CHEST Guidelines-Response

Response_chest (2)

Pdf Summary

A letter to the editor by Drs. Anurag Khera and Taramangalam S. Ramakrishnan expressed skepticism about a June 2014 article in CHEST concerning a meta-analysis. This study analyzed the risk of tuberculosis (TB) and influenza in patients with Chronic Obstructive Pulmonary Disease (COPD) using inhaled corticosteroids, indicating a significant Peto Odds Ratio (OR) of 2.29. The letter raised concerns that none of the primary outcomes from the focal studies considered TB and influenza risks, suggesting the sample size might be inadequate for these outcomes. They also questioned the suitability of the Peto OR method, which typically requires rare events, an OR close to 1, and similar group sizes; the authors only met the first criterion. The letter highlighted the dominance of a single study (Calverley et al.) in influencing the OR for TB risk.<br /><br />In response, Dr. Yaa-Hui Dong and colleagues addressed these points, stating their study's goal was to cover the literature comprehensively on the risk of TB and influenza from inhaled corticosteroids in COPD patients. They acknowledged the original trials weren't designed to pre-identify these risks, which might lead to underreporting or misclassification but argued that close patient monitoring provided reliable data nevertheless. The authors justified using the Peto OR, noting its advantage with rare events. To overcome sample size imbalances, they employed multiple statistical approaches (including Mantel-Haenszel and Bayesian methods), which yielded consistent results, thereby reinforcing their findings. They emphasized the importance of using detailed protocols and appropriate statistical tools for valid drug safety assessments.<br /><br />The exchanges emphasized the need for careful methodology in drug safety evaluations using meta-analysis, particularly the scrutinizing roles of statistical methods and study design in interpreting data on rare adverse events.

Keywords

meta-analysis

tuberculosis

influenza

Chronic Obstructive Pulmonary Disease

inhaled corticosteroids

Peto Odds Ratio

drug safety

statistical methods

COPD

rare adverse events