Confronting the frustrations of negative clinical trials in acute respiratory distress syndrome.

Despite robust successes in trials of mechanical ventilation, pharmacologic interventions in acute respiratory distress syndrome have been disappointing. Although ineffective therapy remains the compelling explanation for these negative trials, other possible explanations exist. These negative trials, better termed "statistically negative trials" or "indeterminate trials," cannot prove that a therapy is ineffective. It is important for clinicians and investigators to appreciate the alternative explanations for negative trials of potentially effective therapies because these indicate options for improving clinical trials in acute respiratory distress syndrome. These options can be organized into strategies that increase sample size, increase the signal from the therapy, and reduce the noise or variation in the study. Each of the strategies to improve the likelihood of a positive clinical trial poses a potential tradeoff in generalizability, cost, sample size, signal, or noise.