Abstract
Acute respiratory distress syndrome (ARDS) is a common and highly heterogeneous condition in the critically ill. The association between hyper- and hypo-inflammatory subphenotypes and clinical outcomes has generated significant interest in precise ARDS management. The value of identifying biomarkers to guide treatment and enrollment in future ARDS trials is undisputable. We describe multiple factors complicating the search for subphenotypes and their treatable traits. The observed heterogeneity seen in the clinical course of ARDS is dynamic and influenced by factors beyond lung pathophysiology, including variations in the delivery of best critical care practices, patient comorbidities, and functional status, and patient or family preferences. Current subphenotype definitions lack strong biological plausibility and without clear evidence of benefit from targeted treatments, their use in clinical practice is currently unwarranted.