Ventilator management of infants before extracorporeal membrane oxygenation.

The aim of this project was to review the course of infants referred for consideration of extracorporeal membrane oxygenation (ECMO) to identify maximal ventilator settings that, when exceeded, did not provide clinical benefit to the patient. These settings might then be used in defining failure of conventional mechanical ventilation. We reviewed referral records and hospital charts of all infants treated for severe respiratory failure due to meconium aspiration syndrome during the 52.5 month period from March 15, 1985, to August 1, 1989. At an inspiratory pressure > 35 cm H2O, 75% (43/57) of patients eventually required ECMO, and 28% (4/14) of the infants who did not receive ECMO died. When the inspiratory pressure was > or = 40 cm H2O, 39/49 patients required ECMO, and 30% (3/10) of those not treated with ECMO died. Once the inspiratory pressure was > 45 cm H2O, 91% (29/32) of patients required ECMO, and only one third of those not treated with ECMO survived. Although the limitations for conventional therapy suggested in this paper may be helpful to clinicians, each center needs to establish guidelines for maximal conventional ventilator support. If these guidelines are clearly defined, alternative methods of therapy can be used once these criteria are achieved.

Unilateral pulmonary agenesis without mediastinal displacement.

A radiographic hallmark in pulmonary agenesis is opacification of the ipsilateral hemithorax with displacement of the mediastinum in the direction of the agenetic lung. We believe this case report is the first in which the mediastinum was not displaced; instead, the liver and intact hemidiaphragm were displaced cephalad, preventing cardiac malposition. Contrast-enhanced computed tomography enabled us to diagnose this anomaly precisely.

Hidden mortality rate associated with extracorporeal membrane oxygenation.

We reviewed the outcome of all infants referred to, and accepted in, our extracorporeal membrane oxygenation (ECMO) program during a 52-month period. One hundred sixty-seven referrals, representing 158 infants and nine mothers who had not yet delivered their infants, were accepted. Eighteen infants (11.3% of all neonates transported) died before leaving the referring hospital, during transport, or shortly after admission to our unit. Contraindications to ECMO excluded 17 (10.1%) of the 167 referrals. Sixty-two infants (37.1%) initially did not meet ECMO criteria. Two died before ECMO could be started. Sixty-eight infants (40.7%) were given ECMO therapy, and 11 died (16.1%). Nine mothers were referred because of fetal conditions that might require ECMO; of these infants, two died during delivery and three had contraindications to the use of ECMO. The four remaining infants were given ECMO therapy; three survived. The overall mortality rate was 27.5% (46/167); 18 (39.1%) of the 46 deaths were associated with transfer. The mortality rate associated with congenital diaphragmatic hernia was 63.6%. We recommend early transport of infants with this type of hernia during the postoperative "honeymoon" or during in utero transport with delivery at an ECMO center. We also recommend that infants with meconium aspiration syndrome be transported to an ECMO center when an oxygenation index of 25 is reached. The mortality rate associated with transport needs to be considered in evaluating ECMO programs. Earlier, expedited transfers may increase the survival rate.

Recent experience with diaphragmatic hernia and ECMO.

In the past 4 years at the Medical College of Georgia, a total of 74 patients underwent extracorporeal membrane oxygenation (ECMO) with 62 (84%) survivors. Forty-seven of these infants had meconium aspiration syndrome and 11 had diaphragmatic hernia. The use of ECMO, when indicated, after reduction and repair of the diaphragmatic hernia, results in normal oxygen delivery, allows time for pulmonary maturation, and increases survival. A total of 27 referrals for diaphragmatic hernia were studied. Six infants had surgical repair and did not require ECMO. Eleven patients, after surgical repair, were treated with ECMO and seven survived. More importantly 10 patients died before the use of ECMO. Six infants died either before or during transport from referring hospitals and four died while in the delivery room or neonatal unit before ECMO. Of these 10 infants, eight were potential candidates for ECMO. Thirteen of the twenty-seven (48%) infants survived. Seven of eleven (64%) infants who received the benefit of ECMO survived. Eight infants who met the criteria for ECMO died before its use. Had ECMO been used in those eight infants, our data suggests that at least four may have survived. The data from this report support the concept that infants undergoing surgical repair of diaphragmatic hernia, when ECMO is not available, should be referred to an ECMO center in the early postoperative period. Furthermore infants with prenatal diagnosis of diaphragmatic hernia should be delivered at a center where surgical as well as ECMO expertise are available.

Incidence of hypertension in infants on extracorporeal membrane oxygenation.

Systemic hypertension has been associated with extracorporeal membrane oxygenation (ECMO) applied in neonatal respiratory failure. To determine the incidence of ECMO-related hypertension, we reviewed blood pressure measurements from indwelling aortic catheters in 31 infants consecutively placed on ECMO. Systemic hypertension (systolic blood pressures greater than 100 mm Hg for 4 or more consecutive hours) developed in 18 of the 31. Causes investigated included the roles of renin secretion, sodium, and colloid loads. There was no evidence of increased plasma renin activities in hypertensive infants (H), when compared with their own pre-ECMO controls or with the nonhypertensive infants (NH). Sodium and colloid loads and their rates of delivery were not different between H and NH. No consistent duration of ECMO was clearly associated with development of hypertension (mean time on ECMO at onset of hypertension, 43.8 +/- 38.5 hours; range, 1 to 142 hours). Demographic information was not statistically significant. Contrary to previous reports, H did not seem predisposed to an increased incidence of intracranial hemorrhage. Development of hypertension during ECMO is not related to increased plasma renin activity, sodium or colloid loads, or their rates of infusion.

Compliance, alveolar-arterial oxygen difference, and oxygenation index changes in patients managed with extracorporeal membrane oxygenation.

Eighteen patients with meconium aspiration syndrome who failed conventional management were treated with extracorporeal membrane oxygenation (ECMO) for reversible respiratory failure. Dynamic lung compliance measurements were made prior to, during, and after ECMO support. P(A-a)O2 and oxygenation index (OI) measurements were calculated prior to and after ECMO support. Lung compliance decreased significantly comparing before-ECMO to during-ECMO, and increased significantly comparing during- to after-ECMO, but not comparing before- to after-ECMO measurements. P(A-a)O2 and OI decreased significantly from before to after ECMO. The improvement in oxygenation allowing removal from ECMO does not appear to be related to improved pulmonary mechanics, but may rather be secondary to increased effective pulmonary capillary blood flow.

Difficulty in establishing diagnosis from lung biopsies and bronchial washing analysis in children with leukemia following bone marrow transplantation.

Three children developed severe respiratory distress at days +12, +11, and +11 following allogeneic bone marrow transplantation from donors. The first child was a 13-year-old Hispanic boy transplanted in relapse of Philadelphia chromosome-positive acute lymphoblastic leukemia (ALL). At day -14, a bronchial washing done for a streaky pulmonary infiltrate was negative for acid-fast bacilli. Miliary tuberculosis was discovered at postmortem examination. A second child, transplanted in remission of null-cell ALL, developed severe hypoxia and hypercarbia on day +11 but recovered fully following prolonged mechanical ventilation. An open-lung biopsy showed a pattern of nonspecific, diffuse alveolar damage compatible with respiratory distress syndrome. The third child was transplanted in remission of B-cell ALL and developed fatal fungal and cytomegalovirus pneumonia on day +12. In these latter two cases, it is likely that open-lung biopsy would have missed the diagnosis because of the uneven pulmonary involvement and multiple etiologies observed. All three children received cyclosporine, granulocyte transfusions, and multiple antimicrobials, including amphotericin B. Hyperfractioned total-body irradiation with lung shielding was used in the latter two patients.