Background: In an accompanying abstract we reported FIO2 achieved with a simple O2 mask (SM) or non-rebreather O2 mask (NRM) varied according to O2 flow and level of minute ventilation (VE). Here we test the premise that air entrainment signified by mismatching of O2 flows with both peak and mean inspiratory flows (VI-peak and VI-mean) result in lower FIO2 values. Methods: As a brief review, our model consisted of 3 VE levels: normal (6 L/min: VT 350 mL x 17), ARDS spontaneous breathing (8 L/min: VT 350 mL x 24), and hyperpnea (14 L/min: VT 550 mL x 26). O2 flows were 10, 12, 15, and 20 L/min. Twelve paired measurements each were made between FIO2 and corresponding measures of: 1) O2 flow, 2) VI-peak, 3) VI-mean, and 4) differences between O2 flow and both VI-peak and mean. Paired data were analyzed by linear regression. Alpha was set at 0.05. Results: Relationships between O2 flow and FIO2 differed markedly between SM (R2: 0.48, P = .013) and NRM (R2:0.28, P = .08), that was reflected in how FIO2 corresponded with differences between O2 flow and both VI-peak and VI-mean (Table). Using a cut-off difference of ≤ 10 L/min for both devices, 67% of SM pairings had a FIO2 ≥ 0.60, whereas 100% of NRM pairings had a FIO2 ≥ 0.70. Conclusions: With SM the association between FIO2 and either O2 flow or mismatches between VI-peak or VI-mean differed little; whereas with NRM the mismatch between VI-peak or mean with O2 flow had a marked impact on FIO2. The stronger association with VI-mean reflects the influence of VT and mismatch duration (ie, VI-mean = VT ÷ TI). The weak association between O2 flow and FIO2 with NRM (and its strong relationship with flow differences) underscores the importance of an O2 reservoir in reducing the impact of air entrainment.Table 1. SMNRMFIO2 vs. ΔVI-peak - O2 FlowrateR2: 0.43, P=0.02R2: 0.91, P<0.001FIO2 vs. Δ VI-mean - O2 FlowrateR2: 0.48, P=0.01R2: 0.94, P<0.001
Kallet et al. (Sun,) studied this question.