Abstract
Objectives
To estimate the risk of venous thromboembolism (VTE) in a nationwide cohort of newly diagnosed patients with SLE and in a cohort of established patients with SLE at Karolinska University Hospital compared with the general population.
Methods
Individuals with SLE were identified from the National Patient Register and matched to general population comparators on age, sex and residence. Follow-up was from diagnosis/matching until incident VTE, death, emigration or study end. The Karolinska cohort was followed from enrolment and additionally stratified by lupus nephritis (LN) and antiphospholipid antibody (aPL) positivity. Incidence rates (IR) of VTE were estimated and adjusted HRs with 95% CIs were estimated using Cox models. Time-dependent adjusted HRs were estimated using flexible parametric survival models.
Results
In the nationwide cohort (N=4335), the mean age at inclusion was 49 years, 83% were female and mean follow-up was 7.4 years. The VTE IR was 8.8 per 1000 person-years compared with 2.8 in comparators, corresponding to an HR of 3.3 (95% CI 2.9 to 3.8). In the Karolinska cohort (N=784), the HR was 4.9 (95% CI 3.7 to 6.5). VTE risk was highest in the first few years after SLE diagnosis. LN was not significantly associated with VTE, and aPL positivity was associated with a 60% higher hazard, which was non-significant and attenuated after excluding individuals with prior VTE.
Conclusion
SLE is associated with an over threefold increased risk of VTE, in particular soon after SLE diagnosis, underscoring the need for individualised VTE risk assessment.
Keywords: Hematology; Lupus Erythematosus, Systemic; Lupus Nephritis; Antibodies, Antiphospholipid; Epidemiology
WHAT IS ALREADY KNOWN ON THIS TOPIC
Patients with SLE are at risk of venous thromboembolism (VTE), but evidence remains limited for SLE-specific patient subgroups.
WHAT THIS STUDY ADDS
SLE is associated with an over threefold higher risk of VTE compared with people without SLE. The risk is especially high in the first few years after SLE diagnosis and in antiphospholipid antibody-positive patients.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These results may support future stratification of patients at high risk of VTE, helping to identify those who may benefit from targeted preventative measures.
Introduction
Venous thromboembolism (VTE), comprising deep vein thrombosis (DVT) and pulmonary embolism (PE), is a common condition with an estimated lifetime risk of approximately 1 in 12 individuals. The disease carries a substantial burden due to the need for long-term anticoagulant therapy and the risk of associated complications.1 2 The incidence increases with age, from around 1 per 10 000 persons per year in young adults (<25 years old) to 1 per 100 persons per year in older adults (>75 years).3 Despite the extensive knowledge on multifactorial risk factors (acute/subacute triggers, hereditary and acquired risk factors), more than one-third of VTE events occur in the absence of an obvious provoking factor.4
Patients with rheumatic diseases have an increased risk for developing cardiovascular disease and VTE. Active inflammation at disease onset as well as during flares increases coagulability and leads to thrombosis.5 SLE is a complex autoimmune disease characterised by B-cell dysfunction, immune complex deposition and inflammation, which carries a particularly high risk of VTE.6 Antiphospholipid antibodies (aPL) are present in about 30%–40% of patients with SLE and are a significant and independent risk factor for thrombotic events.7–9 Lupus nephritis (LN) has been reported to be a significant risk factor for vascular thrombosis in patients with SLE, especially those with heavy proteinuria.10 11 Alterations in the coagulation system including platelet activation, endothelial dysfunction, complement activation and systemic inflammation during active disease promote a prothrombotic state. In addition, potential adverse effects of medications such as corticosteroids may further increase this risk.12
Exploring the relationship between SLE and SLE-specific risk factors for VTE risk is required for identifying patients at the highest risk. Thus, we aimed to estimate the risk of VTE in a nationwide cohort of newly diagnosed patients with SLE compared with the general population matched on age, sex and residential area. Furthermore, we examined the relationship between the presence of aPL, LN and VTE in a prevalent cohort of patients with SLE from the Rheumatology Clinic at Karolinska University Hospital. While previous studies have reported an increased risk of VTE in SLE, important gaps remain, including limited data on risk stratified by time since diagnosis, the role of aPL and clinical subgroups and the need for large, population-based estimates with appropriate comparators. Our study addresses these gaps by combining a nationwide cohort with detailed clinical data from a well-characterised tertiary cohort, allowing us to assess VTE risk over time since diagnosis and across clinically relevant subgroups.
Methods
Swedish national registers and a matched cohort study design were used to compare the risk of VTE in individuals with SLE with those without SLE. Diagnoses are recorded by the treating physician and coded using the International Classification of Diseases (ICD) system.
Study population
We identified individuals with SLE using the National Patient Register (NPR), containing data on inpatient care 1964–2022 (nationwide since 1987) and specialised outpatient visits since 2001, including acute care. We included individuals 18 years or older with at least two discharge diagnoses occurring within 1 year listing an ICD code for SLE, requiring at least one code to be in a clinic, which typically diagnoses SLE (rheumatology, internal medicine, nephrology, dermatology). The median time between these SLE-coded visits was 2 months. This definition has been shown to result in an accurate classification of SLE in Swedish register data.13 As SLE is primarily diagnosed and managed in outpatient settings, the availability of outpatient data is essential for accurate case identification. To best identify newly diagnosed (incident) SLE, we restricted the inclusion to individuals with a first ever SLE ICD code January 2006 or later thus allowing for at least 5 years of outpatient data (from 2001) and at least 19 years of inpatient data (from 1987 when coverage became nationwide) to identify prior SLE diagnoses and thereby reduce misclassification of prevalent cases as incident. Date and cause of death were retrieved from Cause of Death Register, which contains information on all deaths registered using ICD codes since 1961.
We also identified adults with SLE enrolled in a cohort at Karolinska University Hospital from 2004 through 2022, referred to as the Karolinska SLE cohort. This cohort includes both incident and prevalent SLE. All patients with SLE in the Karolinska SLE cohort were investigated in person at inclusion and patients included before 2015 fulfilled the 1982 American College of Rheumatology (ACR) criteria for SLE,14 thereafter we required that either the 1982 ACR criteria or the Systemic Lupus International Collaborating Clinics (SLICC) criteria15 were fulfilled. Investigations included a structured interview, a physical examination and review of medical files.
For each patient with SLE, 10 individuals from the general population, without SLE matched by birth year, sex and residential area, were selected from the Total Population Register. Each population-based comparator was required to be living in Sweden at the start of follow-up (matching date) of his/her corresponding patient with SLE. The nationwide SLE and the Karolinska SLE cohorts each had separate general population comparator groups.
Outcome: VTE
The primary outcome was overall VTE (see online supplemental table S1 for ICD codes). VTE was further categorised into non-obstetric (DVT and PE) and pregnancy-related VTE (VTE event during pregnancy and puerperium). The outcome was defined as the first registration with an ICD-10 code for VTE as main or secondary diagnosis in the NPR plus the requirement of a filled anticoagulant prescription within 30 days of VTE diagnosis, or VTE listed as the underlying cause of death in the Cause of Death Register during follow-up. If there was no anticoagulant prescription, but death occurred from any cause within 30 days of VTE, it was considered a VTE event (online supplemental table S1). When both DVT and PE codes were recorded for the same individual within the same clinical episode, these were considered part of a single VTE event and not counted separately in the overall VTE analysis. In analyses where DVT and PE were examined as separate outcomes, each event type was considered individually. A recent systematic review and meta-analysis of 24 studies evaluating the validity of ICD-10 VTE codes highlighted that combining ICD codes with evidence of anticoagulation therapy significantly improves diagnostic accuracy.16
Covariates
Information from several national registers was linked to the study population using each individual’s unique personal identification number. Age (calculated from birth year and month), sex and country of birth (categorised as Nordic, non-Nordic Europe and outside Europe) were collected from the Total Population Register. Information on comorbidities was collected from the NPR using ICD codes. This included data on any hospitalisation or fractures occurring within the 3 months prior to the start of follow-up as well as history of diabetes, acute coronary syndrome and/or other cardiac diseases, cerebrovascular disease, hypertension, renal disease and hereditary or acquired coagulopathy (including history of VTE) and APS. The Cancer Register was used to identify cancer diagnoses occurring before the index date. We also collected information from the Prescribed Drug Register (PDR) on pharmacy dispensations between July 2005 and 2022. This included anticoagulant treatment, corticosteroids and immunosuppressant therapy with conventional, targeted and biological disease-modifying antirheumatic drugs, menopausal hormonal therapy and oestrogen contraceptives dispensed in the 365 days prior to start of follow-up. All medications were identified using ATC codes (online supplemental table S2).
In the Karolinska SLE cohort, blood samples taken at inclusion were tested for aPL at cohort inclusion using the BioPlex 2200 (Bio-Rad) method. The aPL antibodies assessed were anticardiolipin (aCL) IgG and IgM as well as anti-β2 glycoprotein I (anti-β2GPI) IgG and IgM. Lupus anticoagulant (LA) testing followed the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis guidelines.17 aPL antibody profile was defined as single, double or triple positivity. Due to the limited numbers in each subgroup, we combined aPL-positive patients into a single group to retain statistical power. Specialised rheumatologists reviewed medical records and performed a detailed examination of the individuals at the baseline visit when the patients were enrolled in the Karolinska SLE cohort. Whether the patient had a diagnosis of APS or LN was evaluated at this time point. LN was defined by proteinuria >0.5 g/day as documented in the medical records, reflecting a history of LN rather than active disease at baseline. Most diagnoses of LN were confirmed by kidney biopsy. APS was diagnosed by a rheumatology specialist based on Sidney classification criteria including the persistent presence of any aPL positivity in combination with a documented clinical manifestation, defined as either thrombotic events or pregnancy morbidity.
Follow-up
Start of follow-up was the date the patients with SLE reached criteria for inclusion (ie, the second qualifying SLE-coded visit) or matching date for comparators and ended at VTE event, emigration date (retrieved from the Total Population Register), death or end of the study (31 December 2022), whichever came first. In analyses of the Karolinska SLE cohort, follow-up started at the time of enrolment in the cohort and had the same end of follow-up as the nationwide SLE cohort. Their matched comparators started follow-up on the enrolment date of their matched case.
Statistical analysis
Incidence rates (IRs) of any VTE per 1000 person-years (py) were calculated as well as for respective VTE subtypes (DVT, PE and pregnancy-related VTE) for both the Nationwide and Karolinska cohorts. Age and sex-adjusted Cox models were used to estimate the HR and 95% CI (HR 95% CI) for VTE comparing SLE with non-SLE. We did not adjust for comorbidities or treatments because they are consequences of the exposure (intermediates) and would remove part of the true effect of the exposure and lead to overadjustment. In analyses of PE, the outcome was the first PE event regardless of prior DVT, and similarly, for DVT analyses, the outcome was the first DVT event regardless of prior PE. We also calculated time-dependent adjusted HRs using flexible parametric survival models, which use spline functions for modelling hazard functions. Cumulative incidences of VTE during follow-up were estimated and plotted for SLE and the general population taking the competing risk of death into account. Furthermore, the Karolinska cohort was stratified by LN, APS and aPL status, and additional analyses were conducted by comparing the risk of VTE in subgroups defined by LN, APS and aPL status.
Because the Karolinska cohort includes both incident and prevalent SLE cases, in a secondary analysis, we restricted to individuals diagnosed with SLE within 1 year prior to the inclusion date to make the study population more similar to the nationwide cohort. Given the strong confounding effect of prior VTE on recurrence risk, we conducted a sensitivity analysis restricted to VTE-naïve individuals. Individuals with a history of VTE were excluded, and the primary analyses were rerun to evaluate whether the associations held in this subpopulation. Categories were collapsed if fewer than five events occurred in a stratum and if there still remained strata with fewer than five individuals, numbers were not reported and no IRs or HRs were calculated. All statistical analyses were performed with R software V.4.3.3. This study was presented as a poster abstract at the EULAR conference in 2025.18
Results
The study population included 4 335 individuals with SLE and 43 042 general population comparators. Females comprised 83% of the study population, with a mean age of 49 at the start of follow-up and mean follow-up of 7.5 years in SLE and 8.2 years in the general population (table 1). Prior to SLE diagnosis, 9.0% of SLE patients had a VTE vs 1.9% of matched comparators (table 1). A higher proportion of individuals with SLE had a history of hypertension/cardiovascular disease (27.4% vs 13.7%), hospitalisation in the previous 3 months (28.3% vs 3.1%), renal disease (9.9% vs 0.9%) and coagulopathy (including APS; 3.0% vs 0.1%). Moreover, more than half of the SLE patients were treated with glucocorticoids (61.1% vs 5.0%) and the proportion of patients on acetylsalicylic acid or anticoagulation treatment was two times that of the general population (aspirin 15.1% vs 7.0%, anticoagulation 14.4% vs 3.6%). There were 784 patients with SLE from the Karolinska cohort included in the study. The Karolinska cohort was similar to the nationwide cohort, with slightly younger individuals at inclusion (mean age 45.5 years) of which 86% were women (table 1).
Table 1. Characteristics of the SLE and general population groups at start of follow-up.
| General population matched to nationwide SLE (N=43 042) | Nationwide SLE (N=4335) | General population matched to Karolinska SLE (N=7471) |
Karolinska SLE (N=784) | |
|---|---|---|---|---|
| Female | 35 593 (82.7%) | 3584 (82.7%) | 6382 (85.4%) | 671 (85.6%) |
| Age at start of follow-up, mean (±SD) | 49.2 (±18.2) | 49.2 (±18.3) | 45.2 (±15.1) | 45.5 (±15.2) |
| Follow-up duration in person-years, mean (±SD) | 8.2 (±4.7) | 7.5 (±4.8) | 10.4 (±5.6) | 9.4 (±5.6) |
| Country of birth | ||||
| Nordic | 36 124 (83.9%) | 3393 (78.3%) | 6178 (82.7%) | 627 (80.0%) |
| Non-Nordic Europe | 2513 (5.8%) | 238 (5.5%) | 457 (6.1%) | 28 (3.6%) |
| Outside Europe | 3860 (9.0%) | 624 (14.4%) | 741 (9.9%) | 117 (14.9%) |
| Missing | 545 (1.3%) | 80 (1.8%) | 95 (1.3%) | 12 (1.5%) |
| Comorbidities* | ||||
| Diabetes | 1772 (4.1%) | 228 (5.3%) | 193 (2.6%) | 20 (2.6%) |
| Hypertension/cardiovascular disease | 5883 (13.7%) | 1188 (27.4%) | 584 (7.8%) | 199 (25.4%) |
| Cancer diagnosed within a year | 340 (0.8%) | 65 (1.5%) | 42 (0.6%) | 4 (0.5%) |
| Hospitalisation in the previous 3 months | 1347 (3.1%) | 1227 (28.3%) | 215 (2.9%) | 130 (17.5%) |
| Fractures in the previous 3 months | 229 (0.5%) | 56 (1.3%) | 41 (0.5%) | 6 (0.8%) |
| Coagulopathy | 49 (0.1%) | 129 (3.0%) | 7 (0.1%) | 51 (6.5%) |
| Renal disease | 368 (0.9%) | 430 (9.9%) | 48 (0.6%) | 144 (18.4%) |
| Cerebrovascular disease | 1302 (3.0%) | 239 (5.5%) | 115 (1.5%) | 61 (7.8%) |
| Antiphospholipid syndrome, from medical records | – | – | – | 107 (13.6%) |
| Lupus nephritis, from medical records | – | – | – | 289 (36.9%) |
| aPL positivity, from medical records | – | – | – | 225 (28.7%) |
| Treatments† | ||||
| Anticoagulants | 1182 (2.7%) | 623 (14.4%) | 141 (1.9%) | 128 (16.3%) |
| Aspirin | 3023 (7.0%) | 654 (15.1%) | 203 (2.7%) | 112 (14.3%) |
| MHT or oestrogen contraception | 4953 (11.5%) | 461 (10.6%) | 712 (9.5%) | 68 (8.7%) |
| Glucocorticosteroids | 2152 (5.0%) | 2650 (61.1%) | 230 (3.1%) | 393 (50.1%) |
| Immunosuppressants | 562 (1.3%) | 982 (22.7%) | 61 (0.8%) | 242 (30.9%) |
| Hydroxychloroquine | 50 (0.1%) | 2332 (53.8%) | 7 (0.1%) | 358 (45.7%) |
| History of VTE | 823 (1.9%) | 391 (9.0%) | 113 (1.5%) | 88 (11.2%) |
| VTE | 805 (1.9%) | 389 (9.0%) | 111 (1.5%) | 88 (11.2%) |
| Pregnancy-related VTE | 26 (0.1%) | 17 (4.5%) | 2 (0.0%) | 2 (2.3%) |
Start of follow-up was the date criteria were reached for inclusion in the Nationwide SLE cohort (second SLE-coded visit, with at least one visit to a specialist), date of enrolment in the Karolinska SLE cohort, or matching date in the general population.
Retreived from the National Patient Register except where noted otherwise.
Retrieved from the Prescribed Drug Register, only includes medications received in pharmacies in the year prior to start of follow-up and does not include infusions or treatments given in hospital.
aPL, antiphospholipid antibodies; MHT, menopausal hormone therapy; SLE, systemic lupus erythematosus; VTE, venous thromboembolism.
During the study period, there were 288 incident VTEs in the nationwide SLE cohort versus 970 in the general population and 72 VTEs in Karolinska SLE cohort versus 154 in the general population. Close to 100% of VTEs in both SLE and non-SLE groups were not pregnancy-related, mostly PE and DVT. Mean age at VTE was younger in SLE compared with non-SLE (Nationwide SLE: 58.1 vs 65.8 years; Karolinska SLE: 51.5 vs 63.9 years; table 2). The IR of any VTE was 8.8 per 1000 person-years (95% CI 7.8 to 9.9) in Nationwide SLE and 2.8 per 1000 person-years (95% CI 2.6 to 2.9) in their matched comparators; the IR in Karolinska cohort was 9.7 (95% CI 7.6 to 12.2) in SLE and 2.0 (95% CI 1.7 to 2.3) in comparators. Individuals with SLE had a higher hazard of any VTE compared with the general population (Nationwide SLE: HR 3.3, 95% CI 2.9 to 3.8; Karolinska SLE: HR 4.9, 95% CI 3.7 to 6.5). The same increased hazards for VTE were also found when looking separately at PE and DVT (table 2).
Table 2. Number of VTE events, person-years, crude incidence rates and HRs for incident VTE comparing SLE in nationwide and Karolinska cohorts with the general population in Sweden.
| Nationwide SLE | General population | Karolinska SLE | General population | |
|---|---|---|---|---|
| Any first VTE, n (%) | 288 (6.6) | 970 (2.3) | 72 (9.2) | 154 (2.1) |
| Age at VTE diagnosis, mean (±SD) | 58.1 (18.7) | 65.8 (16.2) | 51.5 (15.8) | 63.9 (14.5) |
| Follow-up duration in person-years, mean (±SD) | 7.5 (4.8) | 8.2 (4.7) | 9.4 (5.6) | 10.4 (5.6) |
| Incidence rate (95% CI)* | 8.8 (7.8–9.9) | 2.8 (2.6–2.9) | 9.7 (7.6–12.2) | 2.0 (1.7–2.3) |
| HR (95% CI)† | 3.3 (2.9 to 3.8) | 1.0 (ref) | 4.9 (3.7 to 6.5) | 1.0 (ref) |
| Pulmonary embolism, n (%) | 146 (3.4) | 477 (1.1) | 34 (4.3) | 82 (1.1) |
| Follow-up duration in person-years, mean (±SD) | 7.7 (4.8) | 8.2 (4.7) | 9.8 (5.5) | 10.4 (5.6) |
| Incidence rate (95% CI)* | 4.4 (3.7 to 5.1) | 1.3 (1.2 to 1.5) | 4.4 (3.1 to 6.2) | 1.1 (0.8 to 1.3) |
| HR (95% CI)† | 3.4 (2.9 to 4.2) | 1.0 (ref) | 4.3 (2.9 to 6.5) | 1.0 (ref) |
| Deep vein thrombosis, n (%) | 168 (3.9) | 573 (1.3) | 44 (5.6) | 92 (1.2) |
| Follow-up duration in person-years, mean (±SD) | 7.7 (4.7) | 8.2 (4.7) | 9.6 (5.6) | 10.4 (5.6) |
| Incidence rate (95% CI)* | 5.1 (4.3 to 5.9) | 1.6 (1.5 to 1.8) | 5.8 (4.2 to 7.8) | 1.2 (1.0 to 1.5) |
| HR (95% CI)† | 3.2 (2.7 to 3.8) | 1.0 (ref) | 4.9 (3.4 to 7.1) | 1.0 (ref) |
Incidence rate per 1000 person-years, CI estimated based on the Poisson distribution.
Proportional hazards Cox regression model estimated HRs and 95% CI adjusted for age and sex.
SLE, systemic lupus erythematosus; VTE, venous thromboembolism.
Time-dependent HRs for VTE were highest soon after start of follow-up (second SLE diagnosis in the nationwide cohort and date of enrolment in the Karolinska SLE cohort) (table 3, figure 1). The HRs declined over time, but remained above 1 throughout the follow-up time.
Table 3. Time-dependent HRs and 95% CIs for VTE, pulmonary embolism and deep vein thrombosis comparing SLE to general population comparators at follow-up times adjusted by age and sex.
| Event | Follow-up time* | Nationwide SLE vs general population time-dependent HR (95% CI)* |
Karolinska SLE vs general population time-dependent HR (95% CI)† |
|---|---|---|---|
| Any first VTE | 1 year | 4.1 (3.3 to 5.1) | 8.8 (5.4 to 14.4) |
| 3 years | 2.7 (2.2 to 3.2) | 5.6 (3.8 to 8.2) | |
| 5 years | 2.5 (2.2 to 3.0) | 4.4 (3.2 to 6.1) | |
| 10 years | 2.2 (1.8 to 2.7) | 3.3 (2.1 to 5.0) | |
| Pulmonary embolism | 1 year | 5.3 (3.9 to 7.2) | 5.4 (2.5 to 11.6) |
| 3 years | 2.7 (2.0 to 3.5) | 4.3 (2.4 to 7.6) | |
| 5 years | 2.4 (1.9 to 3.1) | 4.0 (2.6 to 6.4) | |
| 10 years | 1.8 (1.3 to 2.5) | 3.9 (2.3 to 6.5) | |
| Deep vein thrombosis | 1 year | 3.7 (2.8 to 4.9) | 11.2 (6.0 to 20.9) |
| 3 years | 2.6 (2.1 to 3.2) | 5.6 (3.4 to 9.4) | |
| 5 years | 2.5 (2.1 to 3.1) | 4.1 (2.7 to 6.4) | |
| 10 years | 2.2 (1.7 to 2.9) | 2.1 (1.0 to 4.4) |
Time since the second qualifying SLE-coded visit (nationwide SLE), enrolment (Karolinska SLE) or matching date (general population).
Time-dependent HRs and 95% CIs adjusted for age and sex were estimated using flexible parametric survival models.
SLE, systemic lupus erythematosus; VTE, venous thromboembolism.
Figure 1. Time-dependent hazard ratios (HR) and 95% confidence intervals for venous thromboembolism (VTE) comparing systemic lupus erythematosus (SLE) to general population comparators adjusted by age and sex. HRs associated with nationwide SLE for A) any first VTE, B) pulmonary embolism (PE) and C) deep vein thrombosis (DVT) and HRs associated with the Karolinska Cohort for D) any first VTE, E) PE and F) DVT.
Among the Karolinska SLE cohort, the mean age at VTE diagnosis was lower in patients with LN compared with without (47.0 vs 55.4), and in patients with aPL positivity compared with without (48.7 vs 53.3). Patients with LN had a higher hazard of VTE than those without LN (HR 1.3, 95% CI 0.8 to 2.2; table 4), though the association was not statistically significant. The hazard of VTE was 60% higher in those with aPL antibody positivity compared with those without but the CI included the null (1.6 95% CI 1.0 to 2.6; table 4).
Table 4. Number of VTE events, person-years, crude incidence rates and HR for prevalent VTE (VTE overall, PE and DVT) when stratifying the Karolinska SLE cohort by LN and aPL.
| Karolinska SLE with LN | Karolinska SLE without LN | |
|---|---|---|
| Any first VTE, n (%) | 33 (11.4) | 39 (7.9) |
| Age at VTE diagnosis, mean (±SD) | 47.0 (16.1) | 55.4 (14.6) |
| Follow-up duration in person-years, mean (±SD) | 9.7 (5.5) | 9.3 (5.6) |
| Incidence rate (95% CI)* | 11.8 (8.1 to 16.5) | 8.5 (6.0 to 11.6) |
| HR (95% CI)† | 1.3 (0.8 to 2.2) | 1.0 (ref) |
| Pulmonary embolism, n (%) | 14 (4.8) | 20 (4.0) |
| Follow-up duration in person-years, mean (±SD) | 10.2 (5.4) | 9.6 (5.5) |
| Incidence rate (95% CI)* | 4.8 (2.6 to 8.0) | 4.2 (2.6 to 6.5) |
| HR (95% CI)† | 1.2 (0.6 to 2.4) | 1.0 (ref) |
| Deep vein thrombosis, n (%) | 22 (7.6) | 22 (4.4) |
| Follow-up duration in person-years, mean (±SD) | 10.0 (5.5) | 9.4 (5.6) |
| Incidence rate (95% CI)* | 7.7 (4.8 to 11.6) | 4.7 (2.9 to 7.1) |
| HR (95% CI)† | 1.5 (0.8 to 2.8) | 1.0 (ref) |
| Karolinska SLE with aPL+ | Karolinska SLE with aPL- | |
|---|---|---|
| Any first VTE, n (%) | 27 (12.0) | 45 (8.1) |
| Age at VTE diagnosis, mean (±SD) | 48.7 (14.2) | 53.3 (16.5) |
| Follow-up duration in person-years, mean (±SD) | 8.8 (5.8) | 9.7 (5.5) |
| Incidence rate (95% CI)* | 13.7 (9.0 to 19.9) | 8.3 (6.0 to 11.1) |
| HR (95% CI)† | 1.6 (1.0–2.6) | 1.0 (ref) |
| Pulmonary embolism, n (%) | 11 (4.9) | 23 (4.1) |
| Follow-up duration in person-years, mean (±SD) | 9.4 (5.7) | 10.0 (5.4) |
| Incidence rate (95% CI)* | 5.2 (2.6 to 9.3) | 4.1 (2.6 to 6.2) |
| HR (95% CI)† | 1.2 (0.6 to 2.5) | 1.0 (ref) |
| Deep vein thrombosis, n (%) | 19 (8.4) | 25 (4.5) |
| Follow-up duration in person-years, mean (±SD) | 9.0 (5.8) | 9.9 (5.5) |
| Incidence rate (95% CI)* | 9.4 (5.7 to 14.7) | 4.5 (2.9 to 6.7) |
| HR (95% CI)† | 2.1 (1.1 to 3.8) | 1.0 (ref) |
Incidence rate per 1000 person-years, CI estimated based on the Poisson distribution.
Proportional hazard Cox regression model estimated HR 95% CI adjusted for age and sex.
aPL, antiphospholipid antibody; DVT, deep vein thrombosis; LN, lupus nephritis; SLE, systemic lupus erythematosus; VTE, venous thromboembolism.
Excluding individuals with a history of VTE
Given the importance that a previous VTE event carries for VTE recurrence, we further examined whether the risk profile was consistent in a VTE ‘naïve’ population. After excluding individuals with a history of VTE, the results did not differ from the main analysis (Nationwide SLE: HR 3.0, 95% CI 2.6 to 3.5; Karolinska SLE: HR 4.2, 95% CI 3.0 to 5.9; online supplemental table S3, online supplemental table s2). Within the Karolinska VTE ‘naïve’ population, the association with LN remained elevated but not statistically significant (HR: 1.5, 95% CI 0.8 to 2.6). In contrast, the association between aPL positivity and VTE attenuated from 1.6 to 0.9 (HR: 0.9, 95% CI 0.5 to 1.7; online supplemental table S4).
Restricting the Karolinska SLE cohort to newly diagnosed patients
Because the Karolinska cohort includes both incident and prevalent SLE, we estimated the risk of VTE in newly diagnosed patients by including only those diagnosed with SLE within 1 year prior to the inclusion date. The newly diagnosed patients with SLE in the Karolinska Cohort had a higher rate of any VTE than the overall cohort (15.3 per 1000 person-years; 95% CI 9.2 to 23.9). Furthermore, the HR for VTE in newly diagnosed Karolinska patients with SLE was higher than in the main analysis, with an up to 10.8 times higher hazard compared with general population (HR 10.8; 95% CI 5.7 to 20.2). Any aPL positivity in newly diagnosed SLE Karolinska patients had 2.6-fold higher hazard for any VTE compared with aPL- group (HR 3.0, 95% CI 1.0 to 8.8; online supplemental table S5). Among the patients enrolled in the Karolinska SLE cohort who were newly diagnosed, the LN group had a similar hazard to the group without LN although with wide confidence intervals (HR 0.9, 95% CI 0.3 to 2.5). Results were similar when comparing patients with APS to those without (online supplemental table S6).
Discussion
In this nationwide register matched study, there was a three times elevated risk of VTE in individuals with SLE compared with the general population, with the highest relative risk within the first year after diagnosis. Importantly, the increased risk was similar for both PE and DVT and was not explained by prior VTE history. Among patients followed at Karolinska Hospital, we observed no statistically significant difference in risk by LN status, while the aPL positive group had a 60% higher risk (with a CI including 1), which diminished when excluding those with prior VTE. The incidence of VTE was highest among newly diagnosed patients with SLE.
Large population-based studies have consistently reported a 2-fold to 12-fold increased risk of VTE in SLE, with IRs ranging from approximately 1.8 to 10 per 1000 person-years, compared with 0.7–1.5 per 1000 person-years in matched controls. The risk is highest in the first year after SLE diagnosis, with HRs exceeding 10 during this period.19–22 Our estimate of a 3.3 times higher hazard of VTE in SLE compared with the general population is in line with the majority of other studies.21–23 The higher risk for VTE in the first year after diagnosis indicates that patients with SLE are already at an increased risk of VTE early in the disease course. When analysing DVT and PE separately, we observed the same trends and incident rates consistent with those reported in the literature.19–21
The long observation period in this study allowed for analysis of trends over time since diagnosis. Our examination of VTE rates over time showed a reduction for any VTE as well as DVT and PE, after the first 3 years of SLE diagnosis. However, the risk remained elevated compared with comparators even years later.24 25 The markedly increased risk of VTE in newly diagnosed SLE appears to be multifactorial, involving both disease-related and treatment-related mechanisms. Active inflammation, endothelial dysfunction, complement activation and immune-mediated vascular injury create a prothrombotic state, further combined by risk factors such as nephrotic syndrome and aPL.19 22 26–28 In parallel, glucocorticoid treatment frequently prescribed at high doses in the first year to control disease activity has been independently associated with VTE through adverse effects on coagulation, endothelial function and metabolic parameters.20 22 27 29 Together, these factors may explain the highest VTE risk in the first year after diagnosis compared with the general population.
APS and aPL remain a consistent and robust risk factor for VTE in SLE, with numerous cohort studies and meta-analyses showing that the presence of aPL, particularly LA, double or triple positivity, markedly increases thrombotic risk.30–32 Our findings, showing that aPL positivity is associated with a 60% higher risk of VTE among SLE patients, are in line with several previous studies reporting the presence of APS or aPL as major risk factors for VTE in SLE.28–32 Furthermore, the presence of aPL positivity in newly diagnosed Karolinska SLE patients is associated with a threefold higher risk of any VTE, which is in line with a recent Norwegian study.33
Interestingly, when excluding the patients with history of VTE, the risk of VTE associated with aPL positivity reduced, becoming similar between the aPL positive and aPL negative groups. This indicates that the increased risk associated with aPL in the main analysis was driven by those who had a history of VTE (ie, recurrent VTE). Fifty-nine per cent of the aPL+VTEs had a prior VTE compared with only 18% of the aPL- group. Previous large-scale analyses have demonstrated that SLE-related autoantibodies, including aPL, may be detectable several years prior to the clinical diagnosis of SLE. Accordingly, VTE occurring before SLE diagnosis could potentially reflect the presence of circulating aPL that had not yet been clinically identified.34 Thus, these results may be carefully interpreted. Furthermore, aPL positivity in SLE is a major determinant of adverse outcomes, strongly associated with other thrombotic events than only VTE and its contribution to excess long-term mortality.19 35 The presence of aPL is also linked to organ-specific complications such as microvascular kidney disease, pulmonary hypertension, cytopenias and adverse pregnancy outcomes, all of which accelerate damage accrual and impair quality of life.36
We observed that a history of LN at enrolment into the Karolinska SLE Cohort was associated with an elevated but not statistically significant higher risk of subsequent VTE (HR=1.3; 95% CI 0.8 to 2.2), and this finding remained unchanged after excluding individuals with prior VTE. This is in contrast with earlier studies, which reported that active LN was an independent risk factor for VTE, particularly during periods of high systemic disease activity. However, LN status in our analyses was defined as ever LN, and mostly reflected prior diagnosis rather than concurrent active nephritis. This is in line with more recent longitudinal studies, which also report weaker or null associations, especially among patients with early SLE or those in whom LN developed later in the disease course.10 19 22 Therefore, differences between our findings and earlier reports may reflect variation in the timing of LN relative to SLE onset, differences in nephritis activity and severity and evolving treatment approaches.26 37
Our study has some limitations. We used the date of the second criteria-fulfilling SLE-coded visit as the start of follow-up in the nationwide SLE cohort, missing early VTEs occurring before the diagnosis or between the first and second SLE-coded visits. Of patients with SLE with a history of VTE, one fourth (N=93) had a VTE during this time period. In our study, the median time between first and second SLE diagnosis (when criteria for inclusion were fulfilled) was 2 months. Moreover, some VTEs and their subtypes could have been misclassified, as we relied on ICD codes from the Patient Register and the Cause of Death Register, and the accuracy of some of the codes has not been investigated (especially regarding pregnancy-related VTE). The sensitivities of PE and DVT codes are affected by their location, chronicity and copresentation with other conditions.29 We could not take into account smoking, obesity or immobility in this study because these factors are not reliably captured with ICD-10 codes, leading to residual confounding. How medications and disease activity play a role in VTE occurrence was not addressed, but examining the role of these factors is an important next step. We do not have information on in-hospital treatments since the PDR covers only treatments dispensed at the pharmacy. The Karolinska cohort includes patients with prevalent SLE, who may be subjected to survival bias, potentially reflecting lower disease severity or different risk profiles, which could influence observed VTE IRs and disease activity patterns compared with an inception cohort. The use of two diagnostic criteria sets for the Karolinska patients, and the absence of the 2019 SLE classification criteria due to the historical nature of the data, may limit the generalisability of our findings to patients classified according to contemporary definitions. Due to small numbers in the subgroups of single, double and triple aPL positivity, we combined all aPL-positive patients into one group and were, therefore, unable to assess the impact of different aPL profiles on VTE risk. Moreover, LN was defined as ever diagnosed, reflecting a history of LN rather than active disease at baseline. We acknowledge that LN is a heterogeneous condition and our definition does not capture important aspects such as disease activity, histopathological class or severity. These factors may be relevant to VTE risk and should be examined in future studies.
The current study benefits from the inclusion of high-quality prospectively collected inpatient and outpatient data from all of Sweden. Because these registers have nationwide coverage and allowed for follow-up of a relatively large population of individuals diagnosed with SLE for a number of years, it was possible to provide robust estimates of the risk of VTE associated with SLE. The large sample size also allowed for the examination of VTE subtypes. Our results are further supported by examining the risk of VTE in a subgroup of clinically confirmed SLE cases and the relative risk associated with LN and aPL positivity, patient characteristics that required more detailed clinical information from patient charts. We provide estimates restricted to newly diagnosed patients and those with no history of VTE at start of follow-up. We do not have information on race, but previous studies estimate that more than 90% of patients with SLE treated in Swedish university clinics are Caucasian.38 Our results are generalisable to other Caucasian SLE populations seen in outpatient care in the last decade. In the Karolinska SLE cohort, all the patients were referred to a specialised centre, so they might be at higher risk of VTE because of more severe and complex disease; often mild disease cases are not followed at a university hospital. Therefore, results from the Karolinska cohort should be generalised to similar patient groups.
In conclusion, the risk of VTE associated with SLE was over threefold the risk in the general population in an incident nationwide SLE cohort as well as in a university hospital cohort of patients. The increased risk in the first year after diagnosis highlights a time period where preventative measures like prophylactic anticoagulant treatment could be taken, underscoring the importance of risk stratification. Additional studies are required to clarify the VTE risk, so that the burden of VTE can be reduced in SLE population.
Supplementary material
Acknowledgements
Thank you to Nurgul Batyrbekova for assisting with the final tables and figures and reviewing the R code for this study.
Footnotes
Funding: This project was supported by The Swedish Research Council (grant number 2022-00783), Swedish Heart-Lung Foundation (grant number 20230296), Stockholm County Council (ALF, grant number FoUI-986660), Swedish Society of Medicine (grant number SLS-936450, SLS-713911), King Gustaf V 80th Birthday Fund (grant number FAI-2023-1024, FAI-2023-0971) and Swedish Rheumatism Association (grant number R-981193, R-994250).
Provenance and peer review: Not commissioned; externally peer-reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study involves human participants and was approved by Swedish Ethics Review Authority DNR 2021-01148. Consent from individuals is waived when using Swedish register data, as approved by the Swedish Ethics Authority.
Data availability free text: The data are protected by the Swedish Data Protection Act and GDPR and cannot be made publicly available. The authors may share aggregate‑level, non‑identifiable summary information upon reasonable request.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Data availability statement
Data are available upon reasonable request. Data may be obtained from a third party and are not publicly available.
References
- 1.Bell EJ, Lutsey PL, Basu S, et al. Lifetime Risk of Venous Thromboembolism in Two Cohort Studies. Am J Med. 2016;129:339. doi: 10.1016/j.amjmed.2015.10.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Arnesen CAL, Veres K, Horváth-Puhó E, et al. Estimated lifetime risk of venous thromboembolism in men and women in a Danish nationwide cohort: impact of competing risk of death. Eur J Epidemiol. 2022;37:195–203. doi: 10.1007/s10654-021-00813-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Brækkan SK, Hansen JB. VTE epidemiology and challenges for VTE prevention at the population level. Thrombosis Update . 2023;10:100132. doi: 10.1016/j.tru.2023.100132. [DOI] [Google Scholar]
- 4.Di Nisio M, van Es N, Büller HR. Deep vein thrombosis and pulmonary embolism. Lancet. 2016;388:3060–73. doi: 10.1016/S0140-6736(16)30514-1. [DOI] [PubMed] [Google Scholar]
- 5.Zöller B, Li X, Sundquist J, et al. Risk of pulmonary embolism in patients with autoimmune disorders: a nationwide follow-up study from Sweden. Lancet. 2012;379:244–9. doi: 10.1016/S0140-6736(11)61306-8. [DOI] [PubMed] [Google Scholar]
- 6.Palatinus A, Adams M. Thrombosis in systemic lupus erythematosus. Semin Thromb Hemost. 2009;35:621–9. doi: 10.1055/s-0029-1242716. [DOI] [PubMed] [Google Scholar]
- 7.Erkan D. Lupus and thrombosis. J Rheumatol. 2006;33:1715–7. [PubMed] [Google Scholar]
- 8.Bello N, Meyers KJ, Workman J, et al. Systematic Literature Review and Meta-analysis of Venous Thromboembolism Events in Systemic Lupus Erythematosus. Rheumatol Ther. 2023;10:7–34. doi: 10.1007/s40744-022-00513-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Calvo-Alén J, Toloza SMA, Fernández M, et al. Systemic lupus erythematosus in a multiethnic US cohort (LUMINA). XXV. Smoking, older age, disease activity, lupus anticoagulant, and glucocorticoid dose as risk factors for the occurrence of venous thrombosis in lupus patients. Arthritis Rheum. 2005;52:2060–8. doi: 10.1002/art.21149. [DOI] [PubMed] [Google Scholar]
- 10.Choojitarom K, Verasertniyom O, Totemchokchyakarn K, et al. Lupus nephritis and Raynaud’s phenomenon are significant risk factors for vascular thrombosis in SLE patients with positive antiphospholipid antibodies. Clin Rheumatol. 2008;27:345–51. doi: 10.1007/s10067-007-0721-z. [DOI] [PubMed] [Google Scholar]
- 11.Cooley I, Derebail VK, Gibson KL, et al. Chapel hill alliance promoting excellence in lupus (chapel) group of investigators. association of lupus nephritis histopathologic classification with venous thromboembolism-modification by age at biopsy. Kidney Int Rep. 2021;6:1653–60. doi: 10.1016/j.ekir.2021.02.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Larsen ML, Nørgaard L, Linge P, et al. Molecular mechanisms underlying thrombosis in systemic lupus erythematosus - A Systematic review. Semin Arthritis Rheum. 2025;72:152707. doi: 10.1016/j.semarthrit.2025.152707. [DOI] [PubMed] [Google Scholar]
- 13.Arkema EV, Jönsen A, Rönnblom L, et al. Case definitions in Swedish register data to identify systemic lupus erythematosus. BMJ Open. 2016;6:e007769. doi: 10.1136/bmjopen-2015-007769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1982;25:1271–7. doi: 10.1002/art.1780251101. [DOI] [PubMed] [Google Scholar]
- 15.Petri M, Orbai A-M, Alarcón GS, et al. Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus. Arthritis Rheum. 2012;64:2677–86. doi: 10.1002/art.34473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Liu B, Hadzi-Tosev M, Eisa K, et al. Accuracy of venous thromboembolism ICD-10 codes: A systematic review and meta-analysis. Thrombosis Update . 2024;14:100154. doi: 10.1016/j.tru.2023.100154. [DOI] [Google Scholar]
- 17.Devreese KMJ, de Groot PG, de Laat B, et al. Guidance from the Scientific and Standardization Committee for lupus anticoagulant/antiphospholipid antibodies of the International Society on Thrombosis and Haemostasis. J Thromb Haemost. 2020;18:2828–39. doi: 10.1111/jth.15047. [DOI] [PubMed] [Google Scholar]
- 18.Vojinovic T, Antovic A, Gunnarsson I, et al. POS0979 Occurrence of venous thromboembolism in systemic lupus erythematosus: a Swedish population-based cohort study. Ann Rheum Dis. 2025;84:1095–6. doi: 10.1016/j.ard.2025.06.334. [DOI] [Google Scholar]
- 19.Nossent JC, Keen HI, Preen DB, et al. Long-term incidence, risk factors and complications for venous thromboembolism in patients with systemic lupus erythematosus. Lupus. 2024;33:787–96. doi: 10.1177/09612033241247359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ojeda-Fernandez L, Calvisi S, Torrigiani G, et al. Incidence and risk of arterial and venous thrombotic events in systemic lupus erythematosus patients: a population-based study. Rheumatology (Oxford) 2025;64:2723–30. doi: 10.1093/rheumatology/keae496. [DOI] [PubMed] [Google Scholar]
- 21.Chung W-S, Lin C-L, Chang S-N, et al. Systemic lupus erythematosus increases the risks of deep vein thrombosis and pulmonary embolism: a nationwide cohort study. J Thromb Haemost. 2014;12:452–8. doi: 10.1111/jth.12518. [DOI] [PubMed] [Google Scholar]
- 22.Hinojosa-Azaola A, Romero-Diaz J, Vargas-Ruiz AG, et al. Venous and Arterial Thrombotic Events in Systemic Lupus Erythematosus. J Rheumatol. 2016;43:576–86. doi: 10.3899/jrheum.150506. [DOI] [PubMed] [Google Scholar]
- 23.Aviña-Zubieta JA, Vostretsova K, De Vera MA, et al. The risk of pulmonary embolism and deep venous thrombosis in systemic lupus erythematosus: A general population-based study. Semin Arthritis Rheum. 2015;45:195–201. doi: 10.1016/j.semarthrit.2015.05.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhao Y, Huang C, You H, et al. Prognostic factors of systemic lupus erythematosus patients with pulmonary embolism: An 11-year cohort study. Lupus. 2022;31:885–90. doi: 10.1177/09612033221095149. [DOI] [PubMed] [Google Scholar]
- 25.Yuan W, Guan F. Thrombosis and Anticoagulation Therapy in Systemic Lupus Erythematosus. Autoimmune Dis. 2022;2022:3208037. doi: 10.1155/2022/3208037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kharouf F, Mehta P, Li Q, et al. Venous thromboembolism in patients with lupus nephritis: frequency and risk factors—a cohort study. Rheumatology (Oxford) 2026;65 doi: 10.1093/rheumatology/keaf464. [DOI] [PubMed] [Google Scholar]
- 27.Svenungsson E, Gustafsson JT, Grosso G, et al. Complement deposition, C4d, on platelets is associated with vascular events in systemic lupus erythematosus. Rheumatology (Oxford) 2020;59:3264–74. doi: 10.1093/rheumatology/keaa092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Romero-Díaz J, García-Sosa I, Sánchez-Guerrero J. Thrombosis in systemic lupus erythematosus and other autoimmune diseases of recent onset. J Rheumatol. 2009;36:68–75. doi: 10.3899/jrheum.071244. [DOI] [PubMed] [Google Scholar]
- 29.Park D-J, Yoon C-S, Choi S-E, et al. Risk factors for thrombotic events in Korean patients with systemic lupus erythematosus. Sci Rep. 2021;11:23529. doi: 10.1038/s41598-021-03074-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Farina N, Abdulsalam R, McDonnell T, et al. Antiphospholipid antibody positivity in early systemic lupus erythematosus is associated with subsequent vascular events. Rheumatology (Oxford) 2023;62:2252–6. doi: 10.1093/rheumatology/keac596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Demir S, Li J, Magder LS, et al. Antiphospholipid patterns predict risk of thrombosis in systemic lupus erythematosus. Rheumatology (Oxford) 2021;60:3770–7. doi: 10.1093/rheumatology/keaa857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hisada R, Atsumi T. An Antiphospholipid Antibody Profile as a Biomarker for Thrombophilia in Systemic Lupus Erythematosus. Biomolecules. 2023;13:617. doi: 10.3390/biom13040617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Moe SR, Haukeland H, Brunborg C, et al. Accrual of thromboembolic events and antiphospholipid syndrome in new-onset systemic lupus erythematosus: a population-based inception cohort study. RMD Open. 2025;11:e005795. doi: 10.1136/rmdopen-2025-005795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Arbuckle MR, McClain MT, Rubertone MV, et al. Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N Engl J Med. 2003;349:1526–33. doi: 10.1056/NEJMoa021933. [DOI] [PubMed] [Google Scholar]
- 35.Riancho-Zarrabeitia L, Martínez-Taboada V, Rúa-Figueroa I, et al. Antiphospholipid syndrome (APS) in patients with systemic lupus erythematosus (SLE) implies a more severe disease with more damage accrual and higher mortality. Lupus. 2020;29:1556–65. doi: 10.1177/0961203320950477. [DOI] [PubMed] [Google Scholar]
- 36.Siegel CH, Sammaritano LR. Systemic Lupus Erythematosus: A Review. JAMA. 2024;331:1480–91. doi: 10.1001/jama.2024.2315. [DOI] [PubMed] [Google Scholar]
- 37.Kidney Disease: Improving Global Outcomes (KDIGO) Lupus Nephritis Work Group KDIGO 2024 Clinical Practice Guideline for the management of LUPUS NEPHRITIS. Kidney Int. 2024;105:S1–69. doi: 10.1016/j.kint.2023.09.002. [DOI] [PubMed] [Google Scholar]
- 38.Frodlund M, Reid S, Wetterö J, et al. The majority of Swedish systemic lupus erythematosus patients are still affected by irreversible organ impairment: factors related to damage accrual in two regional cohorts. Lupus. 2019;28:1261–72. doi: 10.1177/0961203319860198. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon reasonable request. Data may be obtained from a third party and are not publicly available.