- Email: [email protected]
Assessment of Dronedarone Utilization Using US Claims Databases
Clinical Therapeutics/Volume 36, Number 2, 2014
Assessment of Dronedarone Utilization Using US Claims Databases Jasmanda Wu, PhD, MPH1; Jane Thammakhoune, BS1; Wanju Dai, MD, DrPH1; Andrew Koren, MD2; Stephanie Tcherny-Lessenot, MD, MSc, MPH3; Chuntao Wu, MD, PhD1; Patrick Caubel, MD1; and Juhaeri Juhaeri, PhD1 1
Global Phamacovigilance and Epidemiology, Sanofi, Bridgewater, New Jersey; 2Cardiovascular Medical Unit, Sanofi, Bridgewater, New Jersey; and 3Global Phamacovigilance and Epidemiology, Sanofi, Chilly-Mazarin, France
ABSTRACT Background: A dronedarone utilization study using US MarketScan and InVision Data Mart databases was conducted to estimate the prevalence of the following: (1) dronedarone use in contraindicated patients with worsening heart failure (HF) or hospitalization for HF within 1 month before dronedarone prescription; (2) concomitant prescribing of contraindicated drugs; and (3) recommended creatinine testing after dronedarone initiation among dronedarone users. Methods: In this retrospective cohort study, data in the MarketScan database between July 20, 2009, and December 31, 2011, and in the InVision Data Mart database between July 20, 2009, and March 31, 2012, were analyzed. The study population included patients who received Z1 dronedarone prescription during the study period. The following variables were reported: worsening of or hospitalization for HF, concomitant prescribing of potent cytochrome P450 CYP 3A4 inhibitors or QT-prolonging drugs, and creatinine testing. Results: There were 31,408 and 7025 dronedarone users identified in the MarketScan and InVision Data Mart databases, respectively. Approximately 86% to 90% of patients had a diagnosis of atrial fibrillation in each database. In the MarketScan database, 40% were women and 54% were aged Z65 years. In the InVision Data Mart database, 31% were women and 32% were aged Z65 years. The corresponding prevalence of worsening or hospitalization for HF was 6.4% (95% CI, 6.2–6.7) and 4.7% (95% CI, 4.2–5.2) in each database, respectively. The corresponding estimates of concomitant prescribing of potent cytochrome P450 CYP 3A4 inhibitors and QT-prolonging drugs within 30 days before initiation or refilling of dronedarone were 2.0% (95% CI, 1.8–2.1) and 10.0% (95% CI, 9.7–10.4), respectively, in the
264
MarketScan database, and 2.3% (95% CI, 2.0–2.7) and 11.2% (95% CI, 10.5–12.0) in the InVision Data Mart database. More than 50% of patients in each database had serum creatinine tests conducted after dronedarone initiation. Conclusions: The results of the present analysis based on a long-term follow-up (nearly 3 years) were consistent with the previous findings that dronedarone has mostly been used appropriately in compliance with US prescribing in the target populations. (Clin Ther. 2014;36:264–272) & 2014 Elsevier HS Journals, Inc. All rights reserved. Key words: dronedarone, drug utilization, Risk Evaluation and Mitigation Strategy.
INTRODUCTION Dronedarone, a multichannel blocker displaying antiarrhythmic properties, has been shown to be effective for reducing the risk of cardiovascular hospitalization or death in patients with atrial fibrillation (AF) and atrial flutter (AFL).1,2 In the United States, dronedarone was approved as “an antiarrhythmic drug indicated to reduce the risk of cardiovascular hospitalization in patients with paroxysmal or persistent AF or AFL, with a recent episode of AF/AFL and associated cardiovascular risk factors (ie, age 470, hypertension, diabetes, prior cerebrovascular accident, left atrial diameter Z50 mm or left ventricular ejection fraction [LVEF] o40%), who are in sinus rhythm or who will be cardioverted.”1 Based on the Accepted for publication January 4, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.01.002 0149-2918/$ - see front matter & 2014 Elsevier HS Journals, Inc. All rights reserved.
Volume 36 Number 2
J. Wu et al. nonclinical and clinical safety information collected throughout an extensive development program, dronedarone is contraindicated in the United States in patients taking potent cytochrome P450 (CYP) 3A4 inhibitors that significantly raise serum levels of dronedarone. It is also contraindicated in patients taking drugs that prolong the QT interval and may induce torsade de pointes. In addition, dronedarone is contraindicated in patients with symptomatic heart failure (HF) with recent decompensation requiring hospitalization or New York Heart Association (NYHA) class IV HF. It is also contraindicated in patients in permanent AF who will not or cannot be cardioverted into normal sinus rhythm. In patients with permanent AF, dronedarone doubles the risk of death, stroke, and hospitalization for HF.3 In clinical trials, an increase in serum creatinine level was observed in patients treated with dronedarone 400 mg BID.1 This increase occurs early after treatment initiation and reaches a plateau after 7 days. The US prescribing information for dronedarone states that renal function should be monitored periodically in patients treated with dronedarone because increases in creatinine and blood urea nitrogen have been reported in the postmarketing setting and seem to be reversible after discontinuation of dronedarone.1 In the United States, in addition to labeling and other pharmacovigilance action plan, a Risk Evaluation and Mitigation Strategy (REMS) program has been developed to prevent dronedarone use in: (1) patients with symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF; and (2) patients with permanent AF who will not or cannot be cardioverted into normal sinus rhythm. The REMS educates prescribers through a comprehensive communication plan about increased mortality when dronedarone is used in these patient populations and informs patients about the serious risks of dronedarone, including increased mortality in patients with severe unstable HF or permanent AF, through a medication guide. To characterize dronedarone utilization patterns in the US population and to assess labeling compliance and the effectiveness of the REMS, a drug utilization study using the InVision Data Mart database has been conducted since 2009.4 The first publication based on data between July 2009 and August 2010 demonstrated labeling compliance and effectiveness of the
February 2014
REMS regarding worsening HF. However, because InVision Data Mart is an employment-based private insurance claims database, it has a lower proportion of patients aged Z65 years (9% in the InVision Data Mart vs 13% in the US population5), whereas AF and AFL are more prevalent in elderly patients than in the general population.2,6 Thus, the MarketScan database, which is composed of MarketScan Commercial Claims and Encounters database, the Medicare Supplemental database, and the Medicaid database, was added in the study to improve the representation of elderly patients. The MarketScan Medicare Supplemental database captures Medicare-eligible retirees with employer-sponsored Medicare supplemental plans. The age and gender distribution in the MarketScan Medicare Supplemental database is compatible to the overall Medicare population (MarketScan vs Medicare: age 65–74 years, 55% vs 45%; age 75–84 years, 29% vs 26%; age Z85 years, 12% vs 12%; male sex, 45% vs 45%).7 The MarketScan database has been used in a number of studies of patients with AF.8–11 The current article presents results from a longterm follow-up (nearly 3 years) for evaluation of dronedarone utilization patterns in the US population by using both the MarketScan and the InVision Data Mart databases.
MATERIALS AND METHODS Data Source The study used the US MarketScan Commercial Claims and Encounters database, the Medicare Supplemental database, and the Medicaid database from Thomson Reuters Corporation (New York, New York). The databases are composed of de-identified patient-level records from 4170 million patients since 1995. The MarketScan Commercial Claims and Encounters database includes private sector health data from 150 employers and 21 health plans. The Medicare Supplemental database captures Medicareeligible retirees with employer-sponsored Medicare supplemental plans. On average, nearly 3.8 million lives (active patients) are covered per year. The Medicaid database contains the pooled health care experience of 7 million Medicaid enrollees from 12 states. On average, nearly 6.8 million lives (active patients) are covered per year. The MarketScan databases are in compliance with the Health Information Portability and Accountability Act of 1996.12
265
Clinical Therapeutics The InVision Data Mart (United HealthCare, formerly LabRx, Eden Prairie, Minnesota) database was also used. The InVision Data Mart database is an integrated medical and prescription claims database augmented with partial laboratory data for an employed, commercially insured population of patients and their dependents. This research database includes 430 million members. The database is compliant with the Health Information Portability and Accountability Act of 1996.12 Because this study did not involve the collection, use, or transmittal of individually identifiable data, institutional review board review or approval was not required.
Study Population The analysis was performed among all patients who were prescribed dronedarone at least once as identified in the MarketScan database between July 20, 2009 (the launch date in the United States), and December 31, 2011 (the date of the latest data update received from the database vendor for the current analyses), and in the InVision Data Mart database between July 20, 2009, and March 31, 2012. The end date differed for these 2 databases because of different lag time in receiving claims. The National Drug Code (NDC) was used to identify dronedarone in the MarketScan and InVision Data Mart databases. The NDC is a unique, 3segment number that serves as a universal product identifier for drugs.13 The start date of the dronedarone treatment (baseline of the study cohort or index date) for each patient was defined as the dispensing date of the first dronedarone prescription. To quantify the treatment period, all prescriptions for dronedarone were counted. A patient’s treatment duration on dronedarone was estimated by using the date of dispensing and the days’ supplied for each prescription. The quantity of prescription 4240 (1 prescription) and daily dose of prescription 41200 mg (2 prescriptions) were replaced by the most common values of the prescriptions in the study participants.
Measurements Descriptive Variables We described the following characteristics of dronedarone users: AF/AFL and comorbid conditions (diabetes, hypertension, stroke/transient ischemic attack, and acute myocardial infarction [AMI]) within 1
266
year before the index date. The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), was used to identify relevant medical conditions. The ICD-9-CM code 427.3x in at least 1 medical claim within 1 year before the index date was used to ascertain AF/AFL. This algorithm has been validated and used to reliably capture the cases of AF/AFL in Veterans Affairs administrative data.14 The validation study demonstrated an observed agreement of 0.98 between chart review and the administrative data–based algorithm, a sensitivity of 80%, and a specificity of 99% when only 1 administrative diagnosis was required. Similar sensitivity (87%) and specificity (100%) was found in another study with 1 diagnosis of AF based on Medicare data.15 The ICD-9-CM codes were also used to identify diabetes (250.x), hypertension (401. x–405.x, 437.2), stroke/transient ischemic attack (433. x1, 434.x1, 435.x, 436, 437.1x, 437.9x, and 438.x), and AMI (410.x).16,17 Dronedarone is contraindicated in patients with symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF. Due to the absence of information on recent decompensation requiring hospitalization and NYHA class in the MarketScan and InVision Data Mart databases, we investigated patients with worsening or hospitalized HF within 30 days before initiation or refilling of dronedarone by using the following algorithm as a surrogate:
Step 1: identify patients with a diagnosis of HF
based on the ICD-9-CM codes (see the Supplemental Appendix I in the online version at http://dx.doi.org/ 10.1016/j.clinthera.2014.01.002); Step 2: of the patients identified during step 1, select those who had HF as 1 in-hospital diagnosis within 30 days before initiation or refilling of dronedarone; Step 3: of the patients identified during step 1 and who did not meet the criteria for step 2, select those with a diagnosis of HF and who initiated or were maintained on intravenous diuretics within 30 days before initiation or refilling of dronedarone. To ensure capture of intravenous diuretic use in an outpatient clinic visit, the following 2 steps were followed: (1) identifying the NDC codes for furosemide and ethacrynic accompanied by the Current Procedural Terminology (CPT) codes 96365, 96366, or 96374, which are for intravenous infusion or push; and (2) identifying the CPT codes for
Volume 36 Number 2
J. Wu et al. injections of diuretics, including J1205 (chlorothiazide sodium), J1940 (furosemide), J2150 (mannitol), and J3265 (torsemide). The NDC codes were used to identify the potent CYP3A4 inhibitors (see the Supplemental Appendix II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.01.002)18 and the QT-prolonging drugs (see the Supplemental Appendix III in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.01.002).19 The prevalence ratio of prescriptions for these drugs within 30 days before and after initiation or refilling of dronedarone was calculated. The CPT codes (see the Supplemental Appendix IV in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.01.002) and the Logical Observation Identifiers Names and Codes (see the Supplemental Appendix V in the online version at http://dx.doi.org/ 10.1016/j.clinthera.2014.01.002) were used to identify creatinine testing. The Logical Observation Identifiers Names and Codes is a universal standard for identifying medical laboratory observations. The prevalence ratio of the first serum creatinine testing after initiation of dronedarone treatment was calculated during the time windows of days 0 to 6, 7 to 13, 14 to 30, day Z31, or never.
Baseline Characteristics Baseline characteristics included age on the index date, sex, geographic region, year of the first dronedarone prescription, and enrollment history.
Statistical Methods Statistical analyses were performed by using SAS version 9.1 (SAS Institute, Inc, Cary, North Carolina). Standard descriptive statistics were used to summarize demographic characteristics of dronedarone users, prescriptions, and treatment duration. Prevalence ratios and their CIs were calculated for AF/AFL, diabetes, hypertension, stroke, AMI, worsening or hospitalized HF, concomitant prescriptions of potent CYP3A4 inhibitors, concomitant prescriptions of QTprolonging drugs, and serum creatinine testing.
RESULTS Baseline Characteristics Table I displays the baseline characteristics of patients using dronedarone in the MarketScan and the InVision Data Mart databases. In the MarketScan
February 2014
database, there were 31,408 dronedarone users; 40% were women and 54% were aged Z65 years. The majority ( 95%) of the patients had at least a 1-year enrollment history. There were 7025 dronedarone users in the InVision Data Mart database; 31% were women and 32% were aged Z65 years. The majority (492%) of the patients had at least a 1-year enrollment history. There was a total of 150,037 dronedarone prescriptions in the MarketScan database, with a mean (SD) supply of 46.6 (27.5) days per prescription and a median daily dose of 800 mg per prescription. On a patient level, the mean number of prescriptions was 4.8 (4.4) per patient, and the mean treatment duration was 223.2 (206.7) days per patient. There was a total of 39,371 dronedarone prescriptions in the InVision Data Mart database, with a mean supply of 37.2 (20.2) days per prescription and a median daily dose of 800 mg per prescription. On a patient level, the mean number of prescriptions was 5.6 (5.5) per patient and the mean treatment duration was 208.6 (207.3) days per patient. Table II presents the diagnosis of AF/AFL and cardiovascular risk factors. In the MarketScan database, 27,105 patients (86.3% [95% CI, 85.9– 86.7]) had a diagnosis of AF/AFL at initiation of dronedarone or within 1 year before. The prevalence of patients with at least 1 cardiovascular risk factor (including hypertension, diabetes, stroke, AMI, and age 470 years) was 78.8% (95% CI, 78.4–79.3). In the InVision Data Mart database, 6300 patients (89.7% [95% CI, 89.0–90.4]) had a diagnosis of AF/AFL at initiation of dronedarone or within 1 year before. The prevalence of patients with at least 1 cardiovascular risk factor (including hypertension, diabetes, stroke, AMI, and age 470 years) was 76.1% (95% CI, 75.1–77.1).
Worsening or Hospitalization for HF The prevalence of worsening or hospitalization for HF within 30 days before initiation and refilling is shown in Table III. In the MarketScan database, the combined prevalence before initiation and before refilling of dronedarone was 6.4% (95% CI, 6.2–6.7). In the InVision Data Mart database, the combined prevalence (before initiation plus before refilling) was 4.7% (95% CI, 4.2–5.2).
Concomitant Medication Use The prevalence of patients receiving concomitant prescriptions for potent CYP3A4 inhibitors and
267
Clinical Therapeutics Table I. Baseline characteristics of dronedarone users (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). MarketScan
Characteristic Age on the index date, y
Female sex Region
Year of the first dronedarone prescription
Enrollment history, mo
o18 18–29 30–49 50–64 65–70 Z71 Northeast Midwest South West Unknown 2009 2010 2011 2012 1–3 4–6 7–12 Z13
InVision Data Mart
No.
%
No.
%
4 77 2050 12,318 4485 12,474 12,565 3802 9541 12,121 5646 298 4528 14,915 11,965 NA 167 418 1002 29,821
0.0 0.3 6.5 39.2 14.3 39.7 40.0 12.1 30.4 38.6 17.9 0.9 14.4 47.5 38.1 NA 0.5 1.3 3.2 94.9
5 39 862 3869 1042 1208 2150 599 1685 3746 995 NA 1079 2999 2506 441 153 75 307 6490
0.1 0.6 12.3 55.1 14.8 17.2 30.6 8.5 23.9 53.3 14.2 NA 15.4 42.7 35.7 6.3 2.2 1.1 4.4 92.4
NA ¼ not available.
Table II. Diagnosis of atrial fibrillation (AF) and atrial flutter (AFL) and cardiovascular risk factors at initiation of dronedarone or within 1 year of initiation (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). MarketScan
Diagnosis No. Diagnosis of AF/AFL Diagnosis of hypertension Diagnosis of diabetes Diagnosis of stroke Diagnosis of acute myocardial infarction Z1 risk factor* *
27,105 19,304 7034 2148 1289 24,752
% (95% CI) 86.3 61.5 22.4 6.8 4.1 78.8
(85.9–86.7) (60.9–62.0) (21.9–22.9) (6.6–7.1) (3.9–4.3) (78.4–79.3)
InVision Data Mart No. 6300 4945 1586 440 283 5345
% (95% CI) 89.7 70.4 22.6 6.3 4.0 76.1
(89.0–90.4) (69.3–71.5) (21.6–23.6) (5.7–6.8) (3.6–4.5) (75.1–77.1)
Including hypertension, diabetes, stroke, acute myocardial infarction, or age 470 years.
268
Volume 36 Number 2
J. Wu et al.
Table III. Prevalence of worsening of or hospitalization for heart failure within 30 days before initiation or refilling of dronedarone (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). Worsening or Hospitalization for Heart Failure Within 30 days before initiation Within 30 days before refilling Total
MarketScan
InVision Data Mart
No. of Cases
% (95% CI)
No. of Cases
% (95% CI)
1533 945 2017
4.9 (4.6–5.1) 3.0 (2.8–3.2) 6.4 (6.2–6.7)
247 158 328
3.5 (3.1–3.9) 2.2 (1.9–2.6) 4.7 (4.2–5.2)
QT-prolonging drugs is shown in Tables IV and V. In the MarketScan database, the prevalence of concomitant medication use for potent CYP3A4 inhibitors within 30 days before initiation or refilling of dronedarone was 2.0% (95% CI, 1.8–2.1), and within 30 days after initiation or refilling of dronedarone, it was 1.9% (95% CI, 1.7–2.0). The prevalence of concomitant medication use for QTprolonging drugs within 30 days before initiation or refilling of dronedarone was 10.0% (95% CI, 9.7– 10.4), and within 30 days after initiation or refilling of dronedarone, it was 9.2% (95% CI, 8.9–9.5) in the MarketScan database.
Serum Creatinine Testing In the MarketScan database, 16,909 patients (53.8%) had a serum creatinine test conducted after initiation of dronedarone. Approximately 19% of the patients had the test performed within 30 days after
initiation of dronedarone. In the InVision Data Mart database, 3849 patients (54.8%) had a serum creatinine test conducted after initiation of dronedarone. Approximately 15% of the patients had the test performed within 30 days after initiation of dronedarone.
DISCUSSION The current article presents the long-term follow-up (nearly 3 years) results of the evaluation of dronedarone utilization patterns in the US population, including the labeling compliance and effectiveness of the REMS by using the MarketScan and InVision Data Mart databases. Compared with 1820 patients identified in the first publication,4 many more patients were included in the current analysis. Using both databases, we assessed 38,433 patients who received a total of 189,408 dronedarone prescriptions.
Table IV. Prevalence of patients receiving concomitant prescriptions for potent cytochrome P450 (CYP) 3A4 inhibitors and QT-prolonging drugs within 30 days surrounding initiation or refilling of dronedarone (N ¼ 31,408 in the MarketScan database as of December 2011). Prescription
Potent CYP3A4 Inhibitors No. of Cases
Within 30 Within 30 Subtotal Within 30 Within 30 Subtotal
days before initiation days before refilling days after initiation days after refilling
February 2014
206 470 618 190 444 596
% (95% CI) 0.7 1.5 2.0 0.6 1.4 1.9
(0.6–0.7) (1.4–1.6) (.8–2.1) (0.5–0.7) (1.3–1.5) (1.7–2.0)
QT-Prolonging Drugs No. of Cases 2630 833 3147 1384 1708 2891
% (95% CI) 8.4 2.7 10.0 4.4 5.4 9.2
(8.1–8.7) (2.5–2.8) (9.7–10.4) (4.2–4.6) (5.2–5.7) (8.9–9.5)
269
Clinical Therapeutics
Table V. Prevalence of patients receiving concomitant prescription for potent cytochrome P450 (CYP) 3A4 inhibitors and the QT-prolonging drugs within 30 days surrounding initiation or refilling of dronedarone (N ¼ 7025 in the InVision Data Mart database as of March 2012). Prescription
Potent CYP3A4 Inhibitors No. of Cases
Within 30 Within 30 Subtotal Within 30 Within 30 Subtotal
days before initiation days before refilling days after initiation days after refilling
44 138 164 41 129 161
Consistent with data from the previous publication,4 most patients had a diagnosis of AF/AFL at initiation or within 1 year before prescription of dronedarone (previous publication, 94.1% [95% CI, 92.9–95.1]; MarketScan database, 86.3% [95% CI, 85.9–86.7]; InVision Data Mart database, 89.7% [95% CI, 89.0–90.4]). Dronedarone is contraindicated in patients who have symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF. The results showed that the prevalence of worsening or inhospital diagnosis of HF within 30 days before initiation of dronedarone plus before refilling of dronedarone was 6.4% (95% CI, 6.2–6.7) and 4.7% (95% CI, 4.2–5.2) in the MarketScan and InVision Data Mart databases, respectively. It suggests that the REMS for dronedarone, including product labeling, medication guide, and comprehensive communication plan, has been effective. Although there is no established threshold for assessing effectiveness of the REMS, 10% seems an acceptable benchmark for suggesting that this program may be effective. The prevalence of concomitant medication use for potent CYP3A4 inhibitors within 30 days before or after initiation or refilling of dronedarone was low in both databases. This finding indicates a higher awareness by health care professionals regarding the potential risk of drug–drug interactions with potent CYP3A4 inhibitors as advertised in each dronedarone container or package, package insert, and/or medication guide. The prevalence of concomitant medication use for QT-prolonging drugs within 30 days before or after initiation or refilling of dronedarone was 10%
270
% (95% CI) 0.6 2.0 2.3 0.6 1.8 2.3
(0.4–0.8) (1.6–2.3) (2.0–2.7) (0.4–0.8) (1.5–2.2) (1.9–2.6)
QT-Prolonging Drugs No. of Cases 687 161 788 304 447 711
% (95% CI) 9.8 2.3 11.2 4.3 6.4 10.1
(9.1–10.5) (1.9–2.6) (10.5–12.0) (3.9–4.8) (5.8–6.9) (9.4–10.8)
in both databases. Most (86.4%) of these QTprolonging drugs dispensed within 30 days before or after initiation or refilling of dronedarone were other antiarrhythmic agents. It is likely that patients switched from other antiarrhythmic drugs to dronedarone or vice versa. Unfortunately, drug switching could not be separated from concomitant use in the claims databases. The US prescribing information of dronedarone states that renal function should be monitored periodically in patients treated with dronedarone.1 We found that 453% of patients had a serum creatinine test conducted after initiation of dronedarone in each database. There is room for improvement regarding serum creatinine testing after initiation of dronedarone. The results of the present analysis were consistent with the previous findings that in compliance with the US prescribing information, dronedarone has been used appropriately in the target population, including commonly prescribing dronedarone to AF/AFL patients with cardiovascular risk factors and not commonly prescribing it to patients with contraindications such as worsening or hospitalization for HF. One may argue how to set a threshold for the effectiveness of risk minimization measures. In our opinion, the time trend may be more important than an absolute value. Time trend analysis would allow us to look for any particular pattern with respect to the effectiveness of risk minimization measures. In addition, drug utilization studies are now commonly conducted for assessment of the effectiveness of risk minimization strategies in the United States and Europe.20,21
Volume 36 Number 2
J. Wu et al. However, findings from this type of research are rarely published as results in peer-reviewed journals. By publishing our results, we encourage broader communication for this type of research in the scientific community. Several limitations of the databases must be noted. Claims data are collected for the purpose of payment and therefore lack the clinical detail of medical charts or prospectively collected data (eg, no information was available on subtype of AF/AFL). Dronedarone is contraindicated in patients with AF who will not or cannot be cardioverted into normal sinus rhythm.1 In patients with permanent AF, dronedarone doubles the risk of death, stroke, and hospitalization for HF.3 Unfortunately, no specific ICD-9-CM code is available for permanent AF diagnosis. In addition, there is no recording of indications for prescriptions; diagnosis of AF/AFL for dronedarone was assumed based on dates of diagnosis and dates of dronedarone prescriptions. Furthermore, the databases do not capture medications prescribed in hospitals, and they also do not capture all laboratory tests. We were not able to separate drug switching from concomitant medication use. Lastly, the MarketScan Medicare Supplemental database captures claims, which are not 100% paid by Medicare and are paid by private-sector health care plans for the remaining part. It thus may not capture 100% of hospital claims.
CONCLUSIONS The results of the present analysis, based on a longterm follow-up (nearly 3 years), are consistent with previous findings that in compliance with the US prescribing information, dronedarone has been used appropriately in the target population. This includes commonly prescribing dronedarone to AF/AFL patients with cardiovascular risk factors and not commonly prescribing it to patients with contraindications such as worsening of or hospitalization for HF.
ACKNOWLEDGMENTS The authors would like to thank Laurent Auclert, MD and Shujun Gao, MD, PhD for their helpful review and comments on the manuscript.
CONFLICTS OF INTEREST The authors are current or former employees of Sanofi, the manufacturer of dronedarone. The authors
February 2014
have indicated that they have no other conflicts of interest regarding the content of this article.
REFERENCES 1. Multaq [prescribing information]. http://products.sanofi. us/multaq/multaq.html. Accessed May 20, 2013. 2. Hohnloser SH, Crijns HJ, van Eickels M, et al, for the ATHENA Investigators. Effect of dronedarone on cardiovascular events in atrial fibrillation. N Engl J Med. 2009;360:668–678. 3. Connolly S, Camm AJ, Halperin JL, et al, for the PALLAS investigators. Dronedarone in high-risk permanent atrial fibrillation. N Engl J Med. 2011;365:2268–2276. 4. Gao S, Juhaeri J, Schiappacasse HA, et al. Evaluation of dronedarone use in the US patient population between 2009 and 2010: a descriptive study using a claims database. Clin Ther. 2011;33:1483–1490. 5. United States Census Bureau. State and county quickfacts. http://quickfacts.census.gov/qfd/states/00000.html. Accessed May 20, 2013. 6. Naccarelli GV, Varker H, Lin J, et al. Increasing prevalence of atrial fibrillation and flutter in the United States. Am J Cardiol. 2009;104:1534–1539. 7. Centers for Medicare & Medicaid Services, Medicare & Medicaid Statistical Supplement. http://www.cms.gov/ Research-Statistics-Data-and-Systems/Statistics-Trends-andReports/MedicareMedicaidStatSupp/2012.html. Accessed May 20, 2013. 8. Naccarelli GV, Johnston SS, Dalal M, et al. Rates and implications for hospitalization of patients Z65 years of age with atrial fibrillation/flutter. Am J Cardiol. 2012;109: 543–549. 9. Reynolds MR, Gunnarsson CL, Hunter TD, et al. Health outcomes with catheter ablation or antiarrhythmic drug therapy in atrial fibrillation, results of a propensity-matched analysis. Circ Cardiovasc Qual Outcomes. 2012;5:171–181. 10. Ladapo JA, David G, Gunnarsson CL, et al. Healthcare utilization and expenditures in patients with atrial fibrillation treated with catheter ablation. J Cardiovasc Electrophysiol. 2012;23:1–8. 11. Kim MH, Johnston SS, Chu BC, et al. Estimation of total incremental health care costs in patients with atrial fibrillation in the United States. Circ Cardiovasc Qual Outcomes. 2011;4:313–320. 12. Health Insurance Portability and Accountability Act of 1996, 42 USC 1320-d2. 13. National Drug Code Directory. http://www.fda.gov/cder/ ndc. Accessed May 20, 2013. 14. Borzecki AM, Wong AT, Hickey EC, et al. Identifying hypertension-related comorbidities from administrative
271
Clinical Therapeutics
15.
16.
17.
18.
19.
20.
21.
data: whatʼs the optimal approach? Am J Med Qual. 2004;19:201–206. Yuan Z, Bowlin S, Einstadter D, et al. Atrial fibrillation as a risk factor for stroke: a retrospective cohort study of hospitalized Medicare beneficiaries. Am J Public Health. 1998;88: 395–400. Birman-Deych E, Waterman AD, Yan Y, et al. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors. Med Care. 2005;43:480–485. Kiyota Y, Schneeweiss S, Glynn RJ, et al. Accuracy of Medicare claimsbased diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records. Am Heart J. 2004;148:99–104. Busti AJ, HJ, Daves J, McKeever GC. What are some common medications classified as weak, moderate and strong inhibitors of CYP3A4? PW Drug Interact Newsl. 2009;1:1–4. http://www.pharmacologyweekly.com/ articles/medications-inhibitors-CYP3 A4-enzyme. Accessed January 22, 2014. Drugs with risk of torsades de pointes. http://crediblemeds.org/ pdftemp/pdf/US-CompositeList.pdf. Accessed May 20, 2013. US Department of Health and Human Services. Guidance for industry format and content of proposed Risk Evaluation and Mitigation Strategies (REMS), REMS assessments, and proposed REMS modifications. http://www.fda.gov/downloads/Drugs/GuidanceComplianceR egulatoryInformation/Guidances/UCM 184128.pdf. Accessed May 20, 2013. European Medicines Agency guidance. Guidance on format of the risk management plan (RMP) in the EU part V: risk minimisation measures. http://www.ema.europa.eu/ loads/Drugs/GuidanceComplianceR docs/en_GB/document_library/Reg ulatory_and_procedural_guideline/ 2012/11/WC500134662.pdf. Accessed May 20, 2013.
272
Address correspondence to: Jasmanda Wu, PhD, MPH, Sanofi, Global Pharmacovigilance and Epidemiology, 55 Corporate Drive, Bridgewater, NJ 08807. E-mail: jasmanda.wu@sanofi.com
Volume 36 Number 2
J. Wu et al.
Supplemental Appendix I. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD9-CM), codes of heart failure. ICD-9-CM Code 428.x 398.91 402.01 402.11 402.91 404.01 404.11 404.91 404.03 404.13 404.93
Description Heart failure Rheumatic heart failure (congestive) Malignant hypertensive heart disease with heart failure Benign hypertensive heart disease with heart failure Unspecified hypertensive heart disease with heart failure Malignant hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Benign hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Unspecified hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Malignant hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease Benign hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease Unspecified hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease
Supplemental Appendix III. Drugs that prolong QT interval (generic names). Amiodarone
Supplemental Appendix II. Potent cytochrome P450 3A4 inhibitors (generic names). Atazanavir
Itraconazole
Ritonavir
Chloramphenicol Clarithromycin Cyclosporin Darunavir Indinavir
Ketoconazole Lopinavir Nefazodone Nelfinavir Posaconazole
Saquinavir Telithromycin Tipranavir Troleandomycin
February 2014
Arsenic trioxide Astemizole Bepridil Chloroquine Chlorpromazine Cisapride Clarithromycin Disopyramide Dofetilide Domperidone Droperidol Erythromycin Halofantrine
Haloperidol Ibutilide Levomethadyl Mesoridazine Methadone Pentamidine Pimozide Probucol Procainamide Quinidine Sotalol Sparfloxacin Terfenadine Thioridazine
272.e1
Clinical Therapeutics
Supplemental Appendix IV. Current Procedural Terminology (CPT) codes for creatinine testing. CPT Code
Description
80047 80048 80050 80053 80069 82565 82570 82575
Basic metabolic panel (calcium, ionized) Basic metabolic panel General health panel Comprehensive metabolic profile Renal function panel Blood creatinine Creatinine, other source Creatinine, clearance
Supplemental Appendix V. Logical Observation Identifiers Names and Codes (LOINC) for creatinine testing. LOINC 14682-9 21232-4 2160-0 35203-9 38483-4 40248-7 40264-4 44784-7 59826-8
272.e2
Description Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine
(moles/volume) in serum or plasma (mass/volume) in arterial blood (mass/volume) in serum or plasma (mass or moles/volume) in serum or plasma (mass/volume) in blood (mass/volume) in serum or plasma, baseline (moles/volume) in serum or plasma, baseline (mass/volume) in maximum serum or plasma (moles/volume) in blood
Volume 36 Number 2
Assessment of Dronedarone Utilization Using US Claims Databases Jasmanda Wu, PhD, MPH1; Jane Thammakhoune, BS1; Wanju Dai, MD, DrPH1; Andrew Koren, MD2; Stephanie Tcherny-Lessenot, MD, MSc, MPH3; Chuntao Wu, MD, PhD1; Patrick Caubel, MD1; and Juhaeri Juhaeri, PhD1 1
Global Phamacovigilance and Epidemiology, Sanofi, Bridgewater, New Jersey; 2Cardiovascular Medical Unit, Sanofi, Bridgewater, New Jersey; and 3Global Phamacovigilance and Epidemiology, Sanofi, Chilly-Mazarin, France
ABSTRACT Background: A dronedarone utilization study using US MarketScan and InVision Data Mart databases was conducted to estimate the prevalence of the following: (1) dronedarone use in contraindicated patients with worsening heart failure (HF) or hospitalization for HF within 1 month before dronedarone prescription; (2) concomitant prescribing of contraindicated drugs; and (3) recommended creatinine testing after dronedarone initiation among dronedarone users. Methods: In this retrospective cohort study, data in the MarketScan database between July 20, 2009, and December 31, 2011, and in the InVision Data Mart database between July 20, 2009, and March 31, 2012, were analyzed. The study population included patients who received Z1 dronedarone prescription during the study period. The following variables were reported: worsening of or hospitalization for HF, concomitant prescribing of potent cytochrome P450 CYP 3A4 inhibitors or QT-prolonging drugs, and creatinine testing. Results: There were 31,408 and 7025 dronedarone users identified in the MarketScan and InVision Data Mart databases, respectively. Approximately 86% to 90% of patients had a diagnosis of atrial fibrillation in each database. In the MarketScan database, 40% were women and 54% were aged Z65 years. In the InVision Data Mart database, 31% were women and 32% were aged Z65 years. The corresponding prevalence of worsening or hospitalization for HF was 6.4% (95% CI, 6.2–6.7) and 4.7% (95% CI, 4.2–5.2) in each database, respectively. The corresponding estimates of concomitant prescribing of potent cytochrome P450 CYP 3A4 inhibitors and QT-prolonging drugs within 30 days before initiation or refilling of dronedarone were 2.0% (95% CI, 1.8–2.1) and 10.0% (95% CI, 9.7–10.4), respectively, in the
264
MarketScan database, and 2.3% (95% CI, 2.0–2.7) and 11.2% (95% CI, 10.5–12.0) in the InVision Data Mart database. More than 50% of patients in each database had serum creatinine tests conducted after dronedarone initiation. Conclusions: The results of the present analysis based on a long-term follow-up (nearly 3 years) were consistent with the previous findings that dronedarone has mostly been used appropriately in compliance with US prescribing in the target populations. (Clin Ther. 2014;36:264–272) & 2014 Elsevier HS Journals, Inc. All rights reserved. Key words: dronedarone, drug utilization, Risk Evaluation and Mitigation Strategy.
INTRODUCTION Dronedarone, a multichannel blocker displaying antiarrhythmic properties, has been shown to be effective for reducing the risk of cardiovascular hospitalization or death in patients with atrial fibrillation (AF) and atrial flutter (AFL).1,2 In the United States, dronedarone was approved as “an antiarrhythmic drug indicated to reduce the risk of cardiovascular hospitalization in patients with paroxysmal or persistent AF or AFL, with a recent episode of AF/AFL and associated cardiovascular risk factors (ie, age 470, hypertension, diabetes, prior cerebrovascular accident, left atrial diameter Z50 mm or left ventricular ejection fraction [LVEF] o40%), who are in sinus rhythm or who will be cardioverted.”1 Based on the Accepted for publication January 4, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.01.002 0149-2918/$ - see front matter & 2014 Elsevier HS Journals, Inc. All rights reserved.
Volume 36 Number 2
J. Wu et al. nonclinical and clinical safety information collected throughout an extensive development program, dronedarone is contraindicated in the United States in patients taking potent cytochrome P450 (CYP) 3A4 inhibitors that significantly raise serum levels of dronedarone. It is also contraindicated in patients taking drugs that prolong the QT interval and may induce torsade de pointes. In addition, dronedarone is contraindicated in patients with symptomatic heart failure (HF) with recent decompensation requiring hospitalization or New York Heart Association (NYHA) class IV HF. It is also contraindicated in patients in permanent AF who will not or cannot be cardioverted into normal sinus rhythm. In patients with permanent AF, dronedarone doubles the risk of death, stroke, and hospitalization for HF.3 In clinical trials, an increase in serum creatinine level was observed in patients treated with dronedarone 400 mg BID.1 This increase occurs early after treatment initiation and reaches a plateau after 7 days. The US prescribing information for dronedarone states that renal function should be monitored periodically in patients treated with dronedarone because increases in creatinine and blood urea nitrogen have been reported in the postmarketing setting and seem to be reversible after discontinuation of dronedarone.1 In the United States, in addition to labeling and other pharmacovigilance action plan, a Risk Evaluation and Mitigation Strategy (REMS) program has been developed to prevent dronedarone use in: (1) patients with symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF; and (2) patients with permanent AF who will not or cannot be cardioverted into normal sinus rhythm. The REMS educates prescribers through a comprehensive communication plan about increased mortality when dronedarone is used in these patient populations and informs patients about the serious risks of dronedarone, including increased mortality in patients with severe unstable HF or permanent AF, through a medication guide. To characterize dronedarone utilization patterns in the US population and to assess labeling compliance and the effectiveness of the REMS, a drug utilization study using the InVision Data Mart database has been conducted since 2009.4 The first publication based on data between July 2009 and August 2010 demonstrated labeling compliance and effectiveness of the
February 2014
REMS regarding worsening HF. However, because InVision Data Mart is an employment-based private insurance claims database, it has a lower proportion of patients aged Z65 years (9% in the InVision Data Mart vs 13% in the US population5), whereas AF and AFL are more prevalent in elderly patients than in the general population.2,6 Thus, the MarketScan database, which is composed of MarketScan Commercial Claims and Encounters database, the Medicare Supplemental database, and the Medicaid database, was added in the study to improve the representation of elderly patients. The MarketScan Medicare Supplemental database captures Medicare-eligible retirees with employer-sponsored Medicare supplemental plans. The age and gender distribution in the MarketScan Medicare Supplemental database is compatible to the overall Medicare population (MarketScan vs Medicare: age 65–74 years, 55% vs 45%; age 75–84 years, 29% vs 26%; age Z85 years, 12% vs 12%; male sex, 45% vs 45%).7 The MarketScan database has been used in a number of studies of patients with AF.8–11 The current article presents results from a longterm follow-up (nearly 3 years) for evaluation of dronedarone utilization patterns in the US population by using both the MarketScan and the InVision Data Mart databases.
MATERIALS AND METHODS Data Source The study used the US MarketScan Commercial Claims and Encounters database, the Medicare Supplemental database, and the Medicaid database from Thomson Reuters Corporation (New York, New York). The databases are composed of de-identified patient-level records from 4170 million patients since 1995. The MarketScan Commercial Claims and Encounters database includes private sector health data from 150 employers and 21 health plans. The Medicare Supplemental database captures Medicareeligible retirees with employer-sponsored Medicare supplemental plans. On average, nearly 3.8 million lives (active patients) are covered per year. The Medicaid database contains the pooled health care experience of 7 million Medicaid enrollees from 12 states. On average, nearly 6.8 million lives (active patients) are covered per year. The MarketScan databases are in compliance with the Health Information Portability and Accountability Act of 1996.12
265
Clinical Therapeutics The InVision Data Mart (United HealthCare, formerly LabRx, Eden Prairie, Minnesota) database was also used. The InVision Data Mart database is an integrated medical and prescription claims database augmented with partial laboratory data for an employed, commercially insured population of patients and their dependents. This research database includes 430 million members. The database is compliant with the Health Information Portability and Accountability Act of 1996.12 Because this study did not involve the collection, use, or transmittal of individually identifiable data, institutional review board review or approval was not required.
Study Population The analysis was performed among all patients who were prescribed dronedarone at least once as identified in the MarketScan database between July 20, 2009 (the launch date in the United States), and December 31, 2011 (the date of the latest data update received from the database vendor for the current analyses), and in the InVision Data Mart database between July 20, 2009, and March 31, 2012. The end date differed for these 2 databases because of different lag time in receiving claims. The National Drug Code (NDC) was used to identify dronedarone in the MarketScan and InVision Data Mart databases. The NDC is a unique, 3segment number that serves as a universal product identifier for drugs.13 The start date of the dronedarone treatment (baseline of the study cohort or index date) for each patient was defined as the dispensing date of the first dronedarone prescription. To quantify the treatment period, all prescriptions for dronedarone were counted. A patient’s treatment duration on dronedarone was estimated by using the date of dispensing and the days’ supplied for each prescription. The quantity of prescription 4240 (1 prescription) and daily dose of prescription 41200 mg (2 prescriptions) were replaced by the most common values of the prescriptions in the study participants.
Measurements Descriptive Variables We described the following characteristics of dronedarone users: AF/AFL and comorbid conditions (diabetes, hypertension, stroke/transient ischemic attack, and acute myocardial infarction [AMI]) within 1
266
year before the index date. The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), was used to identify relevant medical conditions. The ICD-9-CM code 427.3x in at least 1 medical claim within 1 year before the index date was used to ascertain AF/AFL. This algorithm has been validated and used to reliably capture the cases of AF/AFL in Veterans Affairs administrative data.14 The validation study demonstrated an observed agreement of 0.98 between chart review and the administrative data–based algorithm, a sensitivity of 80%, and a specificity of 99% when only 1 administrative diagnosis was required. Similar sensitivity (87%) and specificity (100%) was found in another study with 1 diagnosis of AF based on Medicare data.15 The ICD-9-CM codes were also used to identify diabetes (250.x), hypertension (401. x–405.x, 437.2), stroke/transient ischemic attack (433. x1, 434.x1, 435.x, 436, 437.1x, 437.9x, and 438.x), and AMI (410.x).16,17 Dronedarone is contraindicated in patients with symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF. Due to the absence of information on recent decompensation requiring hospitalization and NYHA class in the MarketScan and InVision Data Mart databases, we investigated patients with worsening or hospitalized HF within 30 days before initiation or refilling of dronedarone by using the following algorithm as a surrogate:
Step 1: identify patients with a diagnosis of HF
based on the ICD-9-CM codes (see the Supplemental Appendix I in the online version at http://dx.doi.org/ 10.1016/j.clinthera.2014.01.002); Step 2: of the patients identified during step 1, select those who had HF as 1 in-hospital diagnosis within 30 days before initiation or refilling of dronedarone; Step 3: of the patients identified during step 1 and who did not meet the criteria for step 2, select those with a diagnosis of HF and who initiated or were maintained on intravenous diuretics within 30 days before initiation or refilling of dronedarone. To ensure capture of intravenous diuretic use in an outpatient clinic visit, the following 2 steps were followed: (1) identifying the NDC codes for furosemide and ethacrynic accompanied by the Current Procedural Terminology (CPT) codes 96365, 96366, or 96374, which are for intravenous infusion or push; and (2) identifying the CPT codes for
Volume 36 Number 2
J. Wu et al. injections of diuretics, including J1205 (chlorothiazide sodium), J1940 (furosemide), J2150 (mannitol), and J3265 (torsemide). The NDC codes were used to identify the potent CYP3A4 inhibitors (see the Supplemental Appendix II in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.01.002)18 and the QT-prolonging drugs (see the Supplemental Appendix III in the online version at http://dx.doi.org/10.1016/j.clinthera. 2014.01.002).19 The prevalence ratio of prescriptions for these drugs within 30 days before and after initiation or refilling of dronedarone was calculated. The CPT codes (see the Supplemental Appendix IV in the online version at http://dx.doi.org/10.1016/j. clinthera.2014.01.002) and the Logical Observation Identifiers Names and Codes (see the Supplemental Appendix V in the online version at http://dx.doi.org/ 10.1016/j.clinthera.2014.01.002) were used to identify creatinine testing. The Logical Observation Identifiers Names and Codes is a universal standard for identifying medical laboratory observations. The prevalence ratio of the first serum creatinine testing after initiation of dronedarone treatment was calculated during the time windows of days 0 to 6, 7 to 13, 14 to 30, day Z31, or never.
Baseline Characteristics Baseline characteristics included age on the index date, sex, geographic region, year of the first dronedarone prescription, and enrollment history.
Statistical Methods Statistical analyses were performed by using SAS version 9.1 (SAS Institute, Inc, Cary, North Carolina). Standard descriptive statistics were used to summarize demographic characteristics of dronedarone users, prescriptions, and treatment duration. Prevalence ratios and their CIs were calculated for AF/AFL, diabetes, hypertension, stroke, AMI, worsening or hospitalized HF, concomitant prescriptions of potent CYP3A4 inhibitors, concomitant prescriptions of QTprolonging drugs, and serum creatinine testing.
RESULTS Baseline Characteristics Table I displays the baseline characteristics of patients using dronedarone in the MarketScan and the InVision Data Mart databases. In the MarketScan
February 2014
database, there were 31,408 dronedarone users; 40% were women and 54% were aged Z65 years. The majority ( 95%) of the patients had at least a 1-year enrollment history. There were 7025 dronedarone users in the InVision Data Mart database; 31% were women and 32% were aged Z65 years. The majority (492%) of the patients had at least a 1-year enrollment history. There was a total of 150,037 dronedarone prescriptions in the MarketScan database, with a mean (SD) supply of 46.6 (27.5) days per prescription and a median daily dose of 800 mg per prescription. On a patient level, the mean number of prescriptions was 4.8 (4.4) per patient, and the mean treatment duration was 223.2 (206.7) days per patient. There was a total of 39,371 dronedarone prescriptions in the InVision Data Mart database, with a mean supply of 37.2 (20.2) days per prescription and a median daily dose of 800 mg per prescription. On a patient level, the mean number of prescriptions was 5.6 (5.5) per patient and the mean treatment duration was 208.6 (207.3) days per patient. Table II presents the diagnosis of AF/AFL and cardiovascular risk factors. In the MarketScan database, 27,105 patients (86.3% [95% CI, 85.9– 86.7]) had a diagnosis of AF/AFL at initiation of dronedarone or within 1 year before. The prevalence of patients with at least 1 cardiovascular risk factor (including hypertension, diabetes, stroke, AMI, and age 470 years) was 78.8% (95% CI, 78.4–79.3). In the InVision Data Mart database, 6300 patients (89.7% [95% CI, 89.0–90.4]) had a diagnosis of AF/AFL at initiation of dronedarone or within 1 year before. The prevalence of patients with at least 1 cardiovascular risk factor (including hypertension, diabetes, stroke, AMI, and age 470 years) was 76.1% (95% CI, 75.1–77.1).
Worsening or Hospitalization for HF The prevalence of worsening or hospitalization for HF within 30 days before initiation and refilling is shown in Table III. In the MarketScan database, the combined prevalence before initiation and before refilling of dronedarone was 6.4% (95% CI, 6.2–6.7). In the InVision Data Mart database, the combined prevalence (before initiation plus before refilling) was 4.7% (95% CI, 4.2–5.2).
Concomitant Medication Use The prevalence of patients receiving concomitant prescriptions for potent CYP3A4 inhibitors and
267
Clinical Therapeutics Table I. Baseline characteristics of dronedarone users (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). MarketScan
Characteristic Age on the index date, y
Female sex Region
Year of the first dronedarone prescription
Enrollment history, mo
o18 18–29 30–49 50–64 65–70 Z71 Northeast Midwest South West Unknown 2009 2010 2011 2012 1–3 4–6 7–12 Z13
InVision Data Mart
No.
%
No.
%
4 77 2050 12,318 4485 12,474 12,565 3802 9541 12,121 5646 298 4528 14,915 11,965 NA 167 418 1002 29,821
0.0 0.3 6.5 39.2 14.3 39.7 40.0 12.1 30.4 38.6 17.9 0.9 14.4 47.5 38.1 NA 0.5 1.3 3.2 94.9
5 39 862 3869 1042 1208 2150 599 1685 3746 995 NA 1079 2999 2506 441 153 75 307 6490
0.1 0.6 12.3 55.1 14.8 17.2 30.6 8.5 23.9 53.3 14.2 NA 15.4 42.7 35.7 6.3 2.2 1.1 4.4 92.4
NA ¼ not available.
Table II. Diagnosis of atrial fibrillation (AF) and atrial flutter (AFL) and cardiovascular risk factors at initiation of dronedarone or within 1 year of initiation (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). MarketScan
Diagnosis No. Diagnosis of AF/AFL Diagnosis of hypertension Diagnosis of diabetes Diagnosis of stroke Diagnosis of acute myocardial infarction Z1 risk factor* *
27,105 19,304 7034 2148 1289 24,752
% (95% CI) 86.3 61.5 22.4 6.8 4.1 78.8
(85.9–86.7) (60.9–62.0) (21.9–22.9) (6.6–7.1) (3.9–4.3) (78.4–79.3)
InVision Data Mart No. 6300 4945 1586 440 283 5345
% (95% CI) 89.7 70.4 22.6 6.3 4.0 76.1
(89.0–90.4) (69.3–71.5) (21.6–23.6) (5.7–6.8) (3.6–4.5) (75.1–77.1)
Including hypertension, diabetes, stroke, acute myocardial infarction, or age 470 years.
268
Volume 36 Number 2
J. Wu et al.
Table III. Prevalence of worsening of or hospitalization for heart failure within 30 days before initiation or refilling of dronedarone (N ¼ 31,408 in the MarketScan database as of December 2011 and N ¼ 7025 in the InVision Data Mart database as of March 2012). Worsening or Hospitalization for Heart Failure Within 30 days before initiation Within 30 days before refilling Total
MarketScan
InVision Data Mart
No. of Cases
% (95% CI)
No. of Cases
% (95% CI)
1533 945 2017
4.9 (4.6–5.1) 3.0 (2.8–3.2) 6.4 (6.2–6.7)
247 158 328
3.5 (3.1–3.9) 2.2 (1.9–2.6) 4.7 (4.2–5.2)
QT-prolonging drugs is shown in Tables IV and V. In the MarketScan database, the prevalence of concomitant medication use for potent CYP3A4 inhibitors within 30 days before initiation or refilling of dronedarone was 2.0% (95% CI, 1.8–2.1), and within 30 days after initiation or refilling of dronedarone, it was 1.9% (95% CI, 1.7–2.0). The prevalence of concomitant medication use for QTprolonging drugs within 30 days before initiation or refilling of dronedarone was 10.0% (95% CI, 9.7– 10.4), and within 30 days after initiation or refilling of dronedarone, it was 9.2% (95% CI, 8.9–9.5) in the MarketScan database.
Serum Creatinine Testing In the MarketScan database, 16,909 patients (53.8%) had a serum creatinine test conducted after initiation of dronedarone. Approximately 19% of the patients had the test performed within 30 days after
initiation of dronedarone. In the InVision Data Mart database, 3849 patients (54.8%) had a serum creatinine test conducted after initiation of dronedarone. Approximately 15% of the patients had the test performed within 30 days after initiation of dronedarone.
DISCUSSION The current article presents the long-term follow-up (nearly 3 years) results of the evaluation of dronedarone utilization patterns in the US population, including the labeling compliance and effectiveness of the REMS by using the MarketScan and InVision Data Mart databases. Compared with 1820 patients identified in the first publication,4 many more patients were included in the current analysis. Using both databases, we assessed 38,433 patients who received a total of 189,408 dronedarone prescriptions.
Table IV. Prevalence of patients receiving concomitant prescriptions for potent cytochrome P450 (CYP) 3A4 inhibitors and QT-prolonging drugs within 30 days surrounding initiation or refilling of dronedarone (N ¼ 31,408 in the MarketScan database as of December 2011). Prescription
Potent CYP3A4 Inhibitors No. of Cases
Within 30 Within 30 Subtotal Within 30 Within 30 Subtotal
days before initiation days before refilling days after initiation days after refilling
February 2014
206 470 618 190 444 596
% (95% CI) 0.7 1.5 2.0 0.6 1.4 1.9
(0.6–0.7) (1.4–1.6) (.8–2.1) (0.5–0.7) (1.3–1.5) (1.7–2.0)
QT-Prolonging Drugs No. of Cases 2630 833 3147 1384 1708 2891
% (95% CI) 8.4 2.7 10.0 4.4 5.4 9.2
(8.1–8.7) (2.5–2.8) (9.7–10.4) (4.2–4.6) (5.2–5.7) (8.9–9.5)
269
Clinical Therapeutics
Table V. Prevalence of patients receiving concomitant prescription for potent cytochrome P450 (CYP) 3A4 inhibitors and the QT-prolonging drugs within 30 days surrounding initiation or refilling of dronedarone (N ¼ 7025 in the InVision Data Mart database as of March 2012). Prescription
Potent CYP3A4 Inhibitors No. of Cases
Within 30 Within 30 Subtotal Within 30 Within 30 Subtotal
days before initiation days before refilling days after initiation days after refilling
44 138 164 41 129 161
Consistent with data from the previous publication,4 most patients had a diagnosis of AF/AFL at initiation or within 1 year before prescription of dronedarone (previous publication, 94.1% [95% CI, 92.9–95.1]; MarketScan database, 86.3% [95% CI, 85.9–86.7]; InVision Data Mart database, 89.7% [95% CI, 89.0–90.4]). Dronedarone is contraindicated in patients who have symptomatic HF with recent decompensation requiring hospitalization or with NYHA class IV HF. The results showed that the prevalence of worsening or inhospital diagnosis of HF within 30 days before initiation of dronedarone plus before refilling of dronedarone was 6.4% (95% CI, 6.2–6.7) and 4.7% (95% CI, 4.2–5.2) in the MarketScan and InVision Data Mart databases, respectively. It suggests that the REMS for dronedarone, including product labeling, medication guide, and comprehensive communication plan, has been effective. Although there is no established threshold for assessing effectiveness of the REMS, 10% seems an acceptable benchmark for suggesting that this program may be effective. The prevalence of concomitant medication use for potent CYP3A4 inhibitors within 30 days before or after initiation or refilling of dronedarone was low in both databases. This finding indicates a higher awareness by health care professionals regarding the potential risk of drug–drug interactions with potent CYP3A4 inhibitors as advertised in each dronedarone container or package, package insert, and/or medication guide. The prevalence of concomitant medication use for QT-prolonging drugs within 30 days before or after initiation or refilling of dronedarone was 10%
270
% (95% CI) 0.6 2.0 2.3 0.6 1.8 2.3
(0.4–0.8) (1.6–2.3) (2.0–2.7) (0.4–0.8) (1.5–2.2) (1.9–2.6)
QT-Prolonging Drugs No. of Cases 687 161 788 304 447 711
% (95% CI) 9.8 2.3 11.2 4.3 6.4 10.1
(9.1–10.5) (1.9–2.6) (10.5–12.0) (3.9–4.8) (5.8–6.9) (9.4–10.8)
in both databases. Most (86.4%) of these QTprolonging drugs dispensed within 30 days before or after initiation or refilling of dronedarone were other antiarrhythmic agents. It is likely that patients switched from other antiarrhythmic drugs to dronedarone or vice versa. Unfortunately, drug switching could not be separated from concomitant use in the claims databases. The US prescribing information of dronedarone states that renal function should be monitored periodically in patients treated with dronedarone.1 We found that 453% of patients had a serum creatinine test conducted after initiation of dronedarone in each database. There is room for improvement regarding serum creatinine testing after initiation of dronedarone. The results of the present analysis were consistent with the previous findings that in compliance with the US prescribing information, dronedarone has been used appropriately in the target population, including commonly prescribing dronedarone to AF/AFL patients with cardiovascular risk factors and not commonly prescribing it to patients with contraindications such as worsening or hospitalization for HF. One may argue how to set a threshold for the effectiveness of risk minimization measures. In our opinion, the time trend may be more important than an absolute value. Time trend analysis would allow us to look for any particular pattern with respect to the effectiveness of risk minimization measures. In addition, drug utilization studies are now commonly conducted for assessment of the effectiveness of risk minimization strategies in the United States and Europe.20,21
Volume 36 Number 2
J. Wu et al. However, findings from this type of research are rarely published as results in peer-reviewed journals. By publishing our results, we encourage broader communication for this type of research in the scientific community. Several limitations of the databases must be noted. Claims data are collected for the purpose of payment and therefore lack the clinical detail of medical charts or prospectively collected data (eg, no information was available on subtype of AF/AFL). Dronedarone is contraindicated in patients with AF who will not or cannot be cardioverted into normal sinus rhythm.1 In patients with permanent AF, dronedarone doubles the risk of death, stroke, and hospitalization for HF.3 Unfortunately, no specific ICD-9-CM code is available for permanent AF diagnosis. In addition, there is no recording of indications for prescriptions; diagnosis of AF/AFL for dronedarone was assumed based on dates of diagnosis and dates of dronedarone prescriptions. Furthermore, the databases do not capture medications prescribed in hospitals, and they also do not capture all laboratory tests. We were not able to separate drug switching from concomitant medication use. Lastly, the MarketScan Medicare Supplemental database captures claims, which are not 100% paid by Medicare and are paid by private-sector health care plans for the remaining part. It thus may not capture 100% of hospital claims.
CONCLUSIONS The results of the present analysis, based on a longterm follow-up (nearly 3 years), are consistent with previous findings that in compliance with the US prescribing information, dronedarone has been used appropriately in the target population. This includes commonly prescribing dronedarone to AF/AFL patients with cardiovascular risk factors and not commonly prescribing it to patients with contraindications such as worsening of or hospitalization for HF.
ACKNOWLEDGMENTS The authors would like to thank Laurent Auclert, MD and Shujun Gao, MD, PhD for their helpful review and comments on the manuscript.
CONFLICTS OF INTEREST The authors are current or former employees of Sanofi, the manufacturer of dronedarone. The authors
February 2014
have indicated that they have no other conflicts of interest regarding the content of this article.
REFERENCES 1. Multaq [prescribing information]. http://products.sanofi. us/multaq/multaq.html. Accessed May 20, 2013. 2. Hohnloser SH, Crijns HJ, van Eickels M, et al, for the ATHENA Investigators. Effect of dronedarone on cardiovascular events in atrial fibrillation. N Engl J Med. 2009;360:668–678. 3. Connolly S, Camm AJ, Halperin JL, et al, for the PALLAS investigators. Dronedarone in high-risk permanent atrial fibrillation. N Engl J Med. 2011;365:2268–2276. 4. Gao S, Juhaeri J, Schiappacasse HA, et al. Evaluation of dronedarone use in the US patient population between 2009 and 2010: a descriptive study using a claims database. Clin Ther. 2011;33:1483–1490. 5. United States Census Bureau. State and county quickfacts. http://quickfacts.census.gov/qfd/states/00000.html. Accessed May 20, 2013. 6. Naccarelli GV, Varker H, Lin J, et al. Increasing prevalence of atrial fibrillation and flutter in the United States. Am J Cardiol. 2009;104:1534–1539. 7. Centers for Medicare & Medicaid Services, Medicare & Medicaid Statistical Supplement. http://www.cms.gov/ Research-Statistics-Data-and-Systems/Statistics-Trends-andReports/MedicareMedicaidStatSupp/2012.html. Accessed May 20, 2013. 8. Naccarelli GV, Johnston SS, Dalal M, et al. Rates and implications for hospitalization of patients Z65 years of age with atrial fibrillation/flutter. Am J Cardiol. 2012;109: 543–549. 9. Reynolds MR, Gunnarsson CL, Hunter TD, et al. Health outcomes with catheter ablation or antiarrhythmic drug therapy in atrial fibrillation, results of a propensity-matched analysis. Circ Cardiovasc Qual Outcomes. 2012;5:171–181. 10. Ladapo JA, David G, Gunnarsson CL, et al. Healthcare utilization and expenditures in patients with atrial fibrillation treated with catheter ablation. J Cardiovasc Electrophysiol. 2012;23:1–8. 11. Kim MH, Johnston SS, Chu BC, et al. Estimation of total incremental health care costs in patients with atrial fibrillation in the United States. Circ Cardiovasc Qual Outcomes. 2011;4:313–320. 12. Health Insurance Portability and Accountability Act of 1996, 42 USC 1320-d2. 13. National Drug Code Directory. http://www.fda.gov/cder/ ndc. Accessed May 20, 2013. 14. Borzecki AM, Wong AT, Hickey EC, et al. Identifying hypertension-related comorbidities from administrative
271
Clinical Therapeutics
15.
16.
17.
18.
19.
20.
21.
data: whatʼs the optimal approach? Am J Med Qual. 2004;19:201–206. Yuan Z, Bowlin S, Einstadter D, et al. Atrial fibrillation as a risk factor for stroke: a retrospective cohort study of hospitalized Medicare beneficiaries. Am J Public Health. 1998;88: 395–400. Birman-Deych E, Waterman AD, Yan Y, et al. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors. Med Care. 2005;43:480–485. Kiyota Y, Schneeweiss S, Glynn RJ, et al. Accuracy of Medicare claimsbased diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records. Am Heart J. 2004;148:99–104. Busti AJ, HJ, Daves J, McKeever GC. What are some common medications classified as weak, moderate and strong inhibitors of CYP3A4? PW Drug Interact Newsl. 2009;1:1–4. http://www.pharmacologyweekly.com/ articles/medications-inhibitors-CYP3 A4-enzyme. Accessed January 22, 2014. Drugs with risk of torsades de pointes. http://crediblemeds.org/ pdftemp/pdf/US-CompositeList.pdf. Accessed May 20, 2013. US Department of Health and Human Services. Guidance for industry format and content of proposed Risk Evaluation and Mitigation Strategies (REMS), REMS assessments, and proposed REMS modifications. http://www.fda.gov/downloads/Drugs/GuidanceComplianceR egulatoryInformation/Guidances/UCM 184128.pdf. Accessed May 20, 2013. European Medicines Agency guidance. Guidance on format of the risk management plan (RMP) in the EU part V: risk minimisation measures. http://www.ema.europa.eu/ loads/Drugs/GuidanceComplianceR docs/en_GB/document_library/Reg ulatory_and_procedural_guideline/ 2012/11/WC500134662.pdf. Accessed May 20, 2013.
272
Address correspondence to: Jasmanda Wu, PhD, MPH, Sanofi, Global Pharmacovigilance and Epidemiology, 55 Corporate Drive, Bridgewater, NJ 08807. E-mail: jasmanda.wu@sanofi.com
Volume 36 Number 2
J. Wu et al.
Supplemental Appendix I. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD9-CM), codes of heart failure. ICD-9-CM Code 428.x 398.91 402.01 402.11 402.91 404.01 404.11 404.91 404.03 404.13 404.93
Description Heart failure Rheumatic heart failure (congestive) Malignant hypertensive heart disease with heart failure Benign hypertensive heart disease with heart failure Unspecified hypertensive heart disease with heart failure Malignant hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Benign hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Unspecified hypertensive heart and chronic kidney disease with heart failure and with chronic kidney disease stage I through stage IV, or unspecified Malignant hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease Benign hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease Unspecified hypertensive heart and chronic kidney disease with heart failure and chronic kidney disease stage V or end-stage renal disease
Supplemental Appendix III. Drugs that prolong QT interval (generic names). Amiodarone
Supplemental Appendix II. Potent cytochrome P450 3A4 inhibitors (generic names). Atazanavir
Itraconazole
Ritonavir
Chloramphenicol Clarithromycin Cyclosporin Darunavir Indinavir
Ketoconazole Lopinavir Nefazodone Nelfinavir Posaconazole
Saquinavir Telithromycin Tipranavir Troleandomycin
February 2014
Arsenic trioxide Astemizole Bepridil Chloroquine Chlorpromazine Cisapride Clarithromycin Disopyramide Dofetilide Domperidone Droperidol Erythromycin Halofantrine
Haloperidol Ibutilide Levomethadyl Mesoridazine Methadone Pentamidine Pimozide Probucol Procainamide Quinidine Sotalol Sparfloxacin Terfenadine Thioridazine
272.e1
Clinical Therapeutics
Supplemental Appendix IV. Current Procedural Terminology (CPT) codes for creatinine testing. CPT Code
Description
80047 80048 80050 80053 80069 82565 82570 82575
Basic metabolic panel (calcium, ionized) Basic metabolic panel General health panel Comprehensive metabolic profile Renal function panel Blood creatinine Creatinine, other source Creatinine, clearance
Supplemental Appendix V. Logical Observation Identifiers Names and Codes (LOINC) for creatinine testing. LOINC 14682-9 21232-4 2160-0 35203-9 38483-4 40248-7 40264-4 44784-7 59826-8
272.e2
Description Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine Creatinine
(moles/volume) in serum or plasma (mass/volume) in arterial blood (mass/volume) in serum or plasma (mass or moles/volume) in serum or plasma (mass/volume) in blood (mass/volume) in serum or plasma, baseline (moles/volume) in serum or plasma, baseline (mass/volume) in maximum serum or plasma (moles/volume) in blood
Volume 36 Number 2