To the Editor.
—In the meta-analysis of Ms Mullen and colleagues concerning the association between β-agonist use and death from asthma, the association that was present for β-agonists administered by nebulizer was referred to as "weak" and "of very small magnitude."1 Our evaluation of the evidence they presented does not concur with this assessment.The authors presented two measures of association, the correlation coefficient and the odds ratio (OR). The correlation coefficient was used to infer that the statistically significant association observed for β-agonists administered via nebulizer was of little substantive import. Correlations suffer from limitations as a measure of association and are best avoided when comparing the influence of an exposure in different populations.2 Employing the OR as the measure of association, users of β-agonists administered via nebulizer were more than twice as likely to die from asthma (pooled OR, 2.47; 95% confidence interval [CI], 1.73 to
Alberg AJ, Comstock GW. β-Agonist Use and Death from Asthma. JAMA. 1994;271(11):821. doi:10.1001/jama.1994.03510350029018