Nosocomial infections are one of the most frequent complications of extracorporeal membrane oxygenation (ECMO) and are associated with increased mortality (¹). Bloodstream infections present a particular problem because of the risk of biofilm formation and the clinical challenges associated with cannula and circuit exchanges in this patient population (²). With their significant impact on patient outcomes, nosocomial infections on ECMO must be promptly recognized to ensure adequate treatment with antimicrobials and effective source control. Diagnosing these infections, however, can be difficult. Clinical and laboratory markers typically used to diagnose infections, such as WBC count, temperature dysregulation, or hemodynamic instability, are altered on ECMO due to the properties of both the ECMO circuit and the patient’s underlying immune response to foreign material and critical illness (³). Furthermore, there are no widely accepted standardized definitions of infections during ECMO, and guidelines to diagnose and treat these infections do not yet exist (⁴). Given the grave consequences of missing a bloodstream infection in an ECMO patient, surveillance cultures are frequently collected to help diagnose these infections as early as possible (^5,6).

In this issue of Pediatric Critical Care Medicine, Schmoke et al (⁷) evaluated the use of daily surveillance blood cultures in neonatal and pediatric ECMO patients. In 111 patients, they reviewed 1059 surveillance blood cultures, which yielded only a 3% positivity rate. Similar to previous studies (^1,8–10), they found that surveillance cultures to diagnose bloodstream infections in ECMO patients are low-yield and this ultimately led to their center stopping the practice.