Discussion
In our experiment, we developed an expiratory gas modification bellow, which modified the expiratory flow during the exhalation phase. This model was developed to sample and measure tracheal oxygen concentrations with various breathing patterns of both VT and frequency under normal, restrictive and obstructive lung mechanics. In addition, we provided equations for estimating inspired oxygen in the trachea for clinical reference.
During oxygen therapy with nasal cannula, the rule of thumb is that for patients with a normal rate and depth of breathing, each 1L/min of oxygen supplied increases FIO2 by approximately 4%;1,2,4 therefore, the expected delivery is a FIO2 of 0.24–0.44. However, if the patient’s f approaches or exceeds 20breaths/min, FIO2 will likely be well below estimates, and the delivered FIO2 will only increase by approximately 0.025 for each 1L/min above the ambient oxygen level.1,31 In those studies, FIO2 was measured by an oxygen analyzer placed in the low airway, such as an oxygen analyzer port of the TTL bellow, which is the distal airway. In our experiment in which we designed the test lung, there was no effect of fresh oxygen flow from the exhaled carbon dioxide because the expiratory gas modification compartment was designed to produce a modified expired gas. In addition, we measured FO2 insp. and FO2 pre-insp. at the carina, which is the proximal airway. Lower FO2 insp. and FO2pre-insp. values than estimated were found at all settings, which may be due to differences in monitoring sites and the experimental design. By contrast, Gibson et al. reported that in healthy subjects with a percutaneously placed tracheal sensing catheter, the highest absolute inspired tracheal oxygen concentration with a nasal cannula was 23.6% at 3L/min and 25.4% at 5L/min during normal breathing [VT:690mL, f:17breaths/min, minute ventilation (MV):11L/min, peak inspiratory flow rate (PIFR):37L/min], 22.4% at 3L/min and 23.8% at 5L/min during quiet breathing (VT:400mL, f:16breaths/min, MV:6.4L/min, PIFR:21L/min), and 22.7% at 3L/min and 25.2% at 5L/min during hyperventilation (VT:1400mL, f:14breaths/min, MV:19.5L/min, PIFR:63L/min) .32 In another report, in patients with 97% oxygen supply via nasal cannula, the effective FIO2 was 22.8 ± 0.1%, 27.6 ± 0.5% and 31.8 ± 0.5% with 1, 3 and 5L/min based on trachea sampling, respectively.33 Our findings concur reasonably well with those data, but contrast with a previously published formula.
Oxygen concentration is also influenced by VT in our model in the same sense as previously described by Chikata et al., who reported that statistically significantly changes in VTaffected measured FIO2 at all flow levels: at 2L/min and VT of 300, 500 and 700mL, FIO2 was 0.37 ± 0.01%, 0.32 ± 0% and 0.29 ± 0%, respectively; at 4L/min, FIO2 was 0.45 ± 0.01%, 0.39 ± 0.01% and 0.34 ± 0%, respectively.34 Indeed, various authors have reported decreasing FIO2 values during increasing MV and f.24 Our model clearly demonstrates that changes in VT affect measured oxygen concentration. Specifically, as VT increases with a fixed f in the normal, restrictive and obstructive lung models, there is a reduction in the effective oxygen concentration due to more room air inhalation and dilution of the oxygen concentration from the nasal cannula at all flow levels. For example, the measured FO2 insp. in normal lung mechanics at a flow rate of 1L/min, VT of 300mL and f of 10breaths/min was 22.7 ± 0.10%, whereas it was 22.33 ± 0.15% and 21.60 ± 0.20% at VT of 500mL and 700mL, respectively.
Our study found that FO2 insp. and FO2pre-insp. differed in different respiratory cycles. During the inspiratory phase, FO2 insp. measured at the carina was significantly lower than that measured during the expiratory period. In the previously described by O’Reilly-Nugent et al, the variability of FIO2 in different respiratory cycles during low-flow nasal cannula at oxygen flow rates of 2–4L/min. Their novel method of FIO2 measurement involved sampling via a catheter placed at the distal trachea; this has clarified the uncertainties of other studies that were sampled near the proximal airway where inadequate gas mixing had occurred. The researchers found that because of the bolus of oxygen from the nasopharynx or relatively low inspiratory flow at the beginning of each breath, the peak FIO2 is significantly higher than the remainder of inspiration and then declines rapidly as the inspiratory flow reaches its peak. After inspiration, FIO2 then increases steadily, coinciding with a reduction in inspiratory flow.35 Conversely, the oxygen delivered from the nasal cannula is inhaled into the lungs for gas exchange, and the oxygen concentration measured at the carina is relatively low. However, during the expiratory period, carbon dioxide in the lungs is exhaled, but oxygen inhalation does not occur. Therefore, during this period, the measured oxygen concentration is higher, which is most pronounced particularly between the end-expiratory period and the next inspiratory period. Thus, any variation in the total expiratory time affects the composition of the next inspiration.4,21,27
Finally, our study provides scientific data that f, VTand oxygen flow rate significantly influence the inspired oxygen concentration under different lung mechanics. In addition, during spontaneous breathing, we have successfully proposed equations for clinical practice for nasal cannula oxygen therapy. These equations are based on different lung mechanics, VT and f to estimate the actual patient oxygen concentration; they will provide clinicians a reference when using nasal cannula among patients with normal, restrictive and obstructive lung diseases.
Although we devised a spontaneously breathing lung model with three lung mechanics conditions and different VT and f, but our study was limited to did not consider open or closed mouth breathing states even though oxygen concentration is influenced by the reservoir space in the oral cavity during nasal cannula. In addition, we did not investigate the effect of water vapours on oxygen concentration.