To assess the impact of diagnostic procedures in identifying residual lesions during extracorporeal membrane oxygenation (ECMO) on survival after pediatric cardiac surgery.

Between January 2012 and December 2017, 74 patients required postcardiotomy ECMO. Patients were retrospectively divided into 2 groups: Group I underwent only echocardiography ([echo only] 46 patients, 62.2%) and group II (echo+) underwent additional diagnostic tests (ie, computed tomography [CT] or cardiac catheterization; 28 patients, 37.8%). Propensity score matching was used to balance the 2 groups by baseline characteristics.

Two equal groups (28 patients in each group) were formed by propensity score matching. Fourteen (50%) patients in the echo-only group and 20 (71%) patients in the echo+ group were successfully weaned from ECMO (P = .17). Four (14.3%) patients survived in the echo-only group and 15 (53.5%) patients survived in the echo+ group (P = .004). Patients in the echo+ group had a lower chance of dying compared to the echo-only group (odds ratio, 0.14.6; 95% CI, 0.039-0.52; P = .003). The residual lesions, which may have served as a mortality factor, were found by autopsy in 8 (40%) patients in the echo-only group, while none were found in the echo+ group (P = .014).

The autopsies of patients who died despite postcardiotomy ECMO support showed that in 40% of cases that had been investigated by echo only, residual lesions that had not been detected by echocardiography were present. The cardiac catheterization and CT during ECMO are effective and safe for identifying residual lesions. Early detection and repair of residual lesions may increase the survival rate of pediatric cardiac patients on ECMO.