Accuracy of oxygen delivery through bubble humidifiers and nasal catheters

Abstract Objective To evaluate the equipment used for nasal insufflation of oxygen and determine its accuracy. Study design Original study. Methods Oxygen delivery assemblies consisting of a flowmeter, bubble humidifier, oxygen delivery tubing and nasal insufflation catheters were assembled. Single and double catheter assemblies were made for four sizes of nasogastric feeding tubes (3.5 Fr, 5.0 Fr, 8.0 Fr and 10.0 Fr) resulting in 64 individual assemblies. A gas flow analyzer measured oxygen flow at the tip of the nasal catheter assemblies and from the pressure relief valve (PRV) of the bubble humidifiers. Statistical analyses were conducted to assess the functionality of assemblies. For functional assemblies, the accuracy of oxygen flow relative to the prescribed flow settings was determined. Results Catheter size was significantly associated with the functionality of assemblies. Probability (95% confidence interval) of 3.5 Fr, 5.0 Fr and 8.0 Fr assemblies being functional was estimated at 0.53 (0.14, 0.89), 0.83 (0.36, 0.98) and 0.98 (0.76, 0.99), respectively. All 10.0 Fr assemblies were functional. Functional assemblies, in general, consistently under-delivered the prescribed flow because a large portion of set flow was diverted through the bubble humidifier PRV. Conclusions Leaks through the PRV cause significant diversion of oxygen prior to it reaching the catheter tips. Smaller patients are particularly susceptible, as small catheters limit oxygen delivery creating proportionally greater leaks through the PRV. Clinical relevance It was not possible to accurately deliver oxygen because of leaks through the PRV. Targeting a specific outcome (e.g., SpO2 > 94%, PaO2 80–120 mmHg; 11–16 kPa) and avoiding unnecessarily high fractions of inspired oxygen cannot be done if flow delivery cannot be accurately assured. One possible solution would be to use a bubble humidifier with a 6 psi PRV that does not leak prior to reaching the opening pressure.