Impact of an Inpatient Anticoagulation Monitoring Program

Vitamin K antagonists (VKAs) have been the mainstay of oral anticoagulant therapy for more than 60 years.¹ Their efficacy has been established by well-designed clinical trials for both primary and secondary prevention of venous thromboembolism; for prevention of systemic embolism in patients with prosthetic heart valves or atrial fibrillation; as an adjunct in the prophylaxis of systemic embolism after myocardial infarction; and for reducing the risk of recurrent myocardial infarction.¹ 

Background

Warfarin is the most common VKA in clinical use today.¹ The goal of anticoagulation therapy with warfarin is to administer the lowest effective dose of the drug to maintain the target international normalized ratio (INR).² Anticoagulants are effective in the prevention and treatment of thrombosis, but they are by nature high-risk medications with narrow therapeutic targets.² Interindividual differences in drug metabolism, comorbidities, and use of interacting medications can significantly influence the intensity of anticoagulant effect and the risk of bleeding complications.³ 

Traditionally, anticoagulation management was thought of as a physician responsibility; however, pharmacists have become more involved in the management of both outpatient and, more recently, inpatient anticoagulation therapy. A meta-analysis conducted by Hammond et al in 2003 reviewed 54 separate studies that compared traditional drug management by physicians to inpatient drug management that had a pharmacist involved.⁴ They found that 85% of the studies showed improvements in patient care when a pharmacist was added to the team compared to traditional management. These improvements included improved effectiveness, efficiency, and safety of the patient’s drug therapy.⁴ Another study done by Mamdani et al found that when pharmacists managed inpatient warfarin, there was a decrease in median costs by 21% compared to usual medical care (i.e., physician management).⁵ 

National hospital accreditation organizations and patient safety groups have highlighted the risks associated with the use of anticoagulants and have actively promoted the need for health care institutions to develop strategies to improve the safe use of these medications.³ One of the major risks of warfarin therapy is bleeding, which correlates with elevated INR values.² A study by Kim et al examined hospitalization costs associated with warfarin bleeding in older community-dwelling adults and found that the mean cost for 126 hospitalizations was $10,819.⁶ This same study suggested that when patients were given instructions for warfarin management from a health care professional, it would likely reduce the number of warfarin-related bleeding hospitalizations and associated costs.⁶ 

The 2010 National Patient Safety Goals (NPSG) were created by the Joint Commission Accreditation Program to promote specific improvements in patient safety.⁷ The goals highlight problematic areas in health care and describe evidence- and expert-based solutions to these problems. NPSG 03.05.01 addresses reducing patient harm associated with the use of anticoagulation therapy.⁷ It describes in detail how hospitals need to use approved protocols for the initiation and maintenance of anticoagulant therapy; use proper resources to manage potential food and drug interactions for patients receiving warfarin; provide education regarding anticoagulant therapy to prescribers, staff, patients, and families; evaluate anticoagulation safety practices; take action to improve practices; and measure the effectiveness of those actions in a time frame determined by the organization.⁷ 

The NPSG 03.05.01 also states that to achieve better patient outcomes, patient education should be a vital component of an anticoagulation therapy program.⁷ Effective anticoagulation patient education includes face-to-face interaction with a trained professional who works closely with patients to be sure that they understand the risks involved with anticoagulation therapy, the precautions they need to take, and the need for regular INR monitoring. The use of standardized practices for anticoagulation therapy that include patient involvement can reduce the risk of adverse drug events associated with warfarin.⁷ 

Study: Anticoagulation Monitoring Program

The traditional standard of practice at Palmetto Health Richland (Main Hospital and Heart Hospital, Columbia, South Carolina) is to have the patient’s physician manage inpatient warfarin therapy. Once an active order for warfarin is profiled in the patient’s medication list and an INR is ordered by the physician, a computer-generated alert (CEM) is sent to the clinical pharmacist with the notation of the drug and INR level. INRs greater than 3.5 with or without warfarin also prompt a CEM to be sent to the clinical pharmacist. Although these CEMs are sent to the clinical pharmacist, the physician is ultimately responsible for changing the dose of warfarin based on the INR. Patient education is only provided to those patients who have a physician order for warfarin education before discharge from the hospital. This traditional way of management lends itself to areas of improvement as a result of inconsistent physician monitoring in addition to inconsistent patient education. 

The Palmetto Health Richland Heart Hospital implemented a new system for monitoring warfarin in August 2009. Physicians still order warfarin for their patients; however, a pharmacist or a student pharmacist monitors the dose and INR daily on every patient. When clinically warranted, the pharmacist makes recommendations and subsequent interventions with the permission of the physician. Every new-start warfarin patient receives face-to-face counseling with a pharmacist or a student pharmacist, and patients with history of a warfarin receive a reinforcement teaching by way of a video and a warfarin education booklet. Heart Hospital pharmacists created the program based on their clinical experience and anticoagulation knowledge gained while in school and during residency (two pharmacists were residency trained). Each student pharmacist (fourth-year pharmacy student on cardiology rotation for the month) was trained by the Heart Hospital pharmacist on the basics of warfarin, how to educate patients, and how to search for drug–drug interactions. Students were also able to draw from the knowledge gained from school. Every day the students met with the pharmacist to discuss and plan interventions for each patient. 

Objectives

The primary objective of this study was to evaluate the efficacy of the pharmacist/student pharmacist–managed warfarin monitoring program as compared to a traditional warfarin monitoring program. 

Study Design

This was an Institutional Review Board (IRB), single-center, retrospective cohort study of 200 inpatients receiving warfarin from August 2009 to August 2010. The primary outcome of the study was to demonstrate the efficacy of the pharmacist/student pharmacist–managed warfarin monitoring program (Heart Hospital) compared to the traditional warfarin monitoring program (Main Hospital) by showing a predetermined 20% difference in the number of days a patient’s INR remained in goal range during the hospital length of stay. Secondary outcomes were as follows: average number of days patients bridged; number of patients with documented warfarin education; number of significant drug interactions; number of INRs drawn; and average time to goal INR. Each patient’s medical chart was screened for the following drug interactions: trimethoprim-sulfamethoxazole (TMP-SMZ), triazole antifungals, HMG-CoA reductase inhibitors (statins), amiodarone, and fluoroquinolones.⁸ 

Inclusion Criteria: Adult patients (³18 years of age) admitted to the Palmetto Health Richland Main Hospital telemetry floor or the Palmetto Health Richland Heart Hospital telemetry floor who were on warfarin therapy and had received at least three doses of warfarin inpatient. 

Exclusion Criteria: Patients were excluded if they had received less than three doses of warfarin inpatient, had an indication of pulmonary hypertension, or were admitted to the following services: intensive care unit (ICU), cardiovascular surgery, pediatrics, or orthopedic surgery. 

Methods

Warfarin dose and INR were collected on each patient for the hospital length of stay. A goal INR was determined by the indication for which the patient was taking warfarin. All data were collected and recorded on the data collection form. 

Missing INRs were estimated based on INRs on previous and subsequent days. If missing INRs were out of target range on either side then the Rosendaal et al formula was applied to estimate the INR.⁹ This equation took into account the patient’s goal INR and the number of days between the two known INRs. 

Time in goal INR range was calculated using two different methods. The first method calculated the percentage of time each individual patient stayed in his or her goal range during the hospital stay. Individual patient percentages were combined to calculate the physician population and the pharmacist/student pharmacist population percentage of time in goal INR range. The second method used to evaluate time in goal was to calculate a percentage by time in goal over the patient length of stay for each of the two patient populations. 

Drug interactions were screened by reviewing the electronic medical chart retrospectively for the following medications: TMP-SMZ, triazole antifungals, statins, amiodarone, and fluoroquinolones. 

Statistical Analysis

The primary outcome was designed to show a 20% difference in the number of days a patient’s INR remained in goal range during the hospital length of stay. The authors had a set power of 80% and a predetermined alpha of 0.05 for statistical significance. Assuming an effect size of 20% difference in the primary outcome, a sample size of 76 was estimated for each group. Given the unpredictable nature of study outcomes, our sample size was set at 200 patients, 100 in each group, to ensure optimal evaluation. Fisher’s exact test was used for the primary endpoint and t-test was used for the secondary endpoints. Descriptive statistics were utilized for secondary endpoints that did not require a t-test and for all demographics.

Results

Two hundred patients were identified as meeting the inclusion criteria at Palmetto Health Richland, 100 from the Main Hospital (traditional monitoring) and 100 from the Heart Hospital (pharmacist/student monitoring). Demographics for the groups are in TABLE 1. The breakdown of indications for anticoagulants are in FIGURE 1.

The majority of patients analyzed in the traditional monitoring group were African-American females who were taking warfarin for a thrombotic disorder as compared to the majority of patients in the pharmacist/student monitoring group who consisted mainly of Caucasian men who were taking warfarin for atrial fibrillation. There were a comparable number of new starts to warfarin and continuation of home medication patients in both groups (approximately 20% and 80%, respectively.) The average length of treatment was 8.62 days for the traditional monitoring group and 9.57 days for the pharmacist/student monitoring group. 

Patients were analyzed for the primary outcome in three different groups: the total population, new starts to warfarin therapy, and patients continued on their home medication. TABLE 2 reports the breakdown of the three groups.

Examination of the two populations shows that the traditional monitoring group maintained their patients within the patients’ goal range 29% of the time during the patients’ hospital stay, while the pharmacist/student monitoring group maintained their patients within the goal range 50% of the time during the patients’ hospital stay (P = .0001). 

The new-start population had a total of 20 and 24 patients in the traditional monitoring group and pharmacist/student monitoring group, respectively. The traditional monitoring group maintained their patients within the goal range 21% of the time during the patients’ hospital stay as compared to the pharmacist/student monitoring group, who maintained their patients within the goal range 38% of time during the patients’ hospital stay (P = .0274). 

The warfarin as a continuation of home medication group had a total of 80 and 76 patients in the traditional monitoring group and pharmacist/student monitoring group, respectively. The traditional monitoring group maintained their patients within the patients’ goal range 31% of the time during the patients’ hospital stay as compared to the pharmacist/student monitoring group, who maintained their patients within the patients’ goal range 54% of the time during the patients’ hospital stay (P = .0001). 

The secondary outcomes were analyzed as one group, as reported in TABLE 3. There was a statistically significant difference in the average time to goal INR (days) when looking at new starts for the traditional monitoring group, where new-start patients had an average of 5.2 days in the traditionally monitored population compared to the pharmacist/student monitoring patients, which had an average of 4.1 days (P = .0274). When patients continued home treatment, there was also a statistically significant difference in the average time to goal INR, where traditional patients reached goal in 4.3 days compared to 2.5 days in the pharmacist/student monitored patients (P = .0001).

Discussion

The results of this study showed that patients had increased time in their INR goal range during their hospital stay when a pharmacist or student pharmacist monitored their warfarin therapy as compared to the traditional monitoring by a physician. This study also showed that when a pharmacist or student pharmacist monitored hospitalized patients on warfarin, patients reached goal INR in a shorter amount of time. This finding was true for new starts as well as patients continuing their home medication. 

There were several limitations to this study. This was a single-center study with a small patient population. The estimation of missing INRs is also a limitation. Fortunately, there were a small number of patients who did not have INRs drawn daily during their hospital stay. If INRs were not drawn daily, the INRs drawn before and after the missing INR were used to estimate the missed INR. Screening for drug interactions is also a limitation given the vast number of drug interactions with warfarin. The medications previously stated were chosen based on the National Institutes of Health (NIH)–supported Web site that aids health care providers with initial and maintenance dosing of warfarin.⁸ The lack of a standard protocol in either the traditional monitoring group or the pharmacist/student monitoring group may also be considered a limitation. All dosage adjustments were left up to the individual health care provider; however, improvements in outcomes were demonstrated in the nontraditional monitoring group. This study also did not look at the number of bleeding events, which may also be a limitation. 

This study has provided our institution with two major areas for future application. The study showed an area for improvement when it comes to documentation of warfarin education. There were only 9 patients in the traditional monitoring group with documentation for warfarin teaching as compared to 100 patients in the pharmacist/student monitoring group. This lack of documentation despite knowing where and how to document has led us to provide education to health care providers at our institution in the future. 

Another future application directly due to the results of this study is to expand the pharmacist/student monitoring group to the entire hospital in hopes of improving patient care for all patients, as well as satisfying NPSG 03.05.01. 

Conclusion

This study supports the addition of pharmacists/student pharmacists to monitor hospitalized patients on warfarin. This addition helps to increase the amount of time  patients are in therapeutic INR goal range during their hospital stay as well as help patients reach their goal INR in a shorter amount of time. Pharmacists/student pharmacists also help to increase the amount of patients educated about warfarin and manage drug interactions among these patients.

Previous studies have shown that the addition of pharmacist/student monitoring of hospitalized patients on warfarin may also be cost-effective.⁷ An inpatient pharmacist/student monitoring program is likely to improve patient care for inpatient anticoagulated patients on warfarin. 

REFERENCES

1. Ansell J, Hirsh J, Hylek E, et al. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition). Chest. 2008;133(6 suppl):160S-198S.
2. Kuruvilla M, Gurk-Turner C. A review of warfarin dosing and monitoring. BUMC Proc. 2001;14:305-306.
3. Biscup-Horn PJ, Streiff MB, Ulbrich TR, et al. Impact of an inpatient anticoagulation management service on clinical outcomes. Ann Pharmacother. 2008;42:777-782.
4. Hammond RW, Schwartz AH, Campbell MJ, et al; American College of Clinical Pharmacy. Collaborative drug therapy management by pharmacists–2003. Pharmacotherapy. 2003;23:1210-1225.
5. Mamdani MM, Racine E, McCreadie S, et al. Clinical and economic effectiveness of an inpatient anticoagulation service. Pharmacotherapy. 1999;19:1064-1074.
6. Kim MM, Metlay J, Cohen A, et al. Hospitalization costs associated with warfarin-related bleeding events among older community-dwelling adults. Pharmacoepidemiol Drug Saf. 2010;19:731-736.
7. The Joint Commission Accreditation Program: Hospital National Patient Safety Goals. Effective January 1, 2011. NPSG.03.05.01. www.jointcommission.org/ assets/1/6/2011_NPSGs_HAP.pdf. Accessed August 15, 2011.
8. Warfarin dosing. www.warfarindosing.org. Accessed August 15, 2011.
9. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briët E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost. 1993;69:236-239. 

To comment on this article, contact rdavidson@uspharmacist.com.