The effectiveness of oxygen delivery and reliability of carbon dioxide waveforms: a crossover comparison of 4 nasal cannulae.

BACKGROUND Effective O2 delivery and accurate end-tidal CO2 (ETCO2) sampling are essential features of nasal cannulae (NCs) in patients with compromised respiratory status. We studied 4 NC designs: bifurcated nasal prongs (NPs) with O2 delivery and CO2 sensing in both NPs (Hudson), separate O2/CO2 NPs (Salter), and CO2 sensing in NPs with cloud O2 delivery outside the NPs via multi vents (Oridion) and dual vents (Medline). We hypothesized that design differences between NCs would influence O2 delivery and ETCO2 detection. METHODS Forty-five healthy volunteers, 18 to 35 years, participated in an unrestricted, randomized block design, each subject serving as their own control in a 4-period crossover study design of 4 NCs during one session. Monitoring included electrocardiogram, posterior pharynx O2 sampling from a Hauge Airway (Sharn Anesthesia Products, Tampa, FL), and NC ETCO2. In 11 volunteers, radial artery blood was sampled from a catheter for partial pressures of O2 and carbon dioxide (PaO2 and PaCO2) determination. Per randomization, each NC was positioned, and data were collected over 2 minutes (ETCO2, pharyngeal O2, PaO2, and PaCO2) during room air and during O2 fresh gas flows (FGFs) of 2, 4, and 6 Lpm. Statistical analyses were performed with SAS Analytics Pro, Version 9.3, and JMP Statistical Software, Version 11 (SAS Institute Inc., Cary, NC), significance at P < 0.05. RESULTS Blood gas analyses indicated PaCO2 during steady state at each experimental time period remained unchanged from physiologic baseline. PaO2 did not differ between NC devices at baseline or 2 Lpm O2. The PaO2 at 4 Lpm from the separate NPs and bifurcated NCs was significantly higher than the multi-vented NC. Pharyngeal O2 with the NC with separate NPs was significantly higher than multivented and dual-vented cloud delivery NCs at 2, 4, and 6 Lpm FGF. Pharyngeal O2 with the NC with bifurcated NPs was significantly higher than the multi-vented NC at 2 Lpm, and higher than cloud delivery NCs at 4 and 6 Lpm FGF. ETCO2 was significantly lower with the NC with bifurcated NPs compared to the other 3 NCs, consistent with errant CO2 tracings at higher FGF. CONCLUSIONS NCs provide supplemental inspired O2 concentrations for patients with impaired pulmonary function. Accurate measures of ETCO2 are helpful in assessing respiratory rate and determining whether CO2 retention is occurring from hypoventilation. These findings suggest the NC with separate NPs was the most effective in delivering O2 and the most consistent at providing reliable CO2 waveforms at higher FGFs.