Real-world association between systemic corticosteroid exposure and complications in US patients with severe asthma

Abstract

Background: Systemic corticosteroid (SCS) use remains widespread among patients with severe asthma, despite associated complications.

Objective: Evaluate the association between cumulative SCS exposure and SCS-related complications in severe asthma.

Methods: This retrospective, longitudinal study used claims data from the Optum Clinformatics Data Mart database (GSK ID: 214469). Eligible patients (≥ 12 years old) had an asthma diagnosis and were divided into two cohorts: SCS use and non/burst-SCS use. Patients in the SCS use cohort had a claim for a daily prednisone-equivalent dose ≥ 5 mg SCS following ≥ 6 months of continuous SCS use; those in the non/burst-SCS cohort had no evidence of continuous SCS use and had a non-SCS controller/rescue medication initiation claim. For each cohort, the date of the qualifying claim was the index date. SCS users were further stratified by SCS use during each quarter of follow-up: low (≤ 6 mg/day), medium (> 6-12 mg/day), high (> 12 mg/day), and continuous high (≥ 20 mg/day for 90 days). SCS-related complications were evaluated in the quarter following SCS exposure. The adjusted odds ratios (OR) of experiencing SCS-related complications during follow-up in each of the SCS use groups versus the non/burst SCS cohort were calculated using generalized estimating equations models.

Results: SCS and non/burst-SCS use cohorts included 7473 and 89,281 patients (mean follow-up: 24.6 and 24.2 months), respectively. Compared with the non/burst-SCS use cohort, medium, high, and continuous high SCS use was associated with greater odds of any SCS-related complication (adjusted OR [95% confidence interval]: 1.30 [1.21, 1.39], 1.49 [1.35, 1.64] and 1.63 [1.40, 1.89], respectively) including increased acute gastrointestinal, cardiovascular, and immune system-related complications, and chronic cardiovascular, metabolic/endocrine, central nervous system, bone-/muscle-related, ophthalmologic, and hematologic/oncologic complications. Low-dose SCS use was also associated with significantly increased odds of acute gastrointestinal and immune system-related complications, and chronic bone-/muscle-related and hematologic/oncologic complications versus the non/burst-SCS use cohort.

Conclusion: SCS use, even at low doses, is associated with increased risk of SCS-related complications among patients with severe asthma.

Keywords: Asthma; Cardiovascular; Central nervous system; Endocrine; Gastrointestinal; Metabolic; Ophthalmologic; Systemic corticosteroid; Systemic corticosteroid-related complication.

Fig. 1

Study design. *First pharmacy/medical claim date for SCS ≥ 5 mg/day following 6 months of continuous (≥ 5 mg/day for 6 months without any gap > 14 days between claims) SCS use (SCS use cohort) or date of initiation of any non-SCS asthma controller/rescue medication (non/burst-SCS use cohort). SCS, systemic corticosteroid

Fig. 2

Odds of any SCS-related complications during the follow-up period due to SCS exposure. Prednisone-equivalent SCS dose; *p-value < 0.05; ^†from multivariate GEE models adjusting for patient demographics and baseline clinical characteristics (see Methods for covariates). CI, confidence interval; GEE, generalized estimating equations; OR, odds ratio; SCS, systemic corticosteroid

Fig. 3

Odds of acute SCS-related complications during the follow-up period due to SCS exposure. Prednisone-equivalent SCS dose; *p-value < 0.05; ^†from multivariate GEE models adjusting for patient demographics/baseline clinical characteristics (see Methods for covariates). CI, confidence interval; GEE, generalized estimating equations; OR, odds ratio; SCS, systemic corticosteroid; SD standard deviation

Fig. 4

Odds of chronic SCS-related complications during the follow-up period due to SCS exposure. Prednisone-equivalent SCS dose; *p-value < 0.05; ^†from multivariate GEE models adjusting for patient demographics and baseline clinical characteristics (see Methods for covariates). CI, confidence interval; GEE, generalized estimating equations; OR, odds ratio; SCS, systemic corticosteroid

Fig. 5

Rate of HCRU due to SCS-related complications during the follow-up period. Prednisone-equivalent SCS dose; *p-value < 0.05; ^†from multivariate GEE models adjusting for patient demographics/baseline clinical characteristics (see Methods for covariates). CI, confidence interval; GEE, generalized estimating equations; HCRU, healthcare resource utilization; RR, rate ratio; SCS, systemic corticosteroid