Reese H. Clark

Prospective, randomized comparison of high-frequency oscillation and conventional ventilation in candidates for extracorporeal membrane oxygenation

OBJECTIVE To compare the safety and efficacy of high-frequency oscillation (HFO) with conventional ventilation in the treatment of neonates with respiratory failure. DESIGN We conducted a multicenter, prospective, randomized trial. Patients were stratified according to pulmonary diagnosis and then were randomly selected for conventional ventilation or HFO. A balanced crossover design offered patients who met criteria of treatment failure a trial of the alternative mode of ventilation. SETTING Four tertiary, level 3 neonatal intensive care units accepting regional referrals for extracorporeal membrane oxygenation. PATIENTS Neonates were eligible for enrollment if their gestational age was > 34 weeks, their birth weight was > or = 2 kg, they were < 14 days of age, they required fractional inspired oxygen > 0.50 and a mean airway pressure > 0.98 kPa (10 cm H2O) to support adequate oxygenation, and they required a peak inspiratory pressure > 2.9 kPa (30 cm H2O) and a rate > 40 breaths per minute to support adequate ventilation. Exclusion criteria were lethal congenital anomalies, profound shock, need for cardiopulmonary resuscitation, and failure to obtain consent. MAIN RESULTS Of 79 patients studied, 40 were assigned to conventional ventilation and 39 to HFO. Neonates randomly assigned to HFO required higher peak pressure (3.8 +/- 0.5 vs 3.3 +/- 0.8 kPa, 39 +/- 5 vs 34 +/- 8 cm H2O; p = 0.004) and more often met extracorporeal membrane oxygenation criteria (67% vs 40%; p = 0.03) at study entry than did those given conventional ventilation. Twenty-four patients (60%) assigned to conventional ventilation met treatment failure criteria compared with 17 (44%) of those assigned to HFO (not significant). Of the 24 patients in whom conventional ventilation failed, 15 (63%) responded to HFO; 4 (23%) of the 17 in whom HFO failed responded to conventional ventilation (p = 0.03). There were no differences between the two groups with respect to outcome, need for extracorporeal membrane oxygenation, or complications. CONCLUSIONS We conclude that HFO is a safe and effective rescue technique in the treatment of neonates with respiratory failure in whom conventional ventilation fails.

Pulmonary interstitial emphysema treated by high-frequency oscillatory ventilation

Twenty-seven low birth weight infants who developed pulmonary interstitial emphysema (PIE) and respiratory failure while on conventional ventilation were treated with high-frequency oscillatory ventilation (HFOV). The mean birth weight was 1.2 kg (range 0.55 to 2) with gestational age of 28 wk (range 25 to 34). Ten patients died, six of whom had documented sepsis with shock and were therefore excluded from analysis. All patients showed initial improvement on HFOV. Surviving patients showed continued improvement in oxygenation and ventilation at increasingly lower fraction of inspired oxygen and proximal airway pressure with resolution of PIE, while nonsurvivors progressively developed chronic respiratory insufficiency with continued PIE from which recovery was not possible. Overall survival in nonseptic patients was 80% (16 of 20). We found HFOV to be effective in the treatment of PIE and hypothesieze that interstitial airleak is decreased during HFOV because adequate ventilation is provided at lower peak distal airway pressures.

Efficacy of venovenous extracorporeal membrane oxygenation for neonates with respiratory and circulatory compromise

We report a 12-month experience at Egleston Childrens Hospital in Atlanta, Ga., with a protocol under which venovenous extracorporeal membrane oxygenation (ECMO) was used instead of venoarterial ECMO. Fifty-five newborn infants were referred for ECMO, four of whom had disqualifying conditions (all four died). Thirty-one infants were supported without recourse to ECMO, one of whom died. Of the 20 remaining patients, three were placed on a venoarterial ECMO regimen because of our early uncertainty about the efficacy of venovenous ECMO or because of technical constraints. All other patients (n = 17), including three with congenital diaphragmatic hernia, were supported with venovenous perfusion. No patient begun on a venovenous ECMO regimen required conversion to venoarterial bypass. Before ECMO, venovenous patients required an average dopamine dose of 16 micrograms/kg per minute and an average dobutamine dose of 6 micrograms/kg per minute. Of 15 patients studied before ECMO, three had significantly impaired contractility, and all had evidence of pulmonary hypertension on an echocardiogram. Mean blood pressure did not change while heart rate fell from 172 to 146 beats/min during the first 2 hours of ECMO and vasoactive drug doses were reduced. Of the 17 venovenous ECMO patients, 15 (88%) survived. We conclude that neonatal patients with severe hypoxia and substantial circulatory compromise can be effectively supported by venovenous ECMO in most cases.

Effects of venovenous extracorporeal membrane oxygenation on cardiac performance as determined by echocardiographic measurements.

We evaluated the effects of venovenous extracorporeal membrane oxygenation (ECMO) on cardiac performance by echocardiographic measurements in 15 infants. Heart rate and blood pressure were also recorded. Echocardiographic measurements included aortic and pulmonary peak blood flow velocities, pulmonary time to peak velocity, left ventricular shortening fraction, velocity of circumferential fiber shortening corrected for heart rate, and peak systolic wall stress before, during, and after venovenous ECMO. Pre-ECMO echocardiograms showed borderline or normal indexes of cardiac function. After initiation of venovenous ECMO, all infants had normalization and no infant had deterioration of cardiac performance. The inotropic agents dopamine and dobutamine were decreased from average doses of 12 and 3.6 micrograms/kg per minute, respectively, to 3.7 and 1.3 micrograms/kg per minute, respectively, within 8.8 hours of the institution of venovenous ECMO. During this time the mean arterial pressure remained stable, and the heart rate decreased (169 +/- 21 vs 136 +/- 15 beats/min; p < 0.001). During the course of ECMO there were no changes in left ventricular shortening fraction, velocity of circumferential fiber shortening corrected for heart rate, or aortic peak blood flow velocities. Pulmonary artery peak blood flow velocity (69 +/- 22 vs 92 +/- 28 cm/sec; p = 0.04) and pulmonary time to peak velocity improved (47 +/- 11 vs 65 +/- 16 msec; p = 0.026). We conclude that venovenous ECMO does not have deleterious effects on cardiac performance.

Frequency of chronic lung disease in infants with severe respiratory failure treated with high-frequency ventilation and/or extracorporeal membrane oxygenation.

ObjectiveTo assess the frequency of chronic lung disease and factors associated with its development in term infants with severe respiratory failure who receive high-frequency oscillatory ventilation, or high-frequency oscillatory ventilation, and extracorporeal membrane oxygenation (ECMO). DesignRetrospective review of pulmonary outcome of all ECMO candidates admitted to Wilford Hall USAF Medical Center between July 1985 and September 1989. SettingA tertiary, level III, neonatal ICU accepting regional referrals for high-frequency ventilation and ECMO. PatientsNinety-four patients who were candidates for ECMO were studied. High-frequency oscillatory ventilation alone was used in 48 infants. Forty-six infants were treated with high-frequency oscillatory ventilation and ECMO. Main ResultsTwenty (24%) of 84 survivors developed chronic lung disease. There were no differences in gestational age, birth weight, or gender between those infants who developed chronic lung disease and those infants who did not. Arterial blood gas and ventilatory settings at initiation of high-frequency oscillatory ventilation were similar between those infants who did and those who did not develop chronic disease. Patients who developed chronic lung disease more often had lung hypoplasia (40% vs. 5%) and more often required ECMO (75% vs. 39%) than those patients who did not. In patients without lung hypoplasia, those patients who developed chronic lung disease were older at initiation of high-frequency oscillatory ventilation rescue than those patients who did not develop chronic lung disease (median 91 vs. 46 hrs). ConclusionsThe frequency of chronic lung disease in ECMO candidates is clinically important. Factors associated with chronic lung disease in ECMO candidates are: the presence of lung hypoplasia, delayed referral, and the need for ECMO to support gas exchange. (Crit Care Med 1992; 20:372–377)

Principles and Practice of Venovenous Extracorporeal Membrane Oxygenation

Over the past several years, the use of venovenous extracorporeal membrane oxygenation (ECMO) has increased. The primary advantage of venovenous (VV) over venoarterial (VA) ECMO is preservation of the carotid artery. Its primary disadvantage is that it does not provide circulatory support. While VV ECMO is technically similar to VA ECMO, clinical application of VV ECMO is quite different from VA ECMO. Recent clinical data show that VV ECMO is safe and effective. The purpose of this review is to discuss these differences between VV and VA ECMO, to review the various forms of VV ECMO, and finally to offer recommendations on the safe clinical use of VV ECMO.

Outcome of Respiratory Failure: A Case-Control Study

Despite the availability of ECMO (extracorporeal membrane oxygenation) services for nearly a decade, the criteria for the institution of ECMO for pediatric respiratory failure are still not clearly defined. Therefore, a chart review was performed on children who were mechanically ventilated more than 48 hours in 1989-1990 in order to evaluate possible predictors of death from pediatric respiratory failure. Twenty-three children died as a consequence of respiratory failure. Nonsurvivors in both years were compared with the 78 survivors in 1990, and potential predictors were subjected to multivariate analysis. After 4 days of mechanical ventilation, an alveolar-arterial oxygen gradient (AaD02) greater than 400 torr (53.3 kPa) was a weak predictor of death due to respiratory failure, and yet an AaD02 less than 400 torr (53.3 kPa) was a stronger predictor of survivability. Combination of variables did not yield a better predictor than any single variable. Early prediction of mortality from respiratory failure in this population was not found.