Quantitative characterization of the relationship between levels of extended corticosteroid use and related adverse events in a US population.

OBJECTIVE: This retrospective study assessed the incidence and timing of adverse events (AEs) among patients prescribed varying dose levels of corticosteroids in the US.
METHODS: Patients with selected autoimmune or inflammatory disease diagnoses between 2006 and 2015 were identified from a privately insured administrative database. Patients were stratified into treatment cohorts based on dosage and length of corticosteroid use: intermittent use with duration <60 days, and three extended use cohorts with duration ≥60 days at low (≤7.5 mg/day), medium (>7.5-≤15 mg/day) or high (>15 mg/day) prednisone-equivalent dosage. The incidence of and time to corticosteroid-related AEs were assessed by cohort.
RESULTS: A total of 78,704 patients met the selection criteria, of whom 9.5%, 11.0% and 8.6% were classified into the high-, medium-, and low-dose extended corticosteroid use cohorts, respectively. Corticosteroid exposure varied across study conditions, from 34% of dermatomyositis/polymyositis to 6% of psoriatic arthritis patients prescribed extended high dose. Hypertension, pneumonia and osteoporosis were the AEs with the highest incidence rates (41.9, 27.4 and 19.8 cases respectively per 1000 patient-months for the high-dose cohort). For most AEs, all levels of extended corticosteroid use exhibited significant risks of increased incidence compared to intermittent use. Some AEs had dose relationships, with higher dose correlated with higher incidence; other AEs had duration relationships with longer duration correlated with higher incidence regardless of dose. Average time to AE onset was relatively short, occurring at 2.3-6.7 months after corticosteroid initiation.
CONCLUSIONS: Through a rigorous quantitative characterization, extended steroid exposure was associated with increased incidence and earlier onset of AEs among privately insured adults in the US.