Oxygen Dilution Among Portable Home Ventilators in Children: The Influence of Prescribed Settings and Leak

Résumé fourni par la source

Children requiring invasive home mechanical ventilation (HMV) often receive respiratory support via tracheostomy tubes and single-limb, passive circuits (SLPC). Advantages of SLPCs—compared to active circuits—include greater simplicity, availability of volume-targeting modes, and superior leak compensation [1]. Two primary disadvantages of SLPCs are the inability to directly measure exhaled tidal volume and the variability in delivered fraction of inspired oxygen (FiO2) [1, 2]. Since portable home ventilators (PHV) typically utilize low-pressure oxygen sources with fixed flow rates, any increase in ventilator output (which occurs through increased ambient airflow) will dilute the effective concentration of oxygen ultimately delivered to the patient. Certain factors leading to increased ventilator output include intentional and unintentional leaks as well as higher levels of baseline respiratory support. Several PHVs are approved for use in infants and children, and while their functionality is generally comparable, variations in ventilation modes and device-specific, proprietary algorithms (e.g., leak compensation) likely influence ventilator monitoring and output [3]. It is not known whether these differences independently affect oxygen delivery to the patient. This study aims to evaluate several different PHVs utilizing SLPCs with low-flow oxygen and characterize the influence of leak and setting prescription on effective FiO2 delivery. We analyzed the effective FiO2 delivery of multiple PHVs when subjected to different leak and respiratory support scenarios. Testing was completed at four pediatric centers; ventilators were chosen based on local availability. Testing was designed to simulate the care of pediatric patients with chronic respiratory failure requiring invasive HMV. Accordingly, PHVs were set in pressure-controlled ventilation using SLPCs with low-flow supplemental oxygen. The test lung was calibrated to mimic a pediatric patient. We tested five home ventilator models: the Trilogy EV300/EVO (Philips Respironics; three centers), Trilogy 100 (Philips Respironics; two centers), Vivo 45 LS (Breas; three centers), Astral 150 (ResMed, one center), and V*Home/V+Pro (React Health, four centers). Each ventilator was connected to a 0.5 L infant test lung (SmartLung; IMT Analytics AG). The test lung was set to a compliance of 5 mL/mbar and a resistance of 50 mbar/L/s. Each ventilator was connected to the test lung with an SLPC (Fischer and Paykel, adult) and a Whisper-Swivel exhalation valve (Respironics). An external oxygen analyzer (Handi+/Max O2+/Max O2 ME by Maxtec) was placed between the Whisper-Swivel and the test lung. Oxygen analyzers were chosen based on center availability; each analyzer was calibrated to 21% FiO2 in room air prior to use. Circuit compliance/resistance calibration (i.e., pre-use testing) was performed, when available, for each ventilator prior to testing. The ventilators were set in a pressure-controlled mode of ventilation. Low: Peak inspiratory pressure of 20 cmH2O, positive end expiratory pressure of 8 cmH2O, inspiratory time of 0.7 s, and mandatory breath rate of 20 breaths per minute. High: Peak inspiratory pressure of 30 cmH2O, positive end expiratory pressure of 8 cmH2O, inspiratory time of 0.7 s, and mandatory breath rate of 40 breaths per minute. Low leak: Intrinsic, fixed leak of exhalation valve; no additional leak. High leak: From the “low” leak setting, the adjustable leak port on the test lung (see Figure 1) was titrated to achieve a roughly 20% increase in ventilator-reported leak (example: Low leak = 35 LPM [liter per minute], high leak titrated to 41−43 LPM). Low settings, low leak (LS, LL) Low settings, high leak (LS, HL) High settings, low leak (HS, LL) High settings, high leak (HS, HL) The FiO2 results were compiled from all testing centers; averages and ranges were calculated in a typical fashion and are reported below. In this simulation of pediatric invasive HMV, PHVs with SLPCs show variability in delivered FiO2 when subjected to different levels of leak and respiratory support (Table 1). This variability appears less clinically relevant with lower flows of supplemental oxygen but becomes more significant as flow increases (Table 2). Branson et al. previously demonstrated that FiO2 delivered during noninvasive ventilation using PHV is influenced by ventilatory mode (pressure vs. volume), site of oxygen entrainment, and the minute ventilation generated [4]. Our findings reaffirm aspects of that analysis; however, to our knowledge, this is the first study to analyze actual delivered FiO2 for PHVs using SLPCs with a pediatric, invasive ventilatory model. All ventilators demonstrated similar performance trends. At low supplemental oxygen levels, increased circuit leak or respiratory support had minimal impact on delivered FiO2. At higher oxygen flows, FiO2 variability became more pronounced due to the greater relative influence of increased ambient minute ventilation/inspiratory flow on entrained oxygen. These findings highlight the need for close monitoring of infants and children sensitive to oxygen fluctuations, especially when high ventilatory support or circuit leak is present. In these situations, the use of ventilators with active circuits or oxygen blenders, which allow for precise and consistent FiO2 delivery, may be indicated. It should be noted that depending on the type of device or level of respiratory support required, use of these ventilators may preclude transition to the outpatient setting. For instance, some durable medical equipment companies or home health providers may lack experience managing children with active circuit configurations at home, and oxygen blenders typically require high-pressure oxygen input not currently available for outpatient use. When comparing ventilator models, there is clear variability in delivered FiO2, even when facing identical clinical circumstances. Similarly, this appears to be less of a concern with lower amounts of supplemental oxygen, but may be more significant at higher flow levels. This finding is especially relevant when transitioning between ventilators. Children requiring greater amounts of supplemental oxygen may need adjustments in oxygen flow following ventilator transition; this should not be assumed to be from changing clinical status. The authors suspect that variations in how ventilator parameters are monitored, circuit calibration programs (i.e., pre-use testing), and leak compensation algorithms affect each ventilator's response to external factors. These differing responses likely influence the degree of ambient airflow output, and in turn, the dilution of supplemental oxygen delivered to the patient. Finally, this analysis provides an important reference for estimating delivered FiO2 in pediatric patients requiring invasive HMV. Our findings reinforce that certain guidelines such as the “4:1 rule” for effective FiO2 in nasal cannula delivery (i.e., a 4% increase in FiO2 for every LPM) have no utility when discussing mechanical ventilation with SLPCs [5]. Children transitioning from active-circuit, intensive care ventilators to low-flow PHVs with SLPCs may benefit from higher than anticipated supplemental oxygen levels to maximize the likelihood of a successful transition. Additionally, consistent clinical use of external oxygen analyzers is encouraged in these situations. Although pressure and respiratory rate were not assessed as independent variables, the evidence suggests that delivered FiO2% is influenced by ventilator-generated minute ventilation. The authors suspect that both pressure and mandatory rate independently contribute to FiO2 dilution, with the magnitude ultimately determined by resultant changes in minute ventilation. Additionally, while certain ventilators trended towards greater internal FiO2 variability across testing sites, directly comparing brands is challenging because certain models were evaluated at o

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.