Neurodevelopmental outcome in neonates after extracorporeal membrane oxygenation: cranial magnetic resonance imaging and ultrasonography correlation.

Irreversible ligation of the right common carotid artery and right internal jugular vein is usual in venoarterial extracorporeal membrane oxygenation (ECMO) for treatment of severe respiratory failure in neonates. Vessel ligation with ECMO may magnify risks of cerebral hemorrhage or infarction (CHI) and adversely affect neurodevelopmental outcome. To correlate CHI after ECMO with neurodevelopmental outcome, we reviewed cranial ultrasonography (US) and magnetic resonance imaging (MRI) scans in 22 consecutive neonatal ECMO survivors and compared these with results of Bayley Scales of Infant Development obtained at 3, 6, 12, and 24 months of follow-up. All patients had US, and 19 had MRI. No US or MRI had focal abnormal findings attributable to ECMO; specifically, there was no evidence of CHI. Two infants had generalized cerebral atrophy, and one of these had an abnormal Bayley examination. One infant with a normal MRI had a single right focal seizure 4 days after ECMO. Of 20 infants with Bayley developmental tests at 3 to 30 months of age (mental index range, 72 to 135; motor index range, 71 to 150), only 3 were abnormal. In our experience, the incidence of CHI secondary to ECMO is less than that reported. After ECMO, the absence of intracranial hemorrhage, cerebral infarct, or cerebral atrophy on US or MRI usually correlates with normal short-term neurodevelopmental outcome.

Extracorporeal membrane oxygenation for gram-negative septic shock in the immature pig.

Extracorporeal membrane oxygenation (ECMO) can provide total cardiopulmonary support via extrathoracic vascular cannulation. We evaluated the effects of ECMO in gram-negative septic shock in immature piglets subjected to fecal-Escherichia coli peritonitis. Group I, SEPSIS CONTROL (n = 10), had an intraperitoneal injection of E. coli but did not receive ECMO. Group II, ECMO CONTROL (n = 5), had 24 hr of ECMO support without the intraperitoneal injection of E. coli, and Group III, ECMO SEPSIS (n = 4), had an intraperitoneal injection of E. coli and was treated with ECMO. In the SEPSIS CONTROL group, mean survival time (+/- SD) was 13.8 +/- 6.3 hr with two survivors to 24 hr. In the ECMO CONTROL group, all five animals survived 24 hr on ECMO and then weaned successfully to conventional ventilatory therapy with return of spontaneous circulation. In the ECMO SEPSIS group, one of the animals died after 22 hr on ECMO and the other three animals died at 24 hr immediately upon withdrawal of ECMO support. Conventional ventilatory therapy was unsuccessful in each. Systemic arterial blood pressure was significantly higher in ECMO CONTROL animals (P less than 0.01) as were leukocyte counts (P less than 0.01). In this model of gram-negative septic shock, ECMO did provide cardiopulmonary support but did not improve mortality.

Decreasing blood donor exposure in neonates on extracorporeal membrane oxygenation.

Extracorporeal membrane oxygenation (ECMO) has been successful treatment (80% survival) in over 2,000 neonates with severe respiratory failure (80% predicted mortality without ECMO). Neonates on ECMO require frequent blood product replacement, which increases donor exposure (DE) and the risk of transfusion related complications. Successful, widespread usage of ECMO in neonatal respiratory failure is placing increased numbers of surviving infants at risk for acute and long-term transfusion related problems. We assessed DE rates in 21 consecutive neonatal ECMO survivors. In the first 12 patients packed red blood cell (PRBC) transfusions were administered as 10 mL/kg body weight for hematocrit less than 45%. PRBC exchange transfusions were used in patients with hematocrit less than 45% and hypervolemia. Fresh frozen plasma (FFP) and cryoprecipitate (CRYO) infusions were used empirically for evidence of hemorrhage. DE rates (donors per ECMO day, mean +/- SD) were: PRBC (2.8 +/- 0.6), FFP/CRYO (0.5 +/- 0.7), and platelet (2.0 +/- 1.0), with a total donor exposure rate of 5.3 +/- 2.0 donors per ECMO day. Mean duration of ECMO was 4.6 +/- 2.0 days and total DE per infant was 22.8 +/- 9.5 donors per ECMO run. In a protocol (n = 9) to minimize DE risks, exchange transfusions were eliminated and PRBC transfusion volumes were increased to 15 mL/kg. Empiric use of FFP and CRYO was discontinued. The blood bank divided standard units of PRBCs into four aliquots and dispensed each aliquot sequentially before dispensing blood from another unit.(ABSTRACT TRUNCATED AT 250 WORDS)

Devices and monitoring during neonatal ECMO: survey results.

A survey of active ECMO centres regarding neonatal ECMO equipment and personnel was obtained by telephone interview in late summer 1989. Forty-seven of the centres in the USA listed in the Ann Arbor ELSO (Extracorporeal Life Support Organization) Registry at the time ( greater than 90%) were contacted and all participated. Nearly all use a roller pump, while less than 5% use a centrifugal pump. All programmes use a SciMed membrane oxygenator and 90% a SciMed heat exchanger. Heat exchanger water sources include the Gaymar T-pump (42%), Seabrook (25%) and Cincinnati Sub-Zero (23%) units. Eighty-seven per cent use a bladder box servo-regulated to the roller pump; these are most often custom-made (69%) but 13% of programmes use a commercially available (Seabrook) bladder box. Ten per cent use a pressure-regulated roller pump rather than a conventional (displacement) bladder box to detect decreases in venous return. Nearly 80% monitor circuit line pressures between the pump and patient. Seventeen per cent use an air bubble detector on the arterial side of the circuit. Only 10% use an arterial bubble trap and 6% an arterial line filter. Seventy-five per cent do not monitor gas line pressures into the membrane lung, but one-third do use a gas line pop-off valve to prevent elevated gas phase pressures. Seventy per cent reported use of continuous in-line measurement of mixed venous oxygen saturation; no programme reported any blood chemistries being monitored in line.(ABSTRACT TRUNCATED AT 250 WORDS)