Oral Corticosteroid Exposure and Increased Risk of Related Complications in Patients with Noninfectious Intermediate, Posterior, or Panuveitis: Real-World Data Analysis

Purpose

This study causally examined the dose-response relationship between oral corticosteroids (OCS) exposure and long-term complications among noninfectious uveitis adult patients in the United States.

Methods

The study design was longitudinal, retrospective cohort using Truven Health MarketScan claims database years 2000-2015. The index date was the first day after diagnosis on which OCS≥ 5 mg prednisone equivalent was administered. The period following the index date was parsed into quarters for tracking OCS-related complications; follow-up time was censored when patients switched off of OCS monotherapy. Each quarter of follow-up was divided into 4 groups based on the mean cumulative daily OCS dose (< 7.5 mg; 7.5 to < 30 mg; 30 to < 60 mg; and ≥ 60 mg) and covariate balancing propensity scoring was used to balance groups on baseline characteristics in the first quarter post-index. Marginal structural models (MSMs) were employed to account for time-varying endogeneity between temporal changes in mean cumulative OCS dose and the risk of complications. Patients with systemic autoimmune conditions at baseline were excluded.

Results

The study sample included 3966 patients with a median follow-up of 2 years. Compared to those receiving < 7.5 mg, patients with higher mean cumulative OCS dose had 10%, 16%, and 28% higher risk, respectively, of any OCS-related complication in any given quarter.

Conclusions

A moderate dose-response relationship was found between the long-term use of OCS monotherapy and the risk of developing complications in noninfectious intermediate, posterior, or panuveitis patients. Future research should examine optimal approaches to achieve inflammation control while minimizing OCS exposure.

AuthorsVV Chirikov, R Shah, Y Kwon, D Patel
JournalOphthalmic Epidemiology
Therapeutic AreaInfectious Diseases
Service AreaReal-World Evidence
Year2019
LinkClick Here