#15 High-frequency vs. conventional ventilation at the time of CDH repair is associated with, but does not contribute to higher mortality and BPD rates - A case-control study
Gabrielle Derraugh, CHRIM; Matt Levesque, CHRIM; Suyin Lum Min, CHRIM; Daryl Shantz, University of Manitoba; Melanie Morris, CHRIM; John Minski, University of Manitoba; John Baier, University of Manitoba; Molly Seshia, CHRIM; Richard Keijzer, CHRIM
The multicentre VICI trial reported that conventional mechanical ventilation (CMV) and high-frequency oscillatory ventilation have similar effects on mortality and bronchopulmonary dysplasia (BPD) in patients with congenital diaphragmatic hernia (CDH). At our centre, we use CMV and high-frequency ventilation (HFV), both jet and oscillatory. The objectives of this study were to compare the outcomes of death or BPD according to the mode of ventilation at the time of surgery, and to compare the characteristics of patients on CMV or HFV.
We performed a retrospective case-control study of CDH patients (n=55) at our centre between 1990 and 2015. Analysis was conducted using Fischer’s exact, Mann-Whitney U and t-tests.
At the time of surgery, 30 (54.55%) patients were on CMV and 25 (45.45%) were on HFV. Four patients (13.33%) on CMV had BPD or died compared to 15 patients (60%) on HFV. The odds ratio was 9.38 (95% CI 2.49-35.25, p<0.01). Patients on CMV had fewer instances of severe pulmonary hypertension (33.33% vs 84%, p<0.01), required less sildenafil (3.33% vs 36%, p<0.05), vasoactive medications (66.67% vs 96%, p<0.05) and inhaled nitric oxide (16.67% vs 76%, p<0.01) compared to patients on HFV. Infants requiring HFV had larger defects (p<0.01), took longer to stabilize before surgery (p<0.05) and had more non-cardiac congenital anomalies (p<0.05). After controlling for confounding variables, ventilation mode was not an independent contributor to BPD or death.
These results show that CDH patients repaired on HFV have higher rates of BPD and mortality but also required more support than patients repaired on CMV. Therefore, the association of BPD or death with HFV may reflect a prognostic association rather than a causal association – HFV indicates a sicker child with a poorer prognosis. We did not find that mode of ventilation contributed to development of BPD or death.