Xenophon Sinopidis

Background: In infants treated with a low-flow nasal cannula (LFNC), the oxygen concentration delivered to the lungs (i.e., the effective FiO2) is difficult to estimate. The existing mathematical formulas rely on important assumptions regarding the values of respiratory parameters and, thus, may be inaccurate. We aimed to assess oxygen delivery by LFNC to small infants using realistic simulations on a mechanical breathing model. Methods: A mechanical breathing simulator (infant upper-airway replica, single-space breathing compartment, electric motor, microcontroller) was developed. Breathing simulations (n = 1200) were performed at various tidal volume (VT), inspiratory time (Ti), and respiratory rate (RR) combinations and different cannula flows. Results: Minute ventilation (MV) was the most significant predictor of effective FiO2. FiO2 was higher at lower VT and higher Ti values. Benaron and Benitz’s formula underestimated the effective FiO2 at lower MV values, while Finer’s formula significantly overestimated it. A set of predictive FiO2 charts was developed based on cannula flow, infant body weight, and RR. Conclusions: The effective FiO2 delivered by LFNC to small infants critically depends on VT, Ti, and RR. However, since VT and Ti values are not available in clinical practice, the existing mathematical formulas may be inaccurate. Our novel predictive FiO2 charts could assist in optimizing oxygen delivery by LFNC using easy-to-obtain parameters, such as infant body weight and RR.