Respiratory Rate Predicts Tidal Volume in a Miniature Continuous Flow Pressure Cycled Resuscitator

Background: Current ventilators are safe with a multitude of settings for optimization of support. When these ventilators fail or are in short supply, manual ventilation is required. Manual ventilation overly relies on the skill of the provider to breathe for the patient with little to no feedback. Providers of manual ventilation regardless of experience can easily give inconsistent breaths causing lung injury and inconsistent alveolar ventilation. Recent advancements in fluidic design and 3D printing have allow for the creation and miniaturization of a pressure cycled automated resuscitator that is simpler to operate and has no moving parts. Since this resuscitator uses continuous flow estimated tidal volumes could be predicted from respiratory rate. To test this theory, we designed and implemented a translational model of common compliances found in adults with and without ARDS. Methods: Hybrid Pietrain/Landrace swine were studied under a protocol approved by the Institutional Animal Care and Use Committee. All swine were sedated, anesthetized, and underwent oral intubation. Breath by breath measurements were obtained using the NM3 Monitor (Phillips). All 3 inVent (fluidIQ) resuscitators modules were used pre and post lung injury (Green 18/6, Yellow 24/10, Red 30/14 cm H2O) to fully characterize how each of the modules would react in differing lung conditions. Lung injury was induced through a repeated modified Lachmann bronchoalveolar lavage to achieve an A-a gradient greater than 200-300 mm Hg and a decreased lung compliance > 30%. ABG analysis was performed using the EPIC system (Siemens). The goal was to expose the inVent to a compliance range of 25-50 mL/cm H2O. Pulmonary mechanics were analyzed using JMP Pro 15 (SAS). The data for frequency and VT were characterized using a third-order polynomial curve fit using a 95% confidence interval and confidence of curve fit. Results: 8 swine with an average weight of 36 (5.8) kg were studied. 7 swine contributed to 39,770 breaths of pulmonary mechanics and 38 ABGs recorded. Mean dynamic compliance was 25.8 (14.7) mL/cm H2O and VT per kg delivered was 6.0 (2.6) mL/kg. An estimation equation was created, and example table formed. Conclusions: We continue to complete this study and aim to enroll larger animals to gain higher compliance ranges. Our translational model provide data to confidently create a predictive equation of VT by frequency. This will be helpful in developing a first in human study.