Cardiac stun in infants undergoing extracorporeal membrane oxygenation

Abstract

Previous studies have shown that cardiac performance decreases in infants undergoing extracorporeal membrane oxygenation (ECMO). Some infants have an exaggerated decrease in cardiac performance during ECMO. This syndrome has been called cardiac stun. To better understand this phenomenon, we reviewed the records of infants with cardiac stun and compared them with infants who did not have the syndrome. Cardiac stun was detected in 12 of 240 infants (5.0%) undergoing ECMO. The diagnoses were congenital diaphragmatic hernia (7/12), meconium aspiration syndrome (3/12), respiratory distress syndrome (1/12), and persistent pulmonary hypertension of the newborn (1/12). The weight, gestational age, inotropic support, and time to start of ECMO were similar to infants without cardiac stun. Arterial oxygen tension was lower, carbon dioxide tension was higher, and pH was lower before ECMO in infants in whom cardiac stun developed (p less than or equal to 0.03). Cardiac arrests were more common, before ECMO, in infants in whom cardiac stun developed (6/12; p less than or equal to 0.01). Cardiac stun began at an average 2 1/2 hours after beginning ECMO (range 0.1 to 7 hours). Pulse pressure decreased from 20 mm Hg (range 10 to 45 mm Hg) before stun to 8 mm Hg (range 4 to 12 mm Hg) after stun. Heart rate did not change. Cardiac stun lasted for 33 hours (range 1 to 64 hours) on ECMO and recurred in three infants. Decreases in pump flow and increases in preload, afterload reduction, and inotropic agents did not improve cardiac performance. Survival was lower in the infants in whom cardiac stun developed (p less than or equal to 0.001). Only 5 of 12 infants (42%) survived ECMO when cardiac stun occurred. Our findings show that cardiac stun occurs infrequently during ECMO and is transient in most infants. Infants in whom cardiac stun develops appear to be more ill before ECMO and have a higher mortality after ECMO.