Differences in time to drug initiation and inactive disease achievement by insurance type in patients with nonsystemic juvenile idiopathic arthritis Authors: Elaine Yung, Xiaoxuan Liu, Minal Aundhia, Chen Chen, Michael Wagner, Bin Huang, Hermine I Brunner Published: 2026 ABSTRACT AbstractIntroductionJuvenile idiopathic arthritis (JIA) is a chronic autoimmune condition that can lead to significant complications if not promptly managed. Current guidelines recommend biologic or targeted synthetic disease modifying antirheumatic drugs (b/tsDMARDs) as part of the treatment algorithm. In the USA, patients are covered by a wide range of insurance plans that can be broadly classified into two categories—public or private. The influence of insurance type on access to JIA therapies and subsequent outcomes remains poorly understood. The objective of this study was to investigate how b/tsDMARD initiation and achievement of clinically inactive disease (CID) may differ by insurance type in patients with nonsystemic forms of JIA.MethodsPatients newly diagnosed with JIA according to International League of Associations (ILAR) for Rheumatology classification criteria between 2009 and 2020 at a large tertiary center were studied. Outcome variables of time to b/tsDMARD initiation and time to CID of JIA were assessed. Descriptive statistics were performed, and patients with public or private insurance were compared cross-sectionally and over the study timeframe using survival analysis.ResultsThere were 968 patients with newly diagnosed JIA during the study period; 753 (78%) were privately insured and 215 (22%) were publicly insured. During 5898 patient-years of follow-up, 562 (58%) patients initiated b/tsDMARDs and 677 (70%) patients achieved CID. Time to b/tsDMARD initiation did not differ by insurance type (both approximately 5 months, p = 0.51). Publicly insured patients had higher baseline disease activity and longer time to CID achievement compared with the privately insured (10 months versus 7 months, p < 0.001).ConclusionWhile time to b/tsDMARD initiation was similar across insurance groups, publicly insured patients achieved CID later than privately insured patients. These findings may be interpreted as insurance-associated differences. Additional studies are needed to clarify the roles of baseline disease severity, healthcare access, adherence, and socioeconomic context factors.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40267-026-01241-0. Key Points Insurance type does not impact the time to biologic or targeted synthetic disease modifying antirheumatic drug therapy initiation in patients with juvenile idiopathic arthritis.Despite similar initiation times to biologic or targeted synthetic disease modifying antirheumatic drug therapy, clinical inactive disease is achieved at a significantly later time in the publicly insured compared with those privately insured.Reasons for the observed differences in outcomes in juvenile idiopathic arthritis among publicly versus privately insured children require further investigation. Introduction Juvenile idiopathic arthritis (JIA) refers to a group of autoimmune chronic inflammatory conditions resulting in arthritis prior to age 16 years. JIA can be classified into subtypes according to the revised criteria of the International League of Association for Rheumatology (ILAR) [1]. If untreated, JIA can result in permanent structural joint damage, leading to functional limitations including limb length discrepancy, retrognathia, ocular issues, psychosocial impact, and many additional complications negatively affecting quality of life [2–4]. The 2021 American College of Rheumatology Guidelines for JIA advise that, in the setting of failure of a conventional disease modifying antirheumatic drug (cDMARD), such as methotrexate, adding a biologic disease modifying antirheumatic drug (bDMARD) is strongly recommended [2]. There exists several bDMARDs with differing mechanism of actions, such as tumor necrosis factor inhibitors (TNFi) and agents that target a variety of interleukins (IL). Although several bDMARDs are commercially available, there is no set guidance on which bDMARD medication class should be utilized [2]. Janus kinase (JAK) inhibitors, such as tofacitinib and upadacitinib, are relatively newer therapies for JIA management that are targeted synthetic disease modifying antirheumatic drugs (tsDMARD) [2]. It is dependent on the provider and patient to determine which b/tsDMARD is suitable for their needs where studies have shown that early initiation of bDMARDs often leads to better JIA outcomes, including a higher likelihood of achieving clinical remission [2, 5, 6]. Within the United States of America (USA), there is a large array of insurance groups and insurance plans that provide coverage of b/tsDMARDs. Broadly, insurance groups and insurance plans can be classified into two categories: public and private. Public insurance includes groups that are government funded, such as Medicaid [5, 6]. Such insurance is based on a variety of factors not limited to income level, household size, family status, and disability [7–9]. Private insurance includes commercial insurance groups, available typically through employers or self-purchased, such as Marketplace groups [6, 7]. Between 1997 and 2018, the percentage of children with private insurance has decreased where the percentage enrolled in public insurance has increased [7]. In the USA, b/tsDMARDs are often of high-cost and require prior authorizations (PAs) [10]. Upon receipt of a PA, the insurance group review occurs with a decision on whether the cost of the medication should be covered based on the individual plan’s formulary [10]. Formulary coverage limitations and the requirement for PAs often result in delays in therapy initiation, with the potential of short- and long-term negative impacts on patient clinical outcomes [5, 10, 11]. Limited evidence exists regarding the impact of public versus private insurance on time to b/tsDMARD initiation and achievement of clinically inactive disease (CID). Given that insurance plans have variable formularies and authorization processes, timing of treatment and outcomes may differ between insurance groups. Investigating insurance-associated differences in time to b/tsDMARD initiation and CID achievement is important to characterize variation in treatment pathways and outcomes among patients with JIA. The primary objective of this study was to investigate how b/tsDMARD initiation and achievement of CID differ by insurance type in patients with nonsystemic forms of JIA. Methods JIA categorization Electronic Health Record (EHR) data collected between 1 January 2009 to 31 August 2020 were retrospectively extracted for Cincinnati Children’s Hospital Medical Center (CCHMC) for patients newly diagnosed with one of the following categories of nonsystemic JIA: polyarticular JIA, rheumatoid factor positive or negative (polyJIA RF+, polyJIA RF−); oligoarticular JIA, persistent or extended (oJIA); juvenile psoriatic arthritis (jPsA); enthesitis-related arthritis (ERA), and undifferentiated arthritis [2, 4]. PolyJIA RF+ and polyJIA RF− subtypes were combined as “polyJIA.” oJIA persistent and extended were combined as “oJIA.” jPsA, ERA and undifferentiated arthritis were grouped as “Other JIA subtype.” The JIA subtype categories assessed were polyJIA, oJIA, and Other JIA subtype. Medications and study variables The disease modifying antirheumatic drugs (DMARDs) considered in the study were abatacept, JAK inhibitors (tsDMARDs), and all marketed bDMARDs that block tumor necrosis factor (TNF), interleukin (IL)6, IL12, IL17, or IL23. The combination of these medications included within the study was termed as b/tsDMARDs. Patients with JIA with missing diagnosis dates, children diagnosed prior to 1 January 2009, lack of a clinical visit within 12 months following diagnosis, lack of insurance information or dual insurance type at diagnosis, and children taking cDMARDs and/or b/tsDMARDs before their JIA diagnosis were excluded (Fig. 1). Information collected prospectively in structured flowsheets included: JIA category, age at diagnosis, physician global assessment (PGA) of JIA activity, enthesitis involvement, active joint count, start dates for the first bDMARD (abatacept, adalimumab, baricitinib, certolizumab-pegol, etanercept, golimumab, guselkumab, infliximab and corresponding biosimilars, ixekizumab, risankizumab-rzaa, secukinumab, tocilizumab, or ustekinumab) or tsDMARD (tofacitinib or upadacitinib), and insurance type. Data was extracted from the CCHMC EHR for eligible patients, starting from the date of JIA diagnosis. The study was approved by the CCHMC Institutional Review Board (#2021-0242). Study inclusion and exclusion criteria. A total of 968 patients were included in the final analysis after excluding patients with missing data, not falling within the age range, being administered b/tsDMARD therapy prior to diagnosis and those with systemic JIA subtype. §Numbers reported in each exclusion criteria are not mutually exclusive. b/tsDMARD biologic or targeted synthetic disease modifying anti-rheumatic drug, JIA juvenile idiopathic arthritis Scoring tools and measurements of CID Extracted EHR data was used to compute the 10-joint clinical Juvenile Arthritis Disease Activity Score (cJADAS10) to measure disease activity. CID was defined using criteria derived from the Wallace framework and adapted for ERA/jPsA: active joint count of 0, PGA score of 0, and absence of enthesitis in those with ERA or jPsA [12]. This approach may be preferable to cJADAS10 thresholds for this study because it relies on objective physician-derived measures consistently captured in the EHR. Time to first CID was defined as the interval from diagnosis to the first follow-up visit at which these criteria were met. Insurance categorization Health insurance information was categorized into public or private on the basis of the type of insurance at the start of diagnosis. Patients without insurance or who had combination insurance types at diagnosis were excluded from the study. Statistical analysis Descriptive analyses were performed with continuous variables summarized as medians and interquartile ranges (IQRs), and categorical variables summarized as frequencies and proportions. Participants were not excluded owing to limited or absent follow-up after diagnosis; participants without follow-up encounters were censored at 12 months. At each follow-up visit, patients were assessed for first b/tsDMARD prescription and CID achievement. Group comparisons used Wilcoxon rank-sum tests (two groups), Kruskal–Wallis tests (> 2 groups), or chi-square tests, as appropriate. For time-to-event outcomes, Kaplan–Meier (KM) curves and log-rank tests were used. Cox proportional hazards models evaluated associations with time to b/tsDMARD initiation and time to CID, adjusting for sex, age at diagnosis, duration of symptoms at diagnosis, JIA subtype, baseline cJADAS10, and insurance type. Cox models were conducted using Python 3.9 with statistical significance at alpha = 0.05 (two-tailed). Results Patients and outcomes The analysis included 968 eligible patients with JIA, corresponding to 5898 patient-years of follow-up. As shown in Table 1, most patients were privately insured (n = 753, 78%) and had a median of nine follow-up rheumatology visits during the study period. The distribution of JIA subtypes was similar between insurance groups (p = 0.73). At baseline, disease activity was significantly higher in publicly insured patients (median cJADAS10 13.5 versus 10.5, p = 0.007). Irrespective of insurance status, 562 (58%) patients initiated a b/tsDMARD after diagnosis and 677 (70%) achieved CID (Table 2). Among patients with oJIA, the proportion classified as extended oJIA was higher in the publicly insured group compared with the privately insured group (19.5% versus 13.3%); however, this difference was not statistically significant (p = 0.211). Table 1Baseline characteristics by insurance type†Baseline characteristicsPublic insurance (n = 215)Private insurance (n = 753)p valueAge at diagnosis (years)10.9 (5.7, 14.5)11.4 (6.3, 15.2)0.22Symptoms to diagnosis (years)0.5 (0.3, 1.4)0.45 (0.2, 1.5)0.58cJADAS10 at diagnosis13.5 (8, 17; n = 125)10.5 (6, 15.5; n = 425)0.007Polyarticular JIA (n = 158)18 (16, 21; n = 32)17 (12, 21; n = 126)0.26Oligoarticular JIA (n = 294)10 (6, 14; n = 40)8 (5, 11; n = 139)0.004Other JIA subtypes (n = 451)13 (8, 16; n = 53)10 (6, 15; n = 160)0.09Follow-up visits8 (4, 14)9 (4, 17)0.12Follow-up duration (years)4.7 (2.9, 7.2)5.1 (2.7, 7.4)0.43RaceWhite or Caucasian164 (77%)711 (95%) < 0.001 Black or African American35 (16%)15 (2%)Other14 (7%)21 (3%)JIA subtypePolyarticular RF (+), Polyarticular RF (−),62 (29%)201 (27%)0.73With uveitis2 (3%)15 (7%)Oligoarticular66 (31%)241 (32%)With uveitis4 (6%)17 (7%)Other87 (40%)311 (41%)With uveitis4 (5%)15 (5%)UveitisNo205 (95%)706 (94%)0.48Yes10 (5%)47 (6%)b/tsDMARD initiated after diagnosisNo81 (38%)325 (43%)0.17Yes134 (62%)428 (57%)CID achievedNo77 (36%)214 (28%)0.45Yes138 (64%)539 (72%)The study included a total of 968 patients with a larger majority having private insurance. There were significant differences between those with public and private insurances noted for baseline cJADAS10 at time of diagnosis and racial groupingsb/tsDMARD biologic or targeted synthetic disease modifying antirheumatic drug, CID clinical inactive disease, cJADAS10 clinical Juvenile Arthritis Disease Activity Score, JIA juvenile idiopathic arthritis, IQR interquartile range, RF rheumatoid factor†p value of χ2 for categorical variables, t test for numerical variables. Continuous variable was represented by median (IQR1, IQR3). Categorical variable was represented by count (percentage) Baseline characteristics by insurance type† The study included a total of 968 patients with a larger majority having private insurance. There were significant differences between those with public and private insurances noted for baseline cJADAS10 at time of diagnosis and racial groupings b/tsDMARD biologic or targeted synthetic disease modifying antirheumatic drug, CID clinical inactive disease, cJADAS10 clinical Juvenile Arthritis Disease Activity Score, JIA juvenile idiopathic arthritis, IQR interquartile range, RF rheumatoid factor †p value of χ2 for categorical variables, t test for numerical variables. Continuous variable was represented by median (IQR1, IQR3). Categorical variable was represented by count (percentage) Table 2Days to b/tsDMARD initiation and CID stratified by JIA subtype and insurance typeJIA subtypePublicPrivatep valuePublicPrivatep valueDays to b/tsDMARDDays to CIDPolyarticular JIA (n = 158)125 (80, 208)126 (91,174)0.44378 (343, 560)328 (273, 385)0.01Oligoarticular JIA (n = 294)725 (312, 2450)2093 (1542, 3329)0.05364 (241, 502)208 (175, 248)< 0.005Other JIA subtypes (n = 451)156 (83, 406)147 (111, 219)0.94196 (149, 308)166 (141, 202)0.06JIA subtype stratification by insurance type, baseline cJADAS10, time to b/tsDMARD initiation (days, represented by median, IQR), and time to CID achievement. Both patient insurance type groups had no significant differences in time to b/tsDMARD initiation for each JIA subtype. Days to CID achievement were consistently lower for each JIA subtype in those with private insurance compared to public insurance with significant differences noted in those with polyarticular and oligoarticular JIA subtypes.b/tsDMARD biologic or targeted synthetic disease modifying anti-rheumatic drug, CID clinical inactive disease, cJADAS10 clinical Juvenile Arthritis Disease Activity Score, JIA juvenile idiopathic arthritis, IQR interquartile range Days to b/tsDMARD initiation and CID stratified by JIA subtype and insurance type JIA subtype stratification by insurance type, baseline cJADAS10, time to b/tsDMARD initiation (days, represented by median, IQR), and time to CID achievement. Both patient insurance type groups had no significant differences in time to b/tsDMARD initiation for each JIA subtype. Days to CID achievement were consistently lower for each JIA subtype in those with private insurance compared to public insurance with significant differences noted in those with polyarticular and oligoarticular JIA subtypes. b/tsDMARD biologic or targeted synthetic disease modifying anti-rheumatic drug, CID clinical inactive disease, cJADAS10 clinical Juvenile Arthritis Disease Activity Score, JIA juvenile idiopathic arthritis, IQR interquartile range Time to bDMARD initiation and time to CID achievement KM analysis revealed that there was no significant difference between the insurance types with both at a median of 5 months to b/tsDMARD initiation [p = 0.51, (HR 0.93, 95% CI 0.73–1.18; p = 0.55)] as shown in Fig. 2A. Conversely, as shown in Fig. 2B, the privately insured patients achieved CID significantly earlier in comparison with publicly insured patients [7 months versus 10 months; p-value < 0.001, (HR 1.38, 95% CI 1.1–1.72; p < 0.005)]. Insurance Comparison of Survival Analysis of Days to b/tsDMARD Initiation and CID Achievement. Kaplan–Meier curves comparing time to biologic (b/tsDMARD) initiation and time to clinically inactive disease (CID) among patients with private versus public insurance. Figure 2A: There was no significant difference between the insurance types in time to b/tsDMARD initiation with both at 5 months. Figure 2B: Following b/tsDMARD initiation, patients with private insurance achieved CID significantly faster than those with public insurance (7 months vs. 10 months, p value < 0.005). Shaded areas represent 95% confidence intervals. b/tsDMARD biologic or targeted synthetic disease modifying anti-rheumatic drug, CID clinically inactive disease, cJADAS10 clinical Juvenile Arthritis Disease Activity Score Determinants of b/tsDMARD initiation In adjusted Cox models, patients with more active disease were escalated to b/tsDMARDs more rapidly (HR 1.07, 95% CI 1.05–1.09, p < 0.001), and older age at diagnosis was associated with faster initiation (HR 1.04, 95% CI 1.01–1.06, p = 0.005). Insurance type (HR 0.99, 95% CI 0.75–1.31, p = 0.936), and duration of symptoms at diagnosis (HR 1.02, 95% CI 0.97–1.08, p = 0.369) were not significantly associated with b/tsDMARD initiation. Full model coefficients are provided in the online supplementary resource, Table 3. Determinants of CID achievement Patients with private insurance were more likely to achieve CID earlier than those with public insurance (HR 1.50, 95% CI 1.15–1.95, p < 0.005). Factors that were associated with lower percentages of patients achieving CID included higher cJADAS10 at diagnosis (HR 0.96, 95% CI 0.94–0.98, p < 0.001) and a longer duration of JIA symptoms prior to diagnosis (HR 0.95, 95% CI 0.90–0.99, p = 0.02). There was not a significant difference between JIA subtypes in ever reaching CID (p = 0.36). No other comparisons, including age at diagnosis (p = 0.21), were significantly associated with time to CID. Full model coefficients are provided in the online supplementary resource, Table 4. Discussion At our tertiary pediatric rheumatology center, insurance type was not associated with time to initiating a b/tsDMARD, whereas time to CID differed between insurance groups. Publicly insured patients also had higher baseline disease activity, which likely contributed, at least in part, to slower CID achievement rather than delays in treatment escalation. Achievement of CID was more likely in patients with shorter symptom duration before diagnosis and lower cJADAS10 at diagnosis. Public insurance was more commonly held by non-White children than private insurance in this cohort; the observed demographic distribution is consistent with previously published JIA cohorts [7, 13]. Our findings are broadly consistent with the Childhood Arthritis and Rheumatology Research Alliance (CARRA) registry analyses reporting similar TNFi use across insurance groups (n = 1086), despite higher baseline disease activity among publicly insured patients [3, 6]. In our cohort, publicly insured patients also had higher baseline cJADAS10, while time to b/tsDMARD initiation remained similar across insurance groups. These findings support interpreting the results as insurance-associated differences in outcomes rather than definitive evidence of treatment initiation bias or inequitable access to therapy. Differences in access to care and other unmeasured factors may also contribute and require further study. We observed that publicly insured patients initiated b/tsDMARDs at similar timeframes to privately insured patients but achieved CID later. Prior studies suggest social and structural determinants may influence disease control in JIA [6]. These factors—including healthcare access, therapy adherence, and financial or administrative barriers, were not directly measured in this study and may contribute to the observed differences in CID timing [14, 15]. Often, patients with public insurance have fewer financial resources compared to those with private insurance, which may contribute to this disparity in achievement in CID, such as through challenges related to medication affordability, insurance coverage limitations, and administrative barriers that can affect treatment continuity and long-term disease management. In addition, disruptions related to coverage changes or eligibility renewal may create intermittent gaps in care that hinder sustained disease control despite appropriate therapeutic escalation. When examining the JIA subtypes individually, we observed patterns consistent to the overall cohort. No significant differences in time to b/tsDMARD initiation were seen between insurance type, but clear differences in time to achieving CID were identified for each JIA subtype. Reflective of the baseline characteristics, all JIA subtypes and public insurance together exhibited higher baseline cJADAS10 scores compared with the privately insured. Owing to higher active joint count involvement at diagnosis and potentially symptom severity, it was expected that patients with the polyJIA subtype would be escalated to b/tsDMARDs therapy more quickly than those with oJIA and Other JIA subtypes. This was recognized in the present study in those with polyJIA initiating b/tsDMARD therapy generally faster compared with children with the other JIA categories. Among patients with oJIA, publicly insured children had a higher proportion of extended oJIA, suggesting that subtype heterogeneity may have contributed to the observed delay in CID achievement. Higher baseline disease activity among publicly insured patients may have partly driven slower CID as cJADAS10 may not fully capture all aspects of disease severity. Therefore, the observed differences likely reflect a combination of baseline severity and subtype rather than insurance status alone. Overall, the data highlighted a consistent pattern in which higher baseline disease activity and longer time to CID disproportionately affect publicly insured children across multiple JIA subtypes. Additional evaluation is required to distinguish the effects of CID achievement stratified by both insurance type and JIA subtype [16]. A key limitation of this study is that the time to b/tsDMARD initiation and CID achievement is influenced by numerous factors beyond solely insurance type. Important confounders not captured in our dataset, such as socioeconomic resources, including financial resources, access to specialty rheumatology care, patient and caregiver education, and adjunctive therapies. Additionally, the absence of detailed active joints affected may obscure clinical scenarios in which more aggressive initial therapy would have been justified. The confounders mentioned and detailed active joint activity were beyond the scope of the present study that was designed specifically to evaluate the impact of insurance type in patients with JIA on their access to treatment and time to achieve clinical outcomes. Another limitation is that the exact date of b/tsDMARD acquisition through prescription dispensing records and true treatment start date were not available. Despite these limitations, strengths of the study were the long observational period that enabled the inclusion of a large population-based cohort. Additionally, the inclusion of a broad range of both bDMARD and tsDMARD therapies strengthens the generalizability of the findings. Overall, while insurance type was not associated with time to b/tsDMARD initiation, children with public insurance experienced longer time to CID achievement. These findings should be interpreted as insurance-associated differences, potentially influenced by baseline disease severity and unmeasured social or healthcare access factors. Conclusions In this large, single-center study of JIA, insurance type was not associated with timing of b/tsDMARD initiation. However, publicly insured children experienced longer time to CID despite similar initiation timeframes. These results indicate insurance-associated differences in outcomes and support further research to clarify the effects of baseline severity, access, adherence, and socioeconomic context on disease control. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 19 kb) Acknowledgements The authors thank the patients, caregivers, investigators, and Cincinnati Children’s Hospital Rheumatology Clinic staff who participated in the study. A special thank you to Ms Angela Merritt for her research regulatory support. Funding The project described was supported by the National Institutes of Arthritis and Musculoskeletal Skin Diseases under Award-Number P30AR076316. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Declarations 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. All authors had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. Conflict of interest Authors have no conflicts of interest relevant to this article to disclose. Availability of data and material Not applicable. Ethics approval The study was approved by the CCHMC Institutional Review Board (#2021-0242). Consent to participate Not applicable. Consent for publication Not applicable. Code availability Not applicable. Footnotes Elaine Yung and Xiaoxuan Liu were dual first authors and contributed equally to this work. References Associated Data Supplementary Materials Supplementary file1 (DOCX 19 kb)