Abstract
Objectives
To estimate the overall risk of venous thromboembolism (VTE), pulmonary embolism (PE) and deep vein thrombosis (DVT) among patients newly diagnosed with RA compared with the general population without RA; and to estimate the risk trends of VTE, PE and DVT after RA diagnosis up to 5 years compared with the general population.
Methods
Using previously validated RA case definition, we conducted a matched cohort study using the population-based administrative health database from the province of British Columbia, Canada. We calculated incidence rates (IRs) and fully adjusted hazard ratios (HRs) for the risk of VTE, DVT and PE after RA index date.
Results
Among 39 142 incident RA patients (66% female, mean age 60), 1432, 543 and 1068 developed VTE, PE and DVT, respectively. IRs for the RA cohort were 3.79, 1.43 and 2.82 per 1000 person-years vs 2.70, 1.03 and 1.94 per 1000 person-years for the non-RA cohort. After adjusting for VTE risk factors, the HRs (95% CI) were 1.28 (1.20, 1.36), 1.25 (1.13, 1.39) and 1.30 (1.21, 1.40) for VTE, PE and DVT, respectively. The fully adjusted HRs for VTE during the first five years after RA diagnosis were 1.60, 1.47, 1.40, 1.30 and 1.28, respectively. A similar trend was shown in PE.
Conclusion
This population-based study demonstrates that RA patients have an increased risk of VTE, PE and DVT after diagnosis compared with the general population. This risk is independent of traditional VTE risk factors and is highest during the first year after RA diagnosis, then progressively declined.
Keywords: rheumatoid arthritis, venous thromboembolism, risk factors, population-based
Rheumatology key messages
RA patients were associated with increased risks of VTE (28%), PE (25%) and DVT (30%).
VTE risks were highest during the first year after RA diagnosis (60%), then progressively declined.
RA-associated inflammation may contribute to VTE, particularly during the first year after RA diagnosis.
Introduction
RA has been associated with an increased risk of cardiovascular diseases, such as myocardial infarction and ischaemic stroke [1]. Venous thromboembolism [VTE, which includes pulmonary embolism (PE) and deep vein thrombosis (DVT)], represents a relatively common cardiovascular event that is associated with increased mortality and affects more than one in 1000 persons a year in the Western population [2]. An increased risk of VTE has been reported among most inflammatory autoimmune disorders including RA [3]. Inflammation is an important process for hypercoagulability, endothelial injury and venous stasis (i.e. Virchow’s triad) for developing VTE [4]. A few cohort studies have examined the risk of VTE among RA patients [3, 5–13]. Some of them used hospital-based data, and reported an increased risk of VTE among hospitalized RA patients [8, 12, 13]. However, the majority of RA patients in contemporary care settings in developed countries are non-hospitalized [10]. So the increased risk from those studies might not be generalizable in the real world. The findings are also limited due to failure to adjust for medications or address the effect of potential unmeasured confounders [10], use of prevalent cohort data (which is associated with survival bias) [3, 5, 8, 9, 12] and a small sample size [7, 11, 12]. Few studies have examined the trend of the VTE risk after RA diagnosis, but some of them used the period-specific hazard ratios (HRs) that might induce a built-in selection bias.
To address those limitations, the aims of our study were to estimate the overall risk of VTE, PE and DVT after RA diagnosis in a population-based incident RA cohort from an entire province in Canada; and to estimate the risk trends of VTE, PE and DVT for the first 5 years following RA diagnosis, compared with the general population.
Methods
Data sources
Universal healthcare is available for all residents of the province of British Columbia (BC), Canada (population∼4.7 million). Population Data BC includes all provincially funded healthcare services data since 1990, including all healthcare professional visits [14], hospitalizations [15], demographic data [16], BC cancer registry [17] and vital statistics [18]. Furthermore, Population Data BC includes the comprehensive prescription drug database PharmaNet [19], which captures all dispensed medications for all residents since 1996. Numerous population-based studies have been successfully conducted using Population Data BC [20–22]. Using Population Data BC, we conducted a matched cohort study design.
RA cohort
All incident RA individuals in BC who met criteria for RA case definition for the first time between 1 January 1997 and 31 December 2009 were selected, using physician billing data from the Ministry of Health, and were followed until 31 March 2015. Using previously published criteria [23], individuals were identified as RA if they had at least two physician visits at least 2 months apart within a 5-year period with an International Classification of Diseases, Ninth Revision (ICD-9) code for RA (714.X) [24]. To improve specificity, individuals were excluded, if over a 5-year period after their second RA visit (i.e. index date) they had at least two subsequent visits, on two different days, with the same ICD code for other forms of inflammatory arthritis (systematic lupus erythematosus, and other connective tissue diseases, psoriatic arthritis, ankylosing spondylitis and other spondyloarthropathies), or if a patient saw a rheumatologist and the diagnosis of RA by a non-rheumatologist was never confirmed by the rheumatologist.
To ensure incident RA patients, we required all newly diagnosed RA individuals to have at least 7 years of prior registration in the databases (i.e. ‘run-in’ period). These criteria have been validated in a subsample who participated in an RA survey. Using the opinion of an independent rheumatologist reviewing medical records from their treating physicians as the gold standard, the positive predictive value was 0.82 [20, 25].
Non-RA cohort
To establish the non-RA comparison cohort, we received data from a random sample of ∼2 000 000 BC residents registered with the provincial medical services plan during the study period, and selected individuals without any history of RA and matched them to RA patients (up to 2:1 ratio) on age, sex and the index date. To be comparable to the RA cohort, we also excluded those non-RA individuals without at least 7 years’ run-in time before the matched index date.
Individuals were followed until they experienced an outcome (VTE, PE or DVT), died, left BC or the follow-up period ended (March 31, 2015), whichever occurred first.
Ascertainment of DVT and PE
The primary outcome was the first ever VTE (PE or DVT), PE or DVT during the follow-up period. Incident PE or DVT outcomes were defined by a corresponding ICD code plus an outpatient prescription for any anticoagulant therapy (heparin, warfarin or a similar agent) between 1 month before and 6 months after the ICD code date. The analysis for VTE (i.e. VTE as the outcome) did censor the follow-up when participants developed DVT or PE. However, the analysis for DVT (i.e. DVT as the outcome) did not censor follow-up when participants developed PE (as it is not the outcome of that particular analysis). For the same reason, the analysis for PE (i.e. PE as the outcome) did not censor the follow-up when participants developed DVT. We identified PE (ICD-9: 415.1, 673.2, 639.6; ICD-10: O88.2, I26) from hospitalization data only, and DVT (ICD-9: 453; ICD-10: I82.4, I82.9) from outpatient or hospitalization data. Deaths from PE or DVT (including out-of-hospital deaths) were identified from vital statistics. Given that VTE is a potentially fatal disease, patients may have died before they received treatment; thus, for those patients who died within 2 months after a VTE, PE or DVT diagnosis, anticoagulant therapy was not needed as part of the case definition. Similar definitions for VTE have been used in previous publications [26–28] and were found to have a positive predictive value of 94% in a general practice database [29].
Covariate assessment
To evaluate the baseline characteristics at index date, available covariates known to be potential risk factors for VTE were assessed within the 12 months prior to the index date. Covariates included health resource utilization (number of outpatient visits and hospitalizations), medication use [as identified through Drug Identification Numbers (DINs): hormone replacement therapy, glucocorticoids, cyclooxygenase-2 (Cox-2) inhibitors, aspirin and oral contraceptives] and comorbidities (hypertension, varicose veins, sepsis, inflammatory bowel disease and alcoholism). In addition, trauma or fracture, a history of surgery, as well as the Romano modification of the Charlson Comorbidity Index for administrative data [30] were also obtained in that period.
Statistical analysis
We compared the baseline characteristics of the RA cohort with the non-RA cohort using a χ2 test for categorical variables and the Wilcoxon rank-sum test for continuous variables. To calculate the risk of VTE, we identified incident PE and DVT events during the follow-up period and calculated the incidence rates (IRs) per 1000 person-years for each outcome, individually and together (as VTE). We used Cox proportional hazard regression models [31] to calculate HRs for the risk of VTE, PE and DVT among RA patients compared with the non-RA cohort, adjusting for baseline covariates. We checked the proportional hazard assumption by including time-dependent covariates in the Cox model. We calculated the cumulative incidence of each outcome event in all models after taking the competing risk of death into account, according to Lau et al. [32], and expressed the results as subdistribution HRs with 95% CIs. We then examined the difference in the overall VTE, DVT and PE risks between RA and non-RA cohorts using an additive hazard model (reported as the rate difference). The estimate generated from this model can be interpreted as the number of excess VTE attributable to RA per 1000 person-years. To evaluate the time trends of VTE risk, we also estimated HRs within 1, 2, 3, 4 and 5 years after the index date.
Sensitivity analysis
To test the robustness of our results, we performed a sensitivity analysis to estimate the effect of unmeasured confounders (i.e. obesity) [33, 34]. We calculated fully adjusted HRs by adding the simulated unmeasured confounder with a prevalence ranging from 10% to 20% in the RA and non-RA cohorts, and odds ratios (ORs) for the association between the unmeasured confounder, RA and VTE ranging from 1.3–2.0.
SAS V.9.4 (SAS Institute, Cary, NC, USA) was used for all analyses.
No personally identifying information was made available as part of this study. Procedures used were in compliance with British Columbia’s Freedom of Information and Privacy Protection Act. Ethics approval was obtained from the University of British Columbia’s Behavioral Research Ethics Board (H15-00887).
Results
Baseline characteristics
After excluding individuals with VTE events before the index date, we identified 39 142 incident RA patients (66% female, mean age 60 years), and 78 078 non-RA individuals (66% female, mean age 60 years). Table 1 summarizes the baseline characteristics of the cohorts. Compared with the non-RA cohort, RA patients had a higher number of outpatient visits and hospitalizations, higher use of glucocorticoids, hormone replacement therapy, oral contraceptives, aspirin and Cox-2 inhibitors, as well as greater Charlson Comorbidity Index, and higher prevalence of alcoholism, hypertension, sepsis, varicose veins, trauma, fracture and surgery.
Table 1.
Baseline characterisitics of individuals with and without RA at the time of index date
| Variable | RA (n = 39 142) | Non-RA (n = 78 078) | P-value |
|---|---|---|---|
| Age, mean (s.d.) | 59.57 (15.73) | 59.51 (15.71) | NS |
| Female, n (%) | 25 732 (65.7) | 51 343 (65.8) | NS |
| Health resources utilization | |||
| Hospitalized, n (%) | 7833 (20.0) | 9359 (12.0) | <0.0001 |
| Number of outpatient visits, median (s.d.) | 10.00 (13.33) | 3.00 (9.87) | <0.0001 |
| Comorbidities, n (%) | |||
| Alcoholism with liver disease | 410 (1.1) | 351 (0.5) | <0.0001 |
| Hypertension | 8333 (21.3) | 12 140 (15.6) | <0.0001 |
| Sepsis | 180 (0.5) | 209 (0.3) | <0.0001 |
| Varicose veins | 341 (0.9) | 522 (0.7) | <0.0001 |
| Inflammatory bowel disease | 69 (0.2) | 205 (0.3) | 0.0039 |
| Trauma | 59 (0.2) | 89 (0.1) | NS |
| Fractures | 647 (1.7) | 767 (1.0) | <0.0001 |
| Surgery | 222 (0.6) | 367 (0.5) | 0.0266 |
| Charlson Comorbidity Index, mean (s.d.) | 0.32 (0.91) | 0.22 (0.82) | <0.0001 |
| Medications, n (%) | |||
| Glucocorticoids | 7789 (19.9) | 2073 (2.7) | <0.0001 |
| Hormone replacement therapy | 3064 (7.8) | 3884 (5.0) | <0.0001 |
| Oral contraceptives | 958 (2.5) | 1797 (2.3) | NS |
| Aspirin | 699 (1.8) | 974 (1.3) | <0.0001 |
| Cox-2 inhibitors | 5274 (13.5) | 1896 (2.4) | <0.0001 |
Baseline characterisitics were measured over 1 year prior to index date. Cox-2: cyclooxygenase-2; NS: non-significant.
Risk of VTE in patients with RA
RA was significantly associated with a higher incidence of overall VTE, PE and DVT events (Table 2 and Fig. 1). There were 1432 incident VTE, 543 PE and 1068 DVT events in the RA cohort, compared with 2059, 791 and 1484 in the non-RA cohort, respectively. The cumulative incidence of VTE, PE and DVT was significantly higher in the RA cohort compared with the non-RA cohort (Fig. 1). The IRs for VTE, PE and DVT in the RA cohort were 3.79, 1.43 and 2.82 cases per 1000 person-years, respectively, compared with 2.70, 1.03 and 1.94 cases per 1000 person-years in the non-RA cohort. After matching for age, sex and the index date, the HRs (95% CI) were 1.39 (1.32, 1.47), 1.37 (1.26, 1.50) and 1.43 (1.34, 1.52) for VTE, PE and DVT, respectively, and the rate differences (95% CI) were 1.10 (0.87, 1.33), 0.40 (0.26, 0.54) and 0.88 (0.69, 1.08) cases per 1000 person-years for VTE, PE and DVT, respectively. The corresponding fully adjusted HRs (95% CI) were 1.28 (1.20, 1.36), 1.25 (1.13, 1.39) and 1.30 (1.21, 1.40) (Table 2).
Table 2.
Overall risk of incident VTE, PE and DVT in RA relative to non-RA
| RA (n = 39 142) | Non-RA (n = 78 078) | |
|---|---|---|
| VTE | ||
| Events, n | 1432 | 2059 |
| Incidence rate/1000 person-years (95% CI) | 3.79 (3.60, 4.00) | 2.70 (2.58, 2.81) |
| Age-, sex-, index date-matched rate difference (95% CI) | 1.10 (0.87, 1.33) | 1.00 |
| Age-, sex-, index date-matched HR (95% CI)a | 1.39 (1.32, 1.47) | 1.00 |
| Fully adjusted HR (95% CI)ab | 1.28 (1.20, 1.36) | 1.00 |
| PE | ||
| Events, n | 543 | 791 |
| Incidence rate/1000 person-years (95% CI) | 1.43 (1.31, 1.55) | 1.03 (0.96, 1.10) |
| Age-, sex-, index date-matched rate difference (95% CI) | 0.40 (0.26, 0.54) | 1.00 |
| Age-, sex-, index date-matched HR (95% CI)a | 1.37 (1.26, 1.50) | 1.00 |
| Fully adjusted HR (95% CI)ab | 1.25 (1.13, 1.39) | 1.00 |
| DVT | ||
| Events, n | 1068 | 1484 |
| Incidence rate/1000 person-years (95% CI) | 2.82 (2.65, 2.99) | 1.94 (1.84, 2.04) |
| Age-, sex-, index date-matched rate difference (95% CI) | 0.88 (0.69, 1.08) | 1.00 |
| Age-, sex-, index date-matched HR (95% CI)a | 1.43 (1.34, 1.52) | 1.00 |
| Fully adjusted HR (95% CI)ab | 1.30 (1.21, 1.40) | 1.00 |
Adjusted for the competing risk of death.
Adjusted for all variables listed in Table 1. DVT: deep vein thrombosis; HR: hazard ratio; PE: pulmonary embolism; VTE: venous thromboembolism.
Fig. 1.
Cumulative incidence of VTE, PE and DVT among RA and non-RA
DVT: deep vein thrombosis; PE: pulmonary embolism; VTE: venous thromboembolism.
When we evaluated the impact of follow-up time after RA diagnosis, during the first, second, third, fourth and fifth year after the index date. The HRs (95% CI) of VTE were largest in the first year after the diagnosis of RA compared with the following years [1.60 (1.27, 2.00), 1.47 (1.26, 1.71), 1.40 (1.23, 1.59), 1.30 (1.16, 1.46) and 1.28 (1.15, 1.42), respectively]. Similar trends were also found in PE; the HRs (95% CI) were 1.86 (1.21, 2.86), 1.54 (1.21, 1.98), 1.34 (1.09, 1.66), 1.22 (1.01, 1.47) and 1.29 (1.09, 1.53), respectively. The HRs (95% CI) for DVT were 1.59 (1.20, 2.10), 1.38 (1.14, 1.68), 1.39 (1.19, 1.62), 1.34 (1.17, 1.53) and 1.27 (1.12, 1.43), respectively (Table 3).
Table 3.
Fully adjusted HRs for VTE, PE and DVT in RA according to follow-up period
|
HR (95% CI)
a
|
|||
|---|---|---|---|
| Variable | VTE | PE | DVT |
| <1 year | 1.60 (1.27, 2.00) | 1.86 (1.21, 2.86) | 1.59 (1.20, 2.10) |
| <2 years | 1.47 (1.26, 1.71) | 1.54 (1.21, 1.98) | 1.38 (1.14, 1.68) |
| <3 years | 1.40 (1.23, 1.59) | 1.34 (1.09, 1.66) | 1.39 (1.19, 1.62) |
| <4 years | 1.30 (1.16, 1.46) | 1.22 (1.01, 1.47) | 1.34 (1.17, 1.53) |
| <5 years | 1.28 (1.15, 1.42) | 1.29 (1.09, 1.53) | 1.27 (1.12, 1.43) |
| Total follow-up | 1.28 (1.20, 1.36) | 1.25 (1.13, 1.39) | 1.30 (1.21, 1.40) |
Adjusted for all variables listed in Table 1 and competing risk of death. DVT: deep vein thrombosis; HR: hazard ratio; PE: pulmonary embolism; VTE: venous thromboembolism.
Sensitivity analysis
To assess the robustness of our results, we further adjusted for the unmeasured confounder (e.g. obesity). HRs remained significant, but attenuated at the values of 20% prevalence in the RA cohort and OR of 1.3 for the association between the unmeasured confounder and VTE, PE and DVT (Table 4).
Table 4.
Sensitivity analyses for VTE, PE and DVT after the index date
|
Simulated confounder
a
|
|||||
|---|---|---|---|---|---|
| Events | Primary analysis a | 10%/OR = 1.3 | 10%/OR = 2.0 | 20%/OR = 1.3 | 20%/OR = 2.0 |
| VTE | 1.28 (1.20, 1.36) | 1.24 (1.16, 1.32) | 1.14 (1.07, 1.22) | 1.21 (1.14, 1.90) | 1.00 (0.94, 1.08) |
| PE | 1.25 (1.13, 1.39) | 1.26 (1.13, 1.41) | 1.19 (1.07, 1.34) | 1.23 (1.10, 1.37) | 1.02 (0.91, 1.15) |
| DVT | 1.30 (1.21, 1.40) | 1.24 (1.15, 1.34) | 1.14 (1.06, 1.24) | 1.22 (1.13, 1.32) | 1.00 (0.92, 1.09) |
Adjusted for all variables listed in Table1 and competing risk of death. DVT: deep vein thrombosis; HR: hazard ratio; OR: odds ratio; PE: pulmonary embolism; VTE: venous thromboembolism.
Discussion
In this general population-based study, using a large incident RA cohort, we found that RA was associated with an increased risk of VTE, PE and DVT events compared with the general population without RA. For example, we found 1.10, 0.40 and 0.88 excess VTE, PE and DVT events per 1000 person-years, respectively, due to RA. These associations were independent of traditional VTE risk factors and the competing risk of death. The VTE risk was the highest during the first year after RA diagnosis, then progressively declined but remained statistically significant even 5 years after the index date. The finding of the increased VTE risk among patients with RA could have important implications for clinical care, both immediately after a diagnosis of RA and in long-term treatment. Clinicians should be aware that a diagnosis of RA might dispose patients to have a higher risk of VTE, particularly in the period soon after diagnosis.
Our findings linking RA to an increased overall risk of VTE, PE and DVT are consistent with previous studies [3, 5, 13], despite the large amount of heterogeneity between those studies and ours (stemming from underlying differences in the study populations, settings/healthcare systems, and methodological approaches). Four studies using population-based administrative data from the UK, Sweden and Taiwan showed the risks ranging from 1.29 to 2.14 for VTE, 1.45 to 2.16 for PE and 1.25 to 3.36 for DVT [3, 6, 9, 10]. Two of them used prevalent cohorts that only included surviving individuals [3, 9], who may be less susceptible to VTE; another study did not adjust for certain risk factors that are known to be strongly associated with VTE, such as surgery, fracture and trauma [6, 35]. Proper adjustment for traditional risk factors or risk modifiers will provide a less biased estimation on the independent risk of VTE among RA patients. Therefore, the risk or benefit of anti-rheumatic therapies such as Janus kinase (JAK) inhibitors can be weighted properly by patients and clinicians. Three studies that used the hospital-based data showed a 1.45–1.99 increased risk of VTE in patients with RA [8, 12, 13]. However, hospital-based data mainly include patients with more severe disease activity and other comorbidities or procedures. Certainly, we do not know if the results using hospital-based data could be generalized to outpatient RA patients, who constitute the majority of cases in contemporary care settings in developed countries [10]. Two other studies found higher risk of VTE in patients with RA using the community-based cohort from Rochester, MN, USA (<1000 incident RA patients for both studies) [7, 11] and one study using the United States insurance claims dataset showed similar results to ours [5]. Previous studies evaluating the impact of follow-up time showed different results; some studies have divided time into discrete consecutive intervals, which could induce the built-in selection bias that is due to the differential selection of less susceptible people over time, especially when using the first ever VTE as the outcome. As recommended by Hernán [36], we estimated the HRs within 1, 2, 3, 4 and 5 years to avoid this bias. Our results confirmed that the largest VTE risk happened during the first year after RA diagnosis, declining thereafter [8–10]. This finding is congruent with the evidence that there are higher levels of inflammation at earlier stages of RA, before achieving the maximum benefits of antirheumatic therapy [8, 10]. Studies need to be confirmatory in different populations and settings using the contemporary data. We have previously demonstrated that type of study, sample type, year of study and quality of study are important factors when assessing risk of cardiovascular disease and cardiovascular mortality in RA in meta-analyses [37, 38]. We believe, that our study contributes significantly to developing a meta-analysis on the risk of VTE in patients with RA with the highest level of evidence, especially among observational studies.
The mechanism for an increased risk of VTE is likely multifactorial. Virchow proposed three main precursors to VTE: damage to the vessel wall, increased blood coagulability and venous stasis [4]. RA patients are prone to increases in blood coagulability and venous stasis through decreased mobility and increased inflammation, which can upregulate procoagulants, downregulate anticoagulants and suppress fibrinolysis [39]. Additionally, inflammation may increase the risk of VTE in RA patients by damaging the vessel wall through venulitis or the presence of antiphospholipid antibodies [7, 10, 40]. Therefore, it stands to reason that VTE risk would be highest in the first year after diagnosis, due to untreated inflammation, and subsequently decline with anti-inflammatory treatment.
We acknowledge the potential limitations in this observational study. Although we used one of the strictest RA case definitions, previously validated with a positive predictive value of 0.82, uncertainty in the accuracy for RA case definitions cannot be completely ruled out [25]. Besides, laboratory test data were not available in this study. Even though our multivariable Cox models adjusted for some VTE risk factors, certain confounders were not available in our dataset, such as body mass index to assess obesity [33, 34]. Nevertheless, the results were still statistically significant in our sensitivity analysis adjusted for unmeasured confounders even when we used values of 20% prevalence of obesity in the RA cohort and an OR of 1.3 for the association between obesity and VTE, PE and DVT. Although the proportional hazard assumption in this study during the entire follow-up is violated, the analysis for calculating the period-specific HRs itself deals with the violation of the assumption [36].
Even though RA patients have an increased risk of VTE compared with the general population, two different JAK inhibitors used to treat RA (baricitiniba and tofacitinib) are also associated with an increased thromboembolic risk [41, 42]. A few baricitiniba trials have demonstrated an increased risk for thromboembolism, and post-marketing surveillance for tofacitinib suggested an increased risk of pulmonary thrombosis [41]. However, the role of the JAK inhibitors in VTE risk was not investigated in our study, and further research on this topic is currently underway.
Notable strengths of this study include the use of a large Canadian administrative dataset that includes the entire population of BC, making our results more generalizable. The large sample size provided sufficient statistical power to study the association between RA and VTE. Moreover, this is also the first study to assess the impact of follow-up time on the risk of VTE after addressing built-in selection bias. Finally, the very strict case definition for VTE allowed us to capture true events.
In conclusion, this is the largest population-based study to date to demonstrate that incident RA patients have an increased risk of VTE, PE and DVT (∼25–30%) following RA diagnosis compared with the general population without RA. This risk is independent of traditional VTE risk factors and highest in the first year after RA diagnosis. These findings are consistent with previous studies and call for increased vigilance of VTE risk factors in patients with RA. Controlling inflammation may serve as a preventive strategy for VTE in patients with RA.
Acknowledgements
We would like to thank the Ministry of Health of British Columbia and Population Data BC for providing access to the administrative data. D.L. holds the Mary Pack Arthritis Chair in Rheumatology Research from the University of British Columbia and the Arthritis Society of Canada. J.A.A.-Z. is the Walter & Marilyn Booth research scholar and the BC Lupus Society research scholar. All inferences, opinions, and conclusions drawn in this article are those of the authors and do not reflect the opinions or policies of the Data Stewards.
Funding: This work was supported by the Canadian Institutes of Health Research [CIHR Grants MOP 125960, THC 135235].
Disclosure statement: The authors have declared no conflicts of interest.
References
- 1. Kitas GD, Gabrie SE.. Cardiovascular disease in rheumatoid arthritis: state of the art and future perspectives. Ann Rheum Dis 2011;70:8–14. [DOI] [PubMed] [Google Scholar]
- 2. Beckman MG, Hooper WC, Critchley SE, Ortel TL.. Venous thromboembolism: a public health concern. Am J Prev Med 2010;38:S495–501. [DOI] [PubMed] [Google Scholar]
- 3. Ogdie A, Kay McGill N, Shin DB. et al. Risk of venous thromboembolism in patients with psoriatic arthritis, psoriasis and rheumatoid arthritis: a general population-based cohort study. Eur Heart J 2018;39:3608–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Previtali E, Bucciarelli P, Passamonti SM, Martinelli I.. Risk factors for venous and arterial thrombosis. Blood Transfus 2011;9:120–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kim SC, Schneeweiss S, Liu J, Solomon DH.. Risk of venous thromboembolism in patients with rheumatoid arthritis. Arthritis Care Res (Hoboken) 2013;65:1600–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Holmqvist ME, Neovius M, Eriksson J, Mantel Ä. et al. Risk of venous thromboembolism in patients with rheumatoid arthritis and association with disease duration and hospitalization. JAMA 2012;308:1350–6. [DOI] [PubMed] [Google Scholar]
- 7. Bacani AK, Gabriel SE, Crowson CS, Heit JA, Matteson EL.. Noncardiac vascular disease in rheumatoid arthritis: increase in venous thromboembolic events? Arthritis Rheum 2012;64:53–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Zöller B, Li X, Sundquist J, Sundquist K.. Risk of pulmonary embolism in patients with autoimmune disorders: a nationwide follow-up study from Sweden. Lancet 2012;379:244–9. [DOI] [PubMed] [Google Scholar]
- 9. Chung WS, Peng CL, Lin CL. et al. Rheumatoid arthritis increases the risk of deep vein thrombosis and pulmonary thromboembolism: a nationwide cohort study. Ann Rheum Dis 2014;73:1774–80. [DOI] [PubMed] [Google Scholar]
- 10. Choi HK, Rho YH, Zhu Y. et al. The risk of pulmonary embolism and deep vein thrombosis in rheumatoid arthritis: a UK population-based outpatient cohort study. Ann Rheum Dis 2013;72:1182–7. [DOI] [PubMed] [Google Scholar]
- 11. Liang KP, Liang KV, Matteson EL. et al. Incidence of noncardiac vascular disease in rheumatoid arthritis and relationship to extraarticular disease manifestations. Arthritis Rheum 2006;54:642–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ramagopalan SV, Wotton CJ, Handel AE, Yeates D, Goldacre MJ.. Risk of venous thromboembolism in people admitted to hospital with selected immune-mediated diseases: record-linkage study. BMC Med 2011;9:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Matta F, Singala R, Yaekoub AY, Najjar R, Stein PD.. Risk of venous thromboembolism with rheumatoid arthritis. Thromb Haemost 2009;101:134–8. [PubMed] [Google Scholar]
- 14.British Columbia Ministry of Health. Medical Services Plan (MSP) Payment Information File. Population Data BC, 2017. https://www.popdata.bc.ca/data/health/msp (25 May 2020, date last accessed). [Google Scholar]
- 15.Canadian Institute for Health Information. Discharge Abstract Database (Hospital Separations). Population Data BC, 2017. http://www.popdata.bc.ca/data/health/dad (25 May 2020, date last accessed). [Google Scholar]
- 16.British Columbia Ministry of Health. Consolidation File (MSP Registration & Premium Billing). Population Data BC, 2017. http://www.popdata.bc.ca/data/population/consolidationfile (25 May 2020, date last accessed). [Google Scholar]
- 17. BC Cancer. BC Cancer Registry Data. Population Data BC, 2017. http://www.popdata.bc.ca/data/health/bccancer (25 May 2020, date last accessed). [Google Scholar]
- 18.BC Vital Statistics Agency. Vital Statistics Deaths. Population Data BC, 2017. http://www.popdata.bc.ca/data/population/vsdeaths (25 May 2020, date last accessed). [Google Scholar]
- 19.BC Ministry of Health. PharmaNet. Population Data BC, 2018. http://www.popdata.bc.ca/data/health/PharmaNet (25 May 2020, date last accessed). [Google Scholar]
- 20. Lacaille D, Avina-Zubieta JA, Sayre EC, Abrahamowicz M.. Improvement in 5-year mortality in incident rheumatoid arthritis compared with the general population-closing the mortality gap. Ann Rheum Dis 2017;76:1057–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Etminan M, Forooghian F, Brophy JM, Bird ST, Maberley D.. Oral fluoroquinolones and the risk of retinal detachment. JAMA 2012;307:1414–9. [DOI] [PubMed] [Google Scholar]
- 22. Solomon DH, Massarotti E, Garg R. et al. Association between disease-modifying antirheumatic drugs and diabetes risk in patients with rheumatoid arthritis and psoriasis. JAMA 2011;305:2525–31. [DOI] [PubMed] [Google Scholar]
- 23. Lacaille D, Anis AH, Guh DP, Esdaile JM.. Gaps in care for rheumatoid arthritis: a population study. Arthritis Rheum 2005;53:241–8. [DOI] [PubMed] [Google Scholar]
- 24. MacLean CH, Louie R, Leake B. et al. Quality of care for patients with rheumatoid arthritis. JAMA 2000;284:984–92. [DOI] [PubMed] [Google Scholar]
- 25. Chang J, Rogers P, Lacaille D.. Can American College of Rheumatology criteria for Rheumatoid Arthritis be assessed using self-report data? – Comparison of self-reported data with chart review. Arthritis Rheum 2011;63:S49. [Google Scholar]
- 26. Carruthers EC, Choi HK, Sayre EC, Aviña-Zubieta JA.. Risk of deep venous thrombosis and pulmonary embolism in individuals with polymyositis and dermatomyositis: a general population-based study. Ann Rheum Dis 2016;75:110–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Li L, McCormick N, Sayre EC. et al. Trends of venous thromboembolism risk before and after diagnosis of gout: a general population-based study. Rheumatology 2020;59:1099–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Aviña-Zubieta JA, Bhole VM, Amiri N, Sayre EC, Choi HK.. The risk of deep venous thrombosis and pulmonary embolism in giant cell arteritis: a general population-based study. Ann Rheum Dis 2016;75:148–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Huerta C, Johansson S, Wallander MA, Garcia Rodriguez LA.. Risk factors and short-term mortality of venous thromboembolism diagnosed in the primary care setting in the United Kingdom. Arch Intern Med 2007;167:935–43. [DOI] [PubMed] [Google Scholar]
- 30. Romano PS, Roos LL, Jollis JG.. Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol 1993;46:1075–9. [DOI] [PubMed] [Google Scholar]
- 31. Cox DR. Regression models and life-tables. J R Stat Soc Ser B 1972;34:187–220. [Google Scholar]
- 32. Lau B, Cole SR, Gange SJ.. Practice of epidemiology competing risk regression models for epidemiologic data. Am J Epidemiol 2009;170:244–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Qin B, Yang M, Fu H. et al. Body mass index and the risk of rheumatoid arthritis: a systematic review and dose-response meta-analysis. Arthritis Res Ther 2015;17:86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Heit JA, Spencer FA, White RH.. The epidemiology of venous thromboembolism. J Thromb Thrombolysis 2016;41:3–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Anderson FA Jr, Spencer FA.. Risk factors for venous thromboembolism. Circulation 2003;107:I9–16. [DOI] [PubMed] [Google Scholar]
- 36. Hernán MA. The hazards of hazard ratios. Epidemiology 2010;21:13–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Aviña-Zubieta JA, Choi HK, Sadatsafavi M. et al. Risk of cardiovascular mortality in patients with rheumatoid arthritis: a meta-analysis of observational studies. Arthritis Rheum 2008;59:1690–7. [DOI] [PubMed] [Google Scholar]
- 38. Ward MM. Interpreting studies of cardiovascular mortality in rheumatoid arthritis: the importance of timing. Arthritis Rheum 2008;59:1687–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Xu J, Lupu F, Esmon CT.. Inflammation, innate immunity and blood coagulation. Hamostaseologie 2010;30:5–9. [PubMed] [Google Scholar]
- 40. Olech E, Merrill JT.. The prevalence and clinical significance of antiphospholipid antibodies in rheumatoid arthritis. Curr Rheumatol Rep 2006;8:100–8. [DOI] [PubMed] [Google Scholar]
- 41. Scott IC, Hider SL, Scott DL.. Thromboembolism with Janus Kinase (JAK) inhibitors for rheumatoid arthritis: how real is the risk? Drug Saf 2018;41:645–53. [DOI] [PubMed] [Google Scholar]
- 42. Harigai M. Growing evidence of the safety of JAK inhibitors in patients with rheumatoid arthritis. Rheumatology (Oxford) 2019;58:i34–42. [DOI] [PMC free article] [PubMed] [Google Scholar]