Abstract
Objective
Several trials suggest that triple therapy (methotrexate, sulfasalazine, and hydroxychloroquine) and biologic disease-modifying antirheumatic drugs (bDMARD) have similar efficacy in rheumatoid arthritis (RA). We investigated intensification to triple therapy after initial non-biologic (nbDMARD) prescription among patients with RA.
Methods
We used US insurance claims data to evaluate triple therapy use from 2009–2014. Patients with a visit for RA and initial nbDMARD prescription were included. Frequencies and rates to intensification to triple therapy or bDMARD were calculated. We evaluated whether sociodemographic, temporal, geographic, clinical, and healthcare utilization factors were associated with triple therapy intensification using Cox regression. Among those who intensified therapy, we investigated factors associated with triple therapy use by logistic regression.
Results
There were 24,576 patients initially with mean age of 50.3 (SD 12.3) years, and 78% were female. During the study period, 2,739 (11.1%) intensified treatment to bDMARD compared to 181 (0.7%) who intensified to triple therapy. There was no significant change in triple therapy use across calendar years. Patients who intensified to triple therapy were more likely to use glucocorticoids (HR 1.91, 95%CI 1.41–2.60) compared to no glucocorticoids and more likely to use nonsteroidal anti-inflammatory drugs (NSAID, HR 1.48, 95%CI 1.10–1.99) compared to no NSAID use within 180 days of initial nbDMARD prescription. Among those who intensified treatment to triple therapy or bDMARD, significant associations for triple therapy use included older age, US region (highest odds for triple therapy use in the West, lowest odds for triple therapy use in the Northeast), glucocorticoid use, and lower number of outpatient visits within 180 days of initial nbDMARD prescription.
Conclusion
Despite reports published during the study period suggesting equivalent efficacy of triple therapy and bDMARDs for RA, the use of triple therapy was infrequent and did not increase over time in this large nationwide study.
Keywords: Rheumatoid arthritis, Triple therapy, Health services, Pharmacoepidemiology
Several randomized controlled trials demonstrate that triple therapy (combination of methotrexate, sulfasalazine, and hydroxychloroquine) is non-inferior to biologic disease-modifying antirheumatic drugs (bDMARDs) for the treatment of rheumatoid arthritis (RA) (1–3). Compared to bDMARDs, triple therapy use has been shown to have superior cost effectiveness and similar quality of life measures (4–6).
Despite the growing number of bDMARDs for RA, triple therapy may have an important role in the treatment of RA by optimizing efficacy and cost. In part due to recent efforts advocating treatment to target of remission for RA, clinicians are now faced with complex treatment decisions for RA balancing efficacy, safety, and cost of a variety of options for DMARDs (7). The 2012 American College of Rheumatology guidelines for RA treatment recommended triple therapy as a therapeutic option for patients with RA after failure of initial non-biologic (nbDMARD) monotherapy or as initial therapy for patients with high disease activity and poor prognosis (8). Initial triple therapy at RA diagnosis may reduce disease activity compared to methotrexate monotherapy (9).
While the evidence supporting triple therapy is accumulating in randomized controlled trials, less is known about the use of triple therapy in typical clinical practice. A prior study using a national claims database evaluated adherence patterns with triple therapy compared to etanercept and methotrexate, but previous studies have not investigated temporal trends for intensification to triple therapy or factors associated with triple therapy receipt (10). Factors including perceived efficacy, age, socioeconomic status, geography, comorbidities, previous infections, and healthcare utilization may influence whether a patient is prescribed triple therapy or other therapies such as bDMARD. Therefore, we investigated the utilization of triple therapy and the factors associated with its use in a large nationwide insurance claims database. We aimed to quantify triple therapy use, assess for temporal trends, and investigate whether factors are associated with triple therapy intensification for RA in typical clinical practice.
METHODS
Data Source
We conducted a study using insurance claims data from Aetna, a commercial US health plan, for the period of July 1, 2009 to June 30, 2014. This database contains longitudinal claims information including medical diagnoses, procedures, hospitalizations, physician visits, inpatient/infusion visits, and pharmacy dispensing on subscribers across the United States. Aggregate data on socioeconomic status and demographic composition were obtained by linking zip code of residence with data from the 2010 US Census. The Partners HealthCare institutional review board approved the study protocol.
Study Cohort
We required analyzed subjects to have at least one prescription for a nbDMARD (methotrexate, sulfasalazine, hydroxychloroquine, leflunomide, azathioprine, mycophenolate mofetil, cyclosporine, gold, or D-penicillamine), have 180 days of continuous eligibility prior to first nbDMARD prescription, and have at least one inpatient or outpatient diagnosis code for RA (International Classification of Diseases, Ninth Revision [ICD-9] code 714.xx) prior to and including the nbDMARD prescription date (11). The index date was the date of the first nbDMARD prescription meeting these criteria. Patients with prevalent RA at the index date were eligible to be in the study cohort. Since we investigated intensification to triple therapy or bDMARD, we did not include patients whose first prescription was for bDMARD.
Intensification to Triple Therapy and bDMARD
We defined triple therapy use as prescriptions for methotrexate, sulfasalazine, and hydroxychloroquine within the same 60-day period. We defined the start of triple therapy as the date when the third of three medications was first filled, overlapping the other two. If a patient was prescribed triple therapy at index or in the baseline period, we considered this an outcome occurring at index date and the patient was excluded. We defined intensification to bDMARD as the first prescription for bDMARD (etanercept, adalimumab, infliximab, certolizumab pegol, golimumab, rituximab, abatacept, anakinra, tocilizumab, or tofacitinib) after index nbDMARD. If a patient received both triple therapy and bDMARD during follow-up, we considered the first intensification event as the outcome.
We categorized index nbDMARD as follows: at least one triple therapy nbDMARD (methotrexate, sulfasalazine, or hydroxychloroquine), other nbDMARD in combination with one of the triple therapy nbDMARDs, or other nbDMARD monotherapy. We further classified nbDMARD as monotherapy or as combination (two or more nbDMARDs prescribed at the index date).
Covariates
We defined sociodemographic, geographic, and temporal factors at the index date: sex, age in years, median annual household income in zip code (<$50,000, $50,000–$99,999, or ≥$100,000), US region (South, Northeast, West, or Midwest), and rural/urban residence. Age was further categorized as <40, 40–49, 50–59, and ≥60 years. We investigated year of cohort entry and time to intensification as covariates.
Clinical and healthcare utilization factors were considered within the 180 days prior to the index date. Drugs considered included nonsteroidal anti-inflammatory drugs (NSAIDs), opioid, antibiotic, antifungal, and antiviral use prior to the index date. We calculated the total number of unique prescriptions (excluding non-systemic formulations). Comorbidities considered were asthma, chronic obstructive pulmonary disease, coronary artery disease, cancer (excluding non-melanoma skin cancer), hypertension, ischemic stroke, congestive heart failure, depression, and diabetes mellitus. We also calculated a combined comorbidity score that combined conditions from the Charlson and Elixhauser measures (12). We also considered any hospitalization for serious infection (as previously validated (13)), appointment with a rheumatologist, number of outpatient visits, and inpatient stays and emergency department visits (categorized as 0, 1, or ≥2).
Statistical Analysis
We performed descriptive statistical analysis and unadjusted comparisons using chi-squared tests for categorical variables and t-tests for continuous variables. We used Cox proportional hazards models to estimate hazard ratios for treatment intensification to triple therapy. We initially performed unadjusted Cox models for each covariate described above to estimate hazard ratios (HR) and 95% confidence intervals (CI).
We built multivariable Cox regression models using stepwise forward selection of variables with p values <0.1 as criteria for entry. The final model for triple therapy utilization included US region of residence, index nbDMARD, combination nbDMARD (yes/no), glucocorticoid use, NSAID use, and number of outpatient visits.
We then restricted the sample to those who intensified treatment to triple therapy or bDMARD to investigate predictors for triple therapy use (as compared to bDMARD use) in this subgroup of intensifiers. We used logistic regression to estimate unadjusted odds ratios (OR) and 95%CIs for triple therapy use based on sociodemographic, temporal, geographic, clinical, and healthcare utilization factors. We built multivariable logistic regression models using stepwise forward selection of variables with p values <0.1 as criteria for entry. The final model for triple therapy use among those who intensified therapy consisted of age group, US region of residence, index nbDMARD, combination nbDMARD, glucocorticoid use, and number of outpatient visits. Since only triple therapy and bDMARD intensifiers were included in this analysis, factors associated with increased use of triple therapy were inversely associated with bDMARD use (and vice versa).
We calculated rates of intensification (per 1,000 person-years) to triple therapy or bDMARD over 6-month periods (January-June and July-December) during the entire 60 months of the study. We plotted trend lines for triple therapy and bDMARD intensification rates using smoothened splines. We performed two sensitivity analyses with identical methods: the first restricted the study sample to those who received at least one triple therapy nbDMARD at index date, and the second restricted the study sample to those with two or more ICD-9 codes for RA (714.xx).
We considered two-sided p values <0.05 as statistically significant for all analyses. All analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC).
RESULTS
A total of 24,576 subjects with nbDMARD prescription for RA were analyzed. Mean age was 50.3 years (standard deviation [SD] 12.3) and 77.8% were female (Table 1). The most frequent US region of residence was the South (39.3%) and the Midwest was least frequent (9.0%). The index nbDMARD was one of the three triple therapy DMARDs for most patients (87.8%) and most were on nbDMARD monotherapy (88.9%). Mean number of outpatient visits in the 180 days prior to index date was 9.8 (SD 8.4). Only 35.9% were seen by a rheumatologist within 180 days of the index date.
Table 1.
Characteristics of study sample at the index date (n=24,576).
| Characteristic | N (%) or mean (SD) |
|---|---|
| Sociodemographic factors | |
| Female | 19,110 (77.8%) |
| Mean age, years | 50.3 (12.3) |
| Age group | |
| <40 years | 4,371 (17.8%) |
| 40–49 years | 5,678 (23.1%) |
| 50–59 years | 8,771 (35.7%) |
| ≥60 years | 5,756 (23.4%) |
| Median annual household income by zip code | |
| <$50,000 | 5,670 (24.0%) |
| $50,000–$99,999 | 15,494 (65.6%) |
| ≥$100,000 | 2,463 (10.4%) |
| Temporal factors | |
| Year of index date | |
| 2009 | 7,594 (30.9%) |
| 2010 | 4,621 (18.8%) |
| 2011 | 3,947 (16.1%) |
| 2012 | 3,696 (15.0%) |
| 2013 | 3,380 (13.8%) |
| 2014* | 1,338 (5.4%) |
| Geographic factors | |
| US region of residence | |
| South | 9,319 (39.4%) |
| Northeast | 6,404 (27.1%) |
| West | 5,726 (24.2%) |
| Midwest | 2,181 (9.2%) |
| Rural residence | 1,884 (8.0%) |
| Clinical factors** | |
| Index nbDMARD | |
| MTX, SSZ, or HCQ | 21,584 (87.8%) |
| Other nbDMARD*** in combination with MTX, SSZ, or HCQ | 271 (1.1%) |
| Other nbDMARD monotherapy | 2,721 (11.1%) |
| Combination nbDMARDs | 4,157 (16.9%) |
| Glucocorticoids | 11,147 (45.4%) |
| NSAIDs or coxibs | 10,672 (43.4%) |
| Opioids | 7,465 (30.4%) |
| Antibiotic | 10,613 (43.2%) |
| Antifungal | 1,169 (4.8%) |
| Antiviral | 3,577 (14.6%) |
| Unique prescriptions | 7.8 (4.8) |
| Hospitalization for a serious infection | 384 (1.6%) |
| Asthma or chronic obstructive pulmonary disease | 2,260 (9.2%) |
| Coronary artery disease | 1,024 (4.2%) |
| Cancer | 2,247 (9.1%) |
| Hypertension | 6,469 (26.3%) |
| Ischemic stroke | 313 (1.3%) |
| Congestive heart failure | 259 (1.1%) |
| Depression | 2,183 (8.9%) |
| Diabetes mellitus | 2,006 (8.2%) |
| Combined comorbidity score | 0.2 (1.0) |
| Healthcare utilization factors** | |
| Rheumatologist appointment | 8,816 (35.9%) |
| Outpatient visits | 9.8 (8.4) |
| Inpatient stays | |
| 0 | 22,814 (92.8%) |
| 1 | 1,457 (5.9%) |
| ≥2 | 305 (1.2%) |
| Emergency department visits | |
| 0 | 23,129 (94.1%) |
| 1 | 930 (3.8%) |
| ≥2 | 517 (2.1%) |
End of study period was June 30, 2014.
Clinical and healthcare utilization factors were assessed within 180 days of index date.
Other non-biologic DMARDs were leflunomide, azathioprine, mycophenolate mofetil, cyclosporine, gold, and D-penicillamine.
HCQ, hydroxychloroquine; MTX, methotrexate; nbDMARD, non-biologic disease-modifying antirheumatic drug; NSAIDs, nonsteroidal anti-inflammatory drugs; SSZ, sulfasalazine; US, United States; ZIP, zone improvement plan.
There were a total of 2,739 (11.1%) patients who intensified treatment to bDMARD and 181 (0.7%) who intensified to triple therapy. The mean time to bDMARD from index date was 320 days (SD 341) and the median time to triple therapy was 278 days (SD 298). The Figure shows temporal trends in intensification to bDMARD (red dashed line) and triple therapy (blue dashed line). There was no significant change in use of triple therapy over calendar periods during the study period (7/1/09–6/30/14). Among the entire study sample, only 2013 approached statistical significance for triple therapy use (unadjusted HR 1.59, 95%CI 0.99–2.55) compared to 2009, but was not included in the final multivariable model.
Figure.
Intensification rate to triple therapy (blue dashed line) and biologic DMARD (red dashed line) after initial non-biologic DMARD prescription for rheumatoid arthritis, 2009–2014.
Patients initially prescribed combination nbDMARD (HR 8.31, 95%CI 5.85–11.81), glucocorticoids (HR 1.91, 95%CI 1.41–2.60), and NSAIDs (HR 1.48, 95%CI 1.10–1.99) were significantly more likely to receive triple therapy (Table 2). Compared to the South, subjects in the Northeast were less likely to intensify to triple therapy (HR 0.58, 95%CI 0.38–0.89), although this was not statistically significant in the multivariable model (HR 0.66, 95%CI 0.43–1.02). Increasing number of outpatient visits was associated with decreased risk of triple therapy use (HR 0.97, 95%CI 0.95–0.99). Sex, age, area-level household income, year of index date, rural residence, serious infections, comorbidities, antimicrobial use, number of prescriptions, and rheumatologist appointment were not associated with intensification to triple therapy.
Table 2.
Factors associated with intensification to triple therapy.
| HRs for triple therapy intensification among the entire study sample (n=24,576) | ORs for triple therapy use among those who intensified to triple therapy or biologic DMARD (n=2,920) | |||
|---|---|---|---|---|
| Characteristic | Unadjusted HR (95% CI) | Multivariable HR (95% CI) | Unadjusted OR (95% CI) | Multivariable OR (95% CI) |
| Female | 1.29 (0.88–1.88) | - | 1.40 (0.95–2.07) | - |
| Age group | ||||
| <40 years | 1.0 (Ref) | - | 1.0 (Ref) | 1.0 (Ref) |
| 40–49 years | 1.25 (0.80–1.94) | - | 1.04 (0.80–1.37) | 1.01 (0.76–1.34) |
| 50–59 years | 1.24 (0.79–1.94) | - | 1.20 (0.95–1.51) | 1.23 (0.96–1.57) |
| ≥60 years | 0.95 (0.57–1.57) | - | 1.22 (0.91–1.63) | 1.29 (0.95–1.75) |
| Median annual household income by zip code | ||||
| <$50,000 | 1.0 (Ref) | - | 1.0 (Ref) | - |
| $50,000–$99,999 | 0.95 (0.67–1.34) | - | 1.13 (0.89–1.44) | - |
| ≥$100,000 | 0.68 (0.37–1.24) | - | 0.79 (0.54–1.15) | - |
| Year of index date | ||||
| 2009 | 1.0 (Ref) | - | 1.0 (Ref) | - |
| 2010 | 0.72 (0.45–1.15) | - | 0.66 (0.44–0.97) | - |
| 2011 | 1.16 (0.76–1.78) | - | 1.11 (0.78–1.59) | - |
| 2012 | 1.25 (0.80–1.94) | - | 0.91 (0.63–1.32) | - |
| 2013 | 1.59 (0.99–2.55) | - | 1.13 (0.76–1.67) | - |
| 2014 | 1.69 (0.60–4.73) | - | 1.17 (0.49–2.75) | - |
| US region of residence | ||||
| South | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| Northeast | 0.58 (0.38–0.89) | 0.66 (0.43–1.02) | 0.62 (0.45–0.84) | 0.67 (0.48–0.92) |
| West | 1.30 (0.92–1.84) | 1.25 (0.88–1.77) | 1.42 (1.09–1.84) | 1.43 (1.09–1.88) |
| Midwest | 1.18 (0.71–1.96) | 1.16 (0.70–1.92) | 1.33 (0.91–1.93) | 1.26 (0.85–1.86) |
| Rural residence | 1.17 (0.70–1.95) | - | 1.01 (0.60–1.72) | - |
| Index nbDMARD | ||||
| MTX, SSZ, or HCQ | 22.46 (3.16–159.80) | 25.69 (3.59–183.73) | 3.23 (1.26–8.28) | 5.24 (2.03–13.54) |
| Other nbDMARD in combination with MTX, SSZ, or HCQ | 8.49 (0.53–135.46) | 2.59 (0.16–41.51) | 1.86 (0.42–8.25) | 0.89 (0.20–3.98) |
| Other nbDMARD monotherapy | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) | 1.0 (Ref) |
| Combination nbDMARDs | 5.17 (3.86–6.92) | 8.31 (5.85–11.81) | 5.30 (3.88–7.23) | 5.56 (4.04–7.65) |
| Glucocorticoids | 2.02 (1.50–2.73) | 1.91 (1.41–2.60) | 1.24 (0.91–1.69) | 1.39 (1.00–1.92) |
| NSAIDs or coxibs | 1.68 (1.25–2.25) | 1.48 (1.10–1.99) | 1.16 (0.86–1.58) | - |
| Opioids | 1.14 (0.83–1.56) | - | 0.80 (0.58–1.11) | - |
| Antibiotic | 0.77 (0.57–1.04) | - | 0.73 (0.53–1.00) | - |
| Antifungal | 1.06 (0.54–2.08) | - | 0.97 (0.49–1.94) | - |
| Antiviral | 0.83 (0.53–1.30) | - | 0.92 (0.58–1.45) | - |
| Unique prescriptions | 1.01 (0.98–1.04) | - | 0.99 (0.96–1.02) | - |
| Hospitalization for a serious infection | * | - | * | - |
| Asthma or chronic obstructive pulmonary disease | 0.78 (0.45–1.38) | - | 0.81 (0.45–1.45) | - |
| Coronary artery disease | 0.91 (0.43–1.93) | - | 1.15 (0.52–2.52) | - |
| Cancer | 0.99 (0.60–1.64) | - | 1.27 (0.75–2.13) | - |
| Hypertension | 0.72 (0.51–1.03) | - | 0.86 (0.59–1.24) | - |
| Ischemic stroke | * | - | * | - |
| Congestive heart failure | 1.02 (0.25–4.12) | - | 1.34 (0.31–5.74) | - |
| Depression | 1.00 (0.59–1.70) | - | 0.94 (0.54–1.62) | - |
| Diabetes mellitus | 1.01 (0.59–1.70) | - | 1.14 (0.66–1.98) | - |
| Combined comorbidity score | 0.99 (0.85–1.15) | - | 1.08 (0.91–1.28) | - |
| Rheumatologist appointment | 1.31 (0.97–1.76) | - | 0.98 (0.72–1.33) | - |
| Outpatient visits | 0.97 (0.95–0.99) | 0.97 (0.95–0.99) | 0.95 (0.93–0.98) | 0.95 (0.92–0.98) |
Model did not converge.
CI, confidence interval; DMARD, disease-modifying antirheumatic drug; HCQ, hydroxychloroquine; HR, hazard ratio; MTX, methotrexate; nbDMARD, non-biologic disease-modifying antirheumatic drug; NSAIDs, nonsteroidal anti-inflammatory drugs; OR, odds ratio; SSZ, sulfasalazine; US, United States; ZIP, zone improvement plan.
We also analyzed these factors among patients who intensified therapy to either triple therapy or bDMARD (n=2,920). US geographic area was associated with triple therapy use among intensifiers: increased risk in the West (OR 1.43, 95%CI 1.09–1.88), decreased risk in the Northeast (OR 0.67, 95% CI 0.48–0.92), and no difference in the Midwest (OR 1.26, 95%CI 0.85–1.86) compared to residence in the South. Intensifiers had a trend to be more likely to receive triple therapy with increasing age (multivariable OR 1.29, 95%CI 0.95–1.75). Significant clinical risk factors for triple therapy receipt among intensifiers were: at least one triple therapy nbDMARD at index date (OR 5.24, 95%CI 2.03–13.54), combination nbDMARD at index date (OR 5.56, 95%CI 4.04–7.65), and glucocorticoid use (OR 1.39, 95%CI 1.00–1.92). Increasing number of outpatient visits was associated with 5% decreased odds of receiving triple therapy (95%CI 2–8%). Year of index date was not significantly associated with triple therapy use among those who intensified therapy.
Results were similar when restricted to subjects who initiated at least one triple therapy nbDMARD (0.8% intensified to triple therapy) as well as when restricted to subjects who had at least two ICD-9 codes for RA (0.9% intensified to triple therapy). There was no increase in triple therapy use over calendar time; number of outpatient visits and glucocorticoid and NSAID use within 180 days of index date or prior were significantly associated with triple therapy use in these analyses.
DISCUSSION
Despite the growing evidence indicating similar efficacy and reduced cost of triple therapy compared to bDMARD, our large nationwide study indicates infrequent use of triple therapy in routine clinical practice. Triple therapy has demonstrated cost effectiveness compared bDMARD (4, 14). Despite this, only 0.7–0.9% of patients with RA were prescribed triple therapy, even though 89% were initially prescribed methotrexate, sulfasalazine, or hydroxychloroquine. Of those who intensified therapy to bDMARD or triple therapy, 93.8% were intensified to bDMARD. Our study period included the timeframe during which seminal randomized controlled trials that reported similar efficacy for triple therapy compared bDMARD were completed and results published (1–3). We found no evidence that triple therapy uptake subsequently increased. Even with guidelines advising to treat RA to the target of remission or low disease activity, triple therapy use remained an infrequent choice for pharmacologic therapy in RA throughout our study.
We evaluated whether there was a temporal change in triple therapy use by several strategies. The Swedish Pharmacotherapy (Swefot) trial was published in August 2009, the Treatment of Early Aggressive RA (TEAR) study was published in September 2012, and the RA Comparison of Active Therapies (RACAT) study was published in July 2013 (1–3). Of note, the Swefot trial reported that infliximab and methotrexate were superior to triple therapy for RA efficacy at month 12, but not at months 6, 9, or 24 so some interpret these results as favoring combination bDMARD and methotrexate over triple therapy (3). We investigated whether calendar year at index date was associated with intensification to triple therapy. The year 2013 approached statistical significance (unadjusted HR 1.59, 95%CI 0.99–2.55) compared to 2009, but was not statistically significant in multivariable models. When only analyzing those who intensified treatment to triple therapy or bDMARD, there was no association in any year compared to 2009. We also investigated rates of intensification at 6-month intervals throughout the entire study period and found no increase rate in triple therapy. Comparing before and after publication dates of the RACAT and TEAR studies also showed no increased in uptake of triple therapy prescription. Many of the studies that reported superior cost-effectiveness and similar quality of life measures of triple therapy compared to bDMARD were published near the end of our study period so may not have been able to influence the clinical uptake of triple therapy (4, 6). Since the index date was the first nbDMARD prescription for each patient, we were able to replicate an early RA cohort on initial nbDMARD therapy where later triple therapy use may be indicated. Since we had longitudinal data, the index date also captures the time period of patients with early RA so that we could accurately investigate secular trends of triple therapy use.
We examined a variety of factors to assess predictors of triple therapy use. We hypothesized that sociodemographic, geographic, and clinical factors may make prescribers less inclined to prescribe triple therapy. US region of residence was a significant factor predicting triple therapy use. In particular among those who intensified treatment, residence in the Northeast US was associated with 33% decreased use of triple therapy use compared to the South. Residence in the West (HR 1.43) and Midwest (HR 1.26) were also associated with increased risk of triple therapy use. Rural residence and area-level household income was not associated with triple therapy use. There was a trend for increasing age to be associated with triple therapy use. However, neither comorbidities nor a combined comorbidity score were associated with triple therapy use. RA patients who received glucocorticoids in the 180 days prior to the index date were more likely to receive triple therapy in the entire study sample as well as those how intensified therapy. Patients who are treated with glucocorticoids around RA diagnosis may have a more severe disease course. However, it is unclear why these patients were more likely to subsequently receive triple therapy than bDMARD. Despite the risk for serious infections with bDMARD, patients who received antimicrobials or had serious infections prior to the index date were not more likely to receive triple therapy (15).
The infrequent use of triple therapy demonstrated in our study may be due to several factors. Regional differences, glucocorticoid use, and number of outpatient visits were significant predictors for prescription of triple therapy in this analysis. Due to low number of triple therapy outcomes, we were limited in analyzing smaller geographic units or time periods. Clinical trials directly comparing triple therapy and bDMARD have primarily focused on efficacy outcomes with secondary analyses comparing triple therapy and bDMARD for other outcomes such as work loss and side effects. In analyses using data from the Swefot trial, those randomized to triple therapy had less work loss compared to those randomized to infliximab and methotrexate (a difference that was not statistically significant) (4). In RACAT, those treated with triple therapy had significantly increased gastrointestinal side effects compared to those on etanercept and methotrexate, but serious gastrointestinal side effects were similar in both groups (2). Subjects treated with etanercept and methotrexate had higher rates of serious infections compared to those treated with triple therapy (2). Research directly comparing triple therapy with bDMARD for outcomes other than efficacy may further elucidate factors that affect preferences for triple therapy or bDMARD among patients and clinicians.
Few previous studies have investigated triple therapy use in the routine clinical setting. A recent study using insurance claims data suggested that patients adhered to combination etanercept and methotrexate better than triple therapy (10). This is in contrast to reports in randomized controlled trials for similar or even enhanced adherence for triple therapy compared to bDMARD and highlights how findings from these trials do not always generalize to patients seen in typical care (2). Our study could not address whether adherence concerns were associated with triple therapy use. However, we did examine a variety of healthcare utilization factors. Increasing number of outpatient visits was inversely associated with intensification to triple therapy. Patients in our study were less likely to receive triple therapy for every outpatient visit. Patients requiring many appointments may have had a more severe RA disease course that prompted bDMARD by the treating clinicians. However, other healthcare utilization factors, such as rheumatologist appointment and emergency department visits, were not associated with triple therapy use.
Several limitations are important to note. Since our study included insurance beneficiaries throughout the entire US, our findings may reflect the routine clinical care for patients with RA. However, all patients in our study were insured so our findings may not be generalizable to underinsured or uninsured patients for whom generic oral drugs may be a preferable option. While we did not have access to clinical notes to verify RA diagnosis, we used methods with previously demonstrated validity (11). We were not able to evaluate some factors that may have influenced the decision to prescribe triple therapy such as functional status, disease activity, erosions, deformities, lifestyle, and previous side effects of nbDMARD. Since we used administrative claims data for analyses, we were unable to determine whether patients were being prescribed triple therapy or switching between nbDMARD drugs. Therefore, our findings may have actually overestimated the rate of triple therapy use. Since we report infrequent use of triple therapy, this limitation would not change the implications of our study.
In conclusion, we demonstrated that triple therapy was used infrequently in this large nationwide study of RA patients seen in routine clinical settings. We identified several factors associated with triple therapy use, such as region of residence, outpatient visits, and glucocorticoid use. The use of triple therapy in typical clinical practices in the US remains low despite the data showing that triple therapy is non-inferior to bDMARD in RA patients with active disease after initial monotherapy treatment with nbDMARDs such as methotrexate. Future studies investigating barriers and facilitators for triple therapy use for RA treatment are warranted.
Acknowledgments
Funding: This work was supported by an unrestricted research grant from CVS Health to Brigham and Women’s Hospital. Dr. Sparks was supported by the Rheumatology Research Foundation Scientist Development Award and the National Institutes of Health Loan Repayment Award (L30 AR066953). Dr. Matlin and Dr. Shrank are employed by CVS Health. Dr. Solomon’s time was funded by the National Institutes of Health (K24 AR055989). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Sparks had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Sparks, Shrank, Solomon.
Acquisition of data. Sparks, Krumme, Shrank, Pezalla, Choudhry, Solomon.
Analysis and interpretation of data. Sparks, Krumme, Shrank, Matlin, Brill, Pezalla, Choudhry, Solomon.
Potential conflicts of interest:
JAS received an honorarium from CVS Minute Clinic for a presentation unrelated to the focus of this manuscript. DHS receives research support through grants to his institution from Amgen, Lilly, AstraZeneca, Pfizer, Genentech, and CORRONA. NKC receives funding from Merck, AstraZeneca, and Sanofi to study topics (medication adherence) unrelated to the focus of the current manuscript. WHS and OSM are employed by CVS Health. EJP is employed by Aetna.
References
- 1.Moreland LW, O’Dell JR, Paulus HE, Curtis JR, Bathon JM, St Clair EW, et al. A randomized comparative effectiveness study of oral triple therapy versus etanercept plus methotrexate in early aggressive rheumatoid arthritis: the treatment of Early Aggressive Rheumatoid Arthritis Trial. Arthritis Rheum. 2012;64(9):2824–35. doi: 10.1002/art.34498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.O’Dell JR, Mikuls TR, Taylor TH, Ahluwalia V, Brophy M, Warren SR, et al. Therapies for active rheumatoid arthritis after methotrexate failure. N Engl J Med. 2013;369(4):307–18. doi: 10.1056/NEJMoa1303006. [DOI] [PubMed] [Google Scholar]
- 3.van Vollenhoven RF, Ernestam S, Geborek P, Petersson IF, Coster L, Waltbrand E, et al. Addition of infliximab compared with addition of sulfasalazine and hydroxychloroquine to methotrexate in patients with early rheumatoid arthritis (Swefot trial): 1-year results of a randomised trial. Lancet. 2009;374(9688):459–66. doi: 10.1016/S0140-6736(09)60944-2. [DOI] [PubMed] [Google Scholar]
- 4.Eriksson JK, Karlsson JA, Bratt J, Petersson IF, van Vollenhoven RF, Ernestam S, et al. Cost-effectiveness of infliximab versus conventional combination treatment in methotrexate-refractory early rheumatoid arthritis: 2-year results of the register-enriched randomised controlled SWEFOT trial. Ann Rheum Dis. 2015;74(6):1094–101. doi: 10.1136/annrheumdis-2013-205060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Karlsson JA, Neovius M, Nilsson JA, Petersson IF, Bratt J, van Vollenhoven RF, et al. Addition of infliximab compared with addition of sulfasalazine and hydroxychloroquine to methotrexate in early rheumatoid arthritis: 2-year quality-of-life results of the randomised, controlled, SWEFOT trial. Ann Rheum Dis. 2013;72(12):1927–33. doi: 10.1136/annrheumdis-2012-202062. [DOI] [PubMed] [Google Scholar]
- 6.Wailoo A, Hernandez MA, Scott IC, Ibrahim F, Scott DL. Cost-effectiveness of treatment strategies using combination disease-modifying anti-rheumatic drugs and glucocorticoids in early rheumatoid arthritis. Rheumatology (Oxford) 2014 doi: 10.1093/rheumatology/keu039. [DOI] [PubMed] [Google Scholar]
- 7.Smolen JS, Breedveld FC, Burmester GR, Bykerk V, Dougados M, Emery P, et al. Treating rheumatoid arthritis to target: 2014 update of the recommendations of an international task force. Ann Rheum Dis. 2015 doi: 10.1136/annrheumdis-2015-207524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Singh JA, Furst DE, Bharat A, Curtis JR, Kavanaugh AF, Kremer JM, et al. 2012 update of the 2008 American College of Rheumatology recommendations for the use of disease-modifying antirheumatic drugs and biologic agents in the treatment of rheumatoid arthritis. Arthritis Care Res (Hoboken) 2012;64(5):625–39. doi: 10.1002/acr.21641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.de Jong PH, Hazes JM, Han HK, Huisman M, van Zeben D, van der Lubbe PA, et al. Randomised comparison of initial triple DMARD therapy with methotrexate monotherapy in combination with low-dose glucocorticoid bridging therapy; 1-year data of the tREACH trial. Ann Rheum Dis. 2014;73(7):1331–9. doi: 10.1136/annrheumdis-2013-204788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bonafede M, Johnson BH, Tang DH, Shah N, Harrison DJ, Collier DH. Etanercept-methotrexate combination therapy initiators have greater adherence and persistence than triple therapy initiators with rheumatoid arthritis. Arthritis Care Res (Hoboken) 2015 doi: 10.1002/acr.22638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kim SY, Servi A, Polinski JM, Mogun H, Weinblatt ME, Katz JN, et al. Validation of rheumatoid arthritis diagnoses in health care utilization data. Arthritis Res Ther. 2011;13(1):R32. doi: 10.1186/ar3260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gagne JJ, Glynn RJ, Avorn J, Levin R, Schneeweiss S. A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–59. doi: 10.1016/j.jclinepi.2010.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Patkar NM, Curtis JR, Teng GG, Allison JJ, Saag M, Martin C, et al. Administrative codes combined with medical records based criteria accurately identified bacterial infections among rheumatoid arthritis patients. J Clin Epidemiol. 2009;62(3):321–7. 7 e1–7. doi: 10.1016/j.jclinepi.2008.06.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bonafede M, Johnson BH, Princic N, Shah N, Harrison DJ. Cost per patient-year in response using a claims-based algorithm for the 2 years following biologic initiation in patients with rheumatoid arthritis. J Med Econ. 2015;18(5):376–89. doi: 10.3111/13696998.2014.1001849. [DOI] [PubMed] [Google Scholar]
- 15.Bongartz T, Sutton AJ, Sweeting MJ, Buchan I, Matteson EL, Montori V. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infections and malignancies: systematic review and meta-analysis of rare harmful effects in randomized controlled trials. JAMA. 2006;295(19):2275–85. doi: 10.1001/jama.295.19.2275. [DOI] [PubMed] [Google Scholar]