Validating a New Mathematical Calculation: High Frequency Oscillatory Ventilation Index (HFOVI) Compared to the Ventilation Index (VI)

Abstract

Background:

The purpose of this study was to validate a new mathematical equation measuring the high frequency oscillatory ventilation index (HFOVI) against the ventilation index (VI). The ventilation index (VI) calculation has been used to determine the severity of lung injury in critically ill patients. There is no similar calculation for high frequency oscillatory ventilation (HFOVI). The equation using HFOVI is a measure of severity of lung injury which may provide an objective predictor for escalation and de-escalation of therapy. The first step toward this is validation of the HFOVI equation compared to the existing ventilation index. VI is not optimized to evaluate patients on HFOV. HFOVI is a newly developed formula to calculate the severity of lung disease in patients on HFOV.

Methods:

This research project assessed the linearity of the HFOVI compared with the VI equation. Inclusion were all invasively ventilated patients who transitioned from pressure control ventilation to high frequency oscillatory ventilation and/or from high frequency to pressure control ventilation in the neonatal intensive care between 4/1/2019-9/1/2019. Each transition was collected as a separate event. HFOVI and VI were calculated from blood gas data acquired within one hour of transition. Underlying assumption was that the patient’s status was similar/same when they were measured. Subjects on specialty gas therapy were excluded.

Results:

22 unique subjects ages 0-38 days, mean 7.23 (SD=9.55) with 43 events were identified. 7 female (31.8%) and 15 males (68.2%). A linear mixed-effects model was used to analyze the relationship between HFOVI and VI. The result shows that HFOVI is a statistically significant predictor for VI (P < 0.001***). Based on the result, an increase by 1 unit in HFOVI corresponds to 4.177 increase in VI. Repeated measures correlation was used to analyze the correlation between VI and HFOVI. The output showed the repeated measures correlation coefficient of 0.552 (95% CI: 0.145, 0.799) and P-value of 0.008**.

Conclusions:

HFOVI is a statistically significant predictor of VI, and may be a predictor of the severity of lung injury. Utilizing the HFOVI as a predictor of the severity of lung injury with neonatal patients on HFOV may be helpful in identifying patients who need alternate modalities of care.