Anticoagulation Patient Self-monitoring in the United States

Considerations for Clinical Practice Adoption

Edith A. Nutescu, Pharm.D.; Sacheeta Bathija, B.Pharm.; Lisa K. Sharp, Ph.D.; Ben S. Gerber, M.D., M.P.H.; Glen T. Schumock, Pharm.D., M.B.A.; Marian L. Fitzgibbon, Ph.D.


Pharmacotherapy. 2011;31(12):1161-1174. 

In This Article

Efficacy, Safety, and Feasibility of Patient Self-monitoring

The emergence of portable instruments (point-of-care devices) for measuring anticoagulant effect by means of the INR in capillary whole blood make it possible for patients receiving warfarin therapy to monitor the effect of their anticoagulant therapy by themselves.[32–36] These instruments are based on clot-detection methodology that uses thromboplastin to initiate clot formation.[34] Data support reliability and analytic accuracy of portable monitors.[37] The instruments are small and light, weighing only 5.3–28.8 ounces. Several portable monitors are available and approved for patient home testing in the United States, such as the CoaguCheck XS System (Roche Diagnostics, Indianapolis, IN), ProTime Microcoagulation System (ITC Medical, Edison, NJ), and INRatio2 (HemoSense, San Jose, CA).[35] The cost of the devices ranges from $1500–2000, and the test strips cost about $7–12/test; however, most insurance providers, including Medicare, now cover the costs of the device and monitoring for major indications such as atrial fibrillation, VTE, and mechanical heart valves.[33] Self-monitoring empowers patients, offers the advantage of more frequent monitoring, and provides increased patient convenience by allowing testing at home and avoiding the need for frequent laboratory and clinic visits.[34,35]

Self-monitoring can entail patient self-testing (PST) and patient self-management (PSM). In the PST model, patients are responsible for testing their INR in the home setting with use of a point-of-care INR device, and they contact a health care professional with the results and for specific instructions on how to adjust their warfarin dose. In the PSM model, in addition to performing their INR test, patients are responsible for adjusting their warfarin dose and frequency of testing based on a predetermined INR-dosing algorithm. In both models, a health care professional still provides clinical oversight and is responsible for the patient.[32–36]

Several studies have evaluated and shown the benefit of both PST and PSM models of care compared with either routine medical care or anticoagulation clinic management of anticoagulation therapy.[38–70]

Meta-analyses and Systematic Reviews

A systematic review and meta-analysis of 14 randomized controlled trials published through 2005 was conducted and compared self-monitoring (PST or PSM) of anticoagulation with standard monitoring (anticoagulation clinic or routine medical care; Table 1).[71] A total of 3049 subjects receiving anticoagulation therapy were included, and the trials were conducted from 2 months up to 2 years. Of the 14 trials included, only two were conducted in the United States.[65,66] Primary outcome measures were the proportion of INRs within the therapeutic range, thromboembolic and major bleeding events, and mortality from all causes. Secondary outcomes included frequency of testing, dropout rates, and minor bleeding.

Eleven trials reported improvements (range 3–20.9%) in the mean INR results within target range in the PSM groups compared with standard monitoring.[71] Pooled data from the trials showed that self-monitoring (PST and PSM) decreased thromboembolic events by nearly half when compared with standard monitoring (2.3% vs 4.6%, odds ratio [OR] 0.45, 95% confidence interval [CI] 0.30–0.68). Self-monitoring (PST and PSM) also resulted in fewer deaths (2.4% vs 3.9%, OR 0.61, 95% CI 0.38–0.98) and major bleeding events (2.5% vs 3.7%, OR 0.65, 95% CI 0.42–0.99) compared with standard monitoring. In the patients who self-tested and self-adjusted their therapy (PSM), the effect was larger for reducing thromboembolic events and death than self-testing alone (PST), but there was a lesser reduction in bleeding. Frequency of testing was 1.69–4.98 times higher in the self-monitoring group compared with standard monitoring. Of the 7579 patients sampled, 5527 were either excluded or denied participation. Overall, 62% (range 31–81%) of patients could not or would not participate. A total of 22% (range 9–43%) of patients assigned to the intervention group were not able to complete self-monitoring.

An updated systematic review by the same group of authors included 18 randomized controlled trials published through 2007 and a total of 4723 patients (Table 1).[36] Outcomes evaluated were thromboembolic events, major and minor bleeding, mortality, INRs within therapeutic range, frequency of testing, and feasibility of PST and PSM. Of the 18 trials, three were conducted in the United States.[65,66,72] Pooled data showed a nearly 50% decrease in thromboembolic events (relative risk [RR] 0.50, 95% CI 0.36–0.69) and mortality (RR 0.64, 95% CI 0.46–0.89), and no significant reduction in major bleeding (RR 0.87, 95% CI 0.66–1.16) in the intervention group (PST and/or PSM) compared with standard monitoring (anticoagulation clinic or routine medical care). Patient self-testing significantly reduced major bleeding (RR 0.56, 95% CI 0.35–0.91), whereas PSM did not (RR 1.12, 95% CI 0.78–1.61). In contrast, PSM alone showed significant reduction in thromboembolic events (RR 0.47, 95% CI 0.31–0.70) and mortality (RR 0.55, 95% CI 0.36–0.84), whereas PST alone did not.

Twelve trials showed an improvement in anticoagulation control as measured by the percentage of the mean INRs in the therapeutic range. Testing frequency was higher by 1.69–4.98-fold in the PST and PSM groups compared with control; PST or PSM was found not to be feasible in about 50% of the patients screened.

Of the 11,738 patients sampled, 7974 were either excluded or declined participation. The average proportion of patients that could not (or would not) participate in the trials was 68% (range 31–88%).[36] The exclusion rates were even higher in older patients. Of those assigned to the intervention, 25% (range 0–57%) stopped PST or PSM before the end of the trial.

The most recent and largest meta-analysis evaluating the effect of PST or PSM included 22 randomized controlled studies published in English through 2010 (Table 1).[35] This analysis also included the largest U.S. trial published to date, The Home INR Study (THINRS).[67] A total of 8413 patients from studies that enrolled outpatient adults receiving oral anticoagulation for more than 3 months were included. The mean age was 65 years (range 42–75 yrs). Duration of follow-up was less than 12 months in 13 studies, and greater than 12 months in 9 studies. Five studies evaluated PST only and 14 evaluated PSM. Only two of the trials were conducted in the United States.[66,67]

Pooled data showed that PST or PSM resulted in lower mortality (OR 0.74, 95% CI 0.63–0.87), lower risk of thromboembolism (OR 0.58, 95% CI 0.45–0.75), and no increase in the risk of major bleeding (OR 0.89, 95% CI 0.75–1.05) when PST or PSM was compared with standard management (anticoagulation clinic or routine medical care).[35] The pooled weighted mean percentage of time in the therapeutic range in the PST or PSM group was 66.1% (range 56–76.5%) compared with 61.9% (range 32–77%) for patients in the standard management group (11 trials). The weighted mean difference of 1.50% (95% CI –0.63–3.63%) was not statistically significant (p=0.168). Likewise, the percentage of INRs in the therapeutic range was not significantly different between the PST and PSM and standard care groups (pooled weighted mean 71% [range 43–87%] in PST or PSM group and 59% [range 22–78%] in the standard care group [11 trials]). The 5.9% weighted mean difference was not statistically significant (95% CI –0.18–12.0%, p=0.057 [6 studies]). Eight of 11 trials found that patient satisfaction, quality of life, or both were improved in the PST or PSM group compared with standard management. More than 50% of the eligible subjects in one half of the trials were not able to complete the training and declined randomization. Of the patients who were randomly assigned to the intervention, 64–98% completed the study.

Studies Performed Outside the United States

As highlighted in the section above, most of the studies that evaluated the efficacy of PST and PSM have been conducted outside the United States.[38–64] Given the extensive body of evidence that comes mostly from several European countries, the adoption of PST and PSM into clinical practice in Europe has been far greater than that seen in the United States. Most of the non-U.S. studies evaluated the feasibility and efficacy of PSM, and this is the most common approach applied today in practice in Europe.[34] Only a few non-U.S. studies[41,54,73] evaluated PST without dose management. In the two studies that compared PST with PSM,[51,74] no major differences in outcomes were noted. However, from the meta-analysis data reviewed in the section above,[36,71] it is apparent that PSM may result in more favorable outcomes with regard to reducing mortality and thromboembolic events compared with the benefit attained in the PST studies. As the population in these non-U.S. studies has been fairly homogenous, generalizability of these results to the U.S. health care system needs to be further evaluated.

Studies Specific to the United States

Compared with Europe, in the United States, only a limited number of studies have been conducted and published that evaluated the efficacy and safety of anticoagulation self-monitoring (Table 2 and Table 3).[65–69] Of the published trials, only two were larger scale, randomized, controlled studies,[66,67] and the remaining trials were smaller nonrandomized observational or cohort studies.[65,68,69] In contrast to the European studies, the self-monitoring intervention evaluated in most of the U.S.-based studies was PST. Only one small retrospective U.S. study evaluated the effect of PSM.[68]

One of the first U.S.-based randomized studies evaluated the efficacy of PST at home compared with management by an anticoagulation clinic.[65] A total of 50 patients just beginning warfarin and not yet on a stable dose were enrolled from a university or a community hospital. Patients had to demonstrate an ability to use the portable monitor and had to have a telephone in the home, live within a 150-mile radius of the hospital, be receiving warfarin therapy for at least 8 weeks, and be willing and able to participate and give consent. Exclusion criteria were previous warfarin use, history of noncompliance, memory impairment, and chronic alcoholism. Patients in the PST group were managed by telephone follow-up. The primary end point was the percentage of time the prothrombin time ratio remained within the target range, and secondary end points included the frequency of thromboembolic and hemorrhagic complications. The INR was not used to define therapeutic range because at the time this study was conducted, only a few laboratories used thromboplastin reagent with a known International Sensitivity Index. Anticoagulation control was evaluated by calculating the percentage of time the prothrombin time ratio was in the therapeutic range for each patient.

Anticoagulation control was significantly better in the PST group compared with the anticoagulation clinic group, with the mean percentage of time that the prothrombin time ratio remained within the therapeutic range of 87% versus 68% (p<0.001). There were no major bleeding or thromboembolic complications in either group. Both groups exhibited good compliance (74% in anticoagulation clinic group vs 83% in the PST group). Poor compliance was associated with poor anticoagulation control in the PST group but not in the anticoagulation clinic group (p=0.004). Most patients were satisfied with PST and wanted to continue this model after the study ended. This was one of the first U.S. studies to show feasibility of PST in the patients' homes, and it found PST to be superior to an anticoagulation clinic in patients who were able to use the point-of-care monitor.

The efficacy of a multicomponent program of warfarin management was evaluated in an elderly population for 6 months.[66] A total of 325 patients aged 65 years or older who started warfarin during hospitalization and were treated for more than 10 days were enrolled from a university teaching hospital. Patients were stratified according to their baseline risk for major bleeding and randomized to the intervention or usual care. The intervention consisted of two main components. The first part entailed a guideline-based consultation, and the second part was training for self-monitoring coupled with patient education about warfarin and coaching. Patients were educated on the use of the portable monitors initially while in the hospital (30 min–1 hr) followed by a home visit within 3 days of discharge. Thereafter, patients called in their results, and dosing was managed by telephone. Education was delivered one-on-one by an educator, and a workbook specifically formatted for older adults was used. In addition, patient coaching was incorporated to increase patient participation in self-care and to enhance their information seeking and communication skills with their providers. The primary outcome was the first major bleeding event during the 6-month intervention period, and secondary outcomes were death, recurrent VTE, and anticoagulation control at 6 months.

Of the 163 patients randomly assigned to the intervention group, 96 (59%) participated in PST as allocated (either alone or by help from a caregiver) and 67 (41%) either declined participation or were monitored conventionally without self-monitoring. The mean age was 75 years (range 65–94 yrs), 67% of patients were Caucasian, and the mean number of school years completed was 12. The cumulative incidence of major bleeding rate was higher in the routine medical care group compared with the PST group (12% vs 5.6%, p=0.0498) Mortality and recurrent VTE rates were similar in both groups at 6 months. The proportion of time spent in goal INR range was significantly higher in the PST group than the routine medical care group (56% vs 32%, p<0.001).

The THINRS study is the largest prospective randomized trial to date comparing weekly self-testing at home in patients competent in the use of a point-of-care monitor, with monthly testing at a high-quality anticoagulation clinic.[67] Compared with other trials, distinctive to this study was the fact that in order to be eligible for randomization, patients had to be qualified as competent in performing self-testing on the basis of formal training and after a 2–4-week, in-home testing period. After randomization to the PST group, meter-use competency was reevaluated every 3 months. Of the 3745 patients recruited for participation, 2922 (78%) showed competency with testing and were enrolled in the trial. Enrolled patients were treated with warfarin for atrial fibrillation (83%) or mechanical heart valve (24%). Patients were randomly assigned to PST or an anticoagulation clinic and followed for an average of 3 years (range 2–4.75 yrs). Patients assigned to the PST group used an interactive voice-response system with Web-based local monitoring to transmit results and get dosing instructions. The primary end point was time to first event: stroke, major bleed, or death. Secondary end points included time within target INR range, patient satisfaction, and quality of life. Although the investigators and subjects were not blinded to treatment allocation, the outcomes were assessed by independent adjudicators.

There was no difference in the primary outcome—time to first major event (death, stroke, or major bleeding)—among patients assigned to PST versus an anticoagulation clinic (hazard ratio 0.88, 95% CI 0.75–1.04, p=0.14). There was a modest, but statistically significant, increase in the time in therapeutic INR range in the PST versus the anticoagulation clinic group (absolute difference 3.8%, 95% CI 2.7–5.05%, p<0.001). Patient satisfaction and quality of life were also significantly improved in the PST group at 2 years of follow-up (p=0.002 and p<0.001, respectively). Adherence to test frequency in the PST group was 87% compared with 52% in the anticoagulation clinic group. Costs were higher in the PST group but not significantly different from that in the anticoagulation clinic group (difference $1249, 95% CI −$1205–3703, p=0.32). As this study enrolled a veterans population, 98% of the patients were male, 92% were Caucasian, and the mean age was 67 years (range 33–99 yrs). A total of 90% of patients assigned to the PST group had an education level of 12th grade or higher. Although this large U.S. trial did not show superiority of PST over a high-quality anticoagulation clinic in preventing clinical outcomes, it did show improvements in anticoagulation control, patient satisfaction, and quality of life.

The only U.S. study that showed feasibility of PSM versus an anticoagulation clinic enrolled 20 patients who were matched based on age, sex, indication, and duration of anticoagulation with 20 control subjects in a retrospective cohort design.[68] Patients were selected based on compliance, stability, and ability to self-manage therapy. Subjects were instructed in the use of the monitor and the dosage adjustment guidelines over 2 weeks by an anticoagulation nurse specialist, with at least two one-on-one sessions with each subject. The prothrombin times in the PSM group were within the recommended therapeutic range in 88.6% (95% CI 87.2–89.9%) of the determinations compared with 68% (95% CI 65.7–70.3%, p<0.001) of the determinations made by the matched control patients. In response to the 2153 prothrombin time self-measurements, study patients made 67 (3.1%) dosage decisions that were considered incorrect based on physician guidelines. None of these changes led to adverse outcomes. There was no significant difference in thromboembolic events and major bleeding between the two groups. The results of this small cohort study showed that patients can successfully measure their prothrombin time, adjust their own warfarin dosages, and achieve a better degree of anticoagulation control compared with anticoagulation clinic management.

More recently, the effect of PST coupled with a Web-based interactive patient management program was also evaluated.[69,70] The Web-based system can act as a virtual health care provider extender, and links the patient to an expert system that provides treatment advice in real time. This approach allows for combining the advantages of PST with the supervision of a specialized anticoagulation clinic and minimizes the concerns of telephone communication or management.[75]

Sixty patients from a single anticoagulation clinic, who were receiving long-term anticoagulation and with access to the Internet and a printer, were enrolled in a prospective, singlegroup, before-after study.[69] The primary outcome was the difference in time in therapeutic range before and after introduction of a Web-supervised program. Patients were given instructions on how to obtain their INR by using a point-of-care monitor and how to log on to a secure Web site and enter required information to the patient interface component of the management program. Patients were instructed to test their INR and access the program in the morning, which allowed the health care provider sufficient time to review the patient data and make recommendations by means of the Internet and, if necessary, contact the patient before the end of the day. Patients were also instructed to print their daily dosing instructions and take their warfarin in the evening. The mean time in therapeutic range increased from 63% in the anticoagulation clinic group (control period) to 74.4% during Websupervised self-monitoring (study period). The mean difference score between control and study periods was 11.4% (95% CI 5.5–17.3%, p=0.004). There were no hemorrhagic or thromboembolic complications. Similar improvements in anticoagulation control were also reported by another group of authors in a recent study that evaluated PST and online automated management.[70]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.