R. Vialet et al., HIGH-FREQUENCY OSCILLATORY VENTILATION VE RSUS CONVENTIONAL POSITIVE-PRESSURE VENTILATION IN NEONATES, Annales de pediatrie, 43(2), 1996, pp. 99-107
High-frequency oscillatory ventilation (HFOV) ensures gas exchange wit
h tidal volumes that are smaller than the anatomic dead space. This ca
n be expected to reduce the risk of barotrauma and volotrauma. However
, proof that HFOV decreases neonatal mortality and morbidity has not b
een obtained. This study is a retrospective evaluation of the first ye
ar of use of HFOV in a neonatal intensive care unit. Infants treated w
ith HFOV were compared with those who received conventional positive-p
ressure ventilation (PPV) during the same period. HFOV was used with h
igh pressures (mean pressure 2-5 cm H2O above those used for PPV) to m
aintain alveolar recruitment. All neonates who were ventilated within
24 hours of birth were included (n = 92, including 38% treated by HFOV
) except those who were ventilated after a surgical procedure. Patient
s were assigned to the HFOV group based on availability of the oscilla
tor (OHFI(R), Dufour). Outcomes in the subgroup of premature babies bo
rn before 32 weeks of gestational age were studied separately. Evaluat
ion criteria were mortality, ventilation-indued complications (pneumot
horax, bronchopulmonary dysplasia, and intracerebral hemorrhage), and
ventilation parameters during the first 48 hours. Severity at baseline
was comparable in the two groups. There was no difference in mortalit
y (22%). HFOV was associated with nonsignificant improvements in time
under ventilation (66 vs 96 hours) and in rates of occurrence of intra
cerebral hemorrhage (7% vs 15%) and bronchopulmonary dysplasia (4% vs
11%). Fractional concentration of inspired oxygen (FiO(2)) decreased f
aster in the HFOV group (p < 0.001). Results were comparable in the pr
emature babies born before 32 weeks and in the overall study populatio
n. The findings from this study are in keeping with the recent literat
ure. They should be interpreted with caution given the small sample si
ze and absence of randomization. Because the mechanisms involved in HF
OV have not yet been fully elucidated, uncertainties remain as to opti
mal use of this technique. A large body of experimental mental and cli
nical data suggest that HFOV should be used with large pulmonary volum
es. Early use of HFOV may be warranted since lesions due to barotrauma
can develop very rapidly under PPV.