Our section holds a monthly evidence-based journal club in which we discuss articles that may alter how we care for our patients. Leukotriene antagonists offer a new method for asthma treatment [1]. We critically review articles by asking three questions: Are the results of the study valid? What are the results? and Will the results help us in caring for our patients [2]?

In a clinical trial, randomization is the crucial method used to determine validity. Despite the title of Suissa and colleagues' article [1], we are concerned that the study was not a randomized, double-blind, placebo-controlled trial. Suissa and colleagues' study describes 150 patients selected from a randomized trial of 762 patients [3]. The randomization technique from the original study is unknown because it has not been published and the report is not a public document (Ziemba L. Personal communication, Zeneca Pharmaceuticals). Patient selection for Suissa and coworkers' study included two nonrandom steps after the initial randomization: elective participation by centers and then by patients. Initial randomization does not assure that randomization was maintained in the two sequential selection steps of the subprotocol. In fact, participants were older than nonparticipants at the first selection step. Of only five baseline characteristics reported, the FEV₁ was significantly worse in the zafirlukast group at the second selection step. The adjustment in the analyses for these baseline differences may not correct for nonrandom assignments.

Will the results help us care for our patients [2]? We do not think so. The authors recognize that their participants needed additional treatment. Patients with persistent mild-to-moderate asthma, as described in the study, require daily use of a controller medication. In our practice, we attempt to adhere to national guidelines that suggest the use of inhaled steroids, cromoglycate, or theophylline in this setting [4]. Suissa and colleagues compared zafirlukast with inadequate care (as-needed inhaled ß-agonists). Until zafirlukast is compared with other controller medications, it has limited utility in the treatment of asthma.