Performance comparison of 4 portable oxygen concentrators.

BACKGROUND Several portable oxygen concentrators (POCs) have become available in the United States. Technical trade-offs are made in designing POCs, so their performance is expected to differ. METHODS We tested 4 POC models (Invacare XPO(2), Respironics EverGo, AirSep FreeStyle, and Inogen One) for oxygen delivery as a function of respiratory rate. We measured oxygen volume per pulse, pulse duration, trigger sensitivity, oxygen concentration in the gas delivered by the POC (measured with a ceramic-based oxygen sensor), and relative fraction of inspired oxygen (relative F(IO(2))), measured with a setup that included an adult nasal cannula, a model nares, and a lung simulator that has a built-in oxygen sensor. We studied respiratory rates of 15, 20, 25, 30, and 35 breaths/min, at a tidal volume of 500 mL. RESULTS The XPO(2) had the highest pulse flow, and FreeStyle had the lowest, which corresponded to the highest and lowest pulse volumes (40.8 mL and 18.2 mL respectively) at POC setting 2, at 15 breaths/min. The range of oxygen purity was 91.9-94.4% (P = .36). The Inogen had the shortest pulse-delivery time, and the XPO(2) had the longest (132 ms vs 281 ms, P < .001). The FreeStyle had the highest trigger sensitivity, and the Inogen the lowest (0.15 cm H(2)O vs 0.21 cm H(2)O, P < .001). At the maximum settings, with all 4 POCs, relative F(IO(2)) decreased as respiratory rate increased (relative F(IO(2)) range 28.6-31.4% at 15 breaths/min, 23.0-25.3% at 30 breaths/min). CONCLUSIONS These 4 POC models have markedly different performance, which emphasizes the need to adjust the POC setting to meet the specific patient's needs at rest and with activity.