Long-Term French Experience in INR Standardization by a Procedure Using Plasma Calibrants

Transcription

1 COAGULATION AND TRANSFUSION MEDICINE Long-Term French Experience in Standardization by a Procedure Using Plasma Calibrants LUCIE L. HOUBOUYAN, MD, AND ALAIN F. GOGUEL, MD The International Normalized Ratio () has not lowered the interlaboratory differences in prothrombin time (PT) values to the extent expected, mainly because of the instrument-dependency of the International Sensitivity Index (ISI) and other factors (eg, accurate determination of the ISI, the normal value used in the PT ratio). The procedure (PPC) using plasma calibrants (reference lyophilized plasmas with assigned activity) has been evaluated since 1977 in nine French national external quality assessment surveys (NEQAS) involving approximately 4,000 laboratories and numerous local thromboplastin technique combinations. The PPC was compared with the conventional procedure (using the manufacturer's ISI), and the efficiency of antivitamin K-calibrated (AK Cal) plasmas from patients receiving oral anticoagulants vs artificially depleted plasma calibrants was also evaluated. The PPC efficiently standardized PTs with AK Cal plasmas, reducing interlaboratory variability (eg, coefficient of variation, 12% to 6% for survey 92 D) and reagent-instrument effects. However, AK Cal plasmas have drawbacks, such as limited supply, cost, and batch-to-batch variability. The artificially depleted plasma calibrants were less efficient, but usable if carefully prepared. The value of this simple procedure is that local practices are considered in the determination of PT, thus correcting for coagulometer effects and avoiding use of the manufacturer's ISI and need for a normal control plasma. These large-scale French surveys have demonstrated the validity of PPC through 15 years of experience and have shown that it offers the best compromise available in PT standardization. (Key words: Prothrombin time; International Normalized Ratio []; Laboratory standardization; Lyophilized plasma; External quality assessment) Am J Clin Pathol 1997;108: The International Normalized Ratio () was introduced to provide an international expression for prothrombin time (PT) that compensates for differences in responsiveness of thromboplastins, 1-3 thereby standardizing PT reporting and improving the safety and effectiveness of oral anticoagulant treatment, 4 ' 5 even at low dosages. However, the use of continues to be problematic. 6-9 Various reports highlight difficulties encountered in the following: (1) determination of an accurate International Sensitivity Index (ISI) related to a problem of calibration precision because of the reference thromboplastin calibration 10 and to the manufacturer's calibration of the commercial thromboplastin ; (2) the choice of the normal plasma value used in the PT ratio 13,14 ; and (3) above all, the instrument-dependency of the ISI, because the ISI is still determined manually according to From the Department of hnmuno-hematology, Hospital Ambroise Pare, Boulogne, and Medical School, Faculty of Paris-Ouest, Garches, France. Manuscript received August 14, 1996; revision accepted January 8,1997. Address reprints requests to Dr Houbouyan: Immuno- Hematology, Hospital Ambroise Pare, 9 avenue Charles de Gaulle, Boulogne, France. World Health Organization recommendations, 1 whereas most laboratories now use coagulometers. The ISI given by the manufacturer is therefore no longer valid for instrument-measured PTs. Furthermore, the considerable influence of local technical practices is reflected by systematic error in data dispersion. 20 Recently published data on the use of lyophilized plasmas in calibration prompt us to report the long-standing experience of French national external quality assurance surveys (NEQAS) 24 in PT standardization by means of the procedure using plasma calibrants (PPC) Since 1977, we have evaluated the PPC through nine large-scale interlaboratory surveys The PPC was used to correct PTs expressed in percentage of activity, and from 1986 onward to correct PTs expressed in. This article focuses mainly on the surveys dealing with. The PPC was compared with the conventional procedure (using the manufacturer's ISI) to assess its reliability and to evaluate its efficiency in two systems of calibration, the antivitamin K-calibrated (AK Cal) plasmas collected from patients undergoing stabilized oral anticoagulant therapy vs artificially depleted calibrated (ART Cal) plasmas. Table 1 gives the different systems of calibration and test plasma sources used since

2 84 COAGULATION AND TRANSFUSION MEDICINE TABLE 1. FRENCH NATIONAL SURVEYS INVOLVED WITH CALIBRATED PLASMA PROCEDURE Plasma Calibrants Test Plasma Survey Expression ofpt ART ART AVK ART AVK AVK 77 B, 80 D, 80 E 77 D, 82 C 861,92 G 78 A, 82 A 86 H,92 D Percentage of activity Percentage of activity Percentage of activity and Percentage of activity Percentage of activity and ART = nrtificially depleted plasma; AVK = nntivitamin K plasmn obtained from patients undergoing long-term oral anticoagulnnt therapy; PT = prothrombin time; = Internntional Normalized Ratio. MATERIALS AND METHODS Principle ofppc PPC assumes a linear relation between logpt (seconds) and loglnr of the corresponding plasmas. 26 The PTs of two or more reference lyophilized plasma calibrants, with assigned equivalent values (encompassing the therapeutic range) are locally determined with the usual combination of thromboplastin and coagulometer, and a double-log calibration curve is plotted using the assigned values of the plasma calibrants and their corresponding PTs (Fig 1, A). The PT of the survey test sample is measured using the same system (thromboplastin and same method and instrument). The of the test plasma is then determined graphically (Fig 1, B) or by computer calculation. Formerly, this principle was applied to percentage activity, using a special semidecimal rather than a double-log calibration curve. There is no need for recalculation of the ISI or for a normal value with the procedure we propose. x 4 * U A t3 PT ~l sec t4 Survey Samples Lyophilized citrated human survey plasmas were obtained from patients undergoing long-term oral anticoagulant therapy (AK plasmas, purchased from Immuno Laboratory, Vienna, Austria) or were artificially depleted (ART plasmas, processed by La Technique Biologique, Paris, France) and prepared as follows: normal blood from 60 to 80 donors was collected in 0.13 mol/l citrated anticoagulant, buffered with HEPES (0.19 mol/l N-(2-hydroxyethyl) piperazine-n'-(2-ethanesulfonic acid) to ensure stability 29 and centrifuged; pooled plasma was adsorbed onto aluminum hydroxide to deplete the vitamin K factors; the desired clotting activity for ART Cal plasmas could be obtained by back-ritration with normal plasma to yield hypocoagulation similar to that of the AK Cal plasmas. All survey samples were reconstituted with 1 ml distilled water and used within 1 hour. Inrralaboratory coagulation factor assays were performed before dispatch. 24 B M *5 X t- ' t 40 "i r r2o_ t3 t5 t4 FIG 1. Principle of the procedure using plasma calibrants. A, Example of a calibration curve plotted with assigned International Normalized Ratio () activities (x3, x4) of two lyophilized plasma calibrants and their corresponding prothrombin times (PTs; t3, t4) determined with the local thromboplastin-coagulometer combination. B, Example of graphic resolution for a test plasma; its PT (t5) was performed under the same local conditions as the PTs of the calibrants. (x5) read from the calibration curve (A). PT AJCP July 1997

3 HOUBOUYAN AND GOGUEL 85 Standardization Using Plasma Calibrants TABLE 2. VALUES (MEAN AND CV) ACCORDING TO THROMBOPLASTINS (SURVEY 92): COMPARISON OF THE USUAL PROCEDURE AND THE PROCEDURE USING PLASMA CALIBRANTS IN A HOMOGENEOUS SYSTEM OF AVK CALIBRANTS AND TEST PLASMA Graphically by Computer Calculation Thromboplastins n CV (%) IL = International Normalized Ratio; CV = coefficient of variation; AVK = antivitamin K plasma., Rueil France; Marcy 1'Etoile, France;, Asnieres, France; IL, Instrumentation Laboratory, Paris, France. Design of the Surveys In the last two surveys (86 HI, 92 DG), participants were divided into two groups representing the two systems of calibration. One group received two AK Cal plasmas; the other group received two ART Cal plasmas prepared with degrees of hypocoagulation similar to the AK Cal plasmas. Before dispatch, plasma calibrants were assigned an equivalent value. Their PT was determined in French expert laboratories under local conditions, in the authors' laboratory with different reagents and instruments, and by A. M. H. P. van den Besselaar, MD, at Haemostasis Department of Thrombosis Research Center (Leiden, Netherlands). 10 Mean s were calculated and assigned for the two sets of plasma calibrants. All participants received the same AK test plasma and expressed its PT activity in time (seconds) and in, according to their usual procedure, eg, using the manufacturer's LSI (conventional procedure). In addition, they plotted the calibration curve using the assigned of the reference plasma calibrants and expressed the test plasma activity as "corrected " (read graphically). Computerized statistical analysis (mean, standard deviation, coefficient of variation [CV]) was performed at the reference center. RESULTS PPC Compared With the Conventional Procedure In survey 92 DG, the PPC using the AK Cal plasmas gave graphically evident lower dispersion (coefficient of variationfcv], 9%, for the overall data) than the conventional procedure (CV, 12%), and a further improvement in the between-laboratory precision was obtained by computer calculation (CV, 6.2%; Table 2). The PPC also considerably reduced the dispersion in each group of thromboplastins and improved the convergence of the mean values. However, no improvement in dispersion was observed with the ART Cal plasmas in this survey (Table 3). In the previous survey (86 HI), the improvement with PPC compared with the conventional procedure was even greater with the AK Cal plasmas than in the 92 DG survey, for the overall data and within each thromboplastin group (Tables 4 and 5). In fact, in this survey, dispersion was lowered with the ART Cal plasmas also, but less obviously than with AK Cal plasmas. Mean values also were similar, especially in the case of one thromboplastin (Simplastin Akzo, Fresnes, France) that always gave a lower with the conventional procedure. Figure 2 shows that the range of values according to coagulometers and reagents could be minimized with PPC and a trend toward uniform LNR values was achieved. DISCUSSION Interlaboratory variation in PT is still a contentious issue. It is due to different reagent sensitivities to clotting defects and to the variety of techniques and local practices in PT measurement. The 1983 recommendation provided a common international scale for PT reporting. 1 Although introduced to reduce the thromboplastinrelated specificity, the is still affected by a number of unresolved problems that have slowed its acceptance by clinical laboratories and clinicians 7 and that could explain a lack of confidence in its use. 8 ' 9 Interlaboratory surveys in different countries reveal the large differences in reported values still encountered with the use of. 5 ' 9 ' 27 Vol. 108 No. 1

4 86 COAGULATION AND TRANSFUSION MEDICINE TABLE 3. VALUES (MEAN AND CV) ACCORDING TO THROMBOPLASTINS (SURVEY 92): COMPARISON OF THE USUAL PROCEDURE AND THE PROCEDURE USING PLASMA CALIBRANTS IN A HETEROGENEOUS SYSTEM OF ART CALIBRANTS AND AVK TEST PLASMA Graphically by Compi iter Calculation Tliromboplastins n IL , , = International Normalized Ratio; CV = coefficient of variation; ART calibrants = artificially depleted calibrated plasmas; AVK = antivitamin K., Rueil France;, Marcy 1'Etoile, France;, Asnieres, France; IL, Instrumentation Laboratory, Paris, France. Although there is a trend in France toward the use of two brands of thromboplastins of the same species ( [Marcy 1'Etoile, France] and [Asnieres, France] in about 80% of laboratories) among the 13 major reagents, no reduction in interlaboratory dispersion vs the conventional method within each of the two groups was observed compared with the overall dispersion (interlaboratory variability expressed in CV). This may be explained by between-batch differences or by the use of an inappropriate manufacturer's ISI value. 12,19 However, in smaller laboratories, determination of the ISI is not always feasible because it requires too many fresh plasma samples from patients with stable levels of anticoagulation. Moreover, the manufacturer's stated ISI of the thromboplastin is derived from World Health Organization recommendations and applies only to manual techniques, whereas automation is rapidly increasing and diversifying. 8 ' 30 ' 31 A number of reports have identified an instrument effect on the ISI value. 15 " 17 ' 30 ' 32 Kazama et al 31 have suggested that manufacturers give the ISI of the thromboplastin-instrurnent system, and some express the ISI according to the type of clot-detecting system (optical or mechanical). Poggio et al 16 showed that the calibration of some commercial reagents tested on coagulometers against the reference material (manual testing) raises further problems in slope determination. Because considerable variation exists in performance among instruments of the same model, a single instrument may not be representative. 17 After the early work of Miale and Kent, 33 Miale and Loeliger, 34 and Koepke 35 on reference plasmas, we proposed the use of plasma calibrants for PT standardization. 20 ' 25 " 28 During the last 15 years, we have gained extensive experience through nine large-scale interlaboratory surveys. 24 These surveys show that the PPC substantially lowered interlaboratory dispersion compared with the conventional procedure, thus emphasizing its efficiency in standardization of PT (expressed as a percentage and also as an from 1986 onward). The ability of the PPC to reduce the reagent and instrument effects was apparent in all surveys by a noteworthy reduction in reported dispersions within each group defined by a given thromboplastin and by the combination of a thromboplastin and a technique. Furthermore, the mean tended toward the same value for the overall data and in each group, regardless of the thromboplastin batch or technique. In particular, the PPC standardized the for which the conventional procedure regularly yielded lower values with some reagents. The mean values were generally similar to those given by the conventional procedure, depending on the values assigned to the plasma calibrants. Our data underscore the constant efficiency of the PPC using the AK Cal plasma. The PPC was less effective with ART Cal plasmas or ineffective, depending on the survey. This could be due to the preparation of the ART plasmas; the additional adsorption step and the delay caused by compulsory virologic testing of the samples before processing could contribute to the poor performance of these ART plasmas. The PPC has many advantages over the conventional procedure and over some other recent proposals. 12 ' 21 " 23 It is simple and theoretically independent of the thromboplastin source, batch, and instrument effect, because local practices are considered in determining the PT of test and calibrated plasmas. Above all, the PPC avoids use of the commercial thromboplastin ISI values, which are sometimes inappropriately assigned by manufacturers, and AJCP July 1997

5 HOUBOUYAN AND GOGUEL 87 Standardization Using Plasma Calibrants TABLE 4. VALUES (MEAN AND CV) ACCORDING TO THROMBOPLASTINS (SURVEY 86): COMPARISON OF THE USUAL PROCEDURE AND THE PROCEDURE USING PLASMA CALIBRANTS IN A HOMOGENEOUS SYSTEM OF AVK CALIBRANTS AND TEST PLASMA Thromboplastins n Akzo Technique Biologique , International Normalized Ratio; CV = coefficient of variation; AVK = antivitamin K. Akzo, Fresnes, France;, Rueil France;, Marcy l'etoile, France;, Asnicres, France; Technique Biologique, Paris, France. TABLE 5. VALUES (MEAN AND CV) ACCORDING TO THROMBOPLASTINS (SURVEY 86): COMPARISON OF THE USUAL PROCEDURE AND THE PROCEDURE USING PLASMA CALIBRANTS IN A HETEROGENEOUS SYSTEM OF ART CALIBRANTS AND AVK TEST PLASMA Thromboplastins Akzo Technique Biologique = International Normalized Ratio; CV = coefficient of variation; ART calibrants = artificially depleted calibrated plasmas; AVK = antivitamin K. Akzo, Fresnes, France;, Rueil France;, Marcy I'Ftoile, France;, Asnieres, France; Technique Biologique, Paris, France. circumvents the instrument effect on the ISI value Unlike the method proposed by Poller et al 21/22 and others, 23 ' 36 the PPC does not require the intermediate recalculation of a specific ISI 23 for the local thromboplastin-instrument system because the is directly obtained from the calibration. Furthermore, the PPC does not require a normal control plasma value for the PT ratio; indeed, the choice of a normal plasma, which is known to introduce additional variability in the, 14 ' 30 especially at high values, is another unresolved problem. The World Health Organization advises taking the mean PT of 20 normal fresh plasmas (2 samples/day x 10 days = 20), but many routine laboratories lack such material and use various commercial lyophilized plasmas for the normal value. The PPC does, however, have some limitations. The use of lyophilized plasmas in calibration has been questioned by some authors. Van den Besselaar 37 recently showed that lyophilized plasmas deviate slightly from fresh plasmas, depending on the thromboplastin used, especially with ART plasmas. Others 16,36/38 have reported that despite this deviation, lyophilized plasmas can be used to calculate the and to calibrate thromboplastins. Clearly, our study shows that calibration is more reliable with AK Cal plasmas than with ART Cal plasmas, 31 as also recently reported by Stevenson et al. 39 However, AK Cal plasmas are limited in supply, costlier, and subject to ethical and regulatory issues and to possible batch-tobatch variability. Second, it is not easy to provide the wide range of values required for accurate ISI calibration on pooled AK Cal plasmas. Their routine use is therefore difficult 4 ' 22 or unrealistic. ART Cal plasmas are less costly and can therefore be used for routine calibration, although the last French national survey indicates their inconsistent reliability in PT standardization. 27 However, a recent French multicenter study 40 showed that ART Cal plasmas could be used, albeit with a slightly lower reliability, provided they are carefully prepared. Poller et al recently reported good agreement between AK Cal and ART Cal plasmas and recommended their use in calibration, but their survey only concerned a limited number of thromboplastins and instruments. Vol. 108 No. 1

6 COAGULATION AND TRANSFUSION MEDICINE Another important problem with PPC relates to the assignment of an value to the calibrated plasmas. Questions are raised about which thromboplastin (an international reference preparation?), which reagents, and which value (the mean?) should be used. Results from the French collaborative study 40 showed that this problem was more specific to ART Cal plasmas, which gave a wider range of values with the different thromboplastins than AK Cal plasmas. This generates problems of accuracy, especially with ART plasmas. An international multicenter study of this procedure under the auspices of the Scientific and Standardization Committee of the International Society for Thrombosis and Hemostasis (ISTH) is under way to establish whether ART Cal plasmas can be used in the PPC instead of AK Cal plasmas. Other problems also must be addressed, such as the number of calibrators needed, the reproducibility of the calibration, and, hence, the minimum number of days of calibration that are required for optimal reliability of the PPC. Indeed, PT measurement of the plasma calibrants may vary from one day to another, depending on isolated handling problems, such as error in the reconstitution of one vial, reagent instability, or instrumental hitches. The accuracy of the procedure should be investigated. The French multicenter study, 40 involving a few laboratories with 12 thromboplastininstrument combinations, has addressed these questions. For instance, this study shows that at least two calibrators are needed, but only marginal improvement is achieved with three (AK) or more calibrators. Most important, at least 3 to 4 days are needed to ensure reproducibility of the calibration. The current practice is that calibration should be done at least for every new batch of reagent and checked periodically (eg, every 3 months). Adaptation of the PPC to the new generation of fully automated coagulometers also calls for careful consideration, because such instruments require entry of the normal value and of the ISI of the thromboplastin to yield the of the test sample. However, some new coagulometers have flexible computer options, which could be exploited to construct the bilogarithrnic calibration curve. Increasing laboratory automation has changed our concept of and highlighted the need for further PT standardization. Now that most research groups accept that plasma calibrants are useful in standardization, we have demonstrated the validity of this simple procedure through 15 years of French surveys involving a large number of laboratories and numerous combinations of thromboplastins and coagulometers. Despite some flaws, the procedure using plasma calibrants represents a positive step toward PT standardization and, in our view, offers the best compromise available. We do not claim that the proposed calibration procedure using lyophilized plasmas is perfect, but it has proved its efficiency on a larger scale than ever before and may be proposed in everyday practice. Acknowledgments: We thank all the French laboratories that participated in the NEQAS, the expert colleagues from the French Groupe d'etude sur l'hemostase et la Thrombose and Dr van den Besselaar, who participated in the assignment of of the plasma calibrants, and the Laboratoire National de la Sante and Agence du Medicament (Saint-Denis, France [Ministry of Health]) that gave permission to..^cmx2 -CM200 Kgl -Tilt -KC10 3; ;N^EMB YFbt Fbm ^-CMX2, CM200.-/.KC10 /Tilt ^Cgm r-fbm X EMB - - CMX2..--Tilt -CMX2 /CM200 ^CMX2 Tilt S^Fbt VFbm \Cgm Kgl CM200 3 <^Fbt ^-Tilt VFbm \EMB Cgm CMX2 Tilt -CMX2 Thpl: biom6rieux A B FIG 2. International Normalized Ratio () mean values. Comparison of A, the conventional procedure, and B, the procedure using plasma calibrants in a homogeneous system of AVK calibrants and AVK test plasma. Example of data according to techniques for three thromboplastins (survey 86). CM = Coag A Mate Akzo (Fresnes, France); Cgm = coagulometer; Fbm = fibrometer; EMB = electromagnetic bath; Fbt = Fibrintimer (Rueil, France); KC 10 Amelung; Kgl = Koagulab Ortho; Tilt = Tilt tube. AJCP July 1997

7 HOUBOUYAN AND GOGUEL 89 Standardization Using Plasma Calibrants reproduce data from the French National quality control program. Professor Goguel was the scientific director of the French NEQAS in Hematology. Ms C Desmarchais prepared the typescript. REFERENCES 1. WHO Expert Committee on Biological Standardization. 33rd Report. Requirements for thromboplastin and plasma used to control oral anticoagulant therapy. WHO Tech Rep Sen 1983;687: International Committee for Standardization in Haematology and International Committee on Thrombosis and Haemostasis. ICSH/ICTH recommendations for reporting prothrombin time in oral anticoagulant control. Thromb Haemost. 1985;53: Morse EE, Panek S, Pisciotto P, et al. Reemergence of the International Normalized Ratio for the standardization of prothrombin time. Ann Clin Lab Sci. 1993;23: Loeliger EA, Poller L, Samama M, et al. Questions and answers on prothrombin time standardisation in oral anticoagulant control. Thromb Haemost. 1985;54: Palareti G, Coccheri S, Poggi M, et al. The Bologna oral anticoagulant control study group: oral anticoagulant therapy control: evidence that expression improves the interlaboratory comparability of results: the Bologna oral anticoagulant control exercise. Thromb Haemost. 1987;58: Hirsh J. Oral anticoagulant therapy: urgent need for standardization. Circulation. 1992;86: Triplett DA, Brandt J. International Normalized Ratios: has their time come. Arch Pathol Lab Med. 1993;117: Swaim WR. Prothrombin time reporting and the International Normalized Ratio system: improvements are needed. Am ] Clin Pathol. 1993;99: Ng VL, Levin J, Corash L, Gottfried EL. Failure of the International Normalized Ratio to generate consistent results within a local medical community. Am ] Clin Pathol. 1993;99: van den Besselaar AMHP, Bertina RM. Multicenter study of thromboplastin calibration precision: influence of reagent species, composition and International Sensitivity Index (ISI). Thromb Haemost. 1993;69: Holland LJ, Lawrie AS, Hunt BJ. Evidence of discrepant commercial ISI assignment. Blood Coagul Fibrinolysis. 1992;3: Cunningham MT, Johnson GF, Pennell BJ, Olson JD. The reliability of manufacturer-determined instrument-specific International Sensitivity Index values for calculating the International Normalized Ratio. Am ) Clin Pathol. 1994;102: Peters RHM, van den Besselaar AMHP, Olthuis FMFG. Determination of the mean normal prothrombin time for assessment of International Normalized Ratios: usefulness of lyophilised plasma. Thromb Haemost. 1991;66: Critchrield GC, Bennett ST. The influence of the reference mean prothrombin time on the International Normalized Ratio. Am J Clin Pathol. 1994;102: Evatt BL, Brogan D, Triplett DA, Waters G. Effect of thromboplastin and instrumentation on the prothrombin time test. Clin Lab Haematol. 1981;3: Poggio M, van den Besselaar AMHP, van der Velde EA, Bertina RM. The effect of some instruments for prothrombin time testing on the International Sensitivity Index (ISI) of two rabbit tissue thromboplastin reagents. Thromb Haemost. 1989;62: Ray MJ, Smith IR. The dependence of the International Sensitivity Index on the coagulometer used to perform the prothrombin time. Thromb Haemost. 1990;63: Chantarangkul V, Tripodi A, Mannucci PM, Bianchi-Bonomi A. The effect of instrumentation on thromboplastin calibration. Thromb Haemost. 1992;67: Becker DM, Humphries JE, Walker FB, et al. Standardizing the prothrombin time: calibrating coagulation instruments as well as thromboplastin. Arch Pathol Lab Med. 1993;117: Goguel AF, Houbouyan LL, Roussi JH. Calibrated plasma procedure and for prothrombin time standardisation: data from the French Etalonorme Quality Control Surveys. Thromb Haemost. 1987;58:296. Abstract. 21. Poller L, Thomson JM, Taberner DA, Clarke DK. The correction of coagulometer effects on International Normalized Ratios: a multicentre evaluation. Br I Haematol. 1994;86: Poller L, Triplett DA, Hirsh J, Carroll J, Clarke K. The value of plasma calibrants in correcting coagulometer effects on International Normalized Ratios: an international multicenter study. Am ) Clin Pathol. 1995;103: Clarke K, Taberner DA, Thomson JM, Morris JA, Poller L. Assessment of value of calibrated lyophilised plasmas to determine International Sensitivity Index for coagulometers. / Clin Pathol. 1992;45: Annates du Controle de Qualite National en Hdmatologie- Etalonorme. Ed. : Laboratoire National de la Sante/Agence du Medicament, Saint Denis, France, 4 issues per year. 25. Goguel AF. French national quality assessment program in coagulation testing: the Etalonorme experience. In: Triplett DA, ed. Advances in Coagulation Testing: An Overvirew. Northfield, 111: College of American Pathologists; 1986: Goguel AF, Roussi JH, Houbouyan LL. A new approach in onestage prothrombin time standardization: use of calibrated control plasmas. In: Coccheri S, ed. Proceedings of the 5th International Congress on Thrombo-embolic Diseases. Bologne, Italy: Quaderni dellacoagulazion; 1980: Houbouyan LL, Goguel AF. Procedures of reference calibrated plasmas for prothrombin time standardisation: data from French interlaboratory surveys. Thromb Haemost. 1993;69:663. Abstract. 28. Houbouyan LL. Les plasmas de reference calibres: un moyen simple et efficace de standardiser le temps de Quick. Rouen, France: University of Rouen; Thesis. 141p. 29. Zucker S, Cathey MJH, West B. Preparation of quality control specimens for coagulation. Am ) Clin Pathol. 1970;53: Pi DW, Raboud JM, Filby C, Carter CJ. Effect of thromboplastin and coagulometer interaction on the precision of the International Normalized Ratio. / Clin Pathol. 1995;48: Kazama M, Suzuki S, Abe T, et al. Evaluation of International Normalized Ratios by a controlled field survey with 4 different reagents. Thromb Haemost. 1990;64: van den Besselaar AMHP, Evatt BL, Brogan DR, Triplett DA. Proficiency testing and standardization of prothrombin time: effect of thromboplastin, instrumentation and plasma. Am ] Clin Pathol. 1984;82: Miale JB, Kent JW. Standardisation of the therapeutic range for oral anticoagulants based on standard reference plasmas. Am J Clin Pathol. 1972;60: Miale JB, Loeliger EA. 'Reference plasmas' in the laboratory control of oral anticoagulants. Thrombos Diathes Haemorrhagica. 1975;33: Koepke J. Use of survey validated plasmas as a means of prothrombin time standardisation of the coagulation assays. In: Triplett DA, ed. Standardization of Coagulation Assays: An Overviezv. Cap Conference/Aspen, Skokie, III: College of American Pathologists; 1982: Tripodi A, Chantarangkul V, Manotti C, et al. A simplified procedure for thromboplastin calibration: the usefulness of lyophilized plasmas assessed in a collaborative study. Thromb Haemost. 1996;75: van den Besselaar AMHP. Comparison of lyophilized plasmas with fresh plasmas for calibration of thromboplastin reagent in oral anticoagulant. Br ] Haematol. 1996;93: Poller L, Triplett DA, Hirsh J, Carroll J, Clarke K. A comparison of lyophilized artificially depleted plasmas and lyophilized plasmas from patients receiving warfarin in correcting for coagulometer effects on International Normalized Ratios. Am ] Clin Pathol. 1995,103: Stevenson K, Craig S, Dufty JMK, Taberner D. System ISI calibration: a universally applicable scheme is possible only when coumarin plasma calibrants are used. Br] Haematol. 1997;96: Houbouyan L, Bonvarlet MN, Comby S, et al. standardization by the calibrated plasma procedure (CPP). Thromb Haemost. 1995;73:1238. Abstract Vol. 108 No. 1