Laboratory reagents and coagulation assay procedures*

STANDARDIZATION OF DIAGNOSTIC MATERIALS 3. Laboratory reagents and coagulation assay procedures* E. A. LOELIGER 1 This paper reviews current methods of standardizing laboratory investigations in blood clotting, points out their shortcomings, and suggests ways of improving them. Procedures requiring standardization include the diagnosis of haemorrhagic diathesis and thrombotic tendency; the quantification of liver function; and monitoring of the effect of haemostatic replacement therapy and therapeutic anticoagulation. Care in collecting, storing, and handling bloodfor investigation is ofprimary importance. Achievements in the standardiza- tion of reagents are reviewed and examples of standardized methods are given. Although knowledge of the haemostatic process and experimental methods for its investigation have reached an advanced stage, their clinical application is still unsatisfactory owing to the inadequacy of national and international regulations. To improve the situation, one or more internationally approved laboratories for standardization and quality control, operat- ing according to rules elaborated by experts in the field, should be designated in each country. The need for international reference laboratories is urgent. FIELD AND LEVEL OF INVESTIGATION Haemorrhagic diathesis In the first instance, the mechanisms of primary haemostasis and blood clotting are screened. The former is accomplished by determining the (primary) bleeding time (Ivy et al., 1941; Mielke et al., 1969), counting the platelets (Ludin, 1952), and inspecting the blood film. The latter involves measuring the partial thromboplastin time, prothrombin time, and fibrinogen level, as well as testing for fibrin(ogen)- degradation products and urea solubility of the clot (clot retraction is of lesser importance). If the results of screening together with the case history yield insufficient information, laboratory in- vestigation is extended by an assessment of separate platelet functions and coagulation factors (Biggs, 1972; WHO Scientific Group on Inherited Blood Clotting Disorders, 1972). These procedures require specialized laboratories. * Presented at the International Conference on Stan- dardization of Diagnostic Materials, Center for Disease Control, Atlanta, Ga., USA, 5-8 June 1973. ' Head, Laboratory Thrombosis Research, Haematology Section, Department of Medicine, University Hospital, Leiden, Netherlands. Thrombotic tendency Conclusive data are not available to substantiate such a tendency by means of laboratory tests. A simple increase in the number of platelets is known to be accompanied by an increased number of thromboembolic events. It has been asserted that a positive ethanol gelation test-an easily performed screening procedure-is a good indicator of immi- nent thrombosis. A strongly increased Factor VIII level or decreased fibrinolytic activity also contri- butes to the formation of thrombi. Studies in hetero- zygotes for the antithrombin Ill deficiency have shown convincingly that an apparently insignificant shortage of this naturally occurring circulating anti- coagulant is associated with recurrent thrombo- embolic episodes. These determinations also call for specialized laboratories. Other procedures Quantification of the liver function and the moni- toring of anticoagulant therapy may be performed at the routine level by means of the prothrombin time test. However, for the laboratory assessment of the patient's response to replacement therapy, special laboratory facilities are mandatory. 3073 -727 STANDARDIZATION UNDESIRABLE EFFECTS OF INSUFFICIENT STANDARDIZATION Collection of venous blood Laboratory investigation of blood clotting begins with the collection of venous blood. Prolonged venous pressure results in a 10-30% increase in the protein-bound enzyme activity and in platelet num- bers. In blood that is slowly drawn or (as in the case of capillary blood) not easily obtained, the clotting mechanism may be activated. This results in exces- sively high coagulation factor activity and in a falsely low number and disturbed function of plate- lets. Therefore, blood for the investigation of blood clotting should be taken only by highly experienced personnel. Decalcification ofblood Sodium citrate should be used at a strength of 110 mm. Salts of oxalic acid display a stronger avidity for calcium ions, resulting in a rapid loss of Factors V and VIII. The use of EDTA-an even stronger calcium binder-gives rise to a rapid loss not only of Factor V and Factor VIII activity but also of the clottability of fibrinogen by thrombin, in addition to the complete disaggregation of blood platelets. Stabilization of the labile factor For the stabilization of the so-called labile factor (Factor V) buffered decalcifying solution is needed. HEPES buffer has proved to be particularly useful (Zucker et al., 1970c). Containers and pipettes Containers and pipettes in which blood and plasma are stored, transported, diluted, and otherwise hand- led, must be contact-free (siliconized or made of plastic) and clean. Only for the final step of pipetting, performed immediately before the recording of the reaction, is it possible to use disposable (micro) glass pipettes. Yet, despite this precaution, slow, although slight, activation of coagulation factors is regularly observed during the first 6-12 hours after blood sampling. Storage temperature The blood to be investigated should be stored at 18-23°C. At low temperatures (0-40C) the activity of Factor VII may increase to many times its normal level through cold-activation of plasma kallikrein (Gj0nnaess, 1972). The addition of kallikrein inhi- bitors does not suppress this activation (Brozovic & Gurd, 1972). For storing blood or plasma at room temperature for several days, the addition of small amounts of sodium azide (0.1 mg/ml plasma) has been shown to prevent bacterial growth (E. A. L., unpublished observations, 1969). Dilution The dilution step in the assay procedure is of particular importance, because it introduces so many unexpected phenomena-e.g., at low ionic strength, the fibrin clot, under the influence of thrombin, forms more rapidly and more easily than when the ionic strength of the reaction medium is " normal ". A second example: the magnitude of the final dilution of the plasma to be investigated has proved to be of special importance in the control of oral anticoagu- lation, under which conditions a competitive inhi- bitor-the so-called protein induced by vitamin K absence and antagonists (PIVKA)-enters the cir- culation. The final concentration is of great impor- tance for the result obtained with PIVKA-sensitive methods, such as the thrombotest and tests using human-brain thromboplastin. When PIVKA-sensi- tive procedures are used to test blood rather than plasma, the correction factor for the venous haema- tocrit is much smaller than might be expected from the differences in the amount of plasma cal- culated for various haematocrits. Another example of the importance of the dilution step is the strength of the phospholipid suspension used in the partial thromboplastin time test, as high-activity suspensions easily suppress the activity of the circulating anti- coagulant occurring in autoimmune diseases. As a consequence of the use of phospholipid suspensions of different strengths for the assessment of individual coagulation factors, acquired isolated coagulation factor deficiencies may be erroneously diagnosed (Largo et al., 1972; Castro et al., 1972). Finally, in the procedure for thromboplastin calibration (Biggs & Denson, 1967b), the final dilution of the thrombo- plastin to be calibrated must be exactly defined, because the sensitivity of a thromboplastin rises as the dilution is increased. Centrifugation The centrifugation step is also highly important, not only in platelet function tests, but also in coagu- lation research-e.g., after adsorption of the factors of the prothrombin complex where high-speed centri- fugation (30 min at 20 000 g) is needed to eliminate the residual aluminium hydroxide (Loeliger et al., 1973). 728 COAGULATION PROCEDURES & REAGENTS Reagents The example of the large differences in the pro- longation ratio (patient's coagulation time: normal coagulation time) obtained with the various thrombo- plastin preparations in use-differences that are far from eliminated by translating the prolongation ratio into a conventional percentage of activity (Table 1)-is sufficient to demonstrate the urgent need for standardization in this respect. Reading of results The type of reading used for the registration of clotting times has been shown to be important. Particularly where the clotting process is relatively slow-e.g., in the method for assessing the so-called prothrombin and proconvertin complex (P & P method according to Owren & Aas, 1951)-well- standardized automatic reading greatly improves the comparability of results obtained in different labora- tories (E. A. L., unpublished observations, 1972). Mechanized coagulometry for Quick's original one- stage test, in skilled hands, does not perform signi- ficantly better than manual testing (Fewell et al., 1972); automatic reading is, however, more econo- mical. Although coagulometers may measure shorter coagulation times than hand reading, prothrombin time ratios (specimen test in s: control test in s) appear to be independent of the type of reading Table 1. Prolongation ratios a obtained with various assay procedures, and their equivalents in conventional percentages (according to Biggs & Denson, 1967a) Assay procedure Prolongation ratio percentage Quick's procedure Simplastin b 1.45-2.1 18-30 human brain (BCT) 2.5 -4.0 15-26 P & P method (Owren & Aas, 1951) 2.8 -5.0 Ware & Stragnell, 1952 1.6 -2.4 10-20 Simplastin A b' 2.5 -4.6 Thrombotest C 2.7 -4.1 5-10 International percentage 7-14 a Prolongation ratio = patient's coagulation time: normal coagulation time. b Warner Lambert; Nyersey; etc. c Nyegaard. (Zucker et al., 1970a). The choice of the most appropriate apparatus from among the large variety of automatic clot timers is difficult. The fibrometer system has become the most frequently used instru- ment in the USA. Photoelectric clot timers have also been found useful (Sibley & Singer, 1972). Schnitt- ger's coagulometer is particularly suitable for the assessment of Factors VIII, IX, XI, and XII (Velt- kamp et al., 1968). Observer bias With regard to the laboratory record of the results obtained-i.e., the observer's bias-the observer should know that results for haemostatic parameters are subject to wide biological variations and experi- mental error. Biological as well as analytical coeffi- cients of variation may amount to 25% (Veltkamp et al., 1968; Goldenfarb, 1971). Thus unexpected results are often obtained, and these must be recorded and taken into account. Other factors Certain other factors, such as the temperature at which the reaction takes place and the successive steps of the test procedure, should be defined exactly, the latter particularly in regard to the amounts of reactants used, the order of addition, and the time of incubation of the different reactants. If the result of an assay is expressed in relation to a normal reference sample, the two should be investigated in parallel because of instabilities of the biological material. STANDARDIZATION OF REAGENTS AND METHODS Relatively poor results have been obtained, even at the expert level, in the standardization of methods and reagents, although it is known that inadequate standardization leads to after-effects that are extreme- ly dangerous for the individual patient. There can be no doubt that the aimost total lack of formal regulations pertaining to standardization and quality control of laboratory investigations is responsible for the present situation. Only reagents that are used as drugs, e.g., heparin, thrombin, and thrombolytic agents, are subject to legal controls (pharmacopoeias). Most of the available coagulation assay procedures are bioassays. Thus the active material to be assessed can be estimated only by observing the response that it evokes in vivo or in vitro and by comparing it with the response caused by a known amount of the active material and dilutions of it. The known 729 730 STANDARDIZATION amount must therefore refer to well-defined standard reference material.1 To illustrate current methods of standardizing clotting reagents, the examples of heparin, blood clotting Factors VIII and IX, and tissue thrombo- plastins are given below. Other tests and reagents will be referred to only briefly. Heparin Heparin is issued with a potency value in units per weight, the formula 100 IU/mg being generally applied, although the activity per mg of recent preparations is as high as 160 IU. Standardization ofthe biological activity ofheparin is based solely on its coagulation-inhibiting effect in vitro. Points of references are: Internationally: a standard established and ap- proved by WHO (the latest in 1973); Nationally: the British Pharmacopoeia and the United States Pharmacopeia standard preparations; Locally: working standards or the so-called house standards used by commercial firms to calibrate the batches that they produce. Heparin appears to have been satisfactorily standardized, judging by its 30 years of widespread use without reports of major difficulties by either clinicians or pathologists. Yet, considering not only that heparin is prepared from different tissues (e.g., bovine lung; intestinal mucosa of various animals, notably the pig) and consists of mixtures of compo- nents with many divergent chemical and physical properties, but also that its calibration is based on an assay procedure devised some 20 years ago, the existence of rather large differences in biological activity between different batches would hardly be surprising. The repeated warnings given during the past 10 years (for references see Jaques, 1972), and observations of marked differences in clinical effectiveness between two preparations (Goldberg et al., 1972), should not be underestimated. Other authors have reported that subjects treated with heparin prepared from porcine gut require 25% less protamin sulfate as an antidote than those treated with lung heparin (Sekhar et al., 1971). These obser- vations call for a reassessment of the standardization procedure (Jaques & Kavanagh, 1972), although a 1lInternational standards approved by WHO are made available by the WHO International Laboratory for Bio- logical Standards, National Institute for Biological Standards and Control, Holly Hill, Hampstead, London, England. Working standards are produced and made available by national institutes; and house standards refer to preparations produced locally. recent international collaborative study, sponsored by WHO, did not disclose major discrepancies, in vitro, between heparin from different sources (Bang- ham & Woodward, 1970). Examples of how to compare heparin from different origins in vivo are to be found in the recent literature (Gomez-Perez, 1972; Baltes et al., 1973). Fortunately, no significant differences between hog intestinal mucosa and bovine lung heparin has been found. Nevertheless the current International (intestinal mucosa) Standard has re- placed the 1958 bovine lung standard, since about 80% of all commercially available heparin prepara- tions are now made from pig intestine (Bangham & Woodward, 1970). Blood coagulation Factors VIII and IX For patients suffering from haemophilia A and B blood products of fairly constant quality are being obtained, provided that standard procedures of pre- paration are followed strictly. The in vivo activity of cryoprecipitate concentrated Factor VIII can easily be predicted from the activity assessed in vitro (Meyer et al., 1967) except that, for Factor IX concentrates, it is dangerous to rely on such an assessment. Many Factor IX concentrates contain activated Factor IX, which disappears rapidly in vivo (Bruning & Loeliger, 1971). Substitution therapy for major surgery, particularly in haemophilia B patients, must be checked once or twice daily to ensure haemostatic levels of Factor IX (Biggs, 1972). Diagnosis, the control of treatment, and genetic counselling for possible carriership of the haemo- philic gene (Veltkamp et al., 1968) require both highly standardized assay methods and reference materials. For the assessment of Factor VIII, natio- nal reference centres can rely on an international reference preparation that has been available since 1970 at the WHO International Laboratory for Bio- logical Standards (Bangham et al., 1971). This re- ference preparation, containing purified Factor VIII derived from cryoprecipitate, has been shown to fulfil the requirements of a long-term standard. It is important to keep in mind that freeze-dried plasma preparations are less stable, but-like any other properly prepared and stored plasma prepara- tion-can serve as a working standard for at least one year. The reason why national standardization is developing so slowly may be that most countries do not have national expert panels and reference laboratories and individual laboratories can rely on pooled normal plasma-so-called house standards- stored at -70°C (Aronson, 1972). COAGULATION PROCEDURES & REAGENTS For haemophilia B-i.e., treatment with Factor IX -a subcommittee of the International Committee on Haemostasis and Thrombosis (ICHT), still active- ly exploring the feasibility of in vitro assessment of Factor IX (concentrates), is attempting to define one or more international reference preparations. The first results were expected to be available by the end of 1973. It may already be concluded that future reference material should be derived from plasma and should not contain activated Factor IX. Hence, a test for activated Factor IX must also be developed to exclude possible contaminations with activated Factor IX in the sample to be tested. A nonactivated partial thromboplastin time might serve this purpose (Aronson, 1972). From the results of the collaborative studies on standardization of the Factor VIII and IX procedure published so far, it appears that reliable methods do exist: a highly standardized (simplified) two- stage assay procedure (Denson, 1967) and an equally well standardized one-stage assay procedure (ori- ginally introduced by Hardisty & Macpherson, 1962, and modified by Veltkamp et al., 1968). Tissue thromboplastins Thromboplastin calibration is one of the main concerns of standardization of the prothrombin time test procedure, since it permits conversion to an international scale of the highly divergent results obtained with the various laboratory methods and thromboplastins in use at present for the control of coumarin-induced hypocoagulability. The ultimate goals of standardization are to achieve an internationally accepted definition of the so-called therapeutic range of hypocoagulability in patients on oral anticoagulation therapy (Loeliger, 1972) and to determine how this is expressed in the results of both a standard thromboplastin and a standard test (and also, if required, in the results obtained by a given thromboplastin in a particular test employed locally-although this might be diffi- cult to achieve, since certain thromboplastins and/or test procedures might prove to be unsuitable for anticoagulant control, e.g., because of the marked insensitivity to the defect produced by anticoagulants in the blood of the treated patients). In the USA, the College of American Pathologists adopted the principle that standardization should be based on the performance, in a strictly defined pro- thrombin time assay procedure, of reagents in re- lation to clearly defined reference materials (Miale & LaFond, 1969). Plasmatic reference material has been developed with which the " therapeutic range " is defined (Miale & Kent, 1972). The Center for Disease Control in Atlanta (CDC) has also contri- buted to the standardization of the one-stage pro- thrombin time procedure: it established that com- mercial rabbit tissue reagents are sufficiently sensitive to provide safe anticoagulant control, although they are less sensitive than the human-brain thrombo- plastin and thrombotest reagent in assessing the anti- coagulant effect (Zucker et al., 1970b). In Great Britain, Poller has been responsible for successful national standardization based on the so- called British comparative thromboplastin (BCT) and referred to as the British system (Poller, 1964). This system also implies a definition of the " thera- peutic range " (Poller & Loeliger, 1969; Bailey et al., 1971). On the international level, the so-called equivalent ratio method (Biggs & Denson, 1967b) has been developed by ICHT. The validity in practice of the assumptions made by Biggs and Denson has been proved (Loeliger & Hemker, 1969). Five ICHT- approved trial reference thromboplastins are main- tained at the Division of Biological Standards of the British National Institute of Medical Research (Biggs & Bangham, 1971) for the calibration of national and commercial thromboplastin preparations. The currently recommended calibration procedure (Denson, 1971) requires freshly prepared plasma from 4 normal individuals and from at least 20 patients stabilized on oral anticoagulants; a cali- bration table is used for the conversion of the results into international terms. Since this procedure is complex, ways of simplifying it are being sought. Lyophilized pooled plasma substituted for fresh plasma gave promising results (Bangham et al., 1973). A given set of such plasma has been shown to re- main useful even after heating at 37°C for several months (E. A. L., unpublished observations, 1972), and artificially prepared lyophilized plasma may also be useful, at least for the calibration of PIVKA- insensitive thromboplastins. Lyophilized plasma is insufficiently stable, however, to be used as long-term reference material (Brozovic et al., 1971a, 1971b), although there is good reason to believe that it can serve as short-term reference material (Miale & Kent, 1972). With the development of standardization by means of thromboplastin and plasma reference material, the ultimate goal of standardization may be at hand, as many centres with long experience already appear to be recommending therapeutic ranges closely simi- 7 731 STANDARDIZATION Table 2. Therapeutic ranges recommended by various research centres (data according to Biggs & Denson, 1 967a) Therapeutic range Centre Biggs/Denson proposed ratio ~~internationalratio percentage Basle (human brain) 1.85-2.6 7-14 Leiden (thrombotest) 2.6 -4.0 7-13 London (rabbit brain) 1.6 -2.0 9-14 Manchester (human brain) 1.86-3.0 8-17 New York (rabbit brain) 1.5 -2.0 7-13 lar to those shown in Table 2 (Biggs & Denson, 1967a). The optimum range appears to be 7-14% (international scale), and this has given excellent clinical results, also in atherothrombosis (Loeliger et al., 1967; Hamming et al., 1965; Meuwissen et al., 1969). As there is insufficient information on the relative merits of the various types of plasma and thrombo- plastin reference materials, an international Colla- borative Study on Prothrombin Time Standardiza- tion was planned for 1973-74 under the auspices of the International Committee on Standardization in Haematology (ICSH), with the CDC, Atlanta, as the organizing centre. It is worth mentioning that the calibration of thromboplastins differs from the classical biological standardization only in that plasma reference mate- rial prepared from patients is used, the response of reference thromboplastin to this material being taken as the baseline. The definition of plasma reference material in terms of coagulation factor activity has been suggested (Loeliger et al., 1970). Other tests and reagents Tests for intravascular coagulation, fibrinolysis, and thrombolysis. One of the first laboratory signs of diffuse intravascular coagulation, particularly in the case of a rapidly growing thrombus, may be soluble fibrin detected by the protamine sulfate and ethanol gelation test (Godal & Abildgaard, 1966; Lipinski & Worowski, 1968; Konttinen et al., 1972). The Biogel column method (Fletcher & Alkjaersig, 1971) seems promising but is not yet suitable for routine use. Imperfect venepuncture leads to positive results. For the diagnosis of overt diffuse intravascular clotting and fibrinolysis, immunological test pro- cedures are usually applied. The qualitative and quantitative immunodiffusion and precipitation tests are widely used (Ouchterlony, 1964; Mancini et al., 1965). More sensitive quantitative tests are a well- standardized tanned red cell haemagglutination in- hibition immunoassay (TRCHII) (Merskey et al., 1969 and 1971) and, to a lesser degree, the staphylo- coccal clumping test (SCT) (Hawiger et al., 1970), the results of which are expressed in fibrinogen equivalents. For the differentiation of intravascular coagulation and fibrinolysis, which might have impoi- tant therapeutic repercussions, turn-over studies and immunochemical tests of fibrin(ogen)-degradation products are under study (G. Izak, unpublished observations, 1973). Standardization of the SCT has been proposed (Niewiarowski & Thomas, 1971). The results ob- tained with these techniques in different laboratories compare rather well, even when the same reagents are not used (Sherry & Johnson, 1971). The Sub- committee on Fibrinolysis and Thrombolysis of ICHT has also considered standardization of fibrin- (ogen)-degradation products (Sherry & Johnson, 1971). In general, increased or decreased fibrinolytic activity can be tested by the euglobulin lysis test. However, this test is of limited value as the result depends on so many imponderables; furthermore, the lytic activity of highly diluted plasma cannot be representative of undiluted blood. In clinical practice the test should be abandoned. If the activity of the naturally occurring activator is to be assessed in experimental work, more specific tests should be used, including the assessment of fibrinolysis-in- hibiting globulins. The standardization of fibrinolytic agents and assay methods was assigned mainly to a working subcommittee of the Committee on Thrombolytic Agents (CTA) of the US National Heart Institute. Streptokinase was standardized first (Bangham & Walton, 1965), and in the early '60s a lyophilized research reference preparation (working standard) of urokinase, the potency of which was defined by CTA, was developed and was subsequently approved by WHO. Unfortunately, a collaborative study car- ried out in 1967 under the auspices ofWHO has not yielded sufficient data to establish the material in question as an international urokinase standard (WHO Expert Committee on Biological Standardi- zation, 1969). 732 COAGULATION PROCEDURES & REAGENTS The National Heart Institute also prepared stan- dard batches of such substrates as plasminogen, fibrinogen, a-casein, and the methyl ester of acetyl- lysine, which can be obtained on request and used in working-standard methods for the assay of plas- min, plasminogen, and urokinase (Johnson et al., 1969). Glycerol-activated plasmin has been under study since 1971. Human plasminogen is assayed by activating it to plasmin with 200 units of strepto- kinase or 2 000 CTA units of urokinase. The amount of plasmin produced is determined by comparison with a standard plasmin curve. The potency of plasmin(ogen) has been defined in CTA terms. Bleeding time, platelet adhesiveness, and platelet aggregation. An international cooperative study on the bleeding time test, designed to set standard conditions (Born & Mason, 1971), has not been carried out yet. The need for such a study has even been doubted, because of the availability of well- defined procedures such as the tests according to Ivy (Ivy et al., 1941) and Borchgrevink (Borchgre- vink, 1958), or one of their modifications (Mielke et al., 1969). The results obtained from a cooperative study on platelet adhesiveness were rather disappointing: the result of the Salzman test for platelet adhesiveness to glass beads does not make it substantially easier to diagnose von Willebrand's disease (Murphy & Salzmann, 1972). The standardization of the turbidimetric method for quantification of platelet aggregation is still in the pilot-study stage. Optimum test performance requires the use of freshly prepared platelet-rich plasma. There seems to be little difference between intraindividual and interindividual variation, both of which are large (Goldenfarb et al., 1970). It has been established that a correction should be made for the anticoagulant when haematocrit values are abnormal. Plasma should be prepared by centrifu- gation at 1 000 rev/min for 15 min, kept at room temperature, and used within one hour (Born & Zucker, 1972). Fibrinogen, thrombin, and antithrombin III. Stan- dardization of the quantification of clottable fibrino- gen (Factor I) constitutes a problem in clinical che- mistry when the amount of washed fibrin threads is assessed gravimetrically or chemically. It should be realized, however, that contamination of the washed fibrin threads with other proteins leads to inconsis- tent results (Beck, 1970). Biological assay procedures (thrombin clotting times of the diluted plasma to be tested) have been developed, and these are especially convenient in clinical emergencies (Clauss, 1957; Vermeylen et al., 1963). Although they are highly useful in practice, biological tests cannot be con- sidered as specific either: falsely low values will be obtained in case of circulating fibrin(ogen)-degrada- tion products. Thrombin activity is currently expressed in NIH units. Research reference preparation 66/305, con- taining 200 units per ampoule, is available from the WHO International Laboratory for Biological Stan- dards. A collaborative study under the auspices of WHO was started in 1971 to develop an international standard preparation (Sherry & Johnson, 1971). This preparation is now ready for approval (A. J. John- son, unpublished observations, 1973). The assessment of antithrombin III activity is of great importance for the diagnosis or exclusion of a thrombotic tendency. The acquisition of a well- standardized and reproducible assay method is therefore most welcome (Abildgaard et al., 1970). Standardization of the test and its comparison with immunological assay procedures are being investi- gated at the WHO International Laboratory for Biological Standards. Determination ofseparate coagulation factors. One- stage assay procedures appear to be highly suitable for standardizing the determination of the separate coagulation factors. So far no attempts have been made to standardize any of these tests. The problem, however, is not too difficult to solve once the princi- ples of thromboplastin and partial thromboplastin standardization have been elucidated. More elabo- rate techniques will be needed to assess activated coagulation factors (particularly Factors VII and IX). Finally, a standardized test for PIVKA has to be developed, for which purified staphylocoagulase should be available. Systems ofstandardization Local. It will be necessary to adopt methods of investigation referred to in this paper and recom- mended by experienced authors (Biggs, 1972) until approved standardized methods become available at the local level. For reagents, calibrated working or home standards should be applied. In addition, a stringent quality-control system (Whitby et al,. 1967) is indispensable. National. Regulations at this level are scarce. In the USA, NIH provides a series of working standards and prescriptions for working-standard methods, and 733 734 STANDARDIZATION prothrombin time proficiency testing programmes are operated. The British system of anticoagulant control (Poller, 1964), maintained under the auspices of the British Anticoagulant Panel of the British Com- mittee for Standardization in Haematology, conti- nues to improve oral anticoagulation on the natio- nal level, and it is hoped that the system will be introduced into other countries (Bailey et al., 1971). Dr Poller's centre 1 has recently been approved by the British Anticoagulant Panel as a national reference laboratory. The Dutch system of standardization, which has been introduced under the auspices of the Dutch Fed- eration of Thrombosis Services, relies on thrombo- plastin calibration with lyophilized plasma pre- parations from healthy individuals and patients (Loeliger et al., 1972). Artificially prepared normal and abnormal blood and plasma will be used for quality control. International. No recommendations concerning standard methods have ever been made at the inter- national level, with the exception of working stan- dard-methods to be used for the assessment of fibrinolytic agents (Johnson et al., 1969). With regard to reagents, a small series of well-standardized reference materials is available at the WHO Inter- national Laboratory for Biological Standards. There is good hope that, for the one-stage prothrombin time test, internationally approved standard material and test prescriptions will be available in due course through the cooperative activity of ICSH and ICHT. CONCLUSIONS Standardization of methods for the investigation of the clotting process and of the reagents used has lagged far behind the rapidly increasing knowledge that has been accumulated in the fields of haemo- stasiologic methodology and coagulation biochemis- try. Many laboratories are unaware that the tests and techniques they use are obsolete and unreliable. The reason for this situation is an almost total lack of national regulations. The situation could be rapidly improved if natio- nal public health inspection services, upon the re- commendations of expert committees of national haematology societies and/or colleges of pathologists with industrial representation, were backed by legis- lation for the use of standard reference material and methods. For the coordination of these activities, private or governmental national reference labora- tories would need to be designated. These would be responsible for providing all the facilities for the standardization of methods and reagents for the indi- vidual laboratories, and would operate according to internationally approved directives. They would have to introduce monitoring systems and proficiency- testing programmes to safeguard the quality of stan- dard preparations and to improve the performance of diagnostic tests. The regular licensing of indivi- dual laboratories should be considered. International facilities for control would need to be created as a consequence of national regula- tions: international reference laboratories should be designated upon the recommendations of expert panels instituted by existing international groups. These international laboratories would calibrate and monitor national reference material and commercial preparations, and organize international proficiency testing trials to define the status of the field at the international level. Finally, an international clearing house could coordinate the standardization activities and promulgate the results achieved. 1 National Anticoagulant Control Reagents Laboratory, Withington Hospital, West Didsbury, Manchester, England. ACKNOWLEDGEMENTS The author expresses his sincere thanks to Professor G. I. C. Ingram, Department of Haematology, Louis Jenner Laboratories, St Thomas's Hospital and Medical School, London, and to many other participants in the WHO/CDC meeting on Standardization of Diagnostic Materials, held in Atlanta, Ga., USA, 5-8 June 1973, for their constructive comments and suggestions. COAGULATION PROCEDURES & REAGENTS 735 R,]SUMt STANDARDISATION DES REACTIFS ET DES TECHNIQUES DE LABORATOIRE DANS LE DOMAINE DE LA COAGULATION SANGUINE La standardisation des techniques de laboratoire dans le domaine de la coagulation sanguine concerne d'une part le diagnostic des diatheses hemorragiques et de la propension aux thromboses et d'autre part lWtude de la fonction h6patique, la surveillance des traitements anti- coagulants et le controle de la r6ponse des malades aux therapeutiques substitutives dans les troubles de la coa- gulation. La standardisation des methodes apparait necessaire aux divers stades des investigations: a) prelevement du sang veineux, materiel i utiliser pour sa conservation et son transport au laboratoire; b) manipulation et traite- ment de l'echantillon au laboratoire meme: risques d'ac- tivation, ajustement de l'equilibre ionique et du pH, dilution, centrifugation, temp6ratures de conservation et de r6action, execution des divers temps de la r6action; c) lecture et interpr6tation des resultats. Quelques resultats ont e obtenus en ce qui regarde la standardisation des reactifs. Un petit nombre d'entre eux, comme l'heparine, la streptokinase, l'urokinase et le facteur VIII, utilis6s non seulement au laboratoire, mais aussi pour le traitement de certaines affections, ont d6ja fait l'objet d'une standardisation a l'echelle inter- nationale. Pour d'autres, les etudes a ce sujet sont actuellement en cours. C'est le cas notamment pour le facteur IX, la thromboplastine tissulaire, la profibro- lysine, les produits de d6gradation de la fibrine et l'anti- thrombine III. L'auteur expose l'etat actuel de la question et plaide en faveur d'une r6vision r6guliere des processus de standardisation allant de pair avec les progres de nos connaissances concernant les ph6nom6nes de coagula- tion sanguine et le perfectionnement des techniques. II met I'accent sur l'importance de la mise en application de systemes de standardisation aux 6chelons international, national et local. Un contr6le international et une regle- mentation nationale sont indispensables pour obtenir et maintenir un niveau de qualite satisfaisant des diagnostics de laboratoire dans la pratique courante. REFERENCES Abildgaard, U. et al. (1970) Thrombos. Diathes. haemorrh. (Stuttg.), 24, 224-229 Aronson, D. L. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 51, 308-311 Bailey, E. L. et al. (1971) Canad. med. Ass. J., 105, 1041-1044 Baltes, B. J. et al. (1973) Clin. Pharmacol. Ther., 14, 287-290 Bangham, D. R. & Walton, P. L. (1965) Bull. Wld Hlth Org., 33, 235-242 Bangham, D. R. & Woodward, P. M. (1970) Bull. Wld Hlth Org., 42, 129-149 Bangham, D. R. et al. (1971) Bull. Wld Hlth Org., 45, 337-351 Bangham, D. R. et al. (1973) Thrombos. Diathes. haemorrh. (Stuttg.), 29, 228-239 Beck, E. A. (1970) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 39, 323-329 Biggs, R. (1972) Human blood coagulation, haemostasis and thrombosis, Oxford, Blackwell Scientific Publica- tions Biggs, R. & Bangham, D. R. (1971) Brit. med. J., 3, 470-471 Biggs, R. & Denson, K. W. E. (1967a) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 26, 445-450 Biggs, R. & Denson, K. W. E. (1967b) Brit. med. J., 1, 84-88 Borchgrevink, C. F. (1958) Acta med. scand., 162, 361-374 Born, G. V. R. & Mason, R. G. (1971) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 45, 413-414 Born, G. V. R. & Zucker, M. B. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 51, 337-340 Brozovi6, M. & Gurd, L. (1972) Scand. J. Haemat., 9, 509-515 Brozovi6, M. et al. (1971a) Brit. J. Haemat., 21, 201-208 Brozovic, M. et al. (1971b) J. clin. Path., 24, 690-693 Bruning, P. F. & Loeliger, E. A. (1971) Brit. J. Haemat., 4, 377-398 Castro, 0. et al. (1972) Ann. intern. Med., 77, 543-548 Clauss, A. (1957) Acta haemat. (Basel), 17, 237-246 Denson, K. W. E. (1967) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 26, 419-421 Denson, K. W. E. (1971) J. clin. Path., 24, 460-463 Fewell, R. G. et al. (1972) J. med. Lab. Technol., 29, 147-151 Fletcher, A. P. & Alkjaersig, N. (1971) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 45, 389-394 Gj0nnaess, H. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), 28, 155-205 Godal, H. C. & Abildgaard, U. (1966) Scand. J. Haemat., 3, 342-350 Goldberg, E. et al. (1972) Lancet, 1, 789 Goldenfarb, P. B. (1971) Amer. J. clin. Path., 55, 561-564 Goldenfarb, P. B. et al. (1970) Atherosclerosis, 12, 335-340 736 STANDARDIZATION Gomez-Perez, F. (1972) J. clin. Pharmacol., 12, 413-416 Hamming, J. J. et al. (1965) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 21, 405 Hardisty, R. M. & Macpherson, J. C. (1962) Thrombos. Diathes. haemorrh. (Stuttg.), 7, 215-229 Hawiger, J. et al. (1970) J. Lab. clin. Med., 75, 93-108 Ivy, A. C. et al. (1941) J. Lab. clin. Med., 26, 1812-1822 Jaques, L. B. (1972) Lancet, 2, 1262 Jaques, L. B. & Kavanagh, L. (1972) Lancet, 2, 1315 Johnson, A. J. et al. (1969) Thrombos. Diathes. haemorrh. (Stuttg.), 21, 259-272 Konttinen, Y. P. et al. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), 28, 342-350 Largo, R. et al. (1972) Schweiz. med. Wschr., 102, 1612-1613 Lipinski, B. & Worowski, K. (1968) Thrombos. Diathes. haemorrh. (Stuttg.), 20, 44-49 Loeliger, E. A. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), 28, 109-119 Loeliger, E. A. & Hemker, H. C. (1969) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 35, 101-105 Loeliger, E. A. et al. (1967) Acta med. scand., 182, 549-566 Loeliger, E. A. et al. (1970) Thrombos. Diathes. haemorrh. (Stuttg.), 23, 569-584 Loeliger, E. A. et al. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 51, 331-334 Loeliger, E. A. et al. (1973) Thrombos. Diathes. haemorrh. (Stuttg.), 29, 211-212 Liudin, H. (1952) Acta haemat. (Basel), 7, 342-360 Mancini, G. et al. (1965) Int. J. Immunochem., 2, 235-254 Merskey, C. et al. (1969) Proc. Soc. exp. Biol. (N.Y.), 131, 871-875 Merskey, C. et al. (1971) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 45, 351-355 Meuwissen, 0. J. A. T. et al. (1969) Acta med. scand., 186, 361-368 Meyer, K. et al. (1967) Folia med. neerl., 10, 40-60 Miale, J. B. & Kent, J. W. (1972) Amer. J. clin. Path., 57, 80-88 Miale, J. B. & LaFond, D. (1969) Amer. J. clin. Path., 52, 154-160 Mielke, C. H. et al. (1969) Blood, 34, 204-215 Murphy, E. A. & Salzman, E. W. (1972) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 51, 341-376 Niewiarowski, S. & Thomas, D. P. (1971) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 45, 373-376 Ouchterlony, 0. (1964) Gel-diffusion techniques. In: Immunological methods, Oxford, Blackwell Scientific Publications, pp. 55-78 Owren, P. A. & Aas, K. (1951) Scand. J. clin. Lab. Invest., 3, 201-208 Poller, L. (1964) Lancet, 1, 491-493 Poller, L. & Loeliger, E. A. (1969) Lancet, 1, 464 Sekhar, N. C. et al. (1971) Clin. Res., 19, 430 (Abstr.) Sherry, S. & Johnson, A. J. (1971) Thrombos. Diathes. haemorrh. (Stuttg.), Suppl. 45, 348-349 Sibley, C. & Singer, J. W. (1972) Amer. J. clin. Path., 57, 369-372 Veltkamp, J. J. et al. (1968) Thrombos. Diathes. haemorrh. (Stuttg.), 19, 279-303 Vermeylen, C. et al. (1963) Acta clin. belg., 18, 200-207 Ware, A. G. & Stragnell, R. (1952) Amer. J. clin. Path., 22, 791-797 Whitby, L. G. et al. (1967) Advanc. clin. Chem., 10, 65- 156 WHO Expert Committee on Biological Standardization (1969) Wld Hlth Org. techn. Rep. Ser., No. 413 WHO Scientific Group on Inherited Blood Clotting Dis- orders (1972) Wld Hlth Org. techn. Rep. Ser., No. 504 Zucker, S. et al. (1970a) Amer. J. clin. Path., 53, 340-347 Zucker, S. et al. (1970b) Amer. J. clin. Path., 53, 348-354 Zucker, S. et al. (1970c) Amer. J. clin. Path., 53, 924-927