Performance Measurement of a Pharmacist-Directed Anticoagulation Management Service

Hospital Pharmacy Volume 36, Number 11, pp 1164 1169 2001 Facts and Comparisons PEER-REVIEWED ARTICLE Performance Measurement of a Pharmacist-Directed Anticoagulation Management Service Jon C. Schommer, PhD*, David A. Mott, PhD, Philip J. Schneider, MS, and Karen R. Knoell, PharmD** Abstract The purpose of this study was to use a performance-driven organizational change framework to measure both the clinical performance and time performance of a pharmacist-directed anticoagulation management service. A sample of 286 consecutive encounters between patients and anticoagulation management service staff was selected over a 2-month period at the anticoagulation management service of a midwestern university medical center. Data were analyzed using descriptive statistics, logistic regression, and linear regression analysis. Regression-based estimates documented that patients with at least one regimen adherence problem were approximately 10 times more likely to be outside their clinical INR range (clinical performance issue) and required an average of 4 minutes more per encounter (time performance issue) compared with patients who exhibited no adherence problems. Performance metrics such as clinical and time performance can be useful for organizational adaptation and organizational growth decisions related to an anticoagulation management service. Key Words anticoagulation management; performance measurement Hosp Pharm 2001;36:1164 1169 Adverse drug events, hospital readmissions, and prolonged hospital stays associated with the use of warfarin are well documented and raise concerns from both health system cost and patient care perspectives. 1 7 Pharmacist-directed anticoagulation management services have been shown to be effective in reducing costs and *Associate Professor of Pharmaceutical Care and Health Systems, College of Pharmacy, University of Minnesota, Minneapolis, MN; Assistant Professor of Social and Administrative Sciences, School of Pharmacy, University of Wisconsin, Madison, WI; Clinical Professor and Director, Latiolais Leadership Program, College of Pharmacy, The Ohio State University, Columbus, OH; **Clinical Pharmacist, McConnell Heart Health Center, Columbus, OH. At the time of this study, Dr. Knoell was Clinical Assistant Professor, College of Pharmacy, Ohio State University, Columbus, OH. Address correspondence to Dr. Schommer at the University of Minnesota, College of Pharmacy, 308 Harvard Street S.E., Minneapolis, MN 55455. E-mail: schom010@tc.umn.edu Support for this study was provided by an unrestricted grant from DuPont Pharma. An earlier version of this manuscript was presented at the 1999 American Association of Pharmaceutical Scientists Annual Meeting, November 17, 1999, New Orleans, LA.* improving patients therapeutic outcomes. 1 10 However, less is known about the organizational performance of anticoagulation management services in accomplishing these outcomes. That is, what resources are required to produce the outcomes? What factors influence the resources that are required to produce desired outcomes? The purpose of this study was to use a performance-driven organizational change framework 11 to measure the performance of a pharmacist-directed anticoagulation management service. This framework suggests that organizations attempt to reach their objectives through the most efficient means possible. 11 12 Thus, the assumption underlying our study was that the creation of an anticoagulation management service is an attempt by a health system organization to not only improve the effectiveness of care (eg, attain better anticoagulation management), but also the efficiency of the process through which that care is given. To determine how much a pharmacist-directed anticoagulation management service can contribute to overall organizational goals, we studied both the service s clinical performance and time performance. For this study, clinical performance was defined as the proportion of patients within a clinical International Normalized Ratio (INR) range (+0.5 target INR units, 0.2 target INR units). Time performance was defined as how time-intensive (in minutes) it was to monitor patients enrolled 1164 Volume 36, November 2001

Patient Demographics TABLE 1 Study Variables Age (in years) Gender (male or female) Insurance type (Medicare, Medicaid, private insurance, or self-pay) Number of medications (number of prescription and nonprescription medications taken daily in addition to warfarin) Diagnoses (number of ICD-9-CM diagnoses) Target INR (as recorded on patient chart) Clinical Outcomes Measured INR (as measured by laboratory staff) Regimen Adherence Recorded adherence problem (alcohol, diet, medication, lab follow-up, other) Time Spent in Anticoagulation Management Service Activities Professional (pharmacist) duties Chart review (time, in seconds) Patient assessment (time, in seconds) Calculations (time, in seconds) Physician consultation (time, in seconds) Patient counseling (time, in seconds) Documentation of pharmacist notes (time, in seconds) Technical (technician) duties Obtain lab results (time, in seconds) Information search (time, in seconds) Documentation of compliance and dose changes (time, in seconds) Faxing (time, in seconds) Report generation (time, in seconds) Other (time, in seconds) in the anticoagulation management service. For this study, we assumed that an anticoagulation management service should meet intermediary clinical objectives that are immediately measurable (patients within their clinical International Normalized Ratio) and do this in a time-efficient manner. The specific objectives of this study were to select a sample of patients enrolled in an anticoagulation management service and: (1) estimate the proportion of patients within a clinical International Normalized Ratio (INR) range defined as +0.5 target INR units, 0.2 target INR units, (2) determine relationships among selected patient demographics with INR measurements, (3) describe how time-intensive it was to monitor patients enrolled in the anticoagulation management service, and (4) identify which patients were most time-intensive for monitoring. Results from this investigation can provide insight about how an anticoagulation management service contributes to an organization s overall goals and could be used for decision-making related to staffing, pricing, contracts, and service expansion. 11 METHODS A sample of 286 consecutive encounters between patients and clinical service staff was selected over a 2- month period at a midwestern university medical center s anticoagulation management service. Patients were referred to the service from both inpatient and outpatient settings. Pharmacists performed initial patient counseling using manuals, videotapes, and compliance aids. Patients were then monitored through telephone consultations or clinic visits at intervals determined by the pharmacists. At each follow-up encounter, patients were evaluated for any changes related to their anticoagulation therapy. Communication with patients primary care physicians and with laboratory staff was maintained by fax, computer, or telephone. Data were collected and recorded by the anticoagulation management service staff as part of their customary documentation requirements. The data included patient demographics, clinical outcomes, regimen adherence, and time staff spent in anticoagulation management service activities for each patient. Specific items that were collected for each of these categories are presented in Table 1. It should be noted that operational definitions for the study variables were based on definitions and guidelines established by the clinic pharmacist as part of her practice. For example, the clinical INR range defined as +0.5 target INR units, 0.2 target INR units was based on the normal practice guidelines in the clinic. Also, regimen adherence was a subjective assessment recorded by the clinic pharmacist based on normal practice guidelines at the clinic. This included adherence to medication regimens, alcohol, diet, and lab follow-up as defined, assessed, and Hospital Pharmacy 1165

monitored by the clinic pharmacist. Data were coded and analyzed using the Statistical Package for Social Scientists (SPSS, Inc., Chicago). Descriptive statistics were used for objective 1, logistic regression analysis for objective 2, descriptive statistics for objective 3, and linear regression analysis for objective 4. A significance level of 0.05 was used for statistical tests. RESULTS A total of 89 patients accounted for the 286 consecutive encounters with an average of 3.2 encounters per patient (range 1 to 12) over the 2-month study period. Results for this study are reported for the 286 encounters as the unit of analysis. For these encounters, the patients ranged in age from 17 to 82 years of age, with an average age of 56. In 51% of the encounters, the patient was female. On average, patients for these encounters were taking seven prescription and nonprescription medications daily in addition to warfarin. The insurance coverage types for these encounters were private insurance (40%), Medicare (28%), Medicaid TABLE 2 Likelihood of Being Outside of the Clinical INR Range a Variable b Odds Ratio, exp (b) 95% CI for Odds Ratio c Patient Age 1.00 0.98 to 1.03 Patient Gender 0.60 0.31 to 1.19 Insurance Type Medicare (reference group) Medicaid 2.68 1.08 to 6.67 Private insurance 1.24 0.58 to 2.61 Self-Pay 1.16 0.27 to 4.88 Number of Medications 0.99 0.92 to 1.06 Number of Diagnoses 1.11 0.61 to 2.01 Adherence Problem 9.89 3.78 to 25.84 a Clinical INR = +0.5 Target INR units, to 0.2 Target INR units. b Patient age (in years), gender (male = 0; female = 1), insurance type (Medicare, Medicaid, private insurance, self-pay), number of medications (number taken daily in addition to warfarin), number of diagnoses (0 = one diagnosis; 1 = two or more diagnoses), adherence problem (0 = no problem; 1 = at least one adherence problem identified and noted). c Confidence intervals for the odds ratio = exp[b 1.96(se b )] to exp[b + 1.96(se b )] (26%) and self-pay (6%). The results showed that for 56% of the encounters (160 out of 286), patients were within a clinical INR range defined as +0.5 target INR units, 0.2 target INR units. Figure 1 shows the proportion of encounters for which patients were within (1) target INR, (2) ± 10% of target INR, (3) ± 20% target INR, and (4) clinical INR. Logistic regression results showed that patient age, gender, number of medications, and number of diagnoses did not significantly influence the likelihood of being outside of the clinical INR range (see Table 2). A slight effect for one insurance type was found. Specifically, patients with Medicaid insurance coverage were 2.7 times more likely to be outside the clinical INR range compared with Medicare-covered patients (reference group). In addition to that effect, patients with at least one regimen adherence problem were 9.9 times more likely to be outside the clinical INR range compared with those with no adherence problems identified (95% confidence interval for odds ratio was 3.8 to 25.8). Figure 2 shows that patients classified as having no regimen adherence problems were within their clinical INR for 65% of the encounters we studied. This is in contrast to patients classified as having at least one regimen adherence problem, who were within their clinical INR range for only 19% of the study encounters. Overall, each patient care encounter required an average of 10.7 minutes of staff time. Professional (pharmacist) duties comprised 4.2 minutes and technician duties comprised 6.5 minutes of this total. Multiple regression results showed that the most time-intensive patients were those with regimen adherence problems (standardized beta coefficient = 0.25, p < 0.001). Patient age, gender, insurance type, number of medications, and number of diagnoses did not significantly influence the time required for a patient encounter in the anticoagulation management service (see Table 3). Figure 3 shows that encounters for patients with at least one adherence problem required an additional 4.0 minutes of staff time, on average (14 minutes compared with 10 minutes). Figure 3 also shows that encounters for patients with at least one regimen adherence problem required increased time for both professional and technical duties in the anticoagulation management service. Professional duties devoted to patients without an adherence problem took 3.9 minutes compared with 5.3 minutes for patients with at least one adherence problem noted at that encounter. Technical duties for patients without an adherence problem took 6.1 minutes compared with 8.7 minutes for patients with at least one adherence problem noted for that encounter. DISCUSSION The results should be interpreted with study limitations in mind. First, costs related to the anticoagulation management service (eg, salaries, space, supplies, training, marketing, 1166 Volume 36, November 2001

Effects on Time Required per Patient Encounter Variable b Standardized Beta Significance Patient Age 0.06 0.40 Patient Gender 0.12 0.12 Insurance Type Medicare (reference group) Medicaid 0.04 0.69 Private insurance 0.11 0.16 Self-pay 0.002 0.97 Number of Medications 0.11 0.11 Number of Diagnoses 0.05 0.47 Adherence Problem 0.25 < 0.001 a Clinical INR = +0.5 target INR units, to 0.2 target INR units. b Patient age (in years), gender (male = 0; female = 1), insurance type (Medicare, Medicaid, private insurance, self-pay), number of medications (number taken daily in addition to warfarin), number of diagnoses (0 = one diagnosis; 1 = two or more diagnoses), adherence problem (0 = no problem; 1 = at least one adherence problem identified and noted). 70% 60% 50% 40% 30% 20% 10% 0% 41% 53% and management) were not investigated. Second, the results are only applicable to the anticoagulation management service used for the study. Other anticoagulation management services likely have characteristics different from the one we studied. However, the process we used for the investigation TABLE 3 67% 57% Target ± 10% ± 20% Clinical INR Target Target INR INR INR % within range FIGURE 1. Proportion of encounters for which patients were within target INR, ± 10% of target INR, ± 20% of target INR, and clinical INR (n = 286) Clinical INR = +0.5 target INR units to 0.2 target INR units can be replicated in other settings. Results from this investigation can provide insight about the performance of the anticoagulation management service and the organizational adaptation (process changes) or organizational growth (expansion of services) that might be warranted. 11 Our findings are discussed next in terms of performance, organizational adaptation, and organizational growth. Performance The results showed that both clinical performance and time performance are affected by regimen nonadherence. Patients with at least one regimen adherence problem were almost 10 times more likely to be outside their clinical INR range and were also the most time-consuming for the anticoagulation management service staff. These findings are not surprising, but we believe that this is the first documentation about the extent of these effects in an anticoagulation management service. We believe that directors of anticoagulation management services should be aware of the extent of the relationship between adherence and INR range, particularly if referrals to the service are based on patient nonadherence. We find it enlightening that the likelihood of being outside the clinical INR range was not directly influenced by patient age, gender, number of medications, or number of diagnoses. Regimen nonadherence can have an insidious effect on the clinical performance of an anticoagulation management service. In terms of time performance, the results suggest that there are clearly defined roles for pharmacists and technicians in the provision of anticoagulation management services. In this study, we documented that pharmacist duties comprised about 4.2 minutes and technician duties comprised about 6.5 minutes for the typical patient encounter. To maximize performance, clear processes should be in place so that pharmacists focus on professional duties and technicians focus on technical duties. It is interesting to note that time performance per encounter (measured in minutes) was not directly influenced by patient age, gender, insurance type, number of medications, or number of Hospital Pharmacy 1167

80% 70% 60% 59% 72% 65% 50% 40% 30% 20% 46% 15% 21% 43% 19% No adherence problem (n = 239) At least one adherence problem (n = 47) 10% 0% Target ± 10% ± 20% Clinical INR Target INR Target INR INR FIGURE 2. Proportion of encounters for which patients with or without regimen adherence problems were within target INR, ± 10% of target INR, ± 20% of target INR, and clinical INR (n = 286) Clinical INR = +0.5 target INR units to 0.2 target INR units 14 14 12 10 8 6 4 10 3.9 5.3 6.1 8.7 No adherence problem (n = 239) At least one adherence problem (n = 47) 2 0 Total Professional Technical Duties Duties FIGURE 3. Anticoagulation management service staff time (in minutes) devoted to patient encounters with or without regimen adherence problems (n = 286) Professional (pharmacist) duties = chart review; patient assessment; calculations; physician consultation; patient counseling; documentation of pharmacist notes Technical (technician) duties = obtaining lab results, information searches; documentation of compliance and dose changes; faxing; report generation, other diagnoses. The only study variable with a direct, significant effect on time performance was regimen adherence. Like its effect on clinical performance, regimen adherence problems exert an important influence on the time performance of an anticoagulation management service. Organizational Adaptation The results suggest that changes in the proportion of patients who have at least one regimen adherence problem 1168 Volume 36, November 2001

could require process change (organizational adaptation) for the anticoagulation management service. That is, patients with regimen nonadherence pose a challenge in that they require significantly more time per encounter. As the anticoagulation management service negotiates contracts for payment and sets pricing guidelines, the extra time per encounter for nonadherent patients should be considered. Also, directors of anticoagulation management services should continuously monitor the proportion of their patients who have identified regimen adherence problems so that staffing and other process adaptations can be made in the service. Organizational Growth In addition, the identification of patients with regimen adherence problems could present an opportunity for service expansion (organizational growth). Patients with at least one regimen adherence problem present an opportunity because they tend to be the individuals who are most likely to be outside of their clinical INR range. If efficient methods are in place for identifying these patients, they could be referred to the anticoagulation management service early in their treatment. Thus potential problems or increased costs that are associated with anticoagulation therapy could be avoided. SUMMARY The performance-driven organizational change framework 11 highlights the need for anticoagulation management services to closely monitor performance metrics (such as clinical and time performance) and to also consider how the service affects organizational adaptation and organizational growth. The basic notion of the performance-driven organizational change framework 11 is that it is important to make changes in services so that the service contributes to the overall goals of the organization. From the perspective of our medical center, it would be prudent to investigate any changes in the pricing and contract structure for the service so that they would reflect the increased resources needed to accommodate the extra time spent with nonadherent patients. Also, it would be prudent in this case to investigate the possible expansion of the service for patients identified as being, or potentially being, nonadherent with their treatment regimens. CONCLUSIONS We documented the extent to which patients with at least one regimen adherence problem present challenges for both clinical performance and time performance in an anticoagulation management service. The provision of anticoagulation management services can be optimized by developing clear roles for pharmacists and technicians, and paying attention to staffing patterns based on function and time requirements. Performance metrics such as clinical and time performance can be useful for organizational adaptation and organizational growth decisions related to an anticoagulation management service. REFERENCES 1. Ansell JE. Empowering patients to monitor and manage oral anticoagulation therapy. JAMA. 1999;281:182 3. 2. Ansell JE, Hughes R. Evolving models of warfarin management: Anticoagulation clinics, patient self-monitoring, and patient self-management. Am Heart J. 1996;132: 1095 100. 3. Chiquette E, Amato MG, Bussey HI. Comparison of an anticoagulation clinic with usual medical care: Anticoagulation control, patient outcomes, and health care costs. Arch Intern Med. 1998;158:1641 7. 4. Labor SL, Finley DS. Cost avoidance from implementing an anticoagulation clinic. Paper presented at ASHP Annual Meeting. Reno, NV; June 6, 1994. 5. Conte RR, Kehoe WA, Nielsen N, et al. Nine-year experience with a pharmacist-managed anticoagulation clinic. Am J Hosp Pharm. 1986;43:2460 4. 6. Calcagno D, Pubentz M, Carey R. Improving patient satisfaction with warfarin management. Am J Manag Care. 1996;2:804 10. 7. Landefeld CS, Anderson PA. Guideline-based consultation to prevent anticoagulant-related bleeding. Ann Intern Med. 1992;116:829 37. 8. Garabedian-Fuffalo SM, Gray DR, Sax MJ, et al. Retrospective evaluation of a pharmacist-managed warfarin anticoagulation clinic. Am J Hosp Pharm. 1985;42:304 8. 9. Gray DR, Garabediam-Fuffalo SM, Chretien SD. Cost-justification of a clinical pharmacist-managed anticoagulation clinic. Drug Intell Clin Pharm. 1985;19:575 80. 10.Lee P, Schommer JC. Effect of a pharmacist-managed anticoagulation clinic on warfarin-related hospital readmissions. Am J Hosp Pharm. 1996;53:1580 3. 11. Donaldson L. Performance-Driven Organizational Change. Thousand Oaks, CA: SAGE Publications; 1999. 12.Schneider PJ, Bennett MS, Casper KA. Re-engineering an ambulatory care pharmacy practice. Am J Hosp Pharm. 2000;57:2091 3. n Hospital Pharmacy 1169