SEARCH

SEARCH BY CITATION

Keywords:

  • INR;
  • coagulometers;
  • linear;
  • orthogonal;
  • regression

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

International sensitivity index calibrations based on the W.H.O. recommended method depend on orthogonal regression analysis. As this is not readily available in statistical packages, comparison has been made with simple linear regression analysis in a study of coagulometer effects on the International Normalized Ratio (INR) at 155 European centres. Sets of seven lyophilized normal and 20 lyophilized artificially depleted abnormal plasmas were provided with five coumarin test plasmas and two European Concerted Action on Anticoagulation reference thromboplastins (low International Sensitivity Index (ISI) human and high ISI rabbit). Local ISI based on the artificially depleted lyophilized plasmas using conventional orthogonal regression gave good correction for local coagulometer effects on the human reagent and minimal correction with the rabbit reagent INR. Results were considerably worse after attempts at correction using calibration based on linear regression analysis with both reagents. The results indicate that calibration of coagulometer prothrombin time systems using simple linear regression is not appropriate.

Since the introduction of the W.H.O. standardization scheme ( W.H.O. Expert Committee on Biological Standardisation, 1983), prothrombin time (PT) methodology has changed as most laboratories use automated coagulometers for PT measuring instead of the manual technique. The International Sensitivity Index (ISI) of thromboplastins may be considerably modified by coagulometers, and consequently the International Normalized Ratio (INR) may be inaccurate ( D'Angelo et al, 1989 ; Poller et al, 1989 ; van Rijn et al, 1989 ; Ray & Smith, 1990; Clarke et al, 1992 ). Furthermore, individual coagulometers from the same manufacturer vary in their effects on the ISI. There is therefore an even greater need for the calibration of local PT systems (coagulometer/thromboplastin combinations). The W.H.O. calibration procedure is complex and demanding, although it may be simplified by the provision of certified lyophilized plasmas ( Clarke et al, 1992 ; Poller et al, 1995a , 1998a, b). These remove the need for manual ISI calibrations with fresh plasmas from coumarin-treated patients with a thromboplastin reference preparation in parallel.

The complexity of the calculation for the orthogonal regression calibration still remains a major problem for most centres. The substitution of a procedure based on linear regression analysis has been proposed as an alternative, as this is more readily available in statistical packages and might be considered appropriate with certified values for the lyophilized plasmas on one axis of the calibration.

Linear regression was shown to give similar results to conventional orthogonal regression in the ECAA calibration of the human recombinant thromboplastin with the manual PT technique but it could not be certain that the same would apply to coagulometers ( Poller et al, 1997 ). This report presents the findings of the ECAA field studies where the linear regression results have been compared with conventional orthogonal regression using a variety of different coagulometers at 143 invited European centres.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

Lyophilized normal plasmas

Sets of seven lyophilized plasmas from healthy blood donors were prepared for this study at the ECAA Central Facility. Their values were certified with the manual PT technique in terms of the two ECAA reference reagents (human and rabbit) by six national control laboratories (‘certifying centres’). The tests were performed in duplicate and the certified values were the average of the mean values of the six centres.

Lyophilized abnormal plasmas

The lyophilized depleted plasmas were manufactured at the ECAA Central Facility in Manchester by artificial depletion of normal human plasma by selective adsorption of vitamin K-dependent clotting factors with barium sulphate to provide a range of values to span the therapeutic interval of 1.5–4.5 INR when tested with the International Council for Standardization in Haematology (ICSH) thromboplastin IRP (BCT/441 human plain). Different INR levels were achieved by variable adsorption. Hepes buffer (0.26 g%), glycine (2.18 g%) and sucrose (2.18 g%) were added to the plasmas as protectives prior to lyophilization.

Because of possible complicating effects of depletion of the non-coumarin-dependent clotting factors, factor V and fibrinogen on the PT results, factor V assays ( Thomson, 1970) and fibrinogen determinations ( Clauss, 1957) were performed on all the artificially depleted plasmas to ensure adequate levels. Samples containing less than a minimum level of 50% factor V or 1.5 g/l fibrinogen were excluded. These minimum levels are regarded as adequate for prothrombin time measurement and only when the individual clotting factors V or fibrinogen fall below these levels do they prolong the PT test. The factor V of all the artificially depleted plasmas included in the study ranged from 50% to 100% and fibrinogen content from 1.5 to 4.0 g/l. It has been shown that when factor V levels are > 40% they do not influence INR results ( Tripodi et al, 1995 ). Inter-vial variation, accelerated degradation and long-term stability studies were also performed on each of the depleted lyophilized plasmas. All plasmas gave a CV of < 3% in inter-vial studies and minimum heat degradation stability of 7 d at 40°C. Long-term stability studies also proved satisfactory up to a minimum of 12 months.

The PT of 60 artificially depleted lyophilized abnormal plasmas were certified in terms of both the ECAA human and rabbit reference thromboplastins using the manual PT technique at the certifying centres. The ISI of the reference thromboplastins had been established by a multicentre exercise at 14 ECAA national laboratories ( Poller et al, 1996 ) employing the conventional W.H.O. ISI calibration procedure ( W.H.O. Expert Committee on Biological Standardisation, 1983).

Two different sets of 20 abnormal lyophilized plasmas were selected in serial order of production from the above 60 certified plasmas for the two serial field studies, performed at intervals of 3 months. The purpose in employing two different sets was in order to assess reproducibility of results with different sets of lyophilized plasmas and different ISI values over a period of time.

Lyophilized coumarin test plasmas

Five coumarin plasmas obtained from patients on long-term oral anticoagulant therapy were lyophilized at the Central Facility in Manchester for inclusion as test plasmas. These plasmas were tested by three of the ECAA ‘certifying centres’ using the manual PT technique. Conventional INR values were calculated for the five lyophilized coumarin test plasmas using the ISI from the lyophilized plasma orthogonal regression slope and the mean normal prothrombin time (MNPT) of the seven lyophilized normals, i.e. INR = (PT/MNPT)ISI, where PT is a lyophilized coumarin test plasma.

Owing to the difficulty of obtaining adequate volumes of plasma in single donations from anticoagulant-treated patients, there was insufficient for assays for factor V and fibrinogen, inter-vial and stability testing.

Assigning INR to lyophilized plasmas

Four replicate tests were performed on each of the lyophilized plasmas with the two ECAA thromboplastins by the ‘certifying centres’. The overall mean PT of the four replicate PT was calculated for each of the plasmas at each centre. The assigned (certified) INR were the mean of the results from these centres using the established manual ISI of the ECAA human (ECAAH) and rabbit (ECAAR) reference thromboplastins (ISIECAAH= 0.95; ISIECAAR= 1.67) and MNPT of the seven lyophilized normal plasmas.

Reference thromboplastins

Either the human ECAA reference thromboplastin or the rabbit ECAA reference reagent were issued to each of the field study centres according to their type of local routine thromboplastin reagent. Users of the bovine commercial reagent, Thrombotest, were provided with the ECAA rabbit reference preparation. Participants were instructed to test each of the 32 plasmas (20 lyophilized depleted, five lyophilized coumarins and seven lyophilized normals) in duplicate with the reference thromboplastin and the local routine thromboplastin. The coagulometer employed locally for routine prothrombin time testing was to be used.

STATISTICAL ANALYSIS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

Local calibration using certified PT values and orthogonal regression model

Orthogonal regression ( Kirkwood, 1983; van der Velde, 1984) has been used according to the W.H.O. protocol to obtain coefficients for the calibration equations and subsequent calibration slopes. The log of the certified manual PT with the ECAA thromboplastins for the seven lyophilized normals and 20 lyophilized abnormal plasmas were placed on the vertical (y-axis) and log PT with the local coagulometer PT system results on the horizontal (x-axis).

Local calibration using certified INR values and linear regression model

To determine the local INR, calibration lines were constructed using the linear regression model and certified INR of the ECAA thromboplastins. These were placed on the horizontal axis and the PT obtained with the ECAA thromboplastin with the local coagulometer placed on the vertical axis, again using a double log scale with the lyophilized normal and abnormal plasmas.

There are a number of important differences between the linear and orthogonal regression models and a number of necessary assumptions must be made for the linear model to be applied. Firstly, the principal difference between this model and the orthogonal regression model is that the values on one axis, the reference values (i.e. the assigned INR), are assumed to be free of error. It is assumed that error present in the observations is entirely contained within the second axis (i.e. locally tested lyophilized plasma PT). This is unlike orthogonal regression where error is considered to be present in both axes. Although error is still present in the observed values of the linear model it is assumed that the certified INR are best estimates of the ‘true’ underlying values and as such have been designated as target values. Secondly, the simplified model provides an estimate of the INR which would have been obtained had the correct model been employed (i.e. orthogonal regression). Thus the INR observed are merely estimates of INR values which would have been calculated had the conventional methodology using ISI and MNPT been employed, and as such are not conventional INR but the best approximation to the INR. Finally, the conventional theory of dependent and independent variables inherent in the ordinary linear regression model is not adhered to for the simplified linear model. For the latter it has been assumed that ordinary linear regression is simply a mechanism for fitting a straight line to the observations, without complying with the principles of the model. Thus linear regression INR may be erroneous and it cannot be assumed that they may be substituted for conventional INR obtained by orthogonal regression.

Accuracy of local calibration schemes

The five lyophilized coumarin test plasmas have been certified with INR values which are assumed to be the best estimates of the true underlying INR, IT, with the relevant ECAA thromboplastin. The accuracy, i.e. deviation from the true INR of the individual centres local calibration schemes, has been determined by comparing local INR, IL, with the ‘true’ INR, IT. The absolute percentage deviation of local system INR from the ‘true’ INR has been calculated for each of the five coumarin test plasmas as

inline image

where i =1, 2, … , 5 for the individual test plasmas and di is the absolute percentage deviation from the ‘true’ INR for the ith plasma. The mean percentage deviation, d, was calculated for each centre for the five plasmas (d = (d1 + … + d5)/5). The ‘true’ INR were calculated using the conventional formula IT = (p/m)ir where p is the PT of the lyophilized coumarin test plasma obtained using the manual technique, m is the MNPT of seven lyophilized normal plasmas, and ir the ISI of the reference preparation.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

One hundred and forty-three out of the 155 centres returned complete results. Four of these had to be excluded because of a change of coagulometer between the two studies. Eighteen other centres were excluded because they either participated in only one field study or returned incomplete results. Of the 121 centres remaining, 45 employed the human ECAA reagent and 76 the rabbit reagent. Six brands of coagulometer were used by five or more laboratories.

INR calculation using local system ISI and lyophilized normal plasma MNPT

Table I shows the INR obtained by the three expert certifying centres using the MNPT of the seven lyophilized normal plasmas and the ISI of the ECAA reference thromboplastins (ISIECAAH= 0.95 or ISIECAAR= 1.67). The ECAA human thromboplastin gave an overall mean INR of 4.04 whereas with the ECAA rabbit reagent this was 3.57. Although there was considerable deviation between INR both within reagent (i.e. between centres) and between reagents, a single factor ANOVA revealed no significant results between centres with the two ECAA reagents.

Table 1. Table I(a). INR obtained from lyophilized coumarin test plasmas, obtained using MNPT of seven lyophilized plasmas and ECAA human thromboplastin (Manual ISI=0.95). Table I(b). INR from lyophilized coumarin test plasmas, obtained using MNPT of seven lyophilized plasmas and ECAA rabbit thromboplastin (Manual ISI = 1.67). Thumbnail image of

All field study centres which tested the 20 lyophilized depleted and seven lyophilized normal plasmas for the local calibration schemes also tested the five lyophilized coumarin test plasmas. The mean INR both before and after local correction at the centres testing the five coumarin plasmas are presented in the form of histograms in Figs 1 and 2. The vertical broken lines represent the mean INR for the coumarin test plasmas obtained by the three expert certifying centres (INR = 4.04 for the ECAA human thromboplastin and INR = 3.57 for the ECAA rabbit thromboplastin). Prior to local system correction there was a general trend towards overestimation of INR for the human route of calibration in both field studies.

image

Figure 1. , human route of calibration (ECAA human). (b) Field study 2, human route of calibration (ECAA human).

Download figure to PowerPoint

image

Figure 2. , rabbit route of calibration (ECAA rabbit).

Download figure to PowerPoint

When the stated manual ISI of the reference preparation was replaced with the local linear regression calibration system the overestimation of INR persisted and there appeared to be no overall improvement in INR. The mean INR prior to correction was approximately 4.7 for the human route of calibration over the two field studies. After correction using the linear model the overall mean was virtually unchanged and mean absolute percentage deviation from the ‘true’ INR was of a similar magnitude (see 2 Table II). INR calculated using the linear model overestimated the assigned INR by approximately 17% after local correction of the human route results.

Table 2. Table II. Mean INR for the five lyophilized coumarin test plasmas for field study centres testing all five plasmas. Before correction INR were determined using stated ISI of ECAA thromboplastins (ISIECHuman = 0.95; ISIECRabbit = 1.67). After correction INR were determined using the linear model and the conventional orthogonal model. Thumbnail image of

In contrast, when the conventional orthogonal model was employed to calculate INR with the human route, the overall mean of 4.1 was considerably closer to the ‘true’ mean INR value of 4.04 after correction and mean absolute percentage deviation was reduced to approximately 7% over the two field studies.

When the linear model was employed to determine INR for the rabbit thromboplastin, results were in contrast considerably worsened after the introduction of a correction for the local effect (see 2 Table II). The results were transformed and considerable overestimation observed. Before the procedure for correction was applied the results underestimated ‘true’ INR by approximately 10%. After correction the certified (assigned) INR were overestimated by approximately 25% in the two field studies. When the orthogonal model was used with the rabbit route of calibration to correct INR there was a minimal reduction in deviation compared with the before correction results.

3 Table III lists the mean INR and mean percentage deviation from the ‘true’ INR for the centres which tested the five lyophilized coumarin test plasmas on their local coagulometer with the ECAA thromboplastin. Groups with as few as three coagulometers are included but less emphasis should be placed on these results because of the small samples. Correction of INR is poor across the range of coagulometers for the human thromboplastin with the linear model but good across the range of coagulometers for the orthogonal model, with the exception of Schnitger & Gross. For the rabbit route of calibration there was some correction with a number of the coagulometers where the orthogonal model was used; however, the improvement, where observed, was not great. When the linear model was used to calculate the INR, overestimation of results was considerable for all instruments for this route of calibration.

Table 3. Table III(a). Mean INR and percentage deviation from assigned INR values for the ECAA human route of calibration. Mean results for centres that tested all five coumarin plasmas in the local coagulometer are shown. Table III(b). Mean INR and percentage deviation from assigned INR values for the ECAA rabbit route of calibration. Mean results for centres that tested all five coumarin plasmas in the local coagulometer are shown.Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

There have been a number of recent reports suggesting that a simplified local calibration scheme using lyophilized plasmas with certified values could be used to determine local INR ( Houbouyan & Goguel, 1997; Hubbard et al, 1997 ; Stevenson et al, 1997 ; Poller et al, 1997 ; Craig et al, 1997 ). In this study it has been possible to compare coagulometer effects using low ISI and high ISI thromboplastins on the PT of the same 20 plasmas certified with the standardized manual PT technique at a number of ECAA national laboratories. In addition, the use of test plasmas with assigned target INR values has assisted in quantifying error, in terms of INR, introduced by the coagulometer.

In the previous study from the ECAA ( Poller et al, 1997 ) it was stressed that the favourable results with the simplified analysis were with the manual technique and a low ISI thromboplastin only. Further study was stated to be required to assess validity on coagulometer PT system calibrations.

The orthogonal regression analysis proposed for the W.H.O. standardization scheme by Kirkwood (1983) was based on the fundamental principle that in thromboplastin calibration variability is present on both axes, whereas linear regression only takes account of variability on one axis. It could be argued that with certified values on the x-axis as provided with lyophilized plasma calibrants, the simpler linear regression calibration could be used. This is based on the assumption that carefully determined certified values are the best estimates of the true values.

INR correction achieved using the proposed linear model and local ISI calibrations based on lyophilized artificially depleted plasmas for local INR determination was generally poor. In both field studies the local system model overestimated the ‘true’ INR for both ECAA human and rabbit thromboplastins. Results from a previous ECAA report ( Poller et al, 1997 ) demonstrated that using the manual PT technique the linear model provided a similar degree of correction to that of the orthogonal regression model for local calibration. The question whether different results would be obtained if lyophilized plasmas from coumarin- treated patients had been employed rather than the artificially depleted plasmas remains to be resolved. This is considered unlikely in view of our previous findings that lyophilized plasmas from coumarin-treated patients and artificially depleted plasmas in ISI calibration carry a similar degree of error (deviation from ‘true’ INR) when employed in local ISI calibration ( Poller et al, 1995b , 1998b). The evidence suggests that calibration of coagulometers using linear regression is inappropriate and leads to biased estimates of the coefficients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

Gratitude is expressed to the following who participated in the collaborative exercise: J. Arnout, Thrombosis and Vascular Research, Campus Gasthuisberg, Leuven, Belgium; H. Beeser, Department of Transfusion Medicine and Coagulation, Universitat Freiburg, Germany; A. M. H. P. van den Besselaar, Haemostasis and Thrombosis Research Centre, Leiden University Hospital, The Netherlands; N. Egberg, Department of Clinical Chemistry, Karolinska Hospital, Stockholm, Sweden; J. A. Iriarte, Hospital Civil de Basurto, Bilbao, Spain; J. Jespersen, Department of Clinical Biochemistry, Ribe County Hospital, Esbjerg, Denmark; I. Kontopoulou-Griva, First Regional Transfusion Centre, Hippocration Hospital, Athens, Greece; B. Lämmle, Central Haematology Laboratory, University Hospital Bern, Switzerland; K. Lechner, First Medizinische Klinik der Universität Wien, Austria; P. M. Mannucci, A. Bianchi Bonomi, Haemophilia and Thrombosis Centre, Milan, Italy; B. Otridge, Mater Misericordiae Hospital, Dublin, Ireland; J. Pina Cabral, Centro de Fisiologia da Hemostase, Universidade do Porto, Portugal; L. Poller, ECAA Central Facility, Department of Pathological Sciences, The University of Manchester, U.K.; M. Samama, Service D'Hématologie, Hôtel-Dieu de Paris, France.

This work was supported by grant number PL931349 of the EC Biomed programme.

References

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix
  • 1
    Clarke, K., Taberner, D.A., Thomson, J.M., Morris, J.A., Poller, L. (1992) Assessment of value of calibrated lyophilized plasmas to determine International Sensitivity Index for coagulometers. Journal of Clinical Pathology, 45, 58 60.
  • 2
    Clauss, A. (1957) Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Acta Haematologica, 17, 237 246.
  • 3
    Craig, S., Stevenson, K.J., Dufty, J.M., Taberner, D.A. (1997) Local INR correction: justification for a simplified approach. Clinical Pathology, 50, 783 789.
  • 4
    D'Angelo, A., Seveso, M.P., D'Angelo, S.V., Gilardoni, F., Macagni, A., Manotti, C. (1989) Comparison of two automated coagulometers and the manual tilt tube method for the determination of prothrombin time. American Journal of Clinical Pathology, 92 , 321 328.
  • 5
    Houbouyan, L. & Goguel, A.F. (1997) Long-term French experience in INR standardization by a procedure using plasma calibrants. American Journal of Clinical Pathology, 108, 83 89.
  • 6
    Hubbard, A.R., Margetts, S.M.L., Barrowcliffe, T.W. (1997) INR determination using calibrated reference plasmas. British Journal of Haematology, 98, 74 78.
  • 7
    Kirkwood, T.B.L. (1983) Calibration of reference thromboplastins and standardisation of the prothrombin time ratio. Thrombosis and Haemostasis, 49, 238 244.
  • 8
    Poller, L., Barrowcliffe, T.W., Van Den Besselaar, A.M.H.P., Jespersen, J., Tripodi, A., Houghton, D. (1996) European Concerted Action on Anticoagulation (ECAA). The multicentre calibration of rabbit and human ECAA reference thromboplastins. Thrombosis and Haemostasis, 76, 977 982.
  • 9
    Poller, L., Barrowcliffe, T.W., Van Den Besselaar, A.M.H.P., Jespersen, J., Tripodi, A., Houghton, D. (1997) European Concerted Action on Anticoagulation. A simplified statistical method for local INR using linear regression. British Journal of Haematology, 98, 640 647.
  • 10
    Poller, L., Barrowcliffe, T.W., Van Den Besselaar, A.M.H.P., Jespersen, J., Tripodi, A., Houghton, D. (1998a) European Concerted Action on Anticoagulation (ECAA). Minimum lyophilized plasma requirement for ISI calibration. American Journal of Clinical Pathology, 109, 196 204.
  • 11
    Poller, L., Thomson, J.M., Taberner, D.A. (1989) Effect of automation on the prothrombin time test in NEQAS surveys. Journal of Clinical Pathology, 42, 97 100.
  • 12
    Poller, L., Triplett, D.A., Hirsh, J., Carroll, J., Clarke, K. (1995a) The value of plasma calibrants in correcting coagulometer effects on International Normalized Ratios: an international multicenter study. American Journal of Clinical Pathology, 103, 358 365.
  • 13
    Poller, L., Triplett, D.A., Hirsh, J., Carroll, J., Clarke, K. (1995b) A comparison of lyophilized artificially depleted plasmas and lyophilized plasmas from patients receiving warfarin in correcting for coagulometer effects on International Normalized Ratios. American Journal of Clinical Pathology, 103, 366 371.
  • 14
    Poller, L., Van Den Besselaar, A.M.H.P., Jespersen, J., Tripodi, A., Houghton, D. (1998b) European Concerted Action on Anticoagulation (ECAA). A comparison of artificially depleted, lyophilized coumarin and fresh coumarin plasmas in thromboplastin calibration. British Journal of Haematology, 101, 462 467.
  • 15
    Ray, M.J. & Smith, I.R. (1990) The dependence of the International Sensitivity Index on the coagulometer used to perform the prothrombin time. Thrombosis and Haemostasis, 63, 424 429.
  • 16
    Stevenson, K.J., Craig, S., Dufty, J.M.K., Taberner, D.A. (1997) System ISI calibration: a universally applicable scheme is possible only when coumarin plasma calibrants are used. British Journal of Haematology, 96, 435 441.
  • 17
    Thomson, J.M. (1970) A Practical Guide to Blood Coagulation and Haemostasis, pp. 178 179. Churchill Livingstone, London.
  • 18
    Tripodi, A., Chantarangul, V., Akkawat, B., Clerici, M., Mannucci, P.M. (1995) A partial factor V deficiency in anticoagulated lyophilized plasmas was identified as a cause of the International Normalized Ratio discrepancy in an External Quality Assessment Scheme. Thrombosis Research, 78, 283.
  • 19
    Van Der Velde, E.A. (1984) Orthogonal regression equation. Thromboplastin Calibration and Oral Anticoagulation (ed. by A. M. H. P. van den Besselaar, H. R. Gralnick and S. M. Lewis), pp. 25 40. Martinus Nijhoff, Boston.
  • 20
    Van Rijn, J.L.M.L., Schmidt Nico, A., Rutten, W. (1989) Correction of instrument and reagent based differences in determination of the international normalised ratio (INR) for monitoring anticoagulant therapy. Clinical Chemistry, 355, 840 843.
  • 21
    W.H.O. Expert Committee on Biological Standardisation (1983) 33rd Report. WHO Technical Series, 687, 81 105.

Appendix

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. STATISTICAL ANALYSIS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  9. Appendix

Field Study Participants (* ECAA National Director)

Belgium: Dr J. Arnout, Leuven*; Dr P. Capel, Brussels; Dr M. de Weer, Leuven; Dr B. Chatelain, Mont-Godinne; Dr A. Criel, Brugge; Dr L. Marcelis, Roeselare; Dr A. Lust, Aalst; Dr M. Vanderplanken, Edegem.

Spain: Professor Dr J. A. Iriarte, Bilbao*; Dr F. Martinez-Brotons, Barcelona; Dr V. Vicente, Murcia; Dr M. A. Fernández, Valencia; Dr J. Fontouberta, Barcelona; Dr A. Ordinas, Barcelona; Dr F. J. Batlle, La Corua; Dr J. Lasierra, Logroño; Dr J. Montero, Madrid; Dr J. L. Navarro, Madrid.

Italy: Professor P. M. Mannucci, Milan*; Dr A. Tosetto, Vicenza; Dr L. Del Buono, Milan; Dr S. Testa, Cremona; Dr N. Erba, Como; Dr A. Gobbi, Milan; Dr A. Galletti, Genoa; Dr M. Pradella, Treviso; Dr N. Ciavarella, Bari; Dr G. Palareti, Bologna; Professor G. Mariani, Rome.

Portugal: Professor J. Pina Cabral, Oporto*; Dr A. R. Araújo, Oporto; Dr F. Crespo, Lisbon; Professor B. Justiça, Oporto; Dr Á. Monteiro, Vila Nova de Gaia; Dr J. F. de Lima, Braga; Dr D. A. M. Dias, V.N. de Famalicão; Dr C. A. F. Antunes, Coimbra; Dr G. Tamagnini, Coimbra; Professsor J. C. de Sousa, Lisbon; A. de Freitas, Aveiro.

Sweden: Professor N. Egberg, Stockholm*; Dr J. Svensson, Danderyd; Dr B. Edlund, Örebro; Dr A. Siegbahn, Uppsala; Dr A. Hillarp, Malmö; Dr G. Larsson, Gothenburg; Dr T. Lindahl, Norrköping; Dr B. Ljungberg, Nyköping.

Norway: Professor U Abildgaard, Oslo*.

Germany: Professor H. Beeser, Freiburg*; Dr H. W. Tomesch, Werdau; Dr S. Appel, München; Dr med Y. Schmitt, Darmstadt; Frau Dipl med B. Lüdtke, Stralsund; Herr Wilhelm, Bad Krozingen; Professor Dr med G. Winckelmann, Wiesbaden; PD Dr B Olgemöller, München; Professor Dr R. Zimmermann, Heidelberg.

Denmark: Professor J. Jespersen, Esbjerg*; Dr A. Bremmelgaard, Næstved; Dr E. Magid, Copenhagen; Dr K. Winther, Kolding; Dr I. Brandslund, Vejle; Dr S. Antonsen, Middlefart; Dr K. Kynde, Roskilde; H. Jelert, Sønderborg.

The Netherlands: Dr A. M. H. P. van den Besselaar, Leiden*; Drs A. P. Anker, Dokkum; Dr Ir A. A. M. Ermens, Eindhoven; J. van der Sloot, Hoorn; Dr R. K. A. van Wermeskerken, The Hague; E. J. Harthoorn-Lasthuizen, s'-Hertogenbosch; Dr H. E. S. J. Hensgens, Amstelveen; Drs R. H. M. Peters, Heerenveen; Dr K. Hamulyak, Maastricht; Dr N. J. Verhoef, Oosterhout.

U.K.: Professor L. Poller, Central Facility*; Dr I. D. Walker, Glasgow; Dr G. Dolan, Nottingham; Professor F. E. Preston, Sheffield; Professor S. J. Machin, London; Dr B. T. Colvin, London; Dr M. Bhavnani, Wigan; Dr C. A. Ludlam, Edinburgh; Dr S. Joseph, Ashton-under-Lyne; Dr T. Flaherty, Preston; Dr P. E. Rose, Warwick; Dr C. R. M. Hay, Manchester; Dr H. Cohen, London.

Switzerland: Professor B. Lämmle, Bern*; Dr U. Bauerfeind, Bern; Dr P. Cornu, Vevey; Dr C. Rufener, Geneva; P. D. Dr Ph De Moerloose, Geneva; Professor Dr A. von Felten, Zürich; Dr U. Gössi, Schwyz; Professor Dr A. Huber, Aarau; PD Dr G. A. Marbet, Basle; Dr P. Pugin, Fribourg.

France: Professor M. Samama, Paris*; Dr I. Houbouyan, Boulogne Billancourt; Professor P. Sie, Toulouse; Dr M. Gouault, Criteil; Dr J. Roussi, Garches; Professor I. Juhan, Marseille; Professor J. Goudemand, Lille; Professor J. Sampol, Marseille; Professor M. C. Guillin, Clichy; Professor T. Lecompte, Vandoeuvre les Nancy.

Greece: Dr I. Kontopoulou-Griva, Athens*; Professor T. Mandalaki, Athens; As. Professor S. Aroni, Athens; Dr D. Zabouli, Thessaloniki; Dr T. Tassopoulou, Athens; Dr M. Antonopoulou, Piraeus; Dr M. Parara, Athens; Dr B. Tsoukanas, Piraeus; Professor A. Maniatis, Patras; R. Stathopoulou, Athens.

Ireland: Dr B. Otridge, Dublin*; Dr P. Cotter, Cork; Dr M. Murray, Galway; Dr D. McCarthy, Dublin; Dr R. O'Donnell, Dublin; Dr O. Smith, Dublin; Dr F. Jackson, Waterford; Ms P. Walsh, Dublin; Dr M. Madden, Cork.

Austria: Professor Dr K. Lechner, Vienna*; Dr G. Aspöck, Wels; Professor Dr M. Fischer, Vienna; Dr P. Hopmeier, Vienna; Dr Odpadlik, Vienna; Professor H. Vinazzer, Linz; Dr Traun, Salzburg; Professor Dr Muntean, Graz; Professor Dr Bauer, Vienna.

Finland: Dr M. Syrjälä, Helsinki*; Dr K. Ivaska, Turku; Dr Eeva-Riitta Savolainen, Oulu; Dr A. Rajamäki, Turku; K. Leisma, Joensuu; Dr E. Mahlamäki, Kuopio; Dr M. Lehtinen, Tampere.