511323

research-article2013

AOPXXX10.1177/1060028013511323Annals of PharmacotherapyShoulders et al

Research Report

Impact of Pharmacists’ Interventions and Simvastatin Dose Restrictions

Annals of Pharmacotherapy 2014, Vol. 48(1) 54­–61 © The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028013511323 aop.sagepub.com

Bethany R. Shoulders, BS1, Andrea S. Franks, PharmD, BCPS1,2, Patrick B. Barlow, BA2, Juli D. Williams, MD2, and Michelle Z. Farland, PharmD, BCPS, CDE1

Abstract Background: On June 8, 2011, the United States Food and Drug Administration (FDA) reported safety concerns regarding statin-related myopathies and advised further restrictions on simvastatin dosing. These restrictions reduced the maximum dose for specific patient characteristics, primarily certain concomitant medications. Objective: The purpose of this study was to compare the effectiveness of 2 different pharmacist-conducted educational interventions on appropriate simvastatin use in the primary care setting. Methods: This retrospective cohort analysis was conducted in 2 academic medical center clinics. Patients prescribed simvastatin before June 8, 2011, requiring dosage adjustment based on labeling changes were evaluated for study inclusion. The pharmacists’ interventions included: 30-minute didactic session for prescribers or patient-specific recommendation communicated with the physician during the patient’s follow-up visit. Primary outcomes were the number of patients prescribed FDA-recommended simvastatin doses after pharmacist intervention and the intervention’s impact on low-density lipoprotein (LDL). Results: Medical record review identified 1173 patients prescribed simvastatin prior to June 8, 2011; 126 patients qualified for study inclusion. After controlling for baseline characteristics, the likelihood of patients being prescribed an appropriate dose postintervention increased if they were in the patientspecific recommendation group (odds ratio [OR] = 10.59; 95% CI = 3.43-32.69; P < .0001). LDL change occurred at a similar rate between intervention groups (P = .652). Conclusion: Following FDA labeling changes for simvastatin, patientspecific recommendations made by pharmacists correlated with a greater likelihood of appropriate simvastatin dosing compared with a one-time didactic education session. Patient-specific recommendations positively affect prescribing habits and making steps to improve patient safety. Keywords simvastatin, medication safety, prescriber education

Background 3-Hydroxy-3-methylglutaryl–coenzyme-A reductase inhibitors (statins) have been a pivotal part of treatment for patients with dyslipidemia for more than 2 decades.1 The benefits of statins in lowering low-density lipoprotein (LDL) and preventing cardiovascular events have been well established.2,3 This knowledge has led to widespread use of statins in the general population for primary and secondary prevention of cardiovascular disease. To illustrate this use, in 5 years, from 2000 to 2005, the number of patients purchasing statins increased from 15.8 to 29.7 million.4 Although statin use continues to increase, the safety of this medication class has become a concern. Simvastatin, in particular, has been associated with a risk of a “dose-related” myopathy with monotherapy as well as in combination with

other medications.1 Most drug-drug interactions with simvastatin are related to its metabolism by the CYP3A4 enzyme.5-12 A growing number of case reports of patients developing myopathy accelerating to rhabdomyolysis while taking

1

University of Tennessee Health Science Center College of Pharmacy, Knoxville, TN, USA 2 University of Tennessee Graduate School of Medicine, Knoxville, TN, USA Corresponding Author: Andrea S. Franks, PharmD, BCPS, Departments of Clinical Pharmacy and Family Medicine, University of Tennessee Health Science Center College of Pharmacy and Graduate School of Medicine, 1924 Alcoa Highway, Box 117, Knoxville, TN 37920, USA. Email: [email protected]

Shoulders et al simvastatin have prompted the United States Food and Drug Administration (FDA) to take action.13 In March 2010, the FDA released communication related to the ongoing review of simvastatin safety.14 This research challenged whether or not higher doses of simvastatin could safely achieve further reductions in vascular disease.15 On June 8, 2011, the FDA reported these safety concerns and advised further restrictions on appropriate doses of simvastatin, eliminating the 80-mg dose for patients being treated for less than a year and altering maximum doses for medications known to increase patients’ exposure to simvastatin.16,17 This modification in statin labeling addressed the growing issue of statin safety and provided specific avenues of intervention to decrease the prevalence of simvastatin-related myopathies. Furthermore, pharmacists participating in medical teams have been able to affect patient outcomes, specifically for those with hypercholesterolemia, by providing information to physicians concerning optimal doses for patients. While pharmacists ensure guidelines are carried out and drug interactions are identified, LDL and total cholesterol levels have been shown to improve.18-20 The purpose of the study was to compare the likelihood of appropriate prescribing of simvastatin in the primary care setting following 2 different physician educational interventions conducted by pharmacists.

Methods This was a retrospective cohort analysis from June 7, 2011 through June 8, 2012. These dates represent a time period immediately prior to the FDA’s updated simvastatin dose restrictions and continuing for 1 year afterward. The study was approved by the University of Tennessee Health Science Center Graduate School of Medicine institutional review board prior to implementation. Patients were identified for the analysis if they were ≥18 years of age and were prescribed a medication containing simvastatin (simvastatin, ezetimibe/simvastatin, and niacin/simvastatin) prior to June 8, 2011, and required dosage adjustment based on simvastatin labeling changes. A list of all patients prescribed any simvastatin product during the study period was generated from the electronic medical record. Eligible patients were assumed to be adherent with use of the medication throughout the study time period. Patients were excluded if simvastatin was discontinued prior to June 8, 2011, or if they did not require a simvastatin dose change based on revised labeling. Patients were excluded from the assessment of the second primary outcome if postintervention LDL was not documented during the study period. Pharmacists conducted interventions for the prescribers at 2 different primary care practice sites within an academic medical center. Providers at each site received only 1 method of educational intervention. The intervention conducted by the pharmacist at the first site was a focused,

55 1-time 30-minute lecture to medical residents, faculty, and students. The objectives were to update the prescribing physicians on the dose restrictions for simvastatin (didactic education). The intervention conducted by the pharmacists at the second site included a patient-specific recommendation following medical record review that the pharmacist presented to the physician during each patient’s regularly scheduled primary care follow-up visit (patient-specific recommendation). Two pharmacists (1 full-time equivalent) at the practice routinely reviewed all patients’ medical records prior to each appointment. The review process occurred daily and focused on patients who had appointments the same day. During the review process, the pharmacists would identify drug-related problems and provide a written or verbal recommendation to the physician prior to the appointment on how to resolve the problem. The pharmacists identified patients taking simvastatin, during the daily medical record review, to determine if a dose change was needed. When making recommendations to change simvastatin, the pharmacists did not follow a specific algorithm but instead provided recommendations based on each patient’s clinical scenario. Factors considered included, but were not limited to, the patient’s most recent LDL; National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) guidelines (NCEP ATPIII)2 LDL goal; existing simvastatin dose; simvastatin drug-drug interactions; past use of other statins; and relative LDL-lowering potency of recommended statin. Pharmacists at this site do not have prescriptive privileges, nor did they work under a collaborative practice agreement that permitted them to make changes to statin doses independently. Therefore, all recommendations were provided to the physicians, who would then determine whether or not to implement the change. In the didactic education practice, there are approximately 15 family medicine faculty physicians and 24 residents who see patients in clinic. Approximately 125 patients are seen daily in the family medicine practice. At the patient-specific recommendation practice, there are 6 internal medicine faculty and 27 residents. At this site, only residents see patients, with about 20 patient visits per day. Pharmacists at both practices had established professional relationships with faculty and resident physicians and were actively engaged in the provision of patient care. There were 2 primary outcomes of the study. The first was to determine the proportion of patients who were prescribed simvastatin in accordance with the current FDA simvastatin dose restrictions following the pharmacist intervention. Data to assess this outcome were controlled for sex, baseline high-density lipoprotein (HDL) and LDL, and postintervention LDL. The second primary outcome assessed the impact of the pharmacist intervention on postintervention LDL. Secondary outcomes included a comparison of patients who were at LDL goal at baseline versus

56 those not at LDL goal at baseline to assess prescribing patterns for simvastatin and postintervention LDL. Data were collected to assess potential drug-drug interactions, patient demographics, and medical history. Simvastatin drug-drug interactions were classified into 3 groups according to FDA maximum dosing restrictions. Group 1 included amiodarone, amlodipine, and ranolazine (maximum recommended simvastatin dose, 20 mg). Group 2 included diltiazem and verapamil (maximum recommended simvastatin dose, 10 mg). Group 3 consisted of cyclosporine, danazol, gemfibrozil, clarithromycin, erythromycin, protease inhibitors, itraconazole, ketoconazole, nefazodone, posaconazole, and telithromycin (concomitant simvastatin contraindicated). Additionally, we recorded each patient’s simvastatin dose prior to June 8, 2011, and the statin medication and dose prescribed after the pharmacist’s intervention. Finally, each patient’s baseline LDL goal was determined by NCEP ATP III guidelines.2 Simvastatin dose appropriateness was evaluated prior to and following June 8, 2011, utilizing FDA recommendations for dose restrictions as well as the LDL-lowering potential of the statin dose to reach the patient’s individualized goal. Baseline comparisons of patients’ past medical history were carried out and demographic data were analyzed using independent t tests, Fisher’s exact test, or χ2 tests of independence as appropriate. A 2 × 2 mixed analysis of variance (ANOVA) was used to analyze the interaction between LDL change from baseline to follow-up and intervention group membership. A 2 × 2 factorial ANOVA was used to test the interaction between those patients who met their LDL goal at baseline and those who did not and intervention group membership on mean follow-up LDL. Significant interaction effects were stratified by intervention group using follow-up dependent t tests to look for specific LDL change differences within each group. Furthermore, hierarchical logistic regression was used to assess the likelihood of appropriate statin dose postintervention after controlling for possible confounding variables (sex, baseline HDL, and postintervention LDL). Statistical significance was determined at P < .05 for all analyses. Data were analyzed using SPSS v.20 (IBM Inc, Chicago, IL).

Results Study Population The electronic medical records for both practices identified 1173 patients who were prescribed simvastatin prior to June 8, 2011. The didactic education group contained 841 patients, and the patient-specific recommendation group contained 332 patients. Utilizing the predetermined exclusion criteria, 126 patients qualified for inclusion in the study (didactic education group, n = 71; patient-specific recommendation group, n = 55).

Annals of Pharmacotherapy 48(1) Baseline characteristics for both intervention groups are listed in Table 1. Of note, baseline LDL was similar between groups, but there was a significant difference between groups with regard to gender and baseline HDL. The patient-specific recommendation group had a higher HDL at baseline (45.98 vs 40.76, P = .028), and the didactic education group contained a higher percentage of men (52.1% vs 25.5%, P = .002). Use of concomitant medications that interact with simvastatin are listed in Table 2. The most common group of interacting medications with simvastatin were in group 1 (amiodarone, amlodipine, and ranolazine), with amlodipine having the highest frequency of use. The statin medication and dose that were prescribed postintervention are listed in Table 3.

Study Outcomes Postintervention, 72 patients (57.14%) were prescribed an inappropriate dose of simvastatin. Of these, 53 (74.65%) were in the didactic education group and 19 (34.55%) were in the patient-specific recommendation group. Results found a higher likelihood of appropriate statin dose prescribing postintervention in the patient-specific recommendation group (odds ratio [OR] = 7.18; 95% CI = 3.26-15.82; P < .0001). After controlling for sex, baseline HDL, and postintervention LDL, the patient-specific recommendation group was more likely to be prescribed an appropriate statin dose postintervention (OR = 10.59; 95% CI = 3.43-32.69; P < .0001; Table 4). The second primary outcome was the impact of the pharmacist intervention on postintervention LDL. Of the 101 patients who had both a baseline and postintervention LDL (didactic education n = 63, patient-specific recommendation n = 38), both recommendation groups experienced a decrease in LDL; however, only the decrease in the didactic education group proved to be statistically significant (−8.27 mg/dL, P = .044). Testing for an interaction effect between intervention groups and LDL change suggested that the change in LDL occurred at a similar rate between both groups (P = .652; Figure 1). In all, 58 patients had achieved their NCEP ATP III LDL goal as of June 8, 2011 (didactic education n = 34; patientspecific recommendation n = 24). Following the intervention, 23 patients in this subset continued to be prescribed an inappropriate simvastatin dose (didactic education n = 6; patientspecific recommendation n = 17). After controlling for baseline LDL being at goal, those in the patient-specific recommendation group continued to have an increased likelihood of the simvastatin dose being adjusted (OR = 7.20; 95% CI = 3.2415.94; P < .0001). Patients’ achievement of LDL goal at baseline did not significantly change the likelihood of the simvastatin dose being adjusted (OR = 0.650; 95% CI = 0.2931.44; P = .29; Table 5). A 2 × 2 factorial ANOVA showed a significant interaction between intervention group

57

Shoulders et al Table 1.  Baseline Characteristics of the Final Study Sample. Frequency (%) or Mean (SD), n = 126 Variable Age Race  Caucasian   African American  Other Male sex Baseline LDL Framingham risk score  Smoker  Hypertension  HDL CHD risk equivalents  Diabetes   Abdominal aortic aneurysm   Peripheral artery disease   Carotid artery disease CHD   Myocardial infarction   Unstable angina   Stable angina   Angioplasty or bypass surgery   Myocardial ischemia Hepatic disease   Transaminases >3× ULN   Hepatitis B  Myalgia

Didactic Education (n = 71)

Patient-Specific Recommendation (n = 55)

64.77 (11.63)

63.45 (11.89)

1.32 (−2.86 to 5.50)

62 (87%) 8 (11.3%) 1 (1.4%) 37 (52.1%) 102.31 (36.00) 0.12 (0.08) 16 (22.5%) 67 (94.4%) 40.76 (12.12) 46 (64.8%) 40 (56.3%) 1 (1.4%) 1 (1.4%) 1 (1.4%) 24 (33.8%) 9 (12.7%) 0 4 (5.6%) 16 (22.5%) 2 (2.8%) 6 (8.5%) 5 (7.0%) 1 (1.4%) 1 (1.4%)

43 (78%) 12 (21.8%) — 14 (25.5%) 113.93 (51.19) 0.15 (0.08) 21 (38.2%) 50 (90.9%) 45.98 (14.12) 37 (67.3%) 32 (58.2%) 0 3 (5.5%) 2 (3.6%) 22 (40.0%) 10 (18.2%) 1 (1.8%) 5 (9.1%) 8 (14.5%) 3 (5.5%) 3 (5.5%) 1 (1.8%) 2 (3.6%) 2 (3.6%)

Referent 2.16 (0.82 to 5.74) — 0.31 (0.15 to 0.67) −11.62 (−27.73 to 4.50) −0.04 (−0.11 to 0.04) 2.1 (0.98 to 4.62) 0.60 (0.15 to 2.34) −5.22 (−9.90 to −0.59) 1.12 (0.53 to 2.35) 1.08 (0.60 to 2.20) NC 4.04 (0.41 to 39.94) 2.64 (0.23 to 29.91) 1.31 (0.63 to 2.71) 1.53 (0.58 to 4.08) NC 1.68 (0.43 to 6.59) 0.59 (0.23 to 1.50) 1.99 (0.32 to 12.35) 0.63 (0.15 to 2.62) 0.24 (0.03 to 2.16) 2.64 (2.3 to 29.91) 2.64 (0.23 to 29.91)

OR (95%) Mean Difference (95% CI)

P Valuea .533   — .164 — .002 .156 .302 .056 .502 .028 .771 .836 NC .317 .58 .474 .456 NC .502 .361 .652 .73 .231 .58 .58

Abbreviations: OR, odds ratio; LDL, low-density lipoprotein; HDL, high-density lipoprotein; CHD, coronary heart disease; NC, not calculated; ULN, upper limit of normal. a Groups compared using independent t test, Fisher’s exact, and chi-square test of independence procedures, as appropriate.

Table 2.  Medication Histories for Education and Individual Intervention Groups. Frequency (%) Medication a

At least 1 group 1  Amlodipine  Ranolazine  Amiodarone Not on group 1 medications At least 1 group 2b  Diltiazem  Verapamil Not on group 2 medications At least 1 group 3c  Gemfibrozil Not on group 3 medications a

Didactic Education (n = 71)

Patient-Specific Recommendation (n = 55)

Total (n = 126)

32 (45.1%) 29 (40.8%) 2 (2.8%) 2 (2.8%) 39 (54.9%) 14 (19.7%) 8 (11.3%) 6 (8.5%) 57 (80.3%) 8 (11.3%) 8 (11.3%) 63 (88.7%)

34 (61.8%) 32 (58.2%) 2 (3.6%) 0 21 (38.2%) 13 (23.6%) 8 (14.5%) 5 (9.1%) 42 (76.4%) 2 (3.6%) 2 (3.6%) 53 (96.4%)

66 (52.4%) 61 (48.4%) 4 (3.2%) 2 (1.6%) 60 (47.6%) 27 (21.4%) 16 (12.7) 11 (8.7%) 99 (78.6%) 10 (7.9%) 10 (7.9%) 116 (92.1%)

Group 1 medications (amiodarone, amlodipine, and ranolazine). Group 2 medications (diltiazem and verapamil). c Group 3 medications (cyclosporine, danazol, gemfibrozil, clarithromycin, erythromycin, HIV protease inhibitors, itraconazole, ketoconazole, nefazodone, posaconazole, and telithromycin). b

58

Annals of Pharmacotherapy 48(1)

Table 3.  Statins Prescribed Preintervention and Postintervention. Study Period Preintervention           Postintervention                            

Didactic Education Patient-Specific (n = 71) Recommendation (n = 55)

Statin

Dose

Simvastatin

  5 mg 10 mg 20 mg 40 mg 80 mg

2 (2.8%) 1 (1.4%) 15 (21.1%) 32 (45.1%) 21 (29.6%)

0 1 (1.8%) 6 (10.9%) 32 (58.2%) 16 (29.1%)

Atorvastatin

20 mg 40 mg   5 mg 10 mg 20 mg 40 mg 80 mg 20 mg 40 mg 80 mg 10 mg 20 mg 40 mg   2 mg

1 (1.4%) 0 1 (1.4%) 3 (4.2%) 11 (15.5%) 28 (39.4%) 15 (21.1%) 3 (4.2%) 5 (7.0%) 0 1 (1.4%) 2 (2.8%) 0 1 (1.4%)

10 (18.2%) 9 (16.4%) 0 0 8 (14.5%) 13 (23.6%) 1 (1.8%) 2 (3.6%) 2 (3.6%) 8 (14.5%) 0 1 (1.8%) 1 (1.8%) 0

Simvastatin

Pravastatin

Rosuvastatin

Pitavastatin

Total (n = 126)   2 (1.6%) 2 (1.6%) 21 (16.7%) 64 (50.8%) 37 (29.4%)   11 (8.7%) 9 (7.1%) 1 (0.8%) 3 (2.4%) 19 (15.1%) 41 (32.5%) 16 (12.7%) 5 (4.0%) 7 (5.6%) 8 (6.3%) 1 (0.8%) 3 (2.4%) 1 (0.8%) 1 (0.8%)

Table 4.  Logistic Regression Predicting Appropriate Simvastatin Prescribing.a Predictor Variable

Adjusted OR (95% CI)

P Value

Patient-specific recommendation Female sex HDL Baseline LDL Postintervention LDL

10.59 (3.43 to 32.69) 4.55 (1.52 to 13.60) 1.01 (0.96 to 1.05) 0.98 (0.965 to 0.996) 1.02 (1.00 to 1.04)