Point-of-Care Prothrombin Time Measurement for Professional and Patient Self-Testing Use A Multicenter Clinical Experience

Transcription

1 Coagulation and Transfusion Medicine / POINT-OF-CARE PROTHROMBIN TIME MEASUREMENT Point-of-Care Prothrombin Time Measurement for Professional and Patient Self-Testing Use A Multicenter Clinical Experience Oral Anticoagulation Monitoring Study Group * Key Words: Prothrombin time; Point-of-care testing; Self-testing; Anticoagulation Abstract We enrolled 386 subjects in a multicenter study of a point-of-care (POC) prothrombin time (PT) testing device. POC tests were performed by health care professionals using venous and finger-stick specimens and by patients using finger-stick specimens. Venous blood also was analyzed in the local hospital laboratory and a national reference laboratory. Accurate POC results were obtained by professionals using both types of specimens. Patients results were equivalent to those of professionals. The identification of the patient s therapeutic status based on the International Normalized Ratio (INR) was equivalent for POC and local hospital laboratory PT results; 75% of local laboratory results and 77% of POC results were within 0.4 INR of reference laboratory results, while 93% of either system (POC or local laboratory) were within 0.7 INR. Patients overwhelmingly reported satisfaction with the self-test, including the finger stick and device operation. The INR from the POC device is clinically equivalent to the laboratory INR for assessment of anticoagulation status and management decisions in professional and self-testing environments. Patients can learn to perform accurate PT testing, and POC PT testing is feasible in patients homes. Oral anticoagulant therapy with warfarin and related compounds is indicated to reduce the risk of thromboembolism in a wide variety of cardiac and vascular disorders, including atrial fibrillation, valvular heart disease, mechanical heart valve prostheses, deep vein thrombosis, pulmonary embolism, and stroke. 1,2 Safe and effective long-term anticoagulation requires strict maintenance of the target therapeutic range 3,4 to avoid life-threatening complications of hemorrhage and thrombosis. 1-5 The prothrombin time (PT) test is used widely in clinical practice to monitor the patient s coagulation status and guide warfarin dosage adjustments required to maintain the therapeutic range. There is preliminary evidence that maintenance of the target range is influenced favorably by the frequency of testing. 6,7 When using a monthly schedule of monitoring, only 50% to 60% of patients are expected to remain within the target range. 2 In contrast, patients monitored on a weekly basis will achieve and maintain the desired level of anticoagulation nearly 85% of the time. 8 Frequent coagulation testing is particularly important after initiation of therapy, changes in (or addition of) concomitant medication known to influence warfarin response, and with changes in the patient s medical condition. 5 Accessibility to testing provides a means to closely monitor patients with an out-of-range value before initiating a dosage change that may not be warranted. PT testing traditionally has been performed on plasma in the clinical laboratory. However, point-of-care (POC) coagulation monitors have been developed with a goal of achieving optimal outcomes by allowing more frequent testing with instrument and reagent consistency. Current POC technology provides a rapid and reliable means to determine anticoagulation status and make the required warfarin dosage adjustments in a timely manner It is a particularly attractive 288 Am J Clin Pathol 2001;115: American Society of Clinical Pathologists

2 Coagulation and Transfusion Medicine / ORIGINAL ARTICLE option for alternative site testing, such as physician-office laboratories, outpatient clinics, and patient self-testing. 12,13 The purpose of our multicenter study was to evaluate a new POC monitor, which uses a standardized reagent and instrument system and built-in integral controls, for providing consistent International Normalized Ratio (INR) results. The study objectives were as follows: (1) to establish the equivalence of the POC result to the laboratory standard, (2) to determine the ability of patients to perform PT tests using the POC system under professional supervision, and (3) to determine the accuracy of results obtained by trained health care professionals and by patients in a controlled professional environment. Materials and Methods Study Design The study was a split sample design to evaluate the performance of the ProTime Microcoagulation System (International Technidyne, Edison, NJ) using the laboratory PT test as the standard of reference. POC testing was performed by health care professionals with the ProTime system using fresh capillary (ie, finger stick) and nonanticoagulated venous blood. The performance of the ProTime system was evaluated by comparing results with those obtained at the local hospital laboratory and at a national reference laboratory. The accuracy of PT test results was assessed using linear regression and a (Bland-Altman 14 ) mean vs difference analysis of the data compared with a hospital and reference laboratory result. Bias and systematic error 15 were assessed for the respective regression data sets. Differences of results were evaluated at critical clinical decision limits 16 to determine the impact of INR difference on therapy management. ProTime Microcoagulation System The ProTime system consists of the device, a disposable reagent cuvette, and a Tenderlett Plus (International Technidyne) finger-stick blood collection device Figure 1. Each ProTime cuvette has 5 test channels, 3 of which are used for triplicate testing of the patient s PT and 2 for the simultaneous analysis of 2 levels of quality control. Each channel contains dried thromboplastin reagent, while the control channels also contain dried purified plasma-extracted (vitamin K dependent) coagulation factors and biologic anticoagulants. The combination of reagents in the control channels yields predetermined clotting ranges for a level I (normal) and level II (abnormal) control. With each patient test, the patient s blood specimen rehydrates the thromboplastin and the control formulations. The level I control contains coagulation factors to normalize the INR of the patient s blood specimen, while the level II contains the same factors plus a precise quantity of a coagulation inhibitor (eg, heparin). In this integral quality control system, a set of preprogrammed quality control range limits are specified for the level I and level II controls and for the relationship between the level I and level II controls. 17 If any of the preset quality criteria are violated, a test error message is generated and displayed instead of a result, thereby minimizing the risk of reporting an erroneous result. The ProTime instrument initiates a self-check procedure to verify electrical and mechanical components before every test and prompts the user through the steps of running the test. The system measures the PT from a whole blood microsample using a combination mechanical-optical method to detect fibrin clot formation. The instrument draws a precise volume of blood into the channels of the cuvette, each of which contains a high sensitivity thromboplastin (International Sensitivity Index [ISI], approximately 1.0). The blood sample in each channel then is pumped back and forth through a precision restriction in the cuvette until a clot begins to form, obstructing the channel and slowing the flow of blood. Clot detection occurs when the blood movement monitored by a series of light sensors decreases below a predetermined rate. The median of the patient s 3 PT test values is reported as an INR. The median PT was selected in preference to the mean PT as a way to screen out random results. The 3 separate PT values replicate within 5%, and the mean and median values are generally within 5%. In approximately 2% of cases (based on a screen of 965 data triplicates), when the mean and median differ by more than 5% owing to a single aberrant PT value, the median is the more accurate result. The whole blood clotting time is converted directly to the INR using a conversion equation that was established at the inception of this trial (n = 247 test specimens). All statistical comparisons for the present study were performed using the INR. For the operator s benefit a plasma-equivalent clotting time in seconds, at an operator-specified ISI, is calculated from the INR and displayed. The Appendix provides the details of this conversion, including the calibration of the cuvette interlot variability. Before the clinical evaluation, extensive performance validation was performed in the laboratory to demonstrate precision of the PT result and integral control values. In this evaluation, reliable results were obtained for the hematocrit range of 20% to 60% ( ). Once installed at the user s site, no additional validation was required. Subjects Study participants (N = 386) were patients receiving oral anticoagulation medication and nonanticoagulated control subjects. All patients attending 1 of the 7 participating hospital-based anticoagulation clinics regardless of age, sex, or medical condition were eligible to participate. Each participating anticoagulation center was associated American Society of Clinical Pathologists Am J Clin Pathol 2001;115:

3 Oral Anticoagulation Monitoring Study Group / POINT-OF-CARE PROTHROMBIN TIME MEASUREMENT A B C Figure 1 A, ProTime Microcoagulation System, International Technidyne, Edison, NJ. B, The Tenderlett Plus is used to perform a finger incision. Blood is collected in the collection cup attached to the Tenderlett Plus. C, The collection cup is then engaged in the test cartridge, and the device start button is pushed, initiating the testing procedure. Used by permission of International Technidyne. with a tertiary care medical facility. (See footnote * for participating centers and principal investigators.) Volunteer subjects who were not taking oral anticoagulants were recruited as control subjects at each clinical site. All adult participants signed informed consent before beginning the study. One center was a primary pediatric referral center for thrombophilia management, which accounts for the large number of pediatric patient participants. The parent or legal guardian of minors (children younger than 18 years) also signed consents. All study sites received internal review board approval before initiating the study. At each site, up to 4 different devices were used. Sample Collection and POC Testing by Health Care Professionals A venous blood specimen was collected in a sterile syringe from each subject. Two drops of whole blood (approximately 0.06 ml) then were placed immediately into the Tenderlett Plus collection cup, which was then attached to the ProTime device, and the PT test was performed. The remainder of the venous sample (approximately 4.5 ml) was placed in a standard evacuated blue top collection tube containing sodium citrate (0.129-mol/L concentration) and transported to the facility s clinical laboratory for plasma processing and PT analysis. After collection of the venous sample test, finger-stick whole blood samples were obtained from the patient and tested in the ProTime system by the health care professional. Patient Self-Testing A subset of the total study population (n = 177 adults) was recruited to a patient self-testing (PST) group. The aim of this aspect of the study was to assess the patient s ability to collect and accurately test a finger-stick blood sample 290 Am J Clin Pathol 2000;115: American Society of Clinical Pathologists

4 Coagulation and Transfusion Medicine / ORIGINAL ARTICLE using the ProTime system. As this study progressed, patient selection criteria were identified and refined for later use in a home evaluation. Patients were selected based on health care professional assessment of their ability to independently perform the test, understand the necessity to adhere to an established testing procedure, and communicate effectively with health care personnel. The properly selected patients were given written and verbal instructions followed by a video tape presentation on the correct method to prepare the finger for the incision, collect blood in the collection apparatus, place the collection apparatus in the ProTime instrument, start the device, and record and report the test results. Patients were required to demonstrate proficiency before independently performing the study test. After training and performance of the ProTime test, patients completed a survey about their ability and comfort with performing the self-test and their willingness to perform an unsupervised ProTime test at home. Hospital Laboratory Testing The hospital s clinical laboratory prepared plasma from the anticoagulated venous whole blood sample, divided the plasma into aliquots, performed a PT assay on 1 aliquot, and stored the remaining plasma sample at 40 C for later transport to a reference laboratory. Each hospital laboratory used a unique instrument and reagent system Table 1. Reference Laboratory An independent reference laboratory (Midwest Hemostasis and Thrombosis Laboratories, Muncie, IN) performed routine PT tests on the frozen plasma samples from all sites. The samples were batched by site and tested in duplicate using 2 different reagent systems: reagent A, Ortho Recomboplastin (Ortho Diagnostics Systems, Raritan, NJ; ISI = 1.0) and reagent B, Dade C+ (Baxter, Miami, FL; ISI = 2.0). All tests were analyzed on an Electra 900 instrument (Medical Laboratories Automation, Pleasantville, NY). On the day of testing, the plasma samples were thawed in a Table 1 PT Test Systems Used at the Local Hospital Laboratory Clinical Site Instrument * Reagent ISI 1 ACL Ortho MLA Dade C MLA Dade C STA Ortho STA Thrombol-S ACL Thrombol-S ACL Thrombol-S 1.07 ISI, International Sensitivity Index; PT, prothrombin time. * ACL, Instrumentation Laboratory, Lexington, MA; MLA, Medical Laboratories Automation, Pleasantville, NY; STA, Stago, Milan, Italy. Dade C+, Baxter, Miami, FL; Ortho, Ortho Diagnostics Systems, Raritan, NJ; Thrombol-S, Boehringer Mannheim, Indianapolis, IN. 37 C water bath and tested with both reagent systems serially by the same technician. The plasma PT (mean of duplicate tests) was recorded, and the INR was computed. Data Management To obtain a reliable database for establishing the accuracy of the POC test, an independent assessment of the agreement of the laboratory plasma PT results was conducted. If these plasma results did not agree, the data were not used to assess the POC test accuracy. Ten samples were eliminated from the analysis because the duplicate PT tests performed in the reference laboratory using reagents A and B fell outside the 99% confidence limit. Complete data sets of 35 samples also were eliminated because the results reported by the reference laboratory and the hospital laboratory differed by more than 1 INR unit. The rationale for this data cleansing was that identification of disagreement of the plasma PT results has no bearing on the performance assessment of the POC test. Without agreement of the plasma tests, it is impossible to select the correct plasma result to which the PST result could be compared. Disagreement of the plasma results could be ascribed to INR variability (see Discussion ) or sample handling errors before or during testing. In addition, for the ProTime results, any individual patient sample that violated the quality control criteria built into the ProTime system was not included in the data set analyzed. These criteria included a blood collection time limit of 2.5 minutes (from the start of collection of blood to performance of test) and a sample size range of 8.75 to µl per channel with a maximum of 1.25-µL variation between channels for each cuvette. The integral controls (level I and level II) had to clot within the defined specifications for the test result to be included in the data set. Results Study Population The mean age of the study population was 45 years (range, 1-85 years). Seventy-three subjects were younger than 18 years (mean, 10 years). The mean adult age was 55 years. The total population was 54.9% females (n = 212) and 45.1% males (n = 174). The indication for anticoagulation in this population is shown is Table 2. Equivalence of the POC and Reference Laboratory INR The correlation of the ProTime (venous sample) and the reference laboratory (reagent A) is shown in Figure 2. The slope of the regression line (0.88) demonstrates the degree to which the results are comparable. The good correlation of American Society of Clinical Pathologists Am J Clin Pathol 2001;115:

5 Oral Anticoagulation Monitoring Study Group / POINT-OF-CARE PROTHROMBIN TIME MEASUREMENT Table 2 Indications for Anticoagulation * Indication Adults (n = 313) Children (n = 73) Valve replacement 105 (33.5) 13 (17.8) Venous thrombosis 56 (17.9) 51 (69.9) Atrial fibrillation 34 (10.9) 1 (1.4) Myocardial infarction 16 (5.1) 0 (0.0) Pulmonary embolism 15 (4.8) 0 (0.0) Other 13 (4.2) 0 (0.0) Nonanticoagulated 74 (23.6) 8 (11.0) control subjects * Data are given as number (percentage). the POC and laboratory results (r = 0.92) demonstrates a high degree of similarity of the 2 tests. Plots of individual clinical site data correlated with the reference laboratory data demonstrated no substantial bias of any site. Figure 3 uses a mean vs difference plot to demonstrate the relationship between the ProTime (venous specimen) and the reference laboratory results. The mean values of the 2 methods represent the best estimate of the true test value. 14 The lines indicated at ±0.4 and ±0.7 INR are used to demonstrate that the majority of the difference between the 2 test methods falls within these boundaries, which have been defined as clinically important precision ranges for the monitoring of oral anticoagulant therapy. 16 The data are distributed evenly around zero with an average mean difference, derived from the Bland-Altman analysis, of 0.03 INR. While ProTime Venous INR Reference Laboratory INR Figure 2 Comparison of the ProTime Microcoagulation System (International Technidyne, Edison, NJ) venous International Normalized Ratio (INR) and the reference laboratory INR. The relationship between the venous specimen tested with the ProTime and the plasma specimen tested in the reference laboratory using reagent A (Ortho Recomboplastin, Ortho Diagnostics Systems, Raritan NJ; International Sensitivity Index = 1.0) is illustrated. SE of the y-intercept = 0.057; SE slope = 0.023; mean SE of the regression = 0.137; df = 262. y = 0.88x ; r = 0.92; n = 264. this supports a nonbiased distribution of the data, a further systematic error analysis demonstrated a tendency for the ProTime (venous) to be slightly less than the reference laboratory result. The systematic error was 0.06 at 2 INR, 0.06 at 3 INR, and 0.18 at 4 INR. An increased scatter was observed at mean INR values of 3.0 or greater. At higher INR values, differences of thromboplastin sensitivity, for which the ISI does not completely account, contribute to the INR difference observed using different test methods for the same sample. Aside from a few outliers, values in the commonly targeted therapeutic range of 2.0 to were clinically similar. Comparison of the ProTime and Hospital Laboratory Results With the Reference Laboratory Result INR results from the local hospital laboratory and ProTime system were compared independently with the reference laboratory to define accuracy. Since imprecision of the INR among different test systems is known, the comparison of 2 independent methods (local laboratory and ProTime) with a standard method (reference laboratory) helps define the clinical equivalence of the ProTime result to the hospital laboratory result when managing anticoagulation therapy. The agreement of the reference laboratory result with the ProTime result for the same venous sample is similar to the agreement found between the reference and hospital laboratories Table 3. Nearly 80% of the hospital laboratory and ProTime results were within 0.4 INR of the reference laboratory results. Within the range of 0.7 INR, more than 90% of the ProTime and hospital laboratory results agree with the reference laboratory results. Individual hospital laboratory results showed varying degrees of slight bias compared with the reference laboratory, the degree of which reflected systematic differences between the local hospital PT-INR test system and the reference PT-INR test system. Equivalence of Finger-Stick INR and Reference Laboratory Result The relationship of the ProTime result using finger-stick blood and the reference laboratory result was similar to that observed between the venous ProTime sample and the reference laboratory. Figure 4 displays the relationship of the finger stick performed by the health care professional with the reference laboratory result, for which the correlation coefficient was The systematic error was 0.07 at 2 INR, 0.11 at 3 INR, and 0.29 at 4 INR. Subsequently, a second finger-stick sample evaluation was performed, permitting patients to collect their own blood and perform the test. This was done to assess the ease of use of the ProTime system and to determine the accuracy of patient self-testing. Figure 5 shows the excellent agreement between the results of 292 Am J Clin Pathol 2000;115: American Society of Clinical Pathologists

6 Coagulation and Transfusion Medicine / ORIGINAL ARTICLE Difference (Lab INR ProTime INR) Mean INR Figure 3 Analysis of mean vs difference reference laboratory International Normalized Ratio (INR) and venous ProTime Microcoagulation System (International Technidyne, Edison, NJ) INR. The horizontal lines represent 0.4 and 0.7 INR differences between the 2 systems (see text for details). Table 3 Percentage Agreement of the ProTime and Hospital Laboratory Results With Reference Laboratory Results Range * ProTime (n = 264) Hospital Laboratory (n = 250) 0.4 or less or less Professional Finger-Stick INR Reference Laboratory INR Figure 4 Comparison of ProTime Microcoagulation System (International Technidyne, Edison, NJ) professional fingerstick International Normalized Ratio (INR) and the reference laboratory INR. The relationship between finger-stick collected specimens tested with the ProTime by the health care professional and the plasma specimens tested in the reference laboratory is illustrated. SE of the y- intercept = 0.063; SE slope = 0.025; mean SE of the regression = 0.166; df = 250. y = 0.82x ; r = 0.90; n = 252. * Range represents the difference of the reference laboratory result and ProTime or hospital laboratory result. ProTime Microcoagulation System, International Technidyne, Edison, NJ. patients self-test and the finger-stick samples obtained by health care professionals. A high degree of correlation was found between the self-test and the health care professional finger-stick results (r = 0.92). Minimal systematic error was observed between the 2 finger-stick tests: zero at 2 INR, 0.05 at 3 INR, and 0.1 at 4 INR. The patient s perception of the self-test method was evaluated by survey. After performance of the test, patients overwhelmingly reported a satisfaction with the ease of the self-test and a willingness to use the ProTime at home Table 4. With training, patients were able to master the finger-stick blood collection and collect the amount of blood required to perform the test. Incidence of Integral Control Test Rejection Failure of either of the 2 POC device integral controls results in no test result being displayed to the operator. This on-board quality control function is designed to eliminate bad test results attributable to a poor specimen or deterioration of the test cartridge. At the onset of the trial with newly trained clinical sites, the incidence of quality control violation was approximately 10%. As the sites gained clinical experience, the incidence of quality control violations dropped to 4% to 8% across the individual sites. Discussion The standard test for the monitoring of oral anticoagulants is the plasma-based PT test. POC PT testing permits immediate assessment of anticoagulation status in a variety of clinical settings. In the professional setting, POC testing eliminates the need for blood specimen transport and processing and permits the clinician the opportunity to immediately adjust the patient s therapeutic regimen. The logical extension of professional POC testing is PST, ultimately providing a vehicle to patient self-management. 12 The ProTime system evaluated in the present study is suited for health care professional and patient use. Our evaluation demonstrates clinical agreement of the ProTime system using venous or finger-stick whole blood specimens with a standard reference laboratory plasma test. The ProTime system produced results comparable to those obtained by the reference laboratory for tests performed by qualified health care professionals and by properly selected, suitably trained patients. We observed a slight bias of the POC result compared with the reference laboratory, yet it was of no greater magnitude than documented laboratory-based INR differences, based on interlaboratory and intralaboratory variability American Society of Clinical Pathologists Am J Clin Pathol 2001;115:

7 Oral Anticoagulation Monitoring Study Group / POINT-OF-CARE PROTHROMBIN TIME MEASUREMENT Patient Self-Test INR Professional Finger-Stick INR Figure 5 Comparison of patient self-testing and the health care professional International Normalized Ratio (INR) using a finger-stick specimen. The relationship between the fingerstick specimens collected and tested on the ProTime Microcoagulation System (International Technidyne, Edison, NJ) by the patient and by the health care professional is illustrated. SE of the y-intercept = 0.08; SE slope = 0.030; mean SE of the regression = 0.180; df = 173. y = 0.95x ; r = 0.92; n = 175. INR monitoring serves to maintain appropriate anticoagulation levels specific for each patient s condition. 1-5 Maintenance of the therapeutic range is the accepted surrogate measure for reducing complications associated with oral anticoagulants. Under ideal conditions, the need for an appropriate therapeutic intervention is quickly identified in properly monitored patients. Our study shows that POC testing accurately assesses the therapeutic status and is an effective method of anticoagulation monitoring. The identification of the patient s therapeutic status was equivalent for the ProTime system and the hospital laboratory when compared with a central reference laboratory using comparative INR differences of 0.4 and 0.7, values defined as clinically equivalent by Lassen et al. 16 The combined (venous plus finger stick) systematic error of approximately 0.1 at 3 INR and 0.2 at 4.0 INR yields a nearly negligent impact on therapeutic clinical decisions and, in fact, is less than the 0.4 Table 4 Patient Satisfaction With PST and ProTime Device * Variable Number (%) of Responses Ease of use performing a self-test (n = 177) Easy 144 (81.4) Moderate 20 (11.3) Difficult 13 (7.3) Willingness to use the self-test at home (n = 177) Yes 153 (86.4) No 24 (13.6) PST, patient self-testing. INR value identified as a meaningful descriptor of system reliability. 16 Ng et al 20 also found a nearly identical incidence of disagreement in INR classification when they compared paired samples tested at 3 hospital laboratories. Analyzing individual matched data points in the correlation graphs presented illustrates the importance of physician knowledge of INR variability when interpreting INR results and deciding appropriate therapeutic action. The clinical reality of any laboratory or PST test is that at the critical therapeutic limits (ie, INR), variability of results, even with the best of test conditions, is expected, and each result must be addressed with caution. Often the best course of action is to closely monitor INR results in the ensuing days (or hours) before initiating treatment changes. The ProTime system is a controlled and standardized POC measurement, the use of which ultimately could lead to better anticoagulation management and patient outcome. The original method of reporting prothrombin time in clotting time seconds for the monitoring of warfarin anticoagulation has been replaced largely by the INR system in which the PT results are adjusted according to the sensitivity of the testing reagent. Despite standardization through assignment of the ISI to thromboplastin, INR variability among different test systems (ie, reagent and instrument combination) is well known Consequently, anticoagulation management becomes problematic when a laboratory changes reagents or a patient uses a different laboratory. 23 The ProTime system, using a sensitive, low ISI thromboplastin, is a dedicated instrument and reagent system, which increases test standardization and promotes accurate long-term monitoring of the patient s anticoagulation status. POC testing eliminates variables associated with delayed sample transportation and specimen handling. The ProTime quality control system, including electronic system verification and integral 2-level control testing with every patient test, serves to ensure consistent instrument performance. By incorporating integral positive and negative controls into each cuvette, the need to use external control materials and procedures is eliminated. Measuring the PT in triplicate and reporting the median INR value, rather than the mean INR value, provides a statistical method to screen results for outliers since a single discrepant test value in the triplicate set will not affect the reported INR. In the physician s office or hospital clinic, a POC system greatly reduces turnaround time, providing the clinician the opportunity for immediate dosage adjustment and patient counseling, without the delay associated with routine laboratory testing. With immediate patient counseling available, less patient follow-up may be required by the health care staff. In a financial analysis, the cost of anticoagulation management includes the direct costs of the assay, the indirect expense of professional staff and resource utilization, 294 Am J Clin Pathol 2000;115: American Society of Clinical Pathologists

8 Coagulation and Transfusion Medicine / ORIGINAL ARTICLE and the management of adverse events. While direct costs, which may vary considerably depending on test frequency, may be higher for POC PT testing, indirect and total costs are better controlled under a comprehensive anticoagulation management program. 24 The present study also demonstrated that patients are able to accurately perform a PT self-test with the device. Results of PST by a demographically diverse patient population were identical to those of the finger-stick test performed by health care professionals. This confirms that properly selected and suitably trained patients are capable of obtaining INR results comparable to laboratory INR values and provides another possible monitoring alternative, utilization of the system for PST in nontraditional settings, such as in the home. PST provides patients with a greater access to test monitoring when attendance at the clinic or physician office is not possible. Regular monitoring has been shown to help maintain the patient s anticoagulation status within the therapeutic range 2,6-8 and reduce the frequency of complications associated with oral anticoagulant therapy. 7,25-27 Accessibility to PT monitoring also may provide a means to reduce unnecessary dosage adjustments. More frequent PT testing of patients with out-of-range results is possible with POC testing, in lieu of dosage adjustment. The present study demonstrates POC results that are comparable to those of the clinical laboratory in the hands of health care professionals and properly selected, suitably trained patients. The ProTime system offers the potential to increase patient compliance with the physician-directed testing protocol, improve maintenance of the therapeutic range, and positively influence the reduction of complications associated with oral anticoagulant use. With accessibility to POC testing, the hypothesis that increased routine PT testing serves to improve therapeutic range maintenance and reduce complications 2,6-8 can be supplemented by a prospective, randomized trial. Ultimately, an improved clinical outcome can be achieved by providing closely monitored anticoagulation therapy through the appropriate use of POC testing in anticoagulation centers and, for a qualified subset of patients, through PST. Appendix: Method of INR and Plasma- Equivalent PT Calculation The whole blood clotting time (WBCT) is converted directly to an INR using the equation INR = WBCT (0.055) 0.35, which was derived from the pilot clinical data, as described in the Materials and Methods section. Once this standard conversion was established, subsequent lots of PT cuvettes could be matched to this standard through minor adjustments of the regression equation slope and offset (eg, y-intercept), based on in vitro calibration with normal and abnormal blood specimens. In the device, a plasma-equivalent PT (PT pt ) subsequently is calculated based on this INR and the ISI of the reagent used in the test cuvette (ISI = 1.0), using the conventional INR equation 4 : INR = (PT pt /PT n ) ISI The device default parameters used to calculate the PT pt are ISI = 1.0, which is the sensitivity of the thromboplastin reagent used in the cuvette, and PT n (mean normal) = 13.1 seconds, also established in the pilot clinical trial. Since some clinicians may prefer to match their PT pt results with local laboratory results, the operator may enter a setup phase in the device to select an alternative ISI from which the plasma PT will be calculated. The PT n value used in the calculation is adjusted automatically when the ISI is changed. For example: With an ISI of 1.0, the PT n and PT pt at each respective INR are: PT n = 13.1 PT pt = 13.1 at INR of 1.0; PT pt = 26.1 at INR of 2.0; and PT pt = 39.2 at INR of 3.0 With an ISI of 2.4, the PT n and PT pt at each respective INR are: PT n = 12.0 PT pt = 12.0 at INR of 1.0; PT pt = 16.0 at INR of 2.0; and PT pt = 19.0 at INR of 3.0 These calculations are intended for the reader to understand the operation of the device. All data have been analyzed and presented in the article as INR. Plasma-equivalent PT was not used for analysis. * The Oral Anticoagulation Monitoring Study Group consists of the following principal investigators: Maureen Andrew, MD, Hamilton Civic Hospital Center, Hamilton, and the Hospital for Sick Children, Toronto, Ontario; Jack L. Ansell, MD, Boston University Medical Center, Boston, MA; Daniel M. Becker, MD, University of Virginia Medical Center, Charlottesville; Richard C. Becker, MD, Thrombosis Research Center, University of Massachusetts Memorial Medical Center, Worcester; and Douglas A. Triplett, MD, Ball Memorial Hospital, Midwest Hemostasis and Thrombosis Laboratories, Muncie, IN. Supported in part by International Technidyne, Edison, NJ. Presented in part at the Annual Meeting of the American College of Cardiology, Orlando, FL, March Address reprint requests to Dr Triplett: Dept of Research, Ball Memorial Hospital, 2401 University Ave, Muncie, IN References 1. Cannegieter SC, Rosendaal FR, Wintzen AR, et al. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med. 1995;333: American Society of Clinical Pathologists Am J Clin Pathol 2001;115:

9 Oral Anticoagulation Monitoring Study Group / POINT-OF-CARE PROTHROMBIN TIME MEASUREMENT 2. Gottlieb LK, Salem-Chatz S. Anticoagulation in atrial fibrillation: does efficacy in clinical trials translate into effectiveness in practice? Arch Intern Med. 1994;154: Hirsh J, Fuster V. Guide to anticoagulant therapy, 2: oral anticoagulants. Circulation. 1994;89: Hirsh J, Dalen JE, Anderson DR, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest. 1998;114(5 suppl):445s-469s. 5. Wells PS, Holbrook AM, Crowther NR, et al. Interactions of warfarin with drugs and food. Ann Intern Med. 1994;121: Hasenkam JM, Knudsen L, Kimose HH, et al. Practicability of patient-self testing on oral anticoagulant therapy by the International Normalized Ratio (INR) using a portable whole blood monitor: a pilot investigation. Thromb Res. 1997;85: Horstkotte D. Improvement of prognosis by home PT estimation in patients with lifelong anticoagulant therapy [abstract]. Eur Heart J. 1996;17(suppl): Bernardo A, Halhuber C, Horstkotte D. Home prothrombin estimation. In: Butchart EG, Bodnar E, eds. Thrombosis, Embolism, and Bleeding. London, England: ICR Publishers; 1992: Kaatz SS, White RH, Hill J, et al. Accuracy of laboratory and portable monitor International Normalized Ratio determinations: comparison with a criterion standard. Arch Intern Med. 1995;155: McCurdy SA, White RH. Accuracy and precision of a portable anticoagulation monitor in a clinical setting. Arch Intern Med. 1992;152: Anderson DR, Harrison L, Hirsh L. Evaluation of a portable prothrombin time monitor for home use by patients requiring long-term oral anticoagulant therapy. Arch Intern Med. 1993;153: Ansell JE, Patel D, Ostrovsky D, et al. Long-term patient selfmanagement of oral anticoagulation. Arch Intern Med. 1995;155: White RH, Becker DM, Gunther-Maher M. Outpatient use of a portable International Normalized Ratio/prothrombin time monitor. South Med J. 1994;87: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1: Westgard JO. Points of care in using statistics in method comparison studies [editorial]. Clin Chem. 1998;44: Lassen JF, Brandslund I, Antonsen S. International Normalized Ratio for prothrombin times in patients taking oral anticoagulants: critical difference and probability of significant change in consecutive measurement. Clin Chem. 1995;41: Cimini C, Huang M, Marcelino E, et al. Novel integral quality control system in unitized prothrombin time testing yields equivalent results as standard hospital testing [abstract]. Presented at: the Oakridge Conference, American Association of Clinical Chemistry; April 1997; St Louis, MO. 18. Craig S, Stevenson KJ, Duffy JMK, et al. Local INR correction: justification for a simplified approach. J Clin Pathol. 1997;50: 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: Ng VL, Levin J, Corash L, et al. Failure of the International Normalized Ratio to generate consistent results within a local community. Am J Clin Pathol. 1993;99: Ng VL, Valdes-Camin R, Gottfried EL, et al. Highly sensitive thromboplastins do not improve INR precision. Am J Clin Pathol. 1998;109: Swaim WR. Prothrombin time reporting and the International Normalized Ratio system: improvements are needed. Am J Clin Pathol. 1993;99: Ansell JE. Imprecision of prothrombin time monitoring of oral anticoagulation: a survey of hospital laboratories. Am J Clin Pathol. 1992;98: Ansell JE, Hughes R. Evolving models of warfarin management: anticoagulation clinics, patient self-monitoring and patient self-management. Am Heart J. 1996;132: Ansell J. Anticoagulation management as a risk factor for adverse events: grounds for improvement. J Thromb Thrombolysis. 1998;5(suppl):S13-S Bussey H, Chiquette E, Amato M. Workshop: anticoagulation clinic care versus routine medical care: a review and interim report. J Thromb Thrombolysis. 1996;2: Palareti G, Leali N, Coccheri S, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Lancet. 1996;348: Am J Clin Pathol 2000;115: American Society of Clinical Pathologists