BackgroundPatients with systemic lupus erythematosus, especially lupus nephritis (LN), have higher risk of thrombosis than the general population 1. Since use of corticosteroids also increase the risk of thrombosis 2, steroid pulse therapy (SPT) may increases the risk of thrombosis in patients with LN. However, few studies examined this association.ObjectivesTo compare risk of thrombosis between patients with and without SPT in LN.MethodsThis retrospective, propensity score-matched cohort study was conducted using claims data provided by Medical Data Vision Co., Ltd (Tokyo, Japan). We defined individuals as LN cases if they met all of the following: 1) were diagnosed as LN; 2) had a dose of corticosteroids (CS) over 30 mg/day during hospitalization between April 2009 and January 2018; 3) were 16 years old or over. Cases with central neurological lupus, alveolar hemorrhage, or pregnancy at baseline were excluded. Cases with plasmapheresis or antiplatelet therapies at the start of observation, warfarin within a year, direct oral anticoagulants within a month, major surgery or lower limbs operation within three months, past thrombosis within a year, and prophylactic treatment of thrombosis from the observation starting month were also excluded from the study population. LN cases were divided into 2 groups; receiving SPT (SPT group, n=692) or not receiving SPT (non-SPT group, n=525). The start of observation was defined as commencement of CS treatment during hospitalization. Observation stopped either on April 2018 or the month cases were withdrawn from the database or developed first thrombosis, whichever came first. Thrombosis was defined as follows: at least one of three disease names (thrombosis, embolisms and infarction) and prescription of thrombolytic agents after the start of observation. After propensity-score matching, the incidence rate of thrombosis at Month 1, 2, 3, 4 was calculated using Kaplan–Meier methods. Univariate analysis were conducted by chi-squared test for categorical data and Mann-Whitney U-test for continuous data. Adjusted odds ratio (OR) was calculated using a multivariate logistic regression model.ResultsThe mean age was 47 years old and the proportion of female was 76%. There were no statistically significant differences in baseline variables between the two groups after propensity-score matching (both groups: n=434). The percentage of cases with thrombosis in both groups at each month were similar (SPT vs non-SPT at Month 1, 2, 3, and 4; 3.0% vs 4.4% (p=0.28), 3.5% vs 5.1% (p=0.24), 3.9% vs 5.3% (p=0.331), and 4.6% vs 5.5% (p=0.536), respectively). There were no significant differences in cumulative incidence rates of thrombosis between the two groups (P=0.265 by log-rank test). Univariate analysis revealed five risk factors of thrombosis; activity of daily living (p=0.004), hepatic failure (p<0.0001), malignancy (p=0.02), and use of methotrexate (p=0.038) and oral contraceptive (p=0.037). After adjusting for covariates, OR of SPT was 0.82 [95%Cl 0.44-1.52], which was not significantly ...