Abstract
For over 60 years, warfarin has been the treatment of choice in the prevention of strokes and other thromboembolic events. In recent years, a new class of Novel Oral Anticoagulant (NOAC) medication has become available, leaving clinicians and health system payors to question whether warfarin continues to have a place in therapy. This article argues that it may not be the medication that should be in question but instead the systems in place to manage anticoagulation for the patients who need it. Usual Care (UC) for warfarin management has traditionally required multiple healthcare visits, blood collection visits, and laboratory analysis of International Normalized Ratio (INR) with results to then later be relayed to the patient along with dosage adjustments. The article reviews a new model of care, Community Pharmacist-led Anticoagulation Management Service (CPAMS), in which patients receive a point-of-care INR test along with a pharmacist assessment at a pharmacy and results within minutes. Pharmacists then prescribe dosage adjustments immediately, counsel patients, and provide supporting adherence tools such as a colourful picture-based dosing calendar, created by the decision support tool, INR Online. The Nova Scotia CPAMS Demonstration Project shows that this model will result in efficiencies for healthcare providers and optimal anticoagulation with improved time in therapeutic range outcomes for patients. In addition, the CPAMS Costing Study finds the model to be a cost-effective solution for health systems when compared to UC for warfarin as well as NOAC patients.
Introduction
Dr. Murray Trusler was tired of learning that the patients in the communities serviced by his hospital in Moose Factory were suffering from strokes and other thromboembolic events. Warfarin, an inexpensive anticoagulant medication that has been used for over 60 years 1 (commonly indicated for thromboembolic disorders, for example, atrial fibrillation, venous thromboembolism, and used in the prevention of stroke), was readily available; however, patients in his region were continuing to have strokes due to challenges with their warfarin management. A unique quality of warfarin is the requirement to regularly monitor the level of anticoagulation achieved with therapy, with the International Normalized Ratio (INR). The INR must be maintained within a range of 2.0-3.0 (or 2.5-3.5 for some indications) to prevent thrombosis or hemorrhage, which requires regular monitoring and dosage adjustments.
Dr. Trusler says, “There were two patients in particular that really stood out to me. These were atrial fibrillation patients who were otherwise healthy and were both able to continue working but had challenges accessing the lab to have their INR checked to monitor their warfarin therapy and receiving phone calls with their results when they did have a test.” Due to the challenges, these two patients were switched from warfarin to acetylsalicylic acid, and ultimately, both had strokes and died unnecessarily in Dr. Trusler’s view. The challenges Dr. Trusler encountered in Moose Factory while working there as the Head of Family Medicine between 2004 and 2008 put a magnifying glass on the weaknesses in the usual model of care for patients on warfarin.
In 2011, Trusler decided he needed to do more to improve anticoagulation care in Canada. He believed warfarin was an effective medication in preventing stroke when well managed. However, the then current model of care, which required a patient to visit a lab and wait for results, demonstrated average Time in Therapeutic Range (TTR) of 50%-60%. 2,3 Surely, Canada could do better. Dr. Trusler soon learned of a New Zealand-based decision support tool, INR Online. Around this time, New Zealand researchers Jeff Harrison, Jenny Harrison, and John Shaw were studying a new model of anticoagulation care, calling it “CPAMS” referring to a Community Pharmacist-led Anticoagulation Management Service. 4 In their study, 41 community pharmacies were selected to participate. The 693 patients went to the pharmacy for a point-of-care test using a finger stick capillary blood sample to measure their INR. The pharmacy used a decision support software called INR Online, followed by pharmacist assessment to communicate dosage changes. The mean TTR in the pharmacist-led care group was 78.6%. The results were compared to pre-enrollment data for these patients under physician-led care, which was 61.8%.
In Canada, in the usual care model, a Primary Care Provider (PCP) generally provides the patient with a warfarin prescription and a requisition for a blood draw at the lab. 5 At a later time, their PCP contacts the patient with the results. This process can be time consuming and inconvenient for both patients and clinicians, leading to lack of adherence to the testing intervals, less frequent testing, and a delay in many cases in delivery of the results. 5
In 2015, we, the Pharmacy Association of Nova Scotia (PANS), were researching new models of care in community pharmacies. We were looking for examples that focused on optimizing the skill set of pharmacists to improve patient outcomes, when we came across the CPAMS New Zealand study. We reached out to the New Zealand researchers to learn more, which led to a discussion with Dr. Trusler and INR Online, now licensed in Canada. The PANS was interested in the project as it provided an opportunity for pharmacists to utilize their expertise in medication management to help their patients but also to improve upon some of the existing challenges with usual care with respect to patient access, convenience, and compliance to both testing and their medication. In Nova Scotia, most residents have a community pharmacy near their home, while blood collection centres can be up to an hour away with limited hours of operation. Pharmacies are open 7 days a week, daytime and evenings, providing further convenience to patients and caregivers who work during the day. Anticoagulation management at a pharmacy seemed like a perfect fit.
The Nova Scotia CPAMS Demonstration Project
The PANS put together a proposal for submission to the Nova Scotia Department of Health and Wellness for a CPAMS Demonstration Project modelled after the NZ project. It was approved. The project was capped at 40 pharmacies and a call for applications went out. Seventy-five pharmacies applied, and the committee selected participants through a process that ensured that all counties within the province and all pharmacy banners or chains that applied had representation within the selected project sites. Each participating location selected a minimum of two pharmacists to complete the Management of Oral Anticoagulation Therapy Certificate Program at the University of Waterloo. Each pharmacy also received a CoaguChek XS Pro, the hospital grade point-of-care monitor used to check INR at the pharmacy. This monitor provides results comparable to those obtained with the use of standard laboratory methods in monitoring patients on anticoagulant therapy. 6 Pharmacies were also provided a subscription to INR Online Canada for decision support.
The project was structured such that PCPs (physicians and nurse practitioners) would refer their patients to the project and request that the pharmacist prescribes dosage adjustments for their patients under the Nova Scotia College of Pharmacists Prescribing Standard “Prescribing with a Diagnosis Provided.” 7 Pharmacists prescribed independently within the INR range of 1.5-4.0 and consulted with the PCPs when it was outside of this range. The 40 pharmacies selected identified that they had approximately 300 PCPs who were prescribing warfarin for their patients. As the project required a referral from these providers, it was important that they were aware of the project and the potential benefits to both the providers and their patients. During the first 2 months of the project, participating pharmacists set up meetings with the PCPs in the community to discuss the project and ask for referrals. The project working group was concerned that there would be a hesitancy to refer to the project and were pleasantly surprised when the referrals started pouring in. In the end, there were 946 patients enrolled by 237 physicians and nurse practitioners. 8 Patients were enrolled in the program between February 1, 2018, and July 31, 2018. Not all providers were initially on board, but some changed their minds as they started to hear about the positive experiences from other physicians and patients.
In some pharmacies, patients were provided an appointment time. Other pharmacies adopted an approach where they provided them an appointment day, and the patient could come when it was convenient for them. When they arrived, the pharmacist would take the patient to a private office to do an assessment and ask a series of screening questions to determine if there was anything, such as a change in diet or medications, health status, missed doses, or other factors, that may impact their INR result. The finger-prick blood test was then performed and the results were available within a few minutes. If the result was within the expected range, the patient was told to continue on the same dose. If it was out of range, the pharmacist would use their clinical judgment to determine if a change in dose was required, taking into consideration the recommendations of the decision support software along with the information collected from the patient during the assessment. The data for the study were collected for all patient visits up to July 31, 2019. At that time, patients enrolled in the study were permitted to continue in the program, pending a final decision from the government on system rollout.
CPAMS Nova Scotia final evaluation
Methodology
The final evaluation draws data from a variety of sources that were both quantitative and qualitative. A document review was used to gather descriptive information about the project. Two focus groups were completed with 11 pharmacists including staff, managers, and owners, as well as one focus group with the Project Working Group. Eight interviews with PCPs were completed to gather the perspectives of participating physicians (7) and nurse practitioners (1). Participating patients (444) completed a survey at the time of enrolment, and 488 patients completed a survey at the end of the study (June/July 2019), and a health card number was included for matching purposes. Seventy-two of the pharmacists also completed a survey in June/July 2019. 8
Quantitative data were collected from the INR Online database, as well as a secure project data portal, which pharmacists used to populate some information. In addition to patient demographics, INR Online captured indication, target INR, test dates, INR results, warfarin dose, and details recorded such as adverse events and hospitalizations. The CPAMS data portal was used to capture results that were not otherwise captured in INR Online such as patient eligibility criteria, reasons for attrition, time spent on activities, and point-of-care testing supply usage.
Results
Forty pharmacies participated in CPAMS, with 106 pharmacists providing patient care to CPAMS patients. Participating pharmacies were distributed around the province, with almost half of the participating pharmacies located in small towns (population of 1,000-10,000) and the remainder in more urban settings. A total of 946 patients were enrolled in CPAMS, with 746 still active at the end of the project. A total of 237 PCPs (225 physicians and 12 NPs) enrolled patients in the project. The majority of patients enrolled had an indication of atrial fibrillation. 8
A total of 21,597 INR tests were conducted, representing an average of 1.7 tests per patient per month during the 12 months of the project (following completion of patient enrolment). Pharmacists prescribed dosage adjustments independently for 93% of tests, with only 7% of tests falling outside of the INR range managed by pharmacists (INR: 1.5-4.0). When it was outside of this range, they collaborated with the patient’s nurse practitioner or family physician. 8
Time in therapeutic range
One of the key metrics was TTR, the proportion of time the patient’s INR was in the therapeutic range. Overall, across all CPAMS patients, the TTR was 71.4%. This result is well above the usual care TTR of 50-60%. 2,3 The TTR for CPAMS patients in the final 5 months of the study was even higher, at 75.7%. (Note: TTR for patients who continued in CPAMS after the study period ended was also later calculated for another 6 months and remained closer to this final 5 months). The TTR was measured for this 5-month time period to reflect a follow-up interval similar to the historical data (where patients have an average of 4 months of testing data) and to exclude test results that could have been affected by a manufacturing error with the test strips. Unfortunately, the manufacturer test strip error was not identified until 9 months into the study. At that time, an Urgent Medical Device Notice 9 was sent by the manufacturer noting that any test results above 4.5 were not valid; therefore, the separate TTR analysis was completed for the time period remaining once this problem was corrected.
The historical data refer to the last three test results for each patient which were collected from PCPs at the time of patient enrolment. There was no method to validate these data, and it represented on average a 4-month period of time, which may not be comparable to the TTR calculated for the duration of the project. As a result, usual care (as referenced from literature) is cited in Table 1 as a comparison.
TTR Results (Rosendaal Method), All Patientsa
Abbreviation: CPAMS, Community Pharmacist-led Anticoagulation Management Service.
a Table used with permission, CPAMS Final Evaluation Results, 2019. https://pans.ns.ca/cpams.
Patient safety
The TTR results indicate that patients spent more time within the therapeutic range, thus reducing their chance of adverse events such as stroke or bleeding events. During the Nova Scotia CPAMS project, patients spent 3.3% of the time more than 1.0 below target compared to 4.8% of the time based on historical data and compared to 3.9% of the time in CPAMS New Zealand. Results greater than 5.0 were similar and occurred in 0.6% and 0.5% of the test results in NS and historical data (0.8% in New Zealand). There were no comparators for adverse events within the study; therefore, conclusions cannot be drawn. It can be noted, however, that there were six deaths due to bleeding or thromboembolic events (0.6% of patients) and six major bleeds (0.6%). 8 When compared to results as discussed by White et al. in Sportif III and V, in the patients with TTR of 75% or greater, death was reported in 1.69% and major bleeds in 1.58%. 10
Summary of successes for patients
Ninety-eight percent of patients indicated they prefer to have their warfarin managed at the pharmacy. Patients preferred the finger-prick blood test compared to a venous blood draw. They also found it more convenient and accessible, with CPAMS patients reporting an average of 27 minutes travelling to and waiting at the pharmacy, compared to an average of 107 minutes to travel and wait at a blood collection lab. One PCP interviewed also stated “Patients really like the idea that they get the result right away and know what they’re doing.” Another PCP pointed out “For patients that have trouble with adherence…this is a plus to having the pharmacist do the testing.” Pharmacists were able to help support the patients with medication adherence for all of their medications but also adherence to testing. Eighty percent of INR tests were completed on or before the recommended date. Patients found it as easier to follow their required dosing schedule for CPAMS, which may have been due to the dosing calendar with colour-coded pictures, generated by INR Online. Overall, patients felt they had improved understanding of their condition and felt empowered to better manage their condition. 8
The goal of the project was to ensure positive health outcomes through improved TTR; 72.5% of patients had a TTR of 65% or greater during the project. Many patients also reported additional benefits such as the time spent with the pharmacist led to addressing other health conditions. They also reported reduced stress and anxiety related to warfarin testing. Pharmacists also felt there was a benefit to patients when new medications were added or doses of existing medications changed as they were able to provide the appropriate INR monitoring at the pharmacy, without having to rely on the patient visiting a lab. 8
Successes for pharmacies and pharmacists
The CPAMS project led to enhanced collaboration between pharmacists and other PCPs. Primary care providers and patients reported a better understanding of the value that pharmacists provide for patients. Pharmacists reported improved job satisfaction and found it rewarding to use their clinical skills in the CPAMS project to work more closely with patients. 8
Successes for PCPs
Primary care providers reported time savings by having their warfarin patients managed by the pharmacy. One of the concerns PCPs had with usual care was the time they or their staff spent tracking patients down to relay test results and dosing information. This was avoided when the pharmacist was able to provide this information in real time. 8
Successes for the health system
Successes for the health system include improved partnerships and collaboration between providers and increased efficiency in the system, including reduced need for blood collection and laboratory services for INR. Improved TTR also suggests increased safety in anticoagulation management and a reduction in costly adverse events. 8
The CPAMS costing study final results
Decision makers and government want to know if CPAMS is cost effective when compared to usual care. Dr. Lori Curtis conducted a CPAMS Costing Study, 11 which states that the only instance NS CPAMS is not at least as cost effective as usual care is when physicians continue to be paid a monthly anticoagulation management fee in addition to CPAMS payments. In Nova Scotia, physicians can bill an anticoagulation management fee up to once per month, when they interact with patients for activities such as relaying test results and recommending dosage adjustments. To simplify the study, that fee continued to be available to physicians during the demonstration project. However, when using the results from just the last 5 months of the study, CPAMS is more cost effective than usual care even if a referral fee or only occasional payments are paid to PCPs when collaboration is required. 11
Anticoagulation management and novel oral anticoagulants
In recent years, four Novel Oral Anticoagulant (NOAC) medications have become available as an alternative to warfarin in the management of some thromboembolic conditions. Novel oral anticoagulants, while more costly, have an advantage in that they do not require the regular INR monitoring required by warfarin. Clinical trials have found NOACs to be non-inferior to warfarin therapy, however, the NOAC randomized controlled trials (RE-LY, ENGAGE, Aristotle, Rocket AF) 12 –15 compared NOACs to warfarin management where the TTR ranged from 55% to 68%. The question remains, “How do NOACs compare to warfarin with a TTR of 75%? What about 85%?”
If a patient has the ability to maintain a higher than usual care TTR, would warfarin be a superior choice? The published comparison by White et al 10 compared INR poor control (defined as 60% or less) to INR good control (defined as 75% or greater) outcomes in the Sportif III and V trials and concluded that in patients with atrial fibrillation taking warfarin, risk of death, stroke, major bleeds, and other adverse events were related to INR control and that increased TTR is important for improving outcomes. Death was reduced by 60% when TTR moved from <60% to >75%.
The CPAMS New Zealand is now a nationally funded program. Patients in New Zealand have maintained a TTR of 75% for 10 years, 16 which is similar to the CPAMS NS results during the study period. CPAMS is an “anticoagulation management program,” with a goal of optimizing treatment for all patients, including deciding whether warfarin or one of the NOACs is the right choice for each patient. The randomized control trials for NOACs that compare their use to warfarin have an average TTR of 64%, so PANS wanted to know what an optimized CPAMS program would look like. An optimized program would include moving those that cannot attain a TTR of at least 65% to a NOAC alternative. The PANS asked researchers to analyse the data for those patients who had attained a TTR of 65% or above. If data for the patients with a TTR less than 65% are removed for the last 5 months of the study, the TTR for the remaining patients becomes 85.7%, 17 a significant improvement when compared to usual care in North America.
This finding suggests that many patients managed by CPAMS could benefit from improved outcomes with warfarin therapy when compared to NOACs and all patients should have a trial period of warfarin managed by the CPAMS model to determine if they can attain a “good” TTR. The final evaluation indicates 47 CPAMS patients were switched to NOACs during the study. 8 In some cases, the physician recommended the change and others were recommended by the pharmacist. An anticoagulation management program needs to take into consideration many factors to ensure that the patient’s treatment is optimized. In addition to TTR, clinicians need to consider indication, comorbidities, patient mobility, and access to a CPAMS pharmacy as well as adherence to their medication and testing schedule. Pharmacists are in a unique position to assess medication adherence with direct access to information about early or late refill dates and discussion with the patient and their caregivers. If non-adherence is identified, pharmacists can collaborate with PCPs to decide if support can be provided to improve adherence. Each situation is unique, and CPAMS provides the opportunity to individualize treatment options and provide careful follow-up.
The CPAMS costing addendum to include NOACs
The original costing study demonstrated that CPAMS was more cost effective than usual care. However, with NOAC pricing declining due to generic options becoming available, PANS wanted to know if the CPAMS, as well as an optimized version of CPAMS, where only patients with a TTR of 65% or greater remained on warfarin, would be cost effective when compared to the scenario where all patients are started on NOACs. An addendum to the original costing study 17 examines this question and determines that using current generic prices for NOACs (as well as expected declines in prices), the optimized CPAMS scenario is more cost effective than placing all patients on NOACs. A program whereby all patients have a trial of warfarin for 3 months and those who cannot maintain a TTR greater than 64 are switched to a NOAC (dabigatran) is the most cost-effective scenario, resulting in better outcomes with lower treatment costs.
What’s next?
Patients who were enrolled in the study are continuing to receive CPAMS care from their pharmacies in Nova Scotia. In addition, an immediate need was identified to support patients who do not have a family physician following their care, and an additional 100 patients in this situation are now also enrolled in the program. The final evaluation identified that pharmacists and PCPs felt comfortable with pharmacists managing warfarin dosing outside of 1.5-4.0 range, and this limitation is no longer in place. Pharmacists are using their professional judgment with regard to when a patient needs to be referred for vitamin K or emergency assessment.
Usual care with warfarin is an outdated system, making management difficult for patients, despite the improvements in technology available such as point-of-care testing. It is time to change the system. The CPAMS model of care has proven to be effective in supporting patients in maintaining a TTR that provides good INR control when compared to usual care in the literature. Patients are overwhelmingly supportive of the CPAMS model, finding it more convenient and accessible in addition to feeling empowered to better manage their condition. 8 Pharmacists and PCPs who participated in the study also had a positive experience with the program. 8 The PANS proposes that CPAMS be expanded to include all residents and be available at additional pharmacies that wish to participate. Our recommendation is that an optimized CPAMS model be implemented whereby pharmacists and physicians collaborate to identify patients who should remain on warfarin (TTR > 64%) versus those who will be better suited to a NOAC. We believe this will optimize patient care, allowing patients to achieve the highest TTRs when possible and those who cannot be considered for NOACs if indicated and appropriate. The optimized CPAMS model also has the potential for even further health system benefits with respect to lower overall event and treatment costs.