travelers to malarious

8 Valeri CR, Ragno G, Pivacek LE, et aI. A multicenter srudy ofin vitro areas. Transfusion 1991; 31: 786-88. I

THE LANCET' Vo1361' January 11, 2003' www.thelancet.com 167

J

REVlEW .

33 Guerrero IC, Weneger BG, Schultz MG. Transfusion malaria in the 63 National Kidney Foundation. K/DOQI. Clinical practice guidelinesUnited States, 1972-1981. Ann lntern Med 1983; 99: 221. for the tteannent of anemia of chronic kidney disease.

34 Mungai M, Tegnneier G, Chamberland M, Parise M. Transfusion- Am J Kidney Dí.s 1.997; 30: S192~237.ttansmitted malaria in the United States from 1963 through 1999. 64 Carson jL, Duff A, Poses RM, et ai. Effect of anemia andN EnglJ Med 2001; 344: 1973-78. cardiovascular disease on surgical mortaliry and morbidiry. Lancet

35 Schmunis GA. Trypanosomo cruzi, the etiologic agent of Chagas' 1996; 348: 1055-60.disease: status in the blood supply in endemic and non-endemic 65 Hebert PC, Wells G, Sweedale M, et ai. Does ttansfusion practicecountries. Transfusion 1991; 31: 547-57. affect mortaliry in critically ill patients? Crit Caro Med 1997; 155:

36 Schmunis GA, Zueker F, Cruz IR, Cuchi P. Safery ofblood supply for 1618-23.infectious disease in Latin American countries. Am J Trop Med Hyg 66 Silverberg OS, Wexler D, Sheps D, et ai. The effect of correction of2001; 65: 924-30. mild anemia in severe, resistant congestive heart failure using

37 Shulman IA, Appelman S, Saxena S, Hiti AL, KirchhoffLV. Specific subcutaneous erythropoietin and inttavenous iron: a randomizedantibodies to Trypansoma cruzi among blood donors in Los Angeles, conttolled study. J Am Coll Cardio12001; 37: 1776-79.Califomia. Transfusion 1997; 37: 727-31. 67 Wu WC, Rathmore SS, Wang Y, Radford MJ, Krumholz HM.

38 Cimo PL, Luper WE, Scouroso MA. Transfusion-associated Chagas' Effectiveness ofblood transfusion in elderly patients with acutedisease in Texas: report ora case. Tex MedJ 1993; 89: 48-50. myocardial infarction. N EnglJ Med 2001,345: 1230-36.

39 Centers for Disease Conttol and Prevention. Chagas disease after 68 Goodnough L T, Bach RG. Anemia, ttansfusion, and mortaliry.organ ttansplantation-United States, 2001.JAMA 2002; 287: N EnglJ Med 2001; 345: 1272-73.1795-96. 69 Lenfant C. Transfusion should be monitored for undertransfusion as

40 McQuiston]H, Childs]E, Chamberland ME, Tabor E. Transmission well as overtransfusion. Transfusion 1992; 32: 873-74.of tick-bome agents of disease by blood ttansfusion: a review of 70 Sazama K Is undertransfusion occutring? Transfusion 2001; 41:known and potential risks in the United States. Transfusion 2000; 40: 577-78.274-84. 71 National Heart, Lung, and Blood Institute. http:!!grants.nih.gov!grants

41 Brown P, Preece MA, Will RG. Friendly tire in medicine: hormones, !guide!rfa-files!rfa-HL-02-001.html (accessed Jan 2, 2002).homografts, and Creutzfeldt-Jacob disease. Lancet 1992; 340: 24-27. 72 Goodnough L T, Brecher ME, AuBuchon JP. Transfusion Medicine.

42 Dodd RY, Sullivan MT. Creutzfeldt-Jacob disease and ttansfusion Part 11. Blood Conservation. N Engl J Med 1999; 340: 525-33.safery: tilting at icebergs? Transfusion 1998; 38: 221-23. 73 Hedner U, Erhardtsen E. Potential role for recombinant factor VIla in

43 FDA. Revised preventative measures to reduce the possible risk of ttansfusion medicine. Transfusion 2002; 42: 114-24.ttansmission of CID and vCID by blood and blood products. 74 Winslow RM. Blood substitutes-a moving target. Nat Med 1995; 1:http:!!www.fda.gov!cber!gdlins!cjdvcjd.hnn (acccessed Jan 8, 2002). 1212-15.

44 Consultant Pathologists Group Committee. Blood ttansfusion 75 Stowell CP, Levin J, Spiess BD, Winslow RM. Progress in theaccidents. BMJ 1953; 2: 390-91. development ofRBC substitutes. Transfusion 2001; 41: 287-99.

45 UndenJV, Wagner K. Voyrovich AE, SheehanJ. Transfusion errors 76 Ma Z, Monk TG, Goodnough LT, et aI. Effect ofhemoglobin- andin New York State: an analysis of 10 years' experience. Transfusion perflubron-based oxygen carriers on common clinicallaboratory tests.2001;40: 1207-13. ClinChem 1997;43: 1732-37.

46 Williamson LM, Lowe S, Love E, et alo Serious hazards of 77 Scott MG, Kucik DF, Goodnough LT, Monk TG. Blood substitutes:ttansfusion.lnitiative: analysis offirst two annual reports. BMJ 1999; evolution and furure applications. Clin Chem 1997; 43: 1724-31.319: 16-19. 78 Rioux F, Drapeau G, Marceau F. Recombinant human hemoglobin

47 Sazama K. Reports of355 ttansfusion-associated deaths: 1976 selectively inhibits vasorelaxation elicited by nitric oxide donors inthrough 1985. Transfusion 1990; 30: 583-90. rabbit isolated aortic rings. J Cardiovasc Pharmaco11995; 25: 587-94.

48 Goldman M, Remy-Prince S, Trepanier A, Decary F. Autologous 79 Winslow RM. Blood substitutes. Curr Opin Hematol2002; 9: 146-51.donation error rates in Canada. Transfusion 1997; 37: 523-27. 80 Kasper SM, Grune F, Walter M, Arnr N, Erasmi H, Buzello W. The

49 A$erican Association ofBlood Bank. Non-infectious serious hazards effects ofincreased doses ofbovinehemoglobin on hemodynamicsof ttansfusion. Bethesda, MD: American Association of Blood Bank and oxygen ttansport in patients undergoing preoperative(Association bulletin 01-4), June 14, 2001. hemodilution for elective abdominal aorric surgery. Anesth Analg

50 Aubuchon JP, Upton KS. Is it time for blood banking to take 0fI? 1998; 87: 284-91.Transfusion 2001; 41: 964-67. 81 Sloan EP, Koenigsberg M, Gens D, et ai. Diaspirin cross-linked

51 Battles IB, Kaplan HS, Van Der SchaffTW, Shea CE. The attributes hemoglobin (DCL-Hgb) in the tteannent ofsevere ttaurnaticof medical event-reporting systems. Experience with a protorype hemorrhagic shock. JAMA 1999; 282: 1857-64.medical event-reporting system for ttansfusion medicine. 82 Keipert PE. Use of Oxygent, a perfluorochemical-based oxygenArch Pathol Lab Med 1988; 122: 231-38. carrier, as an altemative to inttaoperative blood ttansfusion.

52 CallurnjL, Kaplan HS, Merkiey LL, et ai. Reporting ofnear-miss ArtifCells Blood Substit lmmobil Biotechnol1995; 23: 381-94.events for ttansfusion medicine: improving ttansfusion safery. 83 Spahn DR, van Brempt R, Theilmeier G, et ai. Perflubron emulsionTransfusion 2001; 41: 1204-11. delays blood ttansfusions in orthopedic surgery. Anesthesiology 1999;

53 Wenz B, Bums ER. Improvement in blood safery using a new blood 91: 1195-208.unir and patient identification system as part of safe ttansfusion 84 Jones JA. Red blood cell substitutes: current status. Br J Anaesthpractice. Transfusion 1991; 31: 401-03. 1995; 74: 697.

54 Jenson NJ, Crosson JT. An automated system for bedside verification 85 Maguire PJ. Stevens C, Hurnes HD, Shander A, Halpem NA,ofthe match between patient identification and blood group Pastores SM. Bioartificial organ support for hepatic, renal, andidentification. Transfusion 1996; 36: 216-21. hematologic failure. Crit Caro Clin 2000; 16: 681-84.

55 Goodnough LT, Brecher ME, Kanter MH, AubuchonJP. 86 Spahn DR, Leone BJ, RevesJG, Pasch T. Cardiovascular andTransfusion Medicine. Part I. N Engl J Med 1999; 340: 438-47. coronary physiology of acute isovolemic hemodilution: a review of

56 Besarab A, Bolton WK, Browne IK, et al. The effects of normal as nonoxygen-carrying and oxygen-carrying solutions. Anesth Analgcompared with low hematocrit values in patients with cardiac disease 1994; 78: 1000-21.who are receiving hemodialysis and epoetin. N EnglJ Med 1998; 339: 87 Cothren C, Moore EE, OfIner PJ, Haenel IB, JohnsonjL. Blood584-90. substitute and erythropoietin therapy in a severely injured Jehovah's

57 Foley RN, Parfrey PS, Hamett ID, Kent GM, Murray OC, Barre PC. Witness. N Engl J Med 2002; 346: 1097-98.The impact of anemia on cardiomyopathy, morbidiry, and mortaliry in 88 Brecher ME, Goodnough L T, Mónk T. The value of oxygen-carryingend state renal disease. Am J Kidney Dis 1996; 28: 53-61. solutions in the operative setting, as determined by mathematical

58 Canella G, LaCamma G, Sandrini M, et ai. Reversal ofleft ventticular modeling. Transfusion 1999; 39: 396-402.hypertrophy following recombinant hurnan erythropoietin tteannent 89 Mullon J, Giacoppe G, Clagett C, McCurie D, Dillard T.ofanaemic dialysed uraemic patients. Nephrol Dia! Transplant 1991; 6: Transfusions ofpolymerized bovine hemoglobin in a patient with31-37. severe autoimmune hemolytic anemia. N EnglJ Med2000; 342:

59 Collins A, Ma JZ, Ebben J, et alo Impact of hematocrit on morbidiry 1638-43.and mortaliry. Semin Nephro12000; 20: 345-49. 90 Bandarenko N, Brecher ME. United States thrombotic

60 Ma J, Ebben J, Xia H, et aI. Hct levei and associated mortaliry in thromboCytopenic purpura apheresis study group: multicenter surveyhemodialysis patients. J Am Soc Neph 1999; 10: 610-19. and rettospective analysis of current efficacy oftherapeutic plasma

61 Collins A, Xia H, EbbenJ, et ai. Change in Hct and risk ofmortaliry. exchange.JClinApheresis 1998; 13: 133-41.J Am Soc Nephro11998; 9: 204A. 91 Bell WR, Braine HG, Ness PM, Kickier TS. Improved survival in

62 Collins AJ, U S, St Peter W, et alo Death, hospitalization, and thrombotic thrombocyropenic purpura-hemolytic uremic syndrome:economic associations among incident hemodialysis patients with clinical experience in 108 patients. N EnglJ Med 1991; 325: 398-403.hematocrit values of36 to 39%.J Am Soc Nephro12001; 12: 2465-73. 92 Clark WF, Rock GA, Buskard N, et ai. Therapeutic plasma exchange:

168 THE LANCET' Vol361 .January 11, 2003' www.thelancet.com

REVIEW

an update from the Canadian Apheresis Group. Ann Intern Med 1999; 97 Gerritsen HE, Rob1es R, Lammle B, Furlan M. Partial amiDo acid131: 453-62. sequence of purified von Willebrand factor-cleaving protease. Blood

93 Furlan M, Robles R, Galbusera M, et ai. von Willebrand factor- 2001; 98: 1654-61.cleaving protease in thrombotic thrombocytopenic purpura and the 98 Zheng X, Chung D, Takayama TK, Majerus EM, Sadler JE,hemolytic-uremic syndrome. N Engl J Med 1998; 339: 1578-84. Fujikawa K. Struclllre of von Willebrand factor-c1eaving protease

94 Tsai HM, Lian EC. Antibodies to von Wil1ebrand factor-c1eaving (ADAMTS13), a metalloprotease invo1ved in thromboticprotease in acute thrombotic thrombocytopenic purpura. thrombocytopenic purpura. J Biol Chem 2001; 276: 41059-63.N EnglJ Med 1998; 339: 1585-89. 99 Goodnough LT. What is a transfusion medicine specialist? Transjusion

95 Levy GG, Nicho1s WC, Lian EC, et ai. Mutations in a member afilie 1999; 39: 1031-33.ADAMTS gene family cause thrombotic thrombocytopenic purpura. 100Noe1 L. Safe b1ood starts with me, b1ood saves 1ives. Transcripts ofNa!Ure 2001; 413: 488-94. World Health Day 2000, April 7, 2000.http://www.who.int/

96 Fujikawa K, Suzuki H, McMullen B, Chung D. Purification of multimedia/whd2000/#noe1 (accessed Jan 5, 2002).human von Willebrand factor-cleaving protease and its identification 101C Health. Millions need access to safe blood, says UN.as a new member ofthe metalloproteinase family. Blood 2001; 98: http://www.canoe.ca/HealthOO04/07 blood.hnnl (accessed Jan 5,1662-66. 2002).

;

~

1

jTHE LANCET. Vo1361 .January 11, 2003. www.thelancet.com 169