Clinical Therapeutics/Volume ], Number ], 2014
Utilization and Adherence Patterns of Subcutaneously Administered Anti–Tumor Necrosis Factor Treatment Among Rheumatoid Arthritis Patients Joseph Tkacz, MS1; Lorie Ellis, PhD2; Susan C. Bolge, PhD2; Roxanne Meyer, PharmD2; Brenna L. Brady, PhD1; and Charles Ruetsch, PhD1 1
Health Analytics, Columbia, Maryland; and 2Janssen Scientific Affairs, Horsham, Pennsylvania
ABSTRACT Background: Adherence to therapy is a key requirement underlying achievement of clinical outcomes in randomized controlled drug registration trials. In postmarketing studies, comparison of adherence among therapies can become more complicated when drug dosing and administration schedules differ or when methods used to measure adherence are not consistently applied. Objective: The objective of this exploratory study was to investigate a broad range of utilization and adherence outcomes associated with subcutaneous biologic treatments for rheumatoid arthritis (RA). Methods: Adult patients (aged Z18 years) exhibiting Z2 claims with an RA diagnosis (code 714.x), at least 24 months of continuous medical and pharmacy eligibility, and 30-day supplies of adalimumab, etanercept, or golimumab were selected from the Optum Insight Clinformatics database. Adherence and utilization measures were calculated and compared across treatment groups. Results: A total of 1532 adalimumab, 2099 etanercept, and 261 golimumab patients met inclusion criteria. Compared with both adalimumab and etanercept patients, golimumab patients were significantly more likely to have a medication possession ratio of Z0.80 (82% vs 71% vs 62%; P o 0.001) and significantly less likely to have Z4 late medication refills (6.9% vs 17.7% vs 26.1%; P o 0.001 for all). Etanercept patients had significantly greater refill intervals (37.7 vs 34.9 and 35.1 days) and had the lowest proportion of adherent fills (70% vs 77% and 75%) compared with both golimumab and adalimumab patients (P o 0.001 for all). Bivariate effects were reproduced in multivariate models that controlled for treatment duration. Conclusions: A number of statistically significant medication adherence differences were observed
] 2014
among golimumab, adalimumab, and etanercept patients in treatment for RA. Overall, golimumab patients appeared to be the most adherent group. Findings may be partially attributable to golimumab patients’ likely increased disease severity, their prior experience with biologic medication, or golimumab’s once-monthly dosing schedule, which requires fewer administrations than both adalimumab and etanercept. (Clin Ther. 2014;]:]]]–]]]) & 2014 The Authors. Published by Elsevier HS Journals, Inc. All rights reserved. Key words: adalimumab, adherence, claims data, etanercept, golimumab, pharmacoeconomic.
Rheumatoid arthritis (RA) is an autoimmune disease that affects 1% of the global population and is characterized by an inflammation of the joints and surrounding tissues.1 During the past decade, the treatment of RA has improved significantly with the development of a variety of biologic agents targeting tumor necrosis factor (TNF).2 As a result, the use of these agents has significantly increased as either monotherapy or combination therapy with nonbiologic disease-modifying antirheumatic drugs (DMARDs).3 Despite increased utilization, patient adherence to biologic agent use remains unclear.4 Adherence, typically defined as the degree to which a medication is taken as prescribed, encompasses dosage, timing, and frequency.5 A number of recent reviews have highlighted the inconsistent reporting of adherence to biologic treatments,4,6,7 Accepted for publication February 22, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.02.019 0149-2918/$ - see front matter & 2014 The Authors. Published by Elsevier HS Journals, Inc. All rights reserved.
1
Clinical Therapeutics obfuscating any attempt to evaluate and compare the real-world cost or clinical benefits of these agents. For example, one investigation of biologic-treated Medicaid enrollees defined 1-year adherence as a proportion of days covered (PDC) of Z80%, which resulted in 1-year adherence rates of 11% for anakinra patients, 32% for etanercept patients, and 43% for infliximab patients.8 Similarly, another investigation measured adherence using the medication possession ratio (MPR) of Z80% but reported much higher rates of adherence: 63% for new etanercept patients and 65% for new adalimumab patients.9 Brocq et al10 focused on 1-year drug continuation rates and reported rates of 87% for etanercept users, 83% for adalimumab users, and 68% for infliximab users. As these results indicate, current estimates of biologic therapy adherence vary considerably across methods, ranging from 11% to 87%, which can be directly attributable to a lack of uniformity in defining and calculating specific adherence outcomes. An analysis examining a broader range of treatment measures related to patient adherence may serve to ameliorate this issue. Adalimumab,11 etanercept,12 and golimumab13 are 3 anti-TNF biologic agents currently approved for the treatment of RA. Each of these medications is administered via subcutaneous injection and has been reported to be effective in the treatment of RA,14–17 although each has a unique formulation and dosing schedule. The purpose of this exploratory study was to investigate a number of utilization measures associated with adalimumab, etanercept, and golimumab treatment, including dosing, refill intervals, and multiple proxies of adherence, and to compare findings across treatment groups.
METHODS Sample Selection Study data were derived from the Optum Insight Clinformatics database of insured individuals. Data were completely void of identifying information. Medical, pharmacy, and laboratory claims for members with a rheumatic disease diagnosis during quarter 4 of calendar year 2005 through quarter 1 of 2012 were studied. Figure 1 details the study inclusion and exclusion criteria and the sample size remaining after each exclusion criterion imposition. To be eligible for this study, members were required to receive
2
treatment with a biologic during the case finding window: January 1, 2009 through March 31, 2011. Members were retained for analyses if their most recent biologic on record was either adalimumab, etanercept, or golimumab, with the earliest fill for that particular biologic serving as the study index date. Members treated with certolizumab were initially considered to be included in this study, although they were eventually excluded due to underrepresentation relative to other groups (n ¼ 80 [3% of final sample]). This approach ensured that representation of patients was maximized for biologics such as golimumab, which indicated a lower overall prevalence in the data set. All members were required to have at least 2 years of continuous eligibility: 1 year before and 1 year after the index date. Members were also required to have at least 2 claims for rheumatoid arthritis (code 714.x) within 1 year of the index date and no diagnoses of ankylosing spondylitis (code 720.x) or psoriatic arthritis (code 696.x) within that same period. All members must have been at least 18 years of age, and women must not have been pregnant at any time during the study. In addition, members with a comorbid diagnosis of Crohn disease (code 555.x) were excluded because this is an approved indication for adalimumab.11 Because most patients received 30-day supplies of medication (68%), to minimize artifacts, only those members with 30-day supply fills of their index medication across their episode of care were maintained for final analyses. After the imposition of all inclusion and exclusion criteria, the final sample size was 3892 (n ¼ 1532 for the adalimumab group, n ¼ 2099 for the etanercept group, and n ¼ 261 for the golimumab group).
Measures Patient demographic characteristics were summarized from the membership table and included age, sex, geographic region of residence, insurance line of business, and type of benefit plan. In addition, the Charlson Comorbidity Index, an overall measure of health,18 was calculated during the 1-year preindex period, as were the rates for a variety of other comorbidities of interest. Prior biologic utilization during the 1-year preindex period was also calculated. The following section outlines the primary treatment outcomes, which were calculated for all
Volume ] Number ]
J. Tkacz et al.
n = 353,8333 Identify index golimumab, adalimumab, or etanercept fill during the case-finding period (1/1/09 –3/31/11)
n = 24,851
Must have at least two fills for the index biologic treatment n = 22,508 Continuously eligible for 2 years around earliest treatment (index) date
n = 10,632
2 diagnoses of rheumatoid arthritis and the absence of other rheumatic disorders within 1 year of index date
n = 5998
At least 18 years old n = 5848 Absence of pregnancy n = 5801 Absence of Crohn disease diagnosis n = 5737 Fills for 30 days' supply of medication
Final Sample
n = 3892 Place into final study groups Adalimumab
Etanercept
n = 1532
n = 2099
Golimumab n = 261
Figure 1. Sample attrition. *Starting sample consisted of members with a single diagnosis of rheumatoid arthritis, ankylosing spondylitis, or psoriatic arthritis between quarter 4 of 2005 and quarter 1 of 2012.
members who qualified for the analyses. All treatment outcomes were derived from the pharmacy claims table and were calculated for the period the patient continued to take their assigned study medication. Patients were followed up until discontinuation or until a period of Z90 days without a fill for the index medication was observed. Patients were followed up for longer than 1 year if they maintained eligibility and continually received treatment (ie, did not have evidence of a lapse in treatment). Adalimumab, etanercept, and golimumab treatment fills were identified via brand name and were summed for each member. The corresponding days’
] 2014
supply value for these treatments was also summed by member. The MPR was calculated by summing the days’ supply value of the index treatment and dividing this amount by the duration of treatment, which was the sum of days between the index date and the most recent index treatment fill plus the days’ supply value of the last fill.19 The MPR values 41 were recoded into 1, and the proportion of patients with an MPR Z0.60 and Z0.80 were reported. The PDC was also calculated, which is another measure of days in which medication was on hand, but adjusts for doublecounting of covered days.19 As with MPR, the proportion of patients with a PDC Z0.60 and
3
Clinical Therapeutics Z0.80 were reported. Medication dosages were extracted from the strength field, and mean patient dosages were calculated. Similarly, mean refill intervals were also calculated for each patient, in addition to summing the total number of early, late, and adherent fills. An adherent fill interval was defined as a refill appearing Z21 and r38 days from the previous fill. A refill interval of r15 days was categorized as early, whereas a refill interval Z45 days was categorized as late.
Statistical Analyses Bivariate group comparisons were initially conducted for all demographic and treatment measures. χ2 tests of equality of proportions were used for categorical variables, and 1-way analysis of variance was used for continuous variables. Turkey’s post hoc tests were conducted to examine group differences where omnibus test results were statistically significant. Outcomes that resulted in statistically significant bivariate group effects were entered into multivariate analyses that controlled for demographic characteristics, health status, and treatment duration. Ordinary least squares regression models were used for continuous variables, while dichotomous, categorical variables were regressed onto predictors via logistic models. The golimumab group served as the reference treatment category in all models. Data management and analyses were conducted using SPSS statistical software, version 20 (SPSS Inc, Chicago, Illinois).
RESULTS Members were predominantly female, (75.5%), with a mean (SD) age of 51.1 (11.2) years, residing predominantly in the mid-American (39.0%) and Southeastern (39.0%) regions of the United States. Nearly the entire sample was commercially insured (99.5%), with most holding a point-of-service benefit package (69.1%). The mean (SD) Charlson Comorbidity Index score was 0.52 (0.95). Table I lists the demographic characteristics by treatment group. Golimumab members were 43 times as likely to evidence prior biologic exposure as adalimumab members and 46 times as likely as etanercept members (P o 0.001 for all). Nearly 60% of the golimumab group had previous experience with a biologic. The most commonly used prior biologic for adalimumab members was etanercept (12.3%), with adalimumab
4
(6.1%) being the most commonly used for etanercept members. Golimumab members also tended to have the highest comorbid disease burden of the 3 groups, evidenced by a significantly higher proportion with a diagnosis of depression, osteoarthritis, chronic pain, cerebrovascular disease, and hypertension compared with both the adalimumab and etanercept groups (P o 0.05 for all). Both the golimumab and etanercept groups had a significantly higher mean Charlson Comorbidity Index score than the adalimumab patients (P o 0.05). There were no statistically significant treatment group differences for the remaining demographic indicators. Table II gives the group descriptive statistics and results of bivariate comparisons for the primary treatment outcomes. Compared with both the adalimumab and golimumab groups, the etanercept group had a significantly greater number of fills (19.4 vs 18.3 vs 13.1) and cumulative days’ supply of treatment (546 vs 516 vs 384), whereas the adalimumab group had significantly greater fills and days’ supply than the golimumab group (P o 0.001 for all). All 3 groups significantly differed from each other on both the MPR and PDC measures, with the golimumab group (MPR ¼ 0.88, PDC ¼ 0.87) having significantly higher values than both the adalimumab (MPR ¼ 0.85, PDC ¼ 0.84) and etanercept groups (MPR ¼ 0.81, PDC ¼ 0.81) and with the adalimumab group having significantly higher values than the etanercept group (P o 0.001 for all). The same pattern of results was obtained when groups were compared on the proportion with MPR and PDC values Z0.80 (P o 0.001 for all). Further, compared with the etanercept group, the golimumab and adalimumab groups evidenced significantly shorter refill intervals (34.9 and 35.1 vs 37.7 days) and a higher mean proportion of treatment fills categorized as adherent (77% and 75% vs 70%; P o 0.001 for all). Groups did not significantly differ on the proportion with an early refill, but compared with the adalimumab and etanercept groups, the golimumab group was significantly less likely to have at least 1 late refill (62.8% vs 71.5% and 74.4%). Further, the golimumab group was significantly less likely to have Z4 late refills compared with both the adalimumab and etanercept groups, whereas the adalimumab group was significantly less likely than the etanercept group (6.9% vs 17.7% vs 26.1%; P o 0.001 for all).
Volume ] Number ]
J. Tkacz et al.
Table I. Patient demographic characteristics. No. (%) of Patients* Characteristic Female Geographic region Mid-American Northeast Southeast West Benefit plan business type Commercial Medicaid Medicare Type of benefit plan EPO HMO IND OTH POS PPO Age, y Mean (SD) Median Patients o65 years old Prior biologics use Any prior biologic use†,‡,§ 1 Prior treatment 2 Prior treatments 3 Prior treatments Specific treatments Golimumab Adalimumab Etanercept Abatacept Infliximab Certolizumab Anakinra Rituximab Tocilizumab None Select preperiod comorbidities Cerebrovascular disease†,‡ Chronic renal disease Diabetes (without complications)
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
1157 (75.5)
1572 (74.9)
211 (80.8)
600 86 619 227
(39.2) (5.6) (40.4) (14.8)
824 163 798 314
(39.3) (7.8) (38.0) (15.0)
93 16 102 50
P 0.108 0.082
(35.6) (6.1) (39.1) (19.2) 0.475
1524 (99.5) 8 (0.5) 0
2087 (99.4) 12 (0.6) 0
261 (100.0) 0 0
229 163 15 1 1073 51
325 211 46 5 1441 71
(15.5) (10.1) (2.2) (0.2) (68.7) (3.4)
44 (16.9) 22 (8.4) 2 (0.8) 0 176 (67.4) 17 (6.5)
50.7 (11.0) 52 1435 (93.7)
51.4 (11.4) 53 1932 (92.0)
50.6 (10.6) 52 246 (94.3)
249 (16.3) 236 (15.4) 13 (0.8) 0
187 (8.9) 174 (8.3) 13 (0.6)
9 (0.6)
6 (0.3) 128 (6.1)
0.026 (14.9) (10.6) (1.0) (0.1) (70.0) (3.3)
0.176
155 129 23 3
(59.4) (49.4) (8.8) (1.1)
(12.3) (1.2) (2.2) (0.4) (0.2) (0.2) 0 1283 (84.6)
20 (1.0) 37 (1.8) 2 (0.1) 0 7 (0.3) 0 1912 (92.0)
(23.8) (24.5) (10.0) (7.7) (2.3) 0 6 (2.3) 0 106 (39.5)
41 (2.7) 20 (1.3) 127 (8.3)
56 (2.7) 43 (2.0) 199 (9.5)
14 (5.4) 8 (3.1) 29 (11.1)
189 18 34 6 3 3
0.113 o0.001
62 64 26 20 6
0.041 0.077 0.240 (continued)
] 2014
5
Clinical Therapeutics Table I. (continued). No. (%) of Patients* Characteristic Disorders of lipid metabolism Depression†,‡ Osteoarthritis†,‡ Chronic pain†,‡ Hypertension†,‡ Charlson Comorbidity Index score, mean (SD)†,§
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
P
501 (32.7) 162 (10.6) 849 (55.4) 58 (3.8) 512 (33.4) 0.47 (0.86)
696 (33.2) 197 (9.4) 1138 (54.2) 73 (3.5) 707 (33.7) 0.54 (0.99)
97 (37.2) 41 (15.7) 165 (63.2) 25 (9.6) 110 (42.1) 0.59 (1.10)
0.365 0.006 0.022 o0.001 0.018 0.023
EPO ¼ exclusive provider organization; HMO ¼ health maintenance organization; IND ¼ indemnity; OTH ¼ other; POS ¼ point of service; PPO ¼ preferred provider organization. * Data are presented as number (percentage) of patients unless otherwise indicated. † Simponi a adalimumab. ‡ Simponi a etanercept. § Adalimumab a etanercept.
Table III and Table IV give the results of ordinary least squares and logistic models of treatment outcomes that controlled for treatment duration and patient characteristics. Men received significantly more fills and days’ supply of treatment compared with women and were more adherent as measured by both the MPR and PDC (P o 0.05 for all). Age was positively associated with treatment fills, days’ supply, MPR, and PDC but inversely related with mean refill interval (P o 0.05 for all). Treatment duration was positively associated with all outcomes, excluding the numeric MPR value (P o 0.05 for all). Health status, as measured by the Charlson Comorbidity Index, and geographic region of residence did not relate to any treatment outcomes (P 4 0.05 for all). As for the main predictor of interest, treatment group, most bivariate effects were reproduced in multivariate models. Adalimumab members were only half as likely as golimumab members to achieve an MPR or PDC Z0.80, whereas etanercept members were only one-third as likely (P o 0.01 for all). Both the adalimumab and etanercept groups had significantly greater mean refill intervals than the golimumab group (P o 0.05), with etanercept members being more than twice as likely as golimumab members to have Z4 late refills (P ¼ .001). Finally, both the adalimumab and etanercept members were significantly less likely to have prior biologic exposure (P o 0.001 for all).
6
DISCUSSION The objectives of the present study were to examine utilization and adherence measures for 3 subcutaneously administered anti-TNF agents indicated for the treatment of RA (adalimumab, etanercept, and golimumab) and to compare treatment outcomes across groups. Golimumab comprised the smallest study group, accounting for only 7% of the overall sample, with the adalimumab (39%) and etanercept (54%) groups making up most of patients. Overall, the sample appeared to be receiving treatment within recommended prescribing guidelines because mean dosages and refill intervals approximated expected values. In addition, multiple adherence measures were assessed during the study period. Data were adequate because most of the sample received treatment for 41 year and evidenced elevated values for both the MPR and PDC. Interestingly, numerous treatment utilization and adherence differences were found across the study groups. Etanercept use was the most prevalent of the 3 biologics studied, whereas golimumab utilization was, as expected, the lowest because it was the latest to be approved and the last commercially available of the 3 agents. Golimumab patients had significantly higher rates of medication adherence as measured by the MPR and PDC compared with both adalimumab and etanercept patients and were also less likely to have
Volume ] Number ]
J. Tkacz et al.
Table II. Treatment outcomes. No. (%) of Patients* Outcome
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
Total fills 28,058 (100.0) 40,701 (100.0) 3409 (100.0) 18.3 (11.9) 19.4 (12.0) 13.1 (7.8) Fills per member, mean (SD)†,‡,§ Total days’ supply, mean (SD)†,‡,§ 516 (334) 546 (338) 384 (230) 0.85 (0.13) 0.81 (0.15) 0.88 (0.11) MPR, mean (SD)†,‡,§ 1080 (70.5) 1298 (61.8) 214 (82.0) MPR Z0.80†,‡,§ MPR Z0.60‡,§ 1435 (93.7) 1861 (88.7) 252 (96.6) 0.84 (0.13) 0.81 (0.15) 0.87 (0.10) PDC, mean (SD)†,‡,§ PDC Z0.80 1073 (70.0) 1284 (61.2) 212 (81.2) †,‡,§ 1432 (93.5) 1860 (88.6) 252 (96.6) PDC Z0.60 Patient dose, mean (SD), mg‡,§ 87.0 (20.2) 193.3 (24.7) 52.2 (10.5) 35.1 (9.3) 37.7 (12.0) 34.9 (8.0) Patient refill interval, mean (SD), d‡,§ 0.75 (0.24) 0.70 (0.28) 0.77 (0.23) Patient proportion of adherent fills, mean (SD)‡,§ Early refill 47 (3.1) 55 (2.6) 8 (3.1) 1095 (71.5) 1561 (74.4) 164 (62.8) Late refill†,‡ 271 (17.7) 547 (26.1) 18 (6.9) Z4 Late refills†,‡,§
P o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 0.703 o0.001 o0.001
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered. * Data are presented as number (percentage) of patients unless otherwise indicated. † Golimumab a adalimumab. ‡ Golimumab a etanercept. § Adalimumab a etanercept.
multiple, late refills. The golimumab group was also the most experienced with biologic medication, with nearly 60% of this group having prior treatment with a biologic, nearly 5 times greater than adalimumab and etanercept patients. Golimumab and adalimumab patients both had briefer refill intervals and a higher percentage of adherent fills compared with etanercept patients. Group effects were maintained in multivariate models that controlled for the duration of therapy and patient characteristics, indicating that these group differences do not account for differences in adherence measures. Overall, golimumab patients appeared to be the most adherent group during the study period, despite having the poorest overall health. This is the first known study to report greater real-world adherence for golimumab compared with 2 established biologics. Differences in clinical effectiveness across adalimumab, etanercept, and golimumab have not been studied directly, although the approval of these agents was based
] 2014
on meeting similar primary outcomes in clinical trials.20–22 There are a number of plausible interpretations of the present results. Golimumab was the only biologic agent in the present study with a once-monthly dosing schedule. Multiple studies of adherence across a variety of medication classes have reported the inverse relationship between dosing frequency and adherence,23–25 and patients have been found to prefer longer vs shorter dosing intervals when given the option.26–28 In the present study, this relationship appeared evident for a number of adherence measures, including the MPR, PDC, and proportion with Z4 late refills, because patients treated with once-monthly golimumab evidenced significantly improved outcomes on these measures compared with bimonthly treated adalimumab patients, who in turn evidenced significant improvements compared with weeklytreated etanercept patients. In addition, increased medication adherence has been associated with
7
8
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered.
0.07 0.03 0.01 o0.01 0.03 0.01 o0.01 1.89 0.68 0.01 o0.01 0.07 0.01 o0.01 0.06 0.01 o0.01 4.58 0.67 o0.01 o0.01 0.01 0.01 0.02 0.01 0.01 0.03 0.37 0.37 0.32 o0.01 0 0 o0.01 0 0 o0.01 0.12 0.01 o0.01 0.80 0 0 0.74 0 0 0.71 0.03 0.18 0.87 o0.01 0 0 0.12 0 0 0.02 0 0 o0.01 0.46 0.01 0.01 0.38 0.01 0.01 0.48 0.59 0.67 0.38 0.83 0 0.01 0.81 0 0.01 0.75 0.06 0.67 0.92 0.67 0.01 0.01 0.50 0.01 0.01 0.56 0.86 0.74 0.25 0 0.76 0.02 0 0.75 0.02 0 15.57 1.13 0 Adalimumab 0.17 0.23 0.47 11.54 6.42 Etanercept 0.62 0.23 0.01 34.20 6.33 Male 0.43 0.13 o0.01 12.31 3.56 Age 0.05 0.01 o0.01 1.36 0.14 Charlson Comorbidity Index score 0.01 0.06 0.82 0.42 1.67 Treatment duration 0.03 0 o0.01 0.85 0 Mid-American region 0.16 0.23 0.47 4.68 6.35 Southeast region 0.14 0.23 0.54 1.40 6.35 West region 0.13 0.25 0.61 2.99 7.05 Constant 2.65 0.38 0 59.02 10.78
P SE β SE P SE Predictor
β
SE
P
β
SE
P
β
MPR Days’ Supply Treatment Fills
Table III. Multivariate analyses of treatment outcomes in ordinary least squares regression models.
β
PDC
P
Mean Refill Interval
Clinical Therapeutics increased disease severity,29 and although disease severity cannot be accurately measured in claims data alone, the golimumab group’s increased rate of both prior biologic use and overall number of comorbidities indicate a potentially more severe disease state. Further, the belief that medication is necessary for disease treatment is also a positive predictor of adherence,30 and the golimumab group’s increased prior exposure to biologic treatment compared with the other groups may have strengthened their belief in the medical necessity of the current treatment regimen. Regardless, results suggest golimumab may be the ideal treatment option for patients with difficulty maintaining adherence, although findings should be replicated in a biologicnaive sample to solidify the relationship between golimumab treatment and adherence. Although there is evidence suggesting that adherence to biologic medication is less than optimal,31 which may be partly due to inconsistent reporting and defining of relevant measures as discussed in the opening of the article, results of the present study revealed generally positive adherence outcomes for all 3 treatment groups. Adherence has been linked to improved outcomes across a variety of chronic diseases states,32 including RA.33,34 In addition, persistence to biologic therapy among RA patients has also been associated with a reduction in non–pharmacy-related health care expenditure.35 Further study of RA patient motivation toward treatment adherence is needed, particularly for biologics. A number of factors related to improved medication adherence have previously been identified in RA patients, including a satisfactory level of patient-physician communication, an increased knowledge of RA in general, and older age,36 whereas nonwhite race and low educational level have been found to be related to decreased adherence.37 In addition, studies have found that patients typically prefer subcutaneous injection over intravenous and intramuscular injections,38,39 which may promote adherence in itself. A better understanding of the drivers of treatment adherence may reduce the amount of time patients spend on ineffective treatments and, given the higher expenses typically associated with biologic treatment,40,41 may have significant economic implications. Given the increasing amount of clinical and economic data being collected in health care today, these data can be used to promote a patient-centered model of care in which
Volume ] Number ]
J. Tkacz et al.
Table IV. Multivariate analyses of treatment outcomes in logistic models. MPR Z0.80 Predictor Adalimumab Etanercept Male Age Charlson Comorbidity Index score Treatment duration Mid-American region Southeast region West region Constant
β
Exp (B)
P
PDC Z0.80 β
Exp (B)
Z4 Late Refills P
β
Exp (B)
P
Prior Biologic Exposure β
Exp (B)
P
0.72 1.16 0.20 0.03 0.06
0.49 o0.01 0.69 0.50 o0.01 0.45 1.57 0.09 1.90 0.15 o0.01 0.31 o0.01 1.14 0.32 o0.01 0.85 2.34 o0.01 2.55 0.08 o0.01 1.22 0.02 0.20 1.22 0.02 0.33 0.72 o0.01 0.12 0.88 0.28 1.03 o0.01 0.02 1.02 o0.01 0.03 0.97 o0.01 0.01 0.99 0.09 1.06 0.12 0.07 1.07 0.09 0.04 1.04 0.39 0.05 1.05 0.37
0 0.01 0.04 0.03 0.07
1.00 o0.01 0 1.00 o0.01 0.99 0.93 0.03 1.03 0.84 0.96 0.77 0.04 0.96 0.80 0.97 0.84 0 1.00 0.98 1.08 0.78 0.03 1.03 0.90
0 0.02 0.10 0.17 2.97
1.00 o0.01 0 1.00 o0.01 1.02 0.90 0.13 0.88 0.50 1.10 0.61 0.30 0.74 0.12 1.19 0.39 0.19 0.83 0.38 0.05 0 1.38 3.97 0
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered.
much improved disease management is possible.42 Analyses similar to those appearing in the present study can be used by prescribers to monitor patient treatment adherence, in addition to health plans for purposes of avoiding unnecessary spend. The present study has a number of limitations. The primary limitation is the potential endogeneity of the between-groups factor (treatment), meaning that there could be other unmeasured factors that drive utilization and adherence patterns associated with each medication. Treatment groups were not matched to variations in patient characteristics among the groups, and data have indicated that differences in these factors may confound study conclusions. Although multivariate models that control for patient characteristics were conducted, an analysis using propensity score matching or instrumental variables may be of some utility. An additional limitation was the data source (administrative claims data), which does not allow for the assessment of clinical outcomes related to adherence and may also include administrative coding errors.43 Finally, the MPR and PDC were calculated based on the length of each individual patients’ treatment episode, as opposed to a fixed interval (eg, 1 year). Given the cycling-forremission strategy used in RA biologic treatment,44 use of a fixed interval can result in a deflated estimate of
] 2014
adherence because patients may adhere for a period only to discontinue use shortly thereafter because of ineffectiveness or adverse effects. In conclusion, the dosing and scheduling complexities of biologic treatment have been found to be problematic for the standardized measurement and assessment of medication adherence. The present study was designed to measure multiple indicators of biologic administration process indicators among a sample of RA patients treated with subcutaneously administered anti-TNFs. Adalimumab, etanercept, and golimumab patients were found to be generally adherent with their treatment regimens because dosages and refill intervals were in agreement with recommended prescribing guidelines and proxies for medication adherence were in the appropriate range. Utilization of golimumab was significantly lower than that of adalimumab and etanercept, although golimumab patients appeared to be the most adherent during their treatment despite also being in the poorest overall health. Specifically, golimumab patients had significantly higher MPRs and PDCs compared with the adalimumab and etanercept patients and were also less likely to have multiple, late refills. The favorable adherence results achieved by the golimumab group may be a result of these patients’ likely increased
9
Clinical Therapeutics disease severity, their prior experience with biologic medication, or golimumab’s once-monthly dosing schedule. Overall, study findings were encouraging but indicate the need for additional comparative effectiveness studies between golimumab and the established biologics to confirm adherence results and to quantify the precise clinical and economic benefits associated with biologic adherence.
CONFLICTS OF INTEREST Janssen Scientific Affairs funded this study. Drs Ellis, Meyer, and Bolge are employees of Janssen Scientific Affairs and are stockholders in Johnson and Johnson. Study sponsors assisted with study design, interpretation of results, and manuscript development.
REFERENCES 1. Gibofsky A. Overview of epidemiology, pathophysiology, and diagnosis of rheumatoid arthritis. Am J Manag Care. 2012;18(13 Suppl):S295–S302. 2. Scott DL. Biologics-based therapy for the treatment of rheumatoid arthritis. Clin Pharmacol Ther. 2012;91:30–43. 3. Yazici Y, Shi N, John A. Utilization of biologic agents in rheumatoid arthritis in the United States: analysis of prescribing patterns in 16,752 newly diagnosed patients and patients new to biologic therapy. Bull NYU Hosp Jt Dis. 2008;66:77–85. 4. Salt E, Frazier SK. Adherence to disease-modifying antirheumatic drugs in patients with rheumatoid arthritis: a narrative review of the literature. Orthop Nurs. 2010;29:260–275. 5. Cramer JA, Roy A, Burrell A, et al. Medication compliance and persistence: terminology and definitions. Value Health. 2008;11:44–47. 6. de Achaval S, Suarez-Almazor ME. Treatment adherence to disease-modifying antirheumatic drugs in patients with rheumatoid arthritis and systemic lupus erythematosus. Int J Clin Rheumtol. 2010;5:313–326. 7. Malaviya AP, Ostor AJ. Drug adherence to biologic DMARDS with a special emphasis on the benefits of subcutaneous abatacept. Patient Prefer Adherence. 2012;6:589–596. 8. Li P, Blum MA, Von Feldt J, Hennessy S, Doshi JA. Adherence, discontinuation, and switching of biologic therapies in medicaid enrollees with rheumatoid arthritis. Value Health. 2010;13:805–812. 9. Borah BJ, Huang X, Zarotsky V, Globe D. Trends in RA patients’ adherence to subcutaneous anti-TNF therapies and costs. Curr Med Res Opin. 2009;25:1365–1377. 10. Brocq O, Roux CH, Albert C, et al. TNF-α antagonist continuation rates in 442 patients with inflammatory joint disease. Joint Bone Spine. 2007;74:148–154.
10
11. Food and Drug Administration. Highlights of Prescribing Information: Humira. 2011. http://www.accessdata.fda. gov/drugsatfda_docs/label/2011/125057s0276lbl.pdf. Accessed May 16, 2013. 12. Food and Drug Administration. Enbrels (etanercept) For Subcutaneous Injection. 2003. http://www.accessdata. fda.gov/drugsatfda_docs/label/2003/etanimm060503LB. pdf. Accessed May 16, 2013. 13. Food and Drug Administration. Highlights of Prescribing Information: Simponi. 2009. http://www.accessdata. fda.gov/drugsatfda_docs/label/2009/125289s000lbl.pdf. Accessed May 16, 2013. 14. Dhillon S, Lyseng-Williamson KA, Scott LJ. Etanercept: a review of its use in the management of rheumatoid arthritis. Drugs. 2007;67:1211–1241. 15. Keystone EC, Genovese MC, Klareskog L, et al. Golimumab, a human antibody to tumour necrosis factor α given by monthly subcutaneous injections, in active rheumatoid arthritis despite methotrexate therapy: the GO-FORWARD Study. Ann Rheum Dis. 2009;68:789–796. 16. Wiens A, Correr CJ, Venson R, Otuki MF, Pontarolo R. A systematic review and meta-analysis of the efficacy and safety of adalimumab for treating rheumatoid arthritis. Rheumatol Int. 2010;30:1063–1070. 17. Zhou H, Jang H, Fleischmann RM, et al. Pharmacokinetics and safety of golimumab, a fully human anti-TNF-alpha monoclonal antibody, in subjects with rheumatoid arthritis. J Clin Pharmacol. 2007;47:383–396. 18. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40: 373–383. 19. Peterson AM, Nau DP, Cramer JA, Benner J, GwadrySridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007;10:3–12. 20. Emery P, Fleischmann RM, Moreland LW, et al. Golimumab, a human anti-tumor necrosis factor alpha monoclonal antibody, injected subcutaneously every four weeks in methotrexate-naive patients with active rheumatoid arthritis: twenty-four-week results of a phase III, multicenter, randomized, double-blind, placebo-controlled study of golimumab before methotrexate as first-line therapy for earlyonset rheumatoid arthritis. Arthritis Rheum. 2009;60: 2272–2283. 21. Moreland LW, Schiff MH, Baumgartner SW, et al. Etanercept therapy in rheumatoid arthritis: a randomized, controlled trial. Ann Intern Med. 1999;130:478–486. 22. Weinblatt ME, Keystone EC, Furst DE, et al. Adalimumab, a fully human anti-tumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial. Arthritis Rheum. 2003;48:35–45.
Volume ] Number ]
J. Tkacz et al. 23. Coleman CI, Limone B, Sobieraj DM, et al. Dosing frequency and medication adherence in chronic disease. J Manag Care Pharm. 2012; 18:527–539. 24. Iskedjian M, Einarson TR, MacKeigan LD, et al. Relationship between daily dose frequency and adherence to antihypertensive pharmacotherapy: evidence from a meta-analysis. Clin Ther. 2002;24:302–316. 25. Saini SD, Schoenfeld P, Kaulback K, Dubinsky MC. Effect of medication dosing frequency on adherence in chronic diseases. Am J Manag Care. 2009;15:e22–e33. 26. Granger AL, Fehnel SE, Hogue SL, Bennett L, Edin HM. An assessment of patient preference and adherence to treatment with Wellbutrin SR: a web-based survey. J Affect Disord. 2006;90:217–221. 27. Kendler D, Kung AW, Fuleihan Gel H, et al. Patients with osteoporosis prefer once weekly to once daily dosing with alendronate. Maturitas. 2004;48:243–251. 28. Reginster JY, Rabenda V, Neuprez A. Adherence, patient preference and dosing frequency: understanding the relationship. Bone. 2006;38(4 Suppl 1): S2–S6. 29. Gao X, Nau DP, Rosenbluth SA, Scott V, Woodward C. The relationship of disease severity, health beliefs and medication adherence among HIV patients. AIDS Care. 2000;12:387–398. 30. Pasma A, van’t Spijker A, Hazes JM, Busschbach JJ, Luime JJ. Factors associated with adherence to pharmaceutical treatment for rheumatoid arthritis patients: a systematic review. Semin Arthritis Rheum. 2013; 43:18–28. 31. Koncz T, Pentek M, Brodszky V, Ersek K, Orlewska E, Gulacsi L. Adherence to biologic DMARD therapies in rheumatoid arthritis. Expert Opin Biol Ther. 2010;10:1367–1378. 32. DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a
] 2014
33.
34.
35.
36.
37.
meta-analysis. Med Care. 2002;40: 794–811. Contreras-Yanez I, Ponce De Leon S, Cabiedes J, Rull-Gabayet M, Pascual-Ramos V. Inadequate therapy behavior is associated to disease flares in patients with rheumatoid arthritis who have achieved remission with disease-modifying antirheumatic drugs. Am J Med Sci. 2010; 340:282–290. Pascual-Ramos V, Contreras-Yanez I, Villa AR, Cabiedes J, Rull-Gabayet M. Medication persistence over 2 years of follow-up in a cohort of early rheumatoid arthritis patients: associated factors and relationship with disease activity and with disability. Arthritis Res Ther. 2009;11:R26. Tang B, Rahman M, Waters HC, Callegari P. Treatment persistence with adalimumab, etanercept, or infliximab in combination with methotrexate and the effects on health care costs in patients with rheumatoid arthritis. Clin Ther. 2008;30:1375–1384. Viller F, Guillemin F, Briancon S, Moum T, Suurmeijer T, van den Heuvel W. Compliance to drug treatment of patients with rheumatoid arthritis: a 3 year longitudinal study. J Rheumatol. 1999;26:2114–2122. Garcia-Gonzalez A, Richardson M, Garcia Popa-Lisseanu M, et al. Treatment adherence in patients with rheumatoid arthritis and systemic lupus erythematosus. Clin Rheumatol. 2008;27:883–889.
38. Williams EL, Edwards CJ. Patient preferences in choosing anti-TNF therapies-R1. Rheumatology (Oxford). 2006;45:1575–1576. 39. Jeffery M, Stokes A, Johnson K, Anderson D, Harper B. Patient Preferences for Tumor Necrosis Factor (TNF) Antagonist. London, UK: The Research Partnership Ltd, Rheumatology Research International; 2004. 40. Modena V, Bianchi G, Roccatello D. Cost-effectiveness of biologic treatment for rheumatoid arthritis in clinical practice: an achievable target? Autoimmun Rev. 2013;12:835–838. 41. van der Velde G, Pham B, Machado M, et al. Cost-effectiveness of biologic response modifiers compared to disease-modifying antirheumatic drugs for rheumatoid arthritis: a systematic review. Arthritis Care Res (Hoboken). 2011;63:65–78. 42. Chawla NV, Davis DA. Bringing big data to personalized healthcare: a patient-centered framework. J Gen Intern Med. 2013;28(Suppl 3):S660– S665. 43. Tyree PT, Lind BK, Lafferty WE. Challenges of using medical insurance claims data for utilization analysis. Am J Med Qual. 2006;21:269– 275. 44. Ramiro S, Machado P, Singh JA, Landewe RB, da Silva JA. Applying science in practice: the optimization of biological therapy in rheumatoid arthritis. Arthritis Res Ther. 2010;12: 220.
Address correspondence to: Joseph Tkacz, MS, Health Analytics, LLC, 9200 Rumsey Rd, Suite 215, Columbia, MD. E-mail: joseph.tkacz@ healthanalytic.com
11
Utilization and Adherence Patterns of Subcutaneously Administered Anti–Tumor Necrosis Factor Treatment Among Rheumatoid Arthritis Patients Joseph Tkacz, MS1; Lorie Ellis, PhD2; Susan C. Bolge, PhD2; Roxanne Meyer, PharmD2; Brenna L. Brady, PhD1; and Charles Ruetsch, PhD1 1
Health Analytics, Columbia, Maryland; and 2Janssen Scientific Affairs, Horsham, Pennsylvania
ABSTRACT Background: Adherence to therapy is a key requirement underlying achievement of clinical outcomes in randomized controlled drug registration trials. In postmarketing studies, comparison of adherence among therapies can become more complicated when drug dosing and administration schedules differ or when methods used to measure adherence are not consistently applied. Objective: The objective of this exploratory study was to investigate a broad range of utilization and adherence outcomes associated with subcutaneous biologic treatments for rheumatoid arthritis (RA). Methods: Adult patients (aged Z18 years) exhibiting Z2 claims with an RA diagnosis (code 714.x), at least 24 months of continuous medical and pharmacy eligibility, and 30-day supplies of adalimumab, etanercept, or golimumab were selected from the Optum Insight Clinformatics database. Adherence and utilization measures were calculated and compared across treatment groups. Results: A total of 1532 adalimumab, 2099 etanercept, and 261 golimumab patients met inclusion criteria. Compared with both adalimumab and etanercept patients, golimumab patients were significantly more likely to have a medication possession ratio of Z0.80 (82% vs 71% vs 62%; P o 0.001) and significantly less likely to have Z4 late medication refills (6.9% vs 17.7% vs 26.1%; P o 0.001 for all). Etanercept patients had significantly greater refill intervals (37.7 vs 34.9 and 35.1 days) and had the lowest proportion of adherent fills (70% vs 77% and 75%) compared with both golimumab and adalimumab patients (P o 0.001 for all). Bivariate effects were reproduced in multivariate models that controlled for treatment duration. Conclusions: A number of statistically significant medication adherence differences were observed
] 2014
among golimumab, adalimumab, and etanercept patients in treatment for RA. Overall, golimumab patients appeared to be the most adherent group. Findings may be partially attributable to golimumab patients’ likely increased disease severity, their prior experience with biologic medication, or golimumab’s once-monthly dosing schedule, which requires fewer administrations than both adalimumab and etanercept. (Clin Ther. 2014;]:]]]–]]]) & 2014 The Authors. Published by Elsevier HS Journals, Inc. All rights reserved. Key words: adalimumab, adherence, claims data, etanercept, golimumab, pharmacoeconomic.
Rheumatoid arthritis (RA) is an autoimmune disease that affects 1% of the global population and is characterized by an inflammation of the joints and surrounding tissues.1 During the past decade, the treatment of RA has improved significantly with the development of a variety of biologic agents targeting tumor necrosis factor (TNF).2 As a result, the use of these agents has significantly increased as either monotherapy or combination therapy with nonbiologic disease-modifying antirheumatic drugs (DMARDs).3 Despite increased utilization, patient adherence to biologic agent use remains unclear.4 Adherence, typically defined as the degree to which a medication is taken as prescribed, encompasses dosage, timing, and frequency.5 A number of recent reviews have highlighted the inconsistent reporting of adherence to biologic treatments,4,6,7 Accepted for publication February 22, 2014. http://dx.doi.org/10.1016/j.clinthera.2014.02.019 0149-2918/$ - see front matter & 2014 The Authors. Published by Elsevier HS Journals, Inc. All rights reserved.
1
Clinical Therapeutics obfuscating any attempt to evaluate and compare the real-world cost or clinical benefits of these agents. For example, one investigation of biologic-treated Medicaid enrollees defined 1-year adherence as a proportion of days covered (PDC) of Z80%, which resulted in 1-year adherence rates of 11% for anakinra patients, 32% for etanercept patients, and 43% for infliximab patients.8 Similarly, another investigation measured adherence using the medication possession ratio (MPR) of Z80% but reported much higher rates of adherence: 63% for new etanercept patients and 65% for new adalimumab patients.9 Brocq et al10 focused on 1-year drug continuation rates and reported rates of 87% for etanercept users, 83% for adalimumab users, and 68% for infliximab users. As these results indicate, current estimates of biologic therapy adherence vary considerably across methods, ranging from 11% to 87%, which can be directly attributable to a lack of uniformity in defining and calculating specific adherence outcomes. An analysis examining a broader range of treatment measures related to patient adherence may serve to ameliorate this issue. Adalimumab,11 etanercept,12 and golimumab13 are 3 anti-TNF biologic agents currently approved for the treatment of RA. Each of these medications is administered via subcutaneous injection and has been reported to be effective in the treatment of RA,14–17 although each has a unique formulation and dosing schedule. The purpose of this exploratory study was to investigate a number of utilization measures associated with adalimumab, etanercept, and golimumab treatment, including dosing, refill intervals, and multiple proxies of adherence, and to compare findings across treatment groups.
METHODS Sample Selection Study data were derived from the Optum Insight Clinformatics database of insured individuals. Data were completely void of identifying information. Medical, pharmacy, and laboratory claims for members with a rheumatic disease diagnosis during quarter 4 of calendar year 2005 through quarter 1 of 2012 were studied. Figure 1 details the study inclusion and exclusion criteria and the sample size remaining after each exclusion criterion imposition. To be eligible for this study, members were required to receive
2
treatment with a biologic during the case finding window: January 1, 2009 through March 31, 2011. Members were retained for analyses if their most recent biologic on record was either adalimumab, etanercept, or golimumab, with the earliest fill for that particular biologic serving as the study index date. Members treated with certolizumab were initially considered to be included in this study, although they were eventually excluded due to underrepresentation relative to other groups (n ¼ 80 [3% of final sample]). This approach ensured that representation of patients was maximized for biologics such as golimumab, which indicated a lower overall prevalence in the data set. All members were required to have at least 2 years of continuous eligibility: 1 year before and 1 year after the index date. Members were also required to have at least 2 claims for rheumatoid arthritis (code 714.x) within 1 year of the index date and no diagnoses of ankylosing spondylitis (code 720.x) or psoriatic arthritis (code 696.x) within that same period. All members must have been at least 18 years of age, and women must not have been pregnant at any time during the study. In addition, members with a comorbid diagnosis of Crohn disease (code 555.x) were excluded because this is an approved indication for adalimumab.11 Because most patients received 30-day supplies of medication (68%), to minimize artifacts, only those members with 30-day supply fills of their index medication across their episode of care were maintained for final analyses. After the imposition of all inclusion and exclusion criteria, the final sample size was 3892 (n ¼ 1532 for the adalimumab group, n ¼ 2099 for the etanercept group, and n ¼ 261 for the golimumab group).
Measures Patient demographic characteristics were summarized from the membership table and included age, sex, geographic region of residence, insurance line of business, and type of benefit plan. In addition, the Charlson Comorbidity Index, an overall measure of health,18 was calculated during the 1-year preindex period, as were the rates for a variety of other comorbidities of interest. Prior biologic utilization during the 1-year preindex period was also calculated. The following section outlines the primary treatment outcomes, which were calculated for all
Volume ] Number ]
J. Tkacz et al.
n = 353,8333 Identify index golimumab, adalimumab, or etanercept fill during the case-finding period (1/1/09 –3/31/11)
n = 24,851
Must have at least two fills for the index biologic treatment n = 22,508 Continuously eligible for 2 years around earliest treatment (index) date
n = 10,632
2 diagnoses of rheumatoid arthritis and the absence of other rheumatic disorders within 1 year of index date
n = 5998
At least 18 years old n = 5848 Absence of pregnancy n = 5801 Absence of Crohn disease diagnosis n = 5737 Fills for 30 days' supply of medication
Final Sample
n = 3892 Place into final study groups Adalimumab
Etanercept
n = 1532
n = 2099
Golimumab n = 261
Figure 1. Sample attrition. *Starting sample consisted of members with a single diagnosis of rheumatoid arthritis, ankylosing spondylitis, or psoriatic arthritis between quarter 4 of 2005 and quarter 1 of 2012.
members who qualified for the analyses. All treatment outcomes were derived from the pharmacy claims table and were calculated for the period the patient continued to take their assigned study medication. Patients were followed up until discontinuation or until a period of Z90 days without a fill for the index medication was observed. Patients were followed up for longer than 1 year if they maintained eligibility and continually received treatment (ie, did not have evidence of a lapse in treatment). Adalimumab, etanercept, and golimumab treatment fills were identified via brand name and were summed for each member. The corresponding days’
] 2014
supply value for these treatments was also summed by member. The MPR was calculated by summing the days’ supply value of the index treatment and dividing this amount by the duration of treatment, which was the sum of days between the index date and the most recent index treatment fill plus the days’ supply value of the last fill.19 The MPR values 41 were recoded into 1, and the proportion of patients with an MPR Z0.60 and Z0.80 were reported. The PDC was also calculated, which is another measure of days in which medication was on hand, but adjusts for doublecounting of covered days.19 As with MPR, the proportion of patients with a PDC Z0.60 and
3
Clinical Therapeutics Z0.80 were reported. Medication dosages were extracted from the strength field, and mean patient dosages were calculated. Similarly, mean refill intervals were also calculated for each patient, in addition to summing the total number of early, late, and adherent fills. An adherent fill interval was defined as a refill appearing Z21 and r38 days from the previous fill. A refill interval of r15 days was categorized as early, whereas a refill interval Z45 days was categorized as late.
Statistical Analyses Bivariate group comparisons were initially conducted for all demographic and treatment measures. χ2 tests of equality of proportions were used for categorical variables, and 1-way analysis of variance was used for continuous variables. Turkey’s post hoc tests were conducted to examine group differences where omnibus test results were statistically significant. Outcomes that resulted in statistically significant bivariate group effects were entered into multivariate analyses that controlled for demographic characteristics, health status, and treatment duration. Ordinary least squares regression models were used for continuous variables, while dichotomous, categorical variables were regressed onto predictors via logistic models. The golimumab group served as the reference treatment category in all models. Data management and analyses were conducted using SPSS statistical software, version 20 (SPSS Inc, Chicago, Illinois).
RESULTS Members were predominantly female, (75.5%), with a mean (SD) age of 51.1 (11.2) years, residing predominantly in the mid-American (39.0%) and Southeastern (39.0%) regions of the United States. Nearly the entire sample was commercially insured (99.5%), with most holding a point-of-service benefit package (69.1%). The mean (SD) Charlson Comorbidity Index score was 0.52 (0.95). Table I lists the demographic characteristics by treatment group. Golimumab members were 43 times as likely to evidence prior biologic exposure as adalimumab members and 46 times as likely as etanercept members (P o 0.001 for all). Nearly 60% of the golimumab group had previous experience with a biologic. The most commonly used prior biologic for adalimumab members was etanercept (12.3%), with adalimumab
4
(6.1%) being the most commonly used for etanercept members. Golimumab members also tended to have the highest comorbid disease burden of the 3 groups, evidenced by a significantly higher proportion with a diagnosis of depression, osteoarthritis, chronic pain, cerebrovascular disease, and hypertension compared with both the adalimumab and etanercept groups (P o 0.05 for all). Both the golimumab and etanercept groups had a significantly higher mean Charlson Comorbidity Index score than the adalimumab patients (P o 0.05). There were no statistically significant treatment group differences for the remaining demographic indicators. Table II gives the group descriptive statistics and results of bivariate comparisons for the primary treatment outcomes. Compared with both the adalimumab and golimumab groups, the etanercept group had a significantly greater number of fills (19.4 vs 18.3 vs 13.1) and cumulative days’ supply of treatment (546 vs 516 vs 384), whereas the adalimumab group had significantly greater fills and days’ supply than the golimumab group (P o 0.001 for all). All 3 groups significantly differed from each other on both the MPR and PDC measures, with the golimumab group (MPR ¼ 0.88, PDC ¼ 0.87) having significantly higher values than both the adalimumab (MPR ¼ 0.85, PDC ¼ 0.84) and etanercept groups (MPR ¼ 0.81, PDC ¼ 0.81) and with the adalimumab group having significantly higher values than the etanercept group (P o 0.001 for all). The same pattern of results was obtained when groups were compared on the proportion with MPR and PDC values Z0.80 (P o 0.001 for all). Further, compared with the etanercept group, the golimumab and adalimumab groups evidenced significantly shorter refill intervals (34.9 and 35.1 vs 37.7 days) and a higher mean proportion of treatment fills categorized as adherent (77% and 75% vs 70%; P o 0.001 for all). Groups did not significantly differ on the proportion with an early refill, but compared with the adalimumab and etanercept groups, the golimumab group was significantly less likely to have at least 1 late refill (62.8% vs 71.5% and 74.4%). Further, the golimumab group was significantly less likely to have Z4 late refills compared with both the adalimumab and etanercept groups, whereas the adalimumab group was significantly less likely than the etanercept group (6.9% vs 17.7% vs 26.1%; P o 0.001 for all).
Volume ] Number ]
J. Tkacz et al.
Table I. Patient demographic characteristics. No. (%) of Patients* Characteristic Female Geographic region Mid-American Northeast Southeast West Benefit plan business type Commercial Medicaid Medicare Type of benefit plan EPO HMO IND OTH POS PPO Age, y Mean (SD) Median Patients o65 years old Prior biologics use Any prior biologic use†,‡,§ 1 Prior treatment 2 Prior treatments 3 Prior treatments Specific treatments Golimumab Adalimumab Etanercept Abatacept Infliximab Certolizumab Anakinra Rituximab Tocilizumab None Select preperiod comorbidities Cerebrovascular disease†,‡ Chronic renal disease Diabetes (without complications)
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
1157 (75.5)
1572 (74.9)
211 (80.8)
600 86 619 227
(39.2) (5.6) (40.4) (14.8)
824 163 798 314
(39.3) (7.8) (38.0) (15.0)
93 16 102 50
P 0.108 0.082
(35.6) (6.1) (39.1) (19.2) 0.475
1524 (99.5) 8 (0.5) 0
2087 (99.4) 12 (0.6) 0
261 (100.0) 0 0
229 163 15 1 1073 51
325 211 46 5 1441 71
(15.5) (10.1) (2.2) (0.2) (68.7) (3.4)
44 (16.9) 22 (8.4) 2 (0.8) 0 176 (67.4) 17 (6.5)
50.7 (11.0) 52 1435 (93.7)
51.4 (11.4) 53 1932 (92.0)
50.6 (10.6) 52 246 (94.3)
249 (16.3) 236 (15.4) 13 (0.8) 0
187 (8.9) 174 (8.3) 13 (0.6)
9 (0.6)
6 (0.3) 128 (6.1)
0.026 (14.9) (10.6) (1.0) (0.1) (70.0) (3.3)
0.176
155 129 23 3
(59.4) (49.4) (8.8) (1.1)
(12.3) (1.2) (2.2) (0.4) (0.2) (0.2) 0 1283 (84.6)
20 (1.0) 37 (1.8) 2 (0.1) 0 7 (0.3) 0 1912 (92.0)
(23.8) (24.5) (10.0) (7.7) (2.3) 0 6 (2.3) 0 106 (39.5)
41 (2.7) 20 (1.3) 127 (8.3)
56 (2.7) 43 (2.0) 199 (9.5)
14 (5.4) 8 (3.1) 29 (11.1)
189 18 34 6 3 3
0.113 o0.001
62 64 26 20 6
0.041 0.077 0.240 (continued)
] 2014
5
Clinical Therapeutics Table I. (continued). No. (%) of Patients* Characteristic Disorders of lipid metabolism Depression†,‡ Osteoarthritis†,‡ Chronic pain†,‡ Hypertension†,‡ Charlson Comorbidity Index score, mean (SD)†,§
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
P
501 (32.7) 162 (10.6) 849 (55.4) 58 (3.8) 512 (33.4) 0.47 (0.86)
696 (33.2) 197 (9.4) 1138 (54.2) 73 (3.5) 707 (33.7) 0.54 (0.99)
97 (37.2) 41 (15.7) 165 (63.2) 25 (9.6) 110 (42.1) 0.59 (1.10)
0.365 0.006 0.022 o0.001 0.018 0.023
EPO ¼ exclusive provider organization; HMO ¼ health maintenance organization; IND ¼ indemnity; OTH ¼ other; POS ¼ point of service; PPO ¼ preferred provider organization. * Data are presented as number (percentage) of patients unless otherwise indicated. † Simponi a adalimumab. ‡ Simponi a etanercept. § Adalimumab a etanercept.
Table III and Table IV give the results of ordinary least squares and logistic models of treatment outcomes that controlled for treatment duration and patient characteristics. Men received significantly more fills and days’ supply of treatment compared with women and were more adherent as measured by both the MPR and PDC (P o 0.05 for all). Age was positively associated with treatment fills, days’ supply, MPR, and PDC but inversely related with mean refill interval (P o 0.05 for all). Treatment duration was positively associated with all outcomes, excluding the numeric MPR value (P o 0.05 for all). Health status, as measured by the Charlson Comorbidity Index, and geographic region of residence did not relate to any treatment outcomes (P 4 0.05 for all). As for the main predictor of interest, treatment group, most bivariate effects were reproduced in multivariate models. Adalimumab members were only half as likely as golimumab members to achieve an MPR or PDC Z0.80, whereas etanercept members were only one-third as likely (P o 0.01 for all). Both the adalimumab and etanercept groups had significantly greater mean refill intervals than the golimumab group (P o 0.05), with etanercept members being more than twice as likely as golimumab members to have Z4 late refills (P ¼ .001). Finally, both the adalimumab and etanercept members were significantly less likely to have prior biologic exposure (P o 0.001 for all).
6
DISCUSSION The objectives of the present study were to examine utilization and adherence measures for 3 subcutaneously administered anti-TNF agents indicated for the treatment of RA (adalimumab, etanercept, and golimumab) and to compare treatment outcomes across groups. Golimumab comprised the smallest study group, accounting for only 7% of the overall sample, with the adalimumab (39%) and etanercept (54%) groups making up most of patients. Overall, the sample appeared to be receiving treatment within recommended prescribing guidelines because mean dosages and refill intervals approximated expected values. In addition, multiple adherence measures were assessed during the study period. Data were adequate because most of the sample received treatment for 41 year and evidenced elevated values for both the MPR and PDC. Interestingly, numerous treatment utilization and adherence differences were found across the study groups. Etanercept use was the most prevalent of the 3 biologics studied, whereas golimumab utilization was, as expected, the lowest because it was the latest to be approved and the last commercially available of the 3 agents. Golimumab patients had significantly higher rates of medication adherence as measured by the MPR and PDC compared with both adalimumab and etanercept patients and were also less likely to have
Volume ] Number ]
J. Tkacz et al.
Table II. Treatment outcomes. No. (%) of Patients* Outcome
Adalimumab (n ¼ 1532)
Etanercept (n ¼ 2099)
Golimumab (n ¼ 261)
Total fills 28,058 (100.0) 40,701 (100.0) 3409 (100.0) 18.3 (11.9) 19.4 (12.0) 13.1 (7.8) Fills per member, mean (SD)†,‡,§ Total days’ supply, mean (SD)†,‡,§ 516 (334) 546 (338) 384 (230) 0.85 (0.13) 0.81 (0.15) 0.88 (0.11) MPR, mean (SD)†,‡,§ 1080 (70.5) 1298 (61.8) 214 (82.0) MPR Z0.80†,‡,§ MPR Z0.60‡,§ 1435 (93.7) 1861 (88.7) 252 (96.6) 0.84 (0.13) 0.81 (0.15) 0.87 (0.10) PDC, mean (SD)†,‡,§ PDC Z0.80 1073 (70.0) 1284 (61.2) 212 (81.2) †,‡,§ 1432 (93.5) 1860 (88.6) 252 (96.6) PDC Z0.60 Patient dose, mean (SD), mg‡,§ 87.0 (20.2) 193.3 (24.7) 52.2 (10.5) 35.1 (9.3) 37.7 (12.0) 34.9 (8.0) Patient refill interval, mean (SD), d‡,§ 0.75 (0.24) 0.70 (0.28) 0.77 (0.23) Patient proportion of adherent fills, mean (SD)‡,§ Early refill 47 (3.1) 55 (2.6) 8 (3.1) 1095 (71.5) 1561 (74.4) 164 (62.8) Late refill†,‡ 271 (17.7) 547 (26.1) 18 (6.9) Z4 Late refills†,‡,§
P o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 0.703 o0.001 o0.001
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered. * Data are presented as number (percentage) of patients unless otherwise indicated. † Golimumab a adalimumab. ‡ Golimumab a etanercept. § Adalimumab a etanercept.
multiple, late refills. The golimumab group was also the most experienced with biologic medication, with nearly 60% of this group having prior treatment with a biologic, nearly 5 times greater than adalimumab and etanercept patients. Golimumab and adalimumab patients both had briefer refill intervals and a higher percentage of adherent fills compared with etanercept patients. Group effects were maintained in multivariate models that controlled for the duration of therapy and patient characteristics, indicating that these group differences do not account for differences in adherence measures. Overall, golimumab patients appeared to be the most adherent group during the study period, despite having the poorest overall health. This is the first known study to report greater real-world adherence for golimumab compared with 2 established biologics. Differences in clinical effectiveness across adalimumab, etanercept, and golimumab have not been studied directly, although the approval of these agents was based
] 2014
on meeting similar primary outcomes in clinical trials.20–22 There are a number of plausible interpretations of the present results. Golimumab was the only biologic agent in the present study with a once-monthly dosing schedule. Multiple studies of adherence across a variety of medication classes have reported the inverse relationship between dosing frequency and adherence,23–25 and patients have been found to prefer longer vs shorter dosing intervals when given the option.26–28 In the present study, this relationship appeared evident for a number of adherence measures, including the MPR, PDC, and proportion with Z4 late refills, because patients treated with once-monthly golimumab evidenced significantly improved outcomes on these measures compared with bimonthly treated adalimumab patients, who in turn evidenced significant improvements compared with weeklytreated etanercept patients. In addition, increased medication adherence has been associated with
7
8
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered.
0.07 0.03 0.01 o0.01 0.03 0.01 o0.01 1.89 0.68 0.01 o0.01 0.07 0.01 o0.01 0.06 0.01 o0.01 4.58 0.67 o0.01 o0.01 0.01 0.01 0.02 0.01 0.01 0.03 0.37 0.37 0.32 o0.01 0 0 o0.01 0 0 o0.01 0.12 0.01 o0.01 0.80 0 0 0.74 0 0 0.71 0.03 0.18 0.87 o0.01 0 0 0.12 0 0 0.02 0 0 o0.01 0.46 0.01 0.01 0.38 0.01 0.01 0.48 0.59 0.67 0.38 0.83 0 0.01 0.81 0 0.01 0.75 0.06 0.67 0.92 0.67 0.01 0.01 0.50 0.01 0.01 0.56 0.86 0.74 0.25 0 0.76 0.02 0 0.75 0.02 0 15.57 1.13 0 Adalimumab 0.17 0.23 0.47 11.54 6.42 Etanercept 0.62 0.23 0.01 34.20 6.33 Male 0.43 0.13 o0.01 12.31 3.56 Age 0.05 0.01 o0.01 1.36 0.14 Charlson Comorbidity Index score 0.01 0.06 0.82 0.42 1.67 Treatment duration 0.03 0 o0.01 0.85 0 Mid-American region 0.16 0.23 0.47 4.68 6.35 Southeast region 0.14 0.23 0.54 1.40 6.35 West region 0.13 0.25 0.61 2.99 7.05 Constant 2.65 0.38 0 59.02 10.78
P SE β SE P SE Predictor
β
SE
P
β
SE
P
β
MPR Days’ Supply Treatment Fills
Table III. Multivariate analyses of treatment outcomes in ordinary least squares regression models.
β
PDC
P
Mean Refill Interval
Clinical Therapeutics increased disease severity,29 and although disease severity cannot be accurately measured in claims data alone, the golimumab group’s increased rate of both prior biologic use and overall number of comorbidities indicate a potentially more severe disease state. Further, the belief that medication is necessary for disease treatment is also a positive predictor of adherence,30 and the golimumab group’s increased prior exposure to biologic treatment compared with the other groups may have strengthened their belief in the medical necessity of the current treatment regimen. Regardless, results suggest golimumab may be the ideal treatment option for patients with difficulty maintaining adherence, although findings should be replicated in a biologicnaive sample to solidify the relationship between golimumab treatment and adherence. Although there is evidence suggesting that adherence to biologic medication is less than optimal,31 which may be partly due to inconsistent reporting and defining of relevant measures as discussed in the opening of the article, results of the present study revealed generally positive adherence outcomes for all 3 treatment groups. Adherence has been linked to improved outcomes across a variety of chronic diseases states,32 including RA.33,34 In addition, persistence to biologic therapy among RA patients has also been associated with a reduction in non–pharmacy-related health care expenditure.35 Further study of RA patient motivation toward treatment adherence is needed, particularly for biologics. A number of factors related to improved medication adherence have previously been identified in RA patients, including a satisfactory level of patient-physician communication, an increased knowledge of RA in general, and older age,36 whereas nonwhite race and low educational level have been found to be related to decreased adherence.37 In addition, studies have found that patients typically prefer subcutaneous injection over intravenous and intramuscular injections,38,39 which may promote adherence in itself. A better understanding of the drivers of treatment adherence may reduce the amount of time patients spend on ineffective treatments and, given the higher expenses typically associated with biologic treatment,40,41 may have significant economic implications. Given the increasing amount of clinical and economic data being collected in health care today, these data can be used to promote a patient-centered model of care in which
Volume ] Number ]
J. Tkacz et al.
Table IV. Multivariate analyses of treatment outcomes in logistic models. MPR Z0.80 Predictor Adalimumab Etanercept Male Age Charlson Comorbidity Index score Treatment duration Mid-American region Southeast region West region Constant
β
Exp (B)
P
PDC Z0.80 β
Exp (B)
Z4 Late Refills P
β
Exp (B)
P
Prior Biologic Exposure β
Exp (B)
P
0.72 1.16 0.20 0.03 0.06
0.49 o0.01 0.69 0.50 o0.01 0.45 1.57 0.09 1.90 0.15 o0.01 0.31 o0.01 1.14 0.32 o0.01 0.85 2.34 o0.01 2.55 0.08 o0.01 1.22 0.02 0.20 1.22 0.02 0.33 0.72 o0.01 0.12 0.88 0.28 1.03 o0.01 0.02 1.02 o0.01 0.03 0.97 o0.01 0.01 0.99 0.09 1.06 0.12 0.07 1.07 0.09 0.04 1.04 0.39 0.05 1.05 0.37
0 0.01 0.04 0.03 0.07
1.00 o0.01 0 1.00 o0.01 0.99 0.93 0.03 1.03 0.84 0.96 0.77 0.04 0.96 0.80 0.97 0.84 0 1.00 0.98 1.08 0.78 0.03 1.03 0.90
0 0.02 0.10 0.17 2.97
1.00 o0.01 0 1.00 o0.01 1.02 0.90 0.13 0.88 0.50 1.10 0.61 0.30 0.74 0.12 1.19 0.39 0.19 0.83 0.38 0.05 0 1.38 3.97 0
MPR ¼ medication possession ratio; PDC ¼ proportion of days covered.
much improved disease management is possible.42 Analyses similar to those appearing in the present study can be used by prescribers to monitor patient treatment adherence, in addition to health plans for purposes of avoiding unnecessary spend. The present study has a number of limitations. The primary limitation is the potential endogeneity of the between-groups factor (treatment), meaning that there could be other unmeasured factors that drive utilization and adherence patterns associated with each medication. Treatment groups were not matched to variations in patient characteristics among the groups, and data have indicated that differences in these factors may confound study conclusions. Although multivariate models that control for patient characteristics were conducted, an analysis using propensity score matching or instrumental variables may be of some utility. An additional limitation was the data source (administrative claims data), which does not allow for the assessment of clinical outcomes related to adherence and may also include administrative coding errors.43 Finally, the MPR and PDC were calculated based on the length of each individual patients’ treatment episode, as opposed to a fixed interval (eg, 1 year). Given the cycling-forremission strategy used in RA biologic treatment,44 use of a fixed interval can result in a deflated estimate of
] 2014
adherence because patients may adhere for a period only to discontinue use shortly thereafter because of ineffectiveness or adverse effects. In conclusion, the dosing and scheduling complexities of biologic treatment have been found to be problematic for the standardized measurement and assessment of medication adherence. The present study was designed to measure multiple indicators of biologic administration process indicators among a sample of RA patients treated with subcutaneously administered anti-TNFs. Adalimumab, etanercept, and golimumab patients were found to be generally adherent with their treatment regimens because dosages and refill intervals were in agreement with recommended prescribing guidelines and proxies for medication adherence were in the appropriate range. Utilization of golimumab was significantly lower than that of adalimumab and etanercept, although golimumab patients appeared to be the most adherent during their treatment despite also being in the poorest overall health. Specifically, golimumab patients had significantly higher MPRs and PDCs compared with the adalimumab and etanercept patients and were also less likely to have multiple, late refills. The favorable adherence results achieved by the golimumab group may be a result of these patients’ likely increased
9
Clinical Therapeutics disease severity, their prior experience with biologic medication, or golimumab’s once-monthly dosing schedule. Overall, study findings were encouraging but indicate the need for additional comparative effectiveness studies between golimumab and the established biologics to confirm adherence results and to quantify the precise clinical and economic benefits associated with biologic adherence.
CONFLICTS OF INTEREST Janssen Scientific Affairs funded this study. Drs Ellis, Meyer, and Bolge are employees of Janssen Scientific Affairs and are stockholders in Johnson and Johnson. Study sponsors assisted with study design, interpretation of results, and manuscript development.
REFERENCES 1. Gibofsky A. Overview of epidemiology, pathophysiology, and diagnosis of rheumatoid arthritis. Am J Manag Care. 2012;18(13 Suppl):S295–S302. 2. Scott DL. Biologics-based therapy for the treatment of rheumatoid arthritis. Clin Pharmacol Ther. 2012;91:30–43. 3. Yazici Y, Shi N, John A. Utilization of biologic agents in rheumatoid arthritis in the United States: analysis of prescribing patterns in 16,752 newly diagnosed patients and patients new to biologic therapy. Bull NYU Hosp Jt Dis. 2008;66:77–85. 4. Salt E, Frazier SK. Adherence to disease-modifying antirheumatic drugs in patients with rheumatoid arthritis: a narrative review of the literature. Orthop Nurs. 2010;29:260–275. 5. Cramer JA, Roy A, Burrell A, et al. Medication compliance and persistence: terminology and definitions. Value Health. 2008;11:44–47. 6. de Achaval S, Suarez-Almazor ME. Treatment adherence to disease-modifying antirheumatic drugs in patients with rheumatoid arthritis and systemic lupus erythematosus. Int J Clin Rheumtol. 2010;5:313–326. 7. Malaviya AP, Ostor AJ. Drug adherence to biologic DMARDS with a special emphasis on the benefits of subcutaneous abatacept. Patient Prefer Adherence. 2012;6:589–596. 8. Li P, Blum MA, Von Feldt J, Hennessy S, Doshi JA. Adherence, discontinuation, and switching of biologic therapies in medicaid enrollees with rheumatoid arthritis. Value Health. 2010;13:805–812. 9. Borah BJ, Huang X, Zarotsky V, Globe D. Trends in RA patients’ adherence to subcutaneous anti-TNF therapies and costs. Curr Med Res Opin. 2009;25:1365–1377. 10. Brocq O, Roux CH, Albert C, et al. TNF-α antagonist continuation rates in 442 patients with inflammatory joint disease. Joint Bone Spine. 2007;74:148–154.
10
11. Food and Drug Administration. Highlights of Prescribing Information: Humira. 2011. http://www.accessdata.fda. gov/drugsatfda_docs/label/2011/125057s0276lbl.pdf. Accessed May 16, 2013. 12. Food and Drug Administration. Enbrels (etanercept) For Subcutaneous Injection. 2003. http://www.accessdata. fda.gov/drugsatfda_docs/label/2003/etanimm060503LB. pdf. Accessed May 16, 2013. 13. Food and Drug Administration. Highlights of Prescribing Information: Simponi. 2009. http://www.accessdata. fda.gov/drugsatfda_docs/label/2009/125289s000lbl.pdf. Accessed May 16, 2013. 14. Dhillon S, Lyseng-Williamson KA, Scott LJ. Etanercept: a review of its use in the management of rheumatoid arthritis. Drugs. 2007;67:1211–1241. 15. Keystone EC, Genovese MC, Klareskog L, et al. Golimumab, a human antibody to tumour necrosis factor α given by monthly subcutaneous injections, in active rheumatoid arthritis despite methotrexate therapy: the GO-FORWARD Study. Ann Rheum Dis. 2009;68:789–796. 16. Wiens A, Correr CJ, Venson R, Otuki MF, Pontarolo R. A systematic review and meta-analysis of the efficacy and safety of adalimumab for treating rheumatoid arthritis. Rheumatol Int. 2010;30:1063–1070. 17. Zhou H, Jang H, Fleischmann RM, et al. Pharmacokinetics and safety of golimumab, a fully human anti-TNF-alpha monoclonal antibody, in subjects with rheumatoid arthritis. J Clin Pharmacol. 2007;47:383–396. 18. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40: 373–383. 19. Peterson AM, Nau DP, Cramer JA, Benner J, GwadrySridhar F, Nichol M. A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007;10:3–12. 20. Emery P, Fleischmann RM, Moreland LW, et al. Golimumab, a human anti-tumor necrosis factor alpha monoclonal antibody, injected subcutaneously every four weeks in methotrexate-naive patients with active rheumatoid arthritis: twenty-four-week results of a phase III, multicenter, randomized, double-blind, placebo-controlled study of golimumab before methotrexate as first-line therapy for earlyonset rheumatoid arthritis. Arthritis Rheum. 2009;60: 2272–2283. 21. Moreland LW, Schiff MH, Baumgartner SW, et al. Etanercept therapy in rheumatoid arthritis: a randomized, controlled trial. Ann Intern Med. 1999;130:478–486. 22. Weinblatt ME, Keystone EC, Furst DE, et al. Adalimumab, a fully human anti-tumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial. Arthritis Rheum. 2003;48:35–45.
Volume ] Number ]
J. Tkacz et al. 23. Coleman CI, Limone B, Sobieraj DM, et al. Dosing frequency and medication adherence in chronic disease. J Manag Care Pharm. 2012; 18:527–539. 24. Iskedjian M, Einarson TR, MacKeigan LD, et al. Relationship between daily dose frequency and adherence to antihypertensive pharmacotherapy: evidence from a meta-analysis. Clin Ther. 2002;24:302–316. 25. Saini SD, Schoenfeld P, Kaulback K, Dubinsky MC. Effect of medication dosing frequency on adherence in chronic diseases. Am J Manag Care. 2009;15:e22–e33. 26. Granger AL, Fehnel SE, Hogue SL, Bennett L, Edin HM. An assessment of patient preference and adherence to treatment with Wellbutrin SR: a web-based survey. J Affect Disord. 2006;90:217–221. 27. Kendler D, Kung AW, Fuleihan Gel H, et al. Patients with osteoporosis prefer once weekly to once daily dosing with alendronate. Maturitas. 2004;48:243–251. 28. Reginster JY, Rabenda V, Neuprez A. Adherence, patient preference and dosing frequency: understanding the relationship. Bone. 2006;38(4 Suppl 1): S2–S6. 29. Gao X, Nau DP, Rosenbluth SA, Scott V, Woodward C. The relationship of disease severity, health beliefs and medication adherence among HIV patients. AIDS Care. 2000;12:387–398. 30. Pasma A, van’t Spijker A, Hazes JM, Busschbach JJ, Luime JJ. Factors associated with adherence to pharmaceutical treatment for rheumatoid arthritis patients: a systematic review. Semin Arthritis Rheum. 2013; 43:18–28. 31. Koncz T, Pentek M, Brodszky V, Ersek K, Orlewska E, Gulacsi L. Adherence to biologic DMARD therapies in rheumatoid arthritis. Expert Opin Biol Ther. 2010;10:1367–1378. 32. DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a
] 2014
33.
34.
35.
36.
37.
meta-analysis. Med Care. 2002;40: 794–811. Contreras-Yanez I, Ponce De Leon S, Cabiedes J, Rull-Gabayet M, Pascual-Ramos V. Inadequate therapy behavior is associated to disease flares in patients with rheumatoid arthritis who have achieved remission with disease-modifying antirheumatic drugs. Am J Med Sci. 2010; 340:282–290. Pascual-Ramos V, Contreras-Yanez I, Villa AR, Cabiedes J, Rull-Gabayet M. Medication persistence over 2 years of follow-up in a cohort of early rheumatoid arthritis patients: associated factors and relationship with disease activity and with disability. Arthritis Res Ther. 2009;11:R26. Tang B, Rahman M, Waters HC, Callegari P. Treatment persistence with adalimumab, etanercept, or infliximab in combination with methotrexate and the effects on health care costs in patients with rheumatoid arthritis. Clin Ther. 2008;30:1375–1384. Viller F, Guillemin F, Briancon S, Moum T, Suurmeijer T, van den Heuvel W. Compliance to drug treatment of patients with rheumatoid arthritis: a 3 year longitudinal study. J Rheumatol. 1999;26:2114–2122. Garcia-Gonzalez A, Richardson M, Garcia Popa-Lisseanu M, et al. Treatment adherence in patients with rheumatoid arthritis and systemic lupus erythematosus. Clin Rheumatol. 2008;27:883–889.
38. Williams EL, Edwards CJ. Patient preferences in choosing anti-TNF therapies-R1. Rheumatology (Oxford). 2006;45:1575–1576. 39. Jeffery M, Stokes A, Johnson K, Anderson D, Harper B. Patient Preferences for Tumor Necrosis Factor (TNF) Antagonist. London, UK: The Research Partnership Ltd, Rheumatology Research International; 2004. 40. Modena V, Bianchi G, Roccatello D. Cost-effectiveness of biologic treatment for rheumatoid arthritis in clinical practice: an achievable target? Autoimmun Rev. 2013;12:835–838. 41. van der Velde G, Pham B, Machado M, et al. Cost-effectiveness of biologic response modifiers compared to disease-modifying antirheumatic drugs for rheumatoid arthritis: a systematic review. Arthritis Care Res (Hoboken). 2011;63:65–78. 42. Chawla NV, Davis DA. Bringing big data to personalized healthcare: a patient-centered framework. J Gen Intern Med. 2013;28(Suppl 3):S660– S665. 43. Tyree PT, Lind BK, Lafferty WE. Challenges of using medical insurance claims data for utilization analysis. Am J Med Qual. 2006;21:269– 275. 44. Ramiro S, Machado P, Singh JA, Landewe RB, da Silva JA. Applying science in practice: the optimization of biological therapy in rheumatoid arthritis. Arthritis Res Ther. 2010;12: 220.
Address correspondence to: Joseph Tkacz, MS, Health Analytics, LLC, 9200 Rumsey Rd, Suite 215, Columbia, MD. E-mail: joseph.tkacz@ healthanalytic.com
11
