Prolonged oral corticosteroid treatment in patients with systemic lupus erythematosus: An evaluation of 12-month economic and clinical burden

BACKGROUND: Prolonged, high-dose corticosteroid treatment for systemic lupus erythematosus (SLE) is associated with substantial health care costs, health care resource utilization (HCRU), and adverse events (AEs). OBJECTIVE: To compare all-cause health care costs, HCRU, and oral corticosteroid (OCS)–related AEs among patients with prevalent OCS use and patients without OCS use. METHODS: This retrospective, longitudinal cohort study (GSK study 214100) used claims data from the IQVIA Real-World Data Adjudicated Claims – US, IQVIA, Inc, database between January 1, 2006, and July 31, 2019, to identify patients with SLE. Patients with at least 1 OCS pharmacy claim during the study period and continuous OCS use during the 6-month pre-index (baseline) period (index date is the date of the first OCS claim following 6 months’ continuous use) formed the “prevalent OCS use cohort.” This cohort was subdivided based on the level of OCS exposure during the 12-month observation period, ie, the number of 6-month periods of greater than 5 mg/day OCS use (0, 1, or 2). Patients without OCS claims formed the “no OCS use cohort.” All patients had continuous enrollment during the baseline and observation periods, had at least 1 inpatient or at least 2 outpatient SLE diagnosis codes during baseline, and were aged at least 5 years at index. A 2-part model, a generalized linear regression model with a negative binomial distribution, and a multivariate logistic regression model were used to compare health care costs, HCRU, and the odds of developing an OCS-related AE between cohorts, respectively. RESULTS: The no OCS use and prevalent OCS use cohorts included 21,517 and 16,209 patients, respectively. Adjusted health care cost differences (95% CI) were significantly lower for the no OCS use cohort vs all prevalent OCS use exposure categories ($5,439 [$4,537-$6,371] vs $17,856 [$16,368-$19,498]), driven by inpatient stays and outpatient visits; HCRU was also significantly lower (adjusted incidence rate ratios vs no OCS use cohort [95% CI]: 1.20 [1.16-1.23] vs 1.47 [1.41-1.52]). Health care costs and HCRU increased with increasing length of OCS exposure. OCS-related AEs occurred more frequently for all prevalent OCS use exposure categories vs the no OCS use cohort (odds ratio [95% CI]: 1.39 [1.25-1.55] vs 2.32 [2.02-2.68]), driven by hematologic/oncologic and immune system–related AEs. The mean (SD) average daily dose of OCS increased with increasing periods of prevalent OCS use (2.5 [1.3], 6.9 [31.1], and 34.6 [1,717.3] mg/day, respectively, for patients with 0, 1, and 2 periods of OCS use). CONCLUSIONS: Prevalent OCS use incurs a substantial clinical and economic burden, highlighting the need for restricted OCS doses and durations.