No study is perfect, and neither was this one. For ethical reasons, a placebo control was not used. However, this seems a minor shortcoming considering that data from a variety of experimental trials of anti-inflammatory agents for treatment of chronic steroid-dependent asthma find only a 20 to 40% improvement from simply participating in a clinical trial.2
Furthermore, their group with a normal pH test finding really serves as an adequate control with a spontaneous rate of improvement of only 20%. The authors criticized themselves for not using a crossover design, but I believe this study technique is inappropriate in these trials when a placebo is used because order and carryover effects regularly bias the data, making them more difficult to logically analyze. Rather, placebo control studies for asthma and GERD should employ a parallel design with separate groups of treated and control patients. As with all medical studies except one,6
the control of GERD with lansoprazole was not assessed by serial pH testing. However, this is unrealistic in the clinical setting and is very difficult to accomplish in large patient trials. Furthermore, who is to argue with the adequacy of PPI therapy when 100% of treated patients responded well? It is only a factor in the treated patients who did not respond, but this was unlikely to be much of a factor as these patients had normal baseline pH tests. Finally, we do not know if these data only apply to patients with symptomatic GERD or is also useful for the many asthmatics with “silent” GERD. Symptom data about GERD were not given in the patient summaries and the criteria for referral to a gastroenterologist appears to only be the difficult-to-manage quality of their asthma. Addressing this issue in future studies is important as 40 to 60% of asthmatic patients have silent reflux, while a recent study9
found their 24-h pH profiles similar to those of asthmatic patients with symptomatic GERD.