Comparison between international normalized ratio using a portable device and conventional methodology

Abstracts

OBJETIVO: Comparar os resultados do Índice de Normalização Internacional (INR) obtidos pelo teste rápido com os do método convencional nos pacientes em terapia de anticoagulação oral com varfarina sódica. MÉTODOS: Para 383 pacientes tratados com varfarina (idade média: 56,5 anos; 207 mulheres), o INR foi determinado em sangue capilar pelo equipamento Hemochron Jr. e comparado com os resultados de amostras de plasma venoso analisadas pelo teste convencional realizado em equipamento Coag-A-Mate. Foram avaliados os resultados do desempenho global das amostras e dos seguintes subgrupos: INR < 2,0, entre 2,0 a 3,5 e > 3,5. RESULTADOS: A comparação entre os valores de INR dos dois métodos resultou em um coeficiente de correlação (r) de 0,86. Entretanto, a análise das diferenças médias entre os resultados dos dois testes, considerando os três subgrupos, apresentou diferenças estatisticamente significativas (p < 0,001): 0,14 ± 0,21 (INR < 2,0); 0,54 ± 0,31 (2,0 < INR < 3,5) e 1,64 ± 1,10 (INR> 3,5). O cálculo do teste t-pareado de Student resultou em um p < 0,001 para os três subgrupos analisados. CONCLUSÃO: A adoção do teste rápido para monitoramento da anticoagulação oral apresenta restrições. Esse método subestimou a intensidade da anticoagulação nos três subgrupos estudados.

Varfarina; testes de coagulação sangüínea

ORIGINAL ARTICLE

Comparison between international normalized ratio using a portable device and conventional methodology

Luciana Pereira Almeida De Piano; Célia Maria Cássaro Strunz; Antonio de Pádua Mansur; Roberto Abi Rached

Heart Institute (InCor), University of São Paulo Medical School - São Paulo, SP, Brazil

^{Mailing Address}

ABSTRACT

OBJECTIVE: To compare the international normalized ratio (INR) measured by a point-of-care (POC) testing device with that measured by the conventional method in patients undergoing anticoagulation therapy with warfarin sodium.

METHODS: The INR of 383 warfarin-treated patients (mean age: 56.5 years; 207 female) was measured in capillary blood using the Hemochron Jr. device and compared with that of venous plasma samples determined by the conventional method performed in a Coag-A-Mate analyzer. Results were evaluated globally and for the following subgroups: INR < 2.0, from 2.0 to 3.5, and > 3.5.

RESULTS: Using both methods, the comparison between INR values yielded a correlation coefficient (r) of 0.86. However, mean differences in INR in both tests, considering the three subgroups, proved to be statistically significant (p <0.001): 0.14 ± 0.21 (INR< 2.0); 0.54 ± 0.31 (2.0 < INR < 3.5), and 1.64 ± 1.10 (INR> 3.5). Paired Student’s t-test analysis revealed a p value < 0.001 for the three subgroups studied.

CONCLUSION: The use of point-of-care testing for monitoring oral anticoagulation has some limitations. Anticoagulation intensity was underestimated by this method in the three subgroups studied.

Key words: Warfarin, blood coagulation tests/methods.

Vitamin K antagonists, used in oral anticoagulant therapy, have been proven highly effective in the prevention and treatment of thromboembolic events in a variety of clinical situations^1-3.

Some patients require long-term oral anticoagulant therapy, such as those with recurrent venous thromboembolism, chronic atrial fibrillation, mechanical prosthetic valves, and family history of thrombophilia. Warfarin sodium is the oral anticoagulant most widely used nowadays. This drug acts by inhibiting vitamin K-dependent coagulation factors (factors II, V, VII, IX e X)^1,3.

The optimal dosage regimen to provide effective and safe anticoagulation should be established for each patient. Bleeding is the major complication of this therapy^4,5. Therefore, treatment and prevention of thromboembolic disease require periodical laboratory control and the use of standard tests^1,3. It is currently agreed that the best examination for monitoring oral anticoagulation is the international normalized ratio (INR) of prothrombin time (PT).

The INR may be determined daily until an adequate level of anticoagulation is achieved, after which it may be performed at periodic intervals^1-3. This periodicity, however, changes patients’ routine, thus affecting their quality of life.

In order to facilitate anticoagulation monitoring and reduce the waiting time from blood collection to test results, devices have been developed that analyze samples obtained by fingerstick and allow rapid INR measurement (point-of-care testing)^6-9. It has been debated, however, as to whether results from point-of-care testing and standard methods are equivalent.

This study aims, therefore, at evaluating the effectiveness of a point-of-care testing device, Hemochron Jr., in monitoring oral anticoagulant therapy, compared with the conventional method.

Methods

Case series - Patients who came to the Heart Institute Clinical Laboratory for oral anticoagulation monitoring and who had been taking warfarin sodium for at least three months were invited to participate in the study.

Three hundred and eighty-three patients ranging in age from13 to 86 years (56.5±13.9), 54.0% of whom were female, were selected and distributed as follows: 36%, with heart valve prosthesis; 24% with arrhythmia; 7% after an stroke episode; 6% with venous thromboembolism; 5% with a pacemaker; and 22% associated with other causes.

At the time of blood sampling, all patients completed a questionnaire covering the following: name, hospital ID, age, gender, main diagnosis, reason for anticoagulation, weekly dosage of the anticoagulant drug, use of other medications and their respective dosages, and bleeding events within the last 10 days.

Patients that experienced problems during sampling were eliminated from the study.

Conventional prothrombin time measurement - Venous blood samples were collected into tubes containing 3.8% sodium citrate solution. Plasma samples were obtained by centrifugation at 2,500 rpm for 10 minutes and tested using the Coag-A-Mate MTX II automated analyzer, Organon Teknika, Biomerieux, (Marcy-L’Etoile, France), and a specific kit containing rabbit-brain thromboplastin with an ISI (International Sensitivity Index) of 1.2.

A plasma pool prepared with samples collected from normal subjects was assayed daily. External quality controls (international and national) were used to assure test reliability. Results were divided into three groups, according to the INR obtained.

INR < 2.0: inadequately anticoagulated patients

2.0 < INR < 3.5: adequately anticoagulated patients

INR > 3.5: overanticoagulated patients

Point-of-care prothrombin time measurement - Capillary blood samples were obtained by fingerstick, and one blood drop was dispensed into the Hemochron Jr. II portable device, ITC, NJ, USA (Fig. 1). Rabbit-brain thromboplastin (ISI = 1.0) was used.

The device has a cuvette containing lyophilized thromboplastin, stabilizers, and buffers. Clot detection is performed by a series of optical detectors aligned with the channel of the test cuvette. The speed at which the blood sample moves between both detectors is measured. As clot formation begins, blood flow is blocked and the movement slows. The electronic optical detection of a fibrin clot in the blood sample automatically terminates the test, causing the instrument’s timer to display the clotting time in seconds. PT, already corrected for plasma value, and INR results appear alternately on the display.

Calibration was carried out daily, as well as internal control tests provided by same manufacturer. A training course for proper device operation was provided for all laboratory personnel that performed measurements.

Statistical analysis - INR values for both devices were analyzed in the whole group and in the three subgroups, namely, < 2.0; from 2.0 to 3.5, and > 3.5, using the paired t test, the Bland-Altman test, and Pearson’s correlation to evaluate result differences. A p-value < 0,05 was considered statistically significant.

Results

The 383 study participants were distributed as follows: 101 (26%) patients with INR < 2.0 (inadequately anticoagulated), 237 (62%) with INR between 2.0 and 3.5 (adequately anticoagulated) and 45 (12%) with INR > 3.5 (overanticoagulated).

Pearson’s correlation coefficient (r) obtained by comparing global performance of the two methods was 0.86 (Fig. 2). When INR values were divided into the subgroups, coefficients dropped to 0.65 (< 2.0), 0.72 (between 2.0 and 3.5), and 0.55 (> 3.5). The Bland-Altman analysis demonstrated significant differences for both the whole group and subgroups. In addition, an increase in mean INR was accompanied by a significant increase in INR differences (Tab. 1). Frequency distribution of values measured in the three categories showed an agreement of only 66% between both methodologies (Tab. 2), with a high percentage of patients adequately anticoagulated by the conventional method yet inadequately anticoagulated by the POC test.

In the subgroup with INR > 3.5, two aberrant results, that is, showing more than 50% deviation from the true INR, were detected after the point-of-care test: 7.2 and 9.4 by the conventional method versus 3.4 and 2.4, respectively, by the point-of-care test.

Statistical analysis using the paired Student’s t test showed significant differences between both methodologies for the three subgroups analyzed (p<0.001).

Discussion

The number of patients undergoing oral anticoagulant therapy has been increasing sharply in recent years, overloading monitoring centers. Hence, point-of-care testing devices have been used as a means of surpassing this problem. These tests are primarily intended to improve the patient’s quality of life, in that they provide greater flexibility, higher frequency and less discomfort¹⁰.

In a number of studies carried out to validate different portable devices, accuracy levels of POC tests were considered clinically acceptable^6-9. On the other hand, some authors reported problems regarding the reliability of test results, pointing to the need of better calibration and use of external quality controls^11-14.

In the present study, we evaluated the performance of the Hemochron Jr. device by comparing its results with those obtained with the standard method. In spite of the good correlation found in global performance (r = 0.86), paired Student’s t-test and Bland Altman test (Tab. 1) revealed significant differences (p <0.001) between both methodologies.

For a significant number of patients within the therapeutic range (Tab. 2), POC values were lower than 2.0 (24%). These results are clinically relevant, because these patients would be prescribed higher doses of warfarin, thereby becoming prone to bleeding. In addition, overanticoagulated patients had their values underestimated by point-of-care testing. In this overanticoagulated group, two aberrant results were detected, in which patients at high risk of bleeding would be left untreated if the values taken into account were those of the point-of-care testing.

This study was carried out by trained professionals, to reduce the likelihood of errors during the preanalytical phase. Thus, we believe that the differences found should not be entirely related to inadequate sample collection or even management of the sample in the device. One of the possible explanations is related to differences between the matrixes in both methods. Also, the presence of platelet proteins with modulating activities of coagulation in whole blood samples may have contributed to the deviations observed.

Data presented in this study indicate lack of accuracy of point-of-care testing, thereby restricting its use for monitoring oral anticoagulant therapy.

References

Manuscript received October 19, 2005; revised manuscript received February 15, 2006; accepted February 24, 2006.

• 1. Ansell J, Hirsh J, Dalen J, Bussey H, Anderson D, Poller L, et al. Managing oral anticoagulant therapy. Chest. 2001; 119 (Suppl 1): 22S-38S.
• 2. Gage BF, Fihn SD, White RH. Management and dosing of warfarin therapy. Am J Med. 2000; 109: 481-8.
• 3. Hirsh J, Dalen J, Anderson DR, Poller L, Bussey H, Ansell J, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest. 2001; 119 (Suppl 1): 8S-21S.
• 4. Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, DAngelo A, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet. 1996; 48: 423-8.
• 5. Levine MN, Raskob G, Landefeld S, Kearon C. Hemorrhagic complications of anticoagulant treatment. Chest. 2001; 119 (Suppl 1): 108S-121S.
• 6. Biasiolo A, Rampazzo P, Furnari O, Filippi B, Pengo V. Comparison between routine laboratory prothrombin time measurements and fingerstick determinations using a near-patient testing device (Pro-Time). Thromb Res. 2000; 97: 495-8.
• 7. Cosmi B, Palareti G, Moia M, Carpenedo M, Pengo V, Biasiolo A, et al. Accuracy of a portable prothrombin time monitor (Coagucheck) in patients on chronic oral anticoagulant therapy: a prospective multicenter study. Thromb Res. 2000; 100: 279-86.
• 8. Cromheecke ME, Levi M, Colly LP, de Mol BJ, Prins MH, Hutten BA, et al. Oral anticoagulation self-management and management by a specialist anticoagulation clinic: a randomized cross-over comparison. Lancet. 2000; 356: 97-102.
• 9. Douketis JD, Lane A, Milne J, Ginsberg JS. Accuracy of a portable International Normalization Ratio monitor in outpatients receiving long-term oral anticoagulant therapy: comparison with a laboratory reference standard using clinically relevant criteria for agreement. Thromb Res. 1998; 92: 11-7.
• 10. Price CP. Point of care testing. BMJ. 2001; 322: 1285-8.
• 11. Poller L, Keown M, Chauhan N, van den Besselaar AM, Tripodi A, Shiach C, et al. European concerted action on anticoagulation: quality assessment of the CoaguChek Mini and TAS PT-NC point-of-care whole-blood prothrombin time monitors. Clin Chem. 2004; 50: 537-44.
• 12. Poller L, Keown M, Chauhan N, van den Besselaar AM, Tripodi A, Shiach C, et al. Reliability of international normalized ratios from two point of care test systems: comparison with conventional methods. BMJ. 2003; 327: 30.
• 13. Tripodi A, Breukink-Engbers WG, van den Besselaar AM. Oral anticoagulant monitoring by laboratory or near-patient testing: what a clinician should be aware of. Semin Vasc Med. 2003; 3: 243-54.
• 14. van den Besselaar AM. Accuracy, precision, and quality control for point-of-care testing of oral anticoagulation. J Thromb Thrombolysis. 2001; 12: 35-40.

Publication Dates

• Publication in this collection
  07 Mar 2007
• Date of issue
  Jan 2007

History

• Accepted
  24 Feb 2006
• Reviewed
  15 Feb 2006
• Received
  19 Oct 2005