| Related Articles |
Systemic corticosteroid for COPD exacerbations, whether the higher dose is better? A meta-analysis of randomized controlled trials.
Clin Respir J. 2013 Oct;7(4):305-18
Authors: Cheng T, Gong Y, Guo Y, Cheng Q, Zhou M, Shi G, Wan H
Abstract
BACKGROUND: Systemic corticosteroids (SCS) have been shown to improve the outcome of acute exacerbation of chronic obstructive pulmonary disease (AECOPD). However, the optimal dose remains controversial.
OBJECTIVES: We performed a meta-analysis to evaluate whether high-dose SCS is better.
METHODS: We searched PubMed, EMBASE, CPCI-S and CENTRAL databases, and references of reviews or meta-analyses to identify randomized controlled trials using SCS in AECOPD. We performed a routine meta-analysis to evaluate the effects of SCS on treatment failure rate and forced expiratory volume in 1 s (FEV1) improvement compared with placebo in AECOPD. Subgroup analysis was performed by dividing the studies into a high-dose group [initial dose ≥80 mg prednisone equivalent (PE)/day] and a low-dose group (initial dose 30-80 mg PE/day) in all patients and in only inpatients. Meta-regression was performed using initial dose as an independent factor. We classified the suspected adverse effects into several groups and combined them separately.
RESULTS: Our search yielded 12 studies involving 1172 patients. SCS use was associated with a significant reduction in the treatment failure rate [risk ratio 0.58; 95% confidence interval (CI): 0.46-0.73] and improvement in ▵FEV1 (0.11 L; 95% CI: 0.08-0.14 L). The high-dose regimen did not show superiority to the low-dose regimen. No obvious correlation was found between the SCS effect and the initial dose. SCS led to an obvious increase in hyperglycemia risk. However, the high-dose group did not show obviously higher risk of adverse effects.
CONCLUSION: SCS can reduce treatment failure rate and improve lung function in AECOPD. The low-dose regimen (initial dose 30-80 mg/day PE) is proper for treating AECOPD.
PMID: 23072733 [PubMed - indexed for MEDLINE]