Swedish Rheumatology Quality Register Research Articles

Efficacy and safety of rituximab in anti-synthetase antibody positive and negative subjects with idiopathic inflammatory myopathy: a registry-based study.

Post-hoc analyses of the Rituximab in Myositis trial indicate that specific autoantibodies profiles may influence treatment response. We compared the efficacy and safety of rituximab in anti-synthetase antibody (ARS-ab) positive and negative patients. Adult idiopathic inflammatory myopathy (IIM) subjects in the Swedish Rheumatology Quality Register who received ⩾ 1 cycle of rituximab were enrolled. Efficacy assessment was based on the International Myositis Assessment and Clinical Studies (IMACS) core set measures and the 2016 ACR/EULAR definition of improvement for PM and DM. Safety assessment included drug-related adverse event and death during study period. Comparisons were done within and between the ARS-ab defined groups before and after first and last cycles. Associations between selected clinical features and improvement after one rituximab cycle were assessed using logistic regression. Sixty-five subjects were included and 43 had a follow-up visit within 5-10 months. Seventy-eight percent of ARS-ab positive subjects had moderate/major ACR/EULAR improvement after one cycle compared with 50% in the ARS-ab negative group. After several cycles, 79% of the ARS-ab positive and 67% of the ARS-ab negative patients achieved moderate/major improvement. A significant glucocorticoid-sparing effect was only observed in the ARS-ab positive group (P = 0.001). The most frequent adverse events were infections. One ARS-ab positive and two ARS-ab negative patients died during follow-up period. Irrespectively of their autoantibody status, a majority of subjects treated with several rituximab cycles had moderate/major improvement. In addition, ARS-ab positive subjects experienced a significant glucocorticoid-sparing effect.

Performance of the Expanded Cardiovascular Risk Prediction Score for Rheumatoid Arthritis in a geographically distant National Register-based cohort: an external validation

BackgroundCardiovascular (CV) risk stratification for patients with rheumatoid arthritis (RA) should facilitate evidence-based management. Prior work has derived an internally validated a CV risk score, the Expanded Cardiovascular Risk Prediction...

Risk of Thyroxine-Treated Autoimmune Thyroid Disease Associated With Disease Onset in Patients With Rheumatoid Arthritis

ImportanceAutoimmune thyroid disease ([AITD] including hypothyroidism and hyperthyroidism) is the most common organ-specific autoimmune disorder and is more prevalent among patients with rheumatoid arthritis (RA). Real-world studies on when and how this increased risk of AITD develops, in association with the time before or after the onset of RA, are lacking.ObjectiveTo estimate the risk of thyroxine-treated AITD among patients with RA at different time points before and after the diagnosis of RA.Design, Setting, and ParticipantsA nationwide register-based case-control and cohort study was conducted between January 1, 2006, and June 30, 2013, with a maximum follow-up time of 7 years before and 8 years after diagnosis of RA. The study used the Swedish Rheumatology Quality Register and linkage to other nationwide registers to identify 8090 adults with new-onset RA and a random population-based sample of 80 782 referents matched by age, sex, and residential area. Statistical analysis was performed from July 1, 2015, to June 30, 2017.ExposuresPresence of AITD in the participants in the case-control design and RA in the participants in the cohort design.Main Outcomes and MeasuresPrevalence and relative risk of incident AITD before (odds ratios) and after (hazard ratios) diagnosis of RA compared with the population as reference.ResultsThere were 8090 patients with RA (5529 women and 2561 men; mean [SD] age, 58.3 [15.2] years) and 80 782 population-based participants as reference who were identified. By the time of diagnosis of RA, the prevalence of AITD was 10.3% among the patients with RA (n = 832) vs 7.1% among the controls (5725 of 80 350) (odds ratio, 1.5; 95% CI, 1.4-1.7). This increased risk of AITD developed during the 5 years (range, 2-5 years) before diagnosis of RA (odds ratio, 1.5; 95% CI, 1.2-1.8) and peaked by the time of diagnosis of RA (range, 0-3 months before diagnosis of RA) (odds ratio, 5.3; 95% CI, 3.7-7.6). From diagnosis of RA and onward, the risk of developing AITD decreased (range, 2-5 years after diagnosis of RA) (hazard ratio, 0.7; 95% CI, 0.5-1.0).Conclusions and RelevanceCompared with the general population, Swedish patients with RA appear to have a higher prevalence of thyroxine-treated AITD at diagnosis of RA and an increased incidence of AITD during the 5-year period before diagnosis of RA. After diagnosis of RA, the risk of developing AITD is suggested to decrease below the expected rate. Besides temporal changes in diagnostic intensity, this pattern of risk raises the question whether AITD may influence the pathogenesis of RA (or vice versa) and, conversely, the question whether antirheumatic therapies may prevent AITD.

Impact of extra-articular spondyloarthritis manifestations and comorbidities on drug retention of a first TNF-inhibitor in ankylosing spondylitis: a population-based nationwide study

ObjectivesTo assess the impact of extra-articular spondyloarthritis (SpA) manifestations (anterior uveitis, psoriasis and inflammatory bowel disease (IBD)), and of comorbidities, on tumour necrosis factor alpha inhibitor (TNFi) drug retention in...

Associations between fatigue and physical capacity in people moderately affected by rheumatoid arthritis

To explore the contribution of physical capacity in explaining variations in fatigue among people with rheumatoid arthritis (RA). This study included participants recruited for a physical activity intervention. Data were collected from the Swedish Rheumatology Quality Registers, from questionnaires on fatigue, activity limitation, perceived health, pain and anxiety/depression and from physical capacity tests (lower limb function, grip strength, and aerobic capacity). We used logistic regression to estimate the association between severe fatigue (≥ 50, visual analogue scale 0–100) and (A) independent variables related to disease and disease impact and (B) model A plus physical capacity tests. Pooled odds ratio tests compared model fit. Out of the 269 participants (mean age 60 years, mean disease activity score [DAS28] 2.8), severe fatigue was reported by 35%. The three variables which were statistically significantly associated with severe fatigue (p < 0.05) in both models were perceived health, pain and anxiety/depression. Anxiety/depression demonstrated the largest effect size with odds ratios of 2.43 (95% CI 1.20, 4.94) in model A and 2.58 (95% CI 1.25, 5.32) in model B. The likelihood ratio test indicated that model B was a better fit to the data than model A with Χ2 (df 3) = 2.65, p = 0.048. Severe fatigue in people with RA is associated with self-rated health, pain and anxiety/depression rather than with physical capacity. Future studies should be prospective, use multidimensional assessments of fatigue to explore the influence of physical capacity and control for possible influence of comorbidities associated with fatigue.

Uptake of rheumatology biosimilars in the absence of forced switching

ABSTRACTBackground: To describe the uptake and system-level effects of the introduction of biosimilars in a setting without forced switching.Research design and methods: We used data from the Swedish Rheumatology Quality register from start of marketing of infliximab (Remsima® and Inflectra®) and etanercept (Benepali®) biosimilars until 31 December 2016. We compared users of each originator-product and its biosimilar(s) by line of treatment: bDMARD-naïve patients, non-medical switchers (vs. matched patients remaining on originator), and patients switching from a previous bDMARD of another type.Results: From the start of marketing 1343 patients started an infliximab biosimilar (22 months) and 2691 started etanercept (9 months). Overall, the introduction of these biosimilars resulted in an increase of the total number of ongoing infliximab and etanercept treatments (originator + biosimilar) . At the end of the study period, biosimilars accounted for 31% of all infliximab treatments and 31% of all etanercept-treated patients. For each line of therapy, we noted only small differences in patient characteristics between those starting the originator product vs. its biosimilar(s).Conclusions: Introduction of biosimilars have effects beyond replacement of the originator product, in terms of an increased rate of bDMARD initiation. Selection to non-medical switching displayed no particular disease- or patient-characteristics.

Prevalence of sustained remission in rheumatoid arthritis: impact of criteria sets and disease duration, a Nationwide Study in Sweden.

The aims of this national study in Sweden of patients with RA were to: examine the prevalence of sustained remission (SR), that is, remission lasting for at least 6 months; compare the prevalence of SR in patients with early RA and established RA; study the timing of onset of and time spent in SR; and study possible predictors of SR. Adult patients with RA included in the Swedish Rheumatology Quality registry were studied. The registry was searched for patients fulfilling remission criteria: DAS28-ESR, Clinical Disease Activity Index (CDAI), Simplified Disease Activity Index (SDAI) and ACR/EULAR remission for at least 6 months. Early RA was defined as symptom duration ⩽6 months at inclusion in the Swedish Rheumatology Quality. Of 29 084 patients, 12 193 (41.9%) reached DAS28 SR at some time point during follow-up compared with 6445 (22.2%), 6199 (21.3%) and 5087 (17.5%) for CDAI, SDAI and ACR/EULAR SR, respectively. SR was more common in early RA (P < 0.001). The median time from symptom onset to SR was 1.9, 2.4, 2.4 and 2.5 years according to DAS28, CDAI, SDAI and ACR/EULAR criteria, respectively. Lower age, male sex and milder disease characteristics were associated with SR. The majority of patients in this nationwide study never reached SR. Patients with early RA are more likely to reach SR than patients with established RA.

Mortality following new-onset Rheumatoid Arthritis: has modern Rheumatology had an impact?

ObjectiveTo investigate if, and when, patients diagnosed with rheumatoid arthritis (RA) in recent years are at increased risk of death.MethodsUsing an extensive register linkage, we designed a population-based nationwide cohort...

Acute coronary syndrome in new-onset rheumatoid arthritis: a population-based nationwide cohort study of time trends in risks and excess risks

BackgroundAcute coronary syndrome (ACS) and other cardiovascular diseases are the main drivers of the increased morbidity and preterm mortality in rheumatoid arthritis (RA). ACS in RA has been linked to...

The performativity of numbers in illness management: The case of Swedish Rheumatology

While there is a proliferation of numerical data in healthcare, little attention has been paid to the role of numbers in constituting the healthcare reality they are intended to depict. This study explores the performativity of numbers in the microlevel management of rheumatoid disease. We draw on a study of patients' and physicians' use of the numbers in the Swedish Rheumatology Quality Registry, conducted between 2009 and 2014. We show how the numbers performed by constructing the disease across time, and by framing action. The numerical performances influenced patients and physicians in different ways, challenging the former to quantify embodied disease and the latter to subsume the disease into one of many possible trajectory standards. Based on our findings, we provide a model of the dynamic performativity of numbers in the on-going management of illness. The model conceptualises how numbers generate new possibilities; by creating tension and alignment they may open up new avenues for communication between patients and physicians.

Rheumatoid Arthritis and Risk of Malignant Lymphoma: Is the Risk Still Increased?

Patients with rheumatoid arthritis (RA) are at increased risk of malignant lymphomas with a strong correlation with RA disease severity. Given the changes in RA therapy over recent decades, this study was undertaken to assess whether lymphoma risk remains increased, and if so, to explore risk predictors and lymphoma subtypes. We identified 12,656 cases of incident RA in the Swedish Rheumatology Quality Register 1997-2012 and obtained information on therapy and inflammatory activity during the first year after diagnosis. Each patient was matched to 10 population comparator subjects. Through linkage to the Swedish Cancer Register, lymphomas, including subtypes, were identified. We assessed hazard ratios (HRs) using Cox regression. Overall, the HR for lymphoma was increased in RA, to 1.6 (95% confidence interval [95% CI] 1.2-2.1). Taking RA duration into account, risks did not appear to have declined over successive calendar years of RA diagnosis. Neither use of methotrexate the first year after RA diagnosis nor ever use of tumor necrosis factor inhibitors (TNFi) increased lymphoma risk (HR 0.9 [95% CI 0.4-1.9]). Use of oral corticosteroids the first year after RA diagnosis was associated with a reduced risk (HR 0.5 [95% CI 0.3-0.9]). Inflammatory activity during the first year after RA diagnosis did not predict future lymphoma risk. Chronic lymphocytic leukemia occurred less frequently, and Hodgkin's lymphoma occurred more frequently, in RA patients than in the general population. The average lymphoma risk in recently diagnosed RA is similar in magnitude to that reported in historical cohorts. Standard antirheumatic treatment including TNFi did not predict future lymphoma risk. Distribution of lymphoma subtypes warrants further investigation.

Anticollagen type II antibodies are associated with an acute onset rheumatoid arthritis phenotype and prognosticate lower degree of inflammation during 5 years follow-up.

ObjectiveAntifibrillar collagen type II (anti-CII) antibody-positive patients with rheumatoid arthritis (RA) have early but not late signs of increased inflammation and joint erosions. We wanted to replicate this in a large RA cohort, and to relate to human leukocyte antigen (HLA)-DRB1* alleles.MethodsAnti-CII and anti-cyclic citrullinated peptide (CCP)2 were measured at baseline in 773 patients with RA from the Swedish Epidemiological Investigation in Rheumatoid Arthritis (EIRA) study with clinical follow-up data from the Swedish Rheumatology Quality Register (SRQ) registry, and 1476 with HLA-DRB1* information. Comparisons were done concerning C reactive protein (CRP), erythrocyte sedimentation rate (ESR), tender joint count (TJC), swollen joint count (SJC), Disease Activity Score encompassing 28 joints based on ESR (DAS28), DAS28CRP, pain-Visual Analogue Scale (VAS), global-VAS and Health Assessment Questionnaire Score (HAQ) at eight occasions during 5 years, and association with HLA-DRB1* alleles.ResultsAnti-CII associated with elevated CRP, ESR, SJC, DAS28 and DAS28CRP at diagnosis and up to 6 months, whereas anti-CCP2 associated with SJC and DAS28 from 6 months to 5 years, but not earlier. The anti-CII-associated phenotype was strong, and predominated in anti-CII/anti-CCP2 double-positive patients. Anti-CII was associated with improvements in CRP, ESR, SJC, TJC and DAS28, whereas anti-CCP2 was associated with deteriorations in SJC and DAS28 over time. Anti-CII-positive patients achieved European League Against Rheumatism good or moderate response more often than negative patients. Anti-CII was positively associated with HLA-DRB1*01 and HLA-DRB1*03, with significant interaction, and double-positive individuals had >14 times higher mean anti-CII levels than HLA double negatives. Whereas smoking was associated with elevated anti-CCP2 levels, smokers had lower anti-CII levels.ConclusionsAnti-CII seropositive RA represents a distinct phenotype, in many respects representing the converse to the clinical, genetic and smoking associations described for anticitrullinated protein peptide autoantibodies. Although not diagnostically useful, early anti-CII determinations predict favourable inflammatory outcome in RA.

Tumour necrosis factor inhibitor treatment and occurrence of anterior uveitis in ankylosing spondylitis: results from the Swedish biologics register

ObjectivesTumour necrosis factor-α inhibitor (TNFi) treatment has been shown to reduce the rates of anterior uveitis (AU) in patients with ankylosing spondylitis (AS). Our objective was to compare the effect...

Incidence and prevalence of idiopathic inflammatory myopathies in Sweden: a nationwide population-based study

To estimate the incidence rate and prevalence of idiopathic inflammatory myopathies (IIMs) in Sweden across clinical subgroups, age, sex, educational level and place of residence and to assess the robustness of register-based case definitions. IIM was identified from the Swedish National Patient Register and the Swedish Rheumatology Quality Register. The base case definition required ⩾2 visits indicating IIM (first ever with a consecutive visit within 1-12 months for incident cases) and the robustness was tested by applying a more liberal and a stricter definition. Using the base case definition, 558 incident IIM patients were identified between 2007 and 2011. The incidence was estimated to 11 (13 for women and 9.7 for men) per 1 000 000 person years and was stable across case definitions. Incidence increased with age and peaked at the 50-79 years age groups. No differences were observed between different levels of education and place of residence. We identified 1267 IIM patients on 1 January 2012 corresponding to a prevalence of 14 per 100 000. We present nationwide register-based incidence and prevalence estimates for IIM, robust across three different case definitions. In contrast to many other reports, we did not find incidence by age to be bimodal and we found no explanation of incidence variation across education and residency. These register-based case definitions can be included in future population-based studies to better understand disease aetiology, risk factors and comorbidities.

Pain rather than self-reported sedentary time explains variation in perceived health and activity limitation in persons with rheumatoid arthritis: a cross sectional study in Sweden

To investigate (1) the amount of self-reported time spent sedentary among a large cohort of persons with rheumatoid arthritis (RA), and (2) the contribution of sedentary time to explain perceived health and activity limitation in RA beyond that of previously known correlates. This cross-sectional study used data from a postal questionnaire and the Swedish Rheumatology Quality registers (SRQ). The International Physical Activity Questionnaire was used to assess sedentary time (sitting) and moderate, vigorous and walking activity (MVPA). Sociodemographics, pain, fatigue, fear-avoidance beliefs, anxiety/depression, disease duration, MVPA and sedentary time were included in multiple regression models with perceived health (Visual Analogue Scale 0–100) and activity limitation (Stanford Health Assessment Questionnaire) as dependent variables. Results: In all 3152 (59%) of 5391 persons identified as eligible from the SRQ, responded to the questionnaire. 2819 individuals with complete data on all study variables were analysed. Mean time (SD) spent sedentary was 257 (213) minutes per day. Sedentary time did not contribute significantly to explain perceived health and only minimally to explain activity limitation. Instead, variation was mainly explained by pain; for perceived health (Beta = 0.780, p < 0.001) and for activity limitation (Beta = 0.445, p < 0.001).The results indicate a non-significant role of sedentary time and a need for increased focus on pain in the management of RA. Future studies should use prospective designs and objective assessment methods to further investigate the associations between sedentary time and health outcomes in persons with RA.

A genetic risk score composed of rheumatoid arthritis risk alleles, HLA-DRB1 haplotypes, and response to TNFi therapy - results from a Swedish cohort study.

BackgroundTo prevent debilitating and irreversible joint damage, rheumatoid arthritis (RA) is often treated with tumor necrosis factor inhibitor (TNFi), but many patients do not respond to this costly therapy. Few predictors for response are known, and it has been proposed that genetic factors which influence the development of RA may also influence disease severity and response to therapy. Several previous studies have attempted to confirm this but results remain inconclusive. We expand on previous studies by including more RA risk alleles, and maximize power by combining them into a genetic risk score.MethodWe linked genotyped RA patients from the Epidemiological Investigation of Rheumatoid Arthritis study to the Swedish Rheumatology Quality Register, identifying patients who started a TNFi as their first biological disease-modifying anti-rheumatic drug, with a return visit within 2–8 months after treatment start (N = 867). We calculated risk scores from 76 established RA risk SNPs, and four HLA-DRB1 amino acid positions, and tested whether risk scores or individual genetic risk factors could predict the European League Against Rheumatism (EULAR) response.ResultsWe found no association between any of the risk scores or HLA-DRB1 haplotypes and EULAR response, neither overall nor stratified by anti-citrullinated protein/peptide antibody (ACPA) status. When evaluating each of the 76 SNPs, we found that the number of SNPs presenting significant associations was not higher than expected by chance (5/76 SNPs had p < 0.05 in ACPA-positive RA, 4/76 in ACPA-negative RA).ConclusionOverall, known RA risk SNPs do not predict response to TNFi, either individually or when combined into a risk score. This does not support the hypothesis that genes influencing RA onset would also influence its prognosis and treatment response.Electronic supplementary materialThe online version of this article (doi:10.1186/s13075-016-1174-z) contains supplementary material, which is available to authorized users.

The evolution of weak standards: the case of the Swedish rheumatology quality registry.

Research in sociology suggests that the effects of standards are not nearly as straightforward or as homogenising as they first appear. The present study extends these insights by demonstrating how even standards designed simply to collect data can produce extensive and unanticipated effects in medical fields as their uses evolve across actors and contexts. We draw on an embedded case study exploring the multifaceted consequences of the use of a practice-driven voluntary documentation standard: the Swedish rheumatology quality registry from 1995-2014. Data collection included document analysis; 100 interviews with specialists, patients and stakeholders in the field; fieldwork; and observations of physician-patient encounters. Our findings show that the scope and influence of the registry increased over time, and that this standard and its evolution contributed to changes in rheumatologist clinical practice, research practice, and governmental practice. These findings suggest that even initially 'weak', voluntary forms of standardisation can generate far-reaching and unpredictable consequences for the performance and delivery of care as well as for the development of a medical field. Future work about how standards can contribute both to uniformity and diversity is warranted.

Remaining Pain in Early Rheumatoid Arthritis Patients Treated With Methotrexate

ObjectiveTo investigate the frequency of remaining pain in early rheumatoid arthritis (RA) after 3 months of treatment with methotrexate as the only disease modifying antirheumatic drug, with a special focus on patients with a good clinical response.MethodsThe study base was cases reported to a population‐based early RA cohort who had followup data from the Swedish Rheumatology Quality Register (n = 1,241). The Disease Activity Score in 28 joints European League Against Rheumatism (EULAR) response criteria were used to evaluate clinical response to treatment as good, moderate, and no response. The primary end point was remaining pain at the 3‐months followup visit, defined as pain >20 mm on a 100‐mm visual analog scale (VAS).ResultsRemaining pain in spite of a EULAR good response at followup was associated with higher baseline disability, using the Health Assessment Questionnaire (adjusted odds ratio [OR] 2.2 [95% confidence interval (95% CI) 1.4–3.4] per unit increase), and less baseline inflammation, using the erythrocyte sedimentation rate (adjusted OR 0.81 [95% CI 0.70–0.93] per 10‐mm increase). Similar associations were detected for remaining pain at followup in spite of low inflammatory activity, defined as a C‐reactive protein level <10. Increase in VAS pain during the treatment period was observed in 19% of the whole cohort, with frequencies in the EULAR response groups of 9% (good response), 15% (moderate response), and 45% (no response).ConclusionThese results are in line with the hypothesis that a subgroup of early RA patients exhibits pain that is not inflammatory mediated, where alternative treatment strategies to traditional antiinflammatory medications need to be considered.

Parity influences the severity of ACPA-negative early rheumatoid arthritis: a cohort study based on the Swedish EIRA material.

BackgroundIn women with rheumatoid arthritis (RA) it has been observed that during pregnancy a majority of patients experience amelioration, but after delivery a relapse of the disease is common. However, there are few studies, with diverging results, addressing the effect of parity on the severity of RA over time. Our aim was to explore the impact of parity, with stratification for anti-citrullinated protein antibody (ACPA) status as well as for onset during reproductive age or not.MethodsFemale RA cases aged 18–70 years were recruited for the Epidemiological Investigation of Rheumatoid Arthritis (EIRA). Information on disease severity (the health assessment questionnaire (HAQ) and the disease activity score 28 (DAS28)) was retrieved from the Swedish Rheumatology Quality Register at inclusion and 3, 6, 12 and 24 months after diagnosis. Mixed models were used to compare mean DAS28 and HAQ scores over time in parous and nulliparous women. Mean differences at individual follow-up visits were compared using analysis of covariance. The odds of having DAS28 or HAQ above the median in parous verus nulliparous women were estimated in logistic regression models.ResultsA total of 1237 female cases (mean age 51 years, 65 % ACPA-positive) were included. ACPA-negative parous women, aged 18–44 years, had on average 1.17 units higher DAS28 (p < 0.001) and 0.43 units higher HAQ score (p < 0.001) compared to nulliparous women during the follow-up time, adjusted for age. In this subgroup, the average DAS28 and HAQ scores were significantly higher in parous women at all follow-up time points. Younger parous ACPA-negative women were significantly more likely to have DAS28 and HAQ values above the median compared to nulliparous women at all follow-up visits. No association between parity and severity of ACPA-positive disease was observed.ConclusionsParity was a predictor of a more severe RA among ACPA-negative younger women, which might indicate that immunomodulatory changes during and after pregnancy affect RA severity, in particular for the ACPA-negative RA phenotype.

Physician Preferences and Variations in Prescription of Biologic Drugs for Rheumatoid Arthritis: A Register-Based Study of 4,010 Patients in Sweden.

The prescription of biologic drugs for rheumatoid arthritis (RA) patients has varied considerably across different regions. Previous studies have shown physician preferences to be an important determinant in the decision to select biologic disease-modifying antirheumatic drugs (bDMARDs) rather than nonbiologic, synthetic DMARDs (sDMARDs) alone. The aim of this study was to test the hypothesis that physician preferences are an important determinant for prescribing bDMARDs for RA patients in Sweden. Using data from the Swedish Rheumatology Quality Register, we identified 4,010 RA patients who were not prescribed bDMARDs during the period 2008-2012, but who, on at least 1 occasion, had an sDMARD prescription and changed treatment for the first time to either a new sDMARD or a bDMARD. Physician preference for the use of bDMARDs was calculated using data on each physician's prescriptions during the study period. The relationship between prescription of a bDMARD and physician preference, controlling for patient characteristics, disease activity, and the physician's local context was evaluated using multivariate logistic regression. When adjusting for patient characteristics, disease activity, and the physician's local context, physician preference was an important predictor for prescription of bDMARDs. Compared with patients of a physician in the lowest preference tertile, patients of physicians in the highest and middle tertiles had an odds ratio for receiving bDMARDs of 2.8 (95% confidence interval [95% CI] 2.13-3.68) and 1.28 (95% CI 1.05-1.57), respectively. Physician preference is an important determinant for prescribing bDMARDs.