THE ASSOCIATION BETWEEN BETA-AGONIST USE AND DEATH FROM ASTHMA - A METAANALYTIC INTEGRATION OF CASE-CONTROL STUDIES

Objectives.-The purpose of this investigation was to provide an empirical summary of the evidence regarding the association between beta-agonist use and death from asthma. This effort integrated the results of case-control studies that examined the use of beta-agonists among asthmatic patients who died and the use of beta-agonists among asthmatic patients who did not die. The possible moderating effects of patient sample age and mode of delivery (oral, metered-dose inhaler, and nebulizer) were also examined. Data Sources.-An on-line computer search (using MEDLINE) was conducted using the key words beta-agonist and asthma. This search was supplemented by ancestry and descendency approach searches. Studies that were available as of April 1992 were eligible for inclusion in this integration. Study Selection.-Studies were included if they reported the precise numbers of cases and controls who did and did not use a beta-agonist. A total of six case-control studies comprising 15 separate tests of the relation between beta-agonist use and death from asthma and data for 364 cases and 1388 controls were included. Data Extraction.-The 2 (case vs control) X 2 (did vs did not use beta-agonist) designs allowed for direct derivation of a chi2 statistic that tested the association between beta-agonist use and death from asthma. Mode of delivery and average age of sample were also coded. Data Synthesis.-Statistical integration revealed a significant, although extremely weak, relation between beta-agonist use and death from asthma (z=3.996; P=.000075; mean r=.055). This relation emerged only when beta-agonists were administered with a nebulizer (z=4.481; P=.0000038; mean r=.103). There was no association between beta-agonist use and death when beta-agonists were administered by metered-dose inhaler (z=1.194; P=.11; mean r=.031) or orally (z=1.247; P=.1; mean r=.031). Adults were more likely than adolescents to evidence the association between beta-agonist use and death. Conclusions.-These results document the extremely small magnitude of the relation between beta-agonist use and death from asthma. Furthermore, these results specify that the weak relation between beta-agonist use and death from asthma may really be restricted to the delivery of beta-agonists with a nebulizer. These findings suggest that the headlines that followed the report by Spitzer et al (1992) were misleading.