BACKGROUND:The cuff leak test is commonly performed before extubation in critically ill children to estimate the risk of postextubation upper-airway obstruction (UAO). Therefore, documentation of cuff leak test results is important for clinical decision-making. We noted delays in extubation related to the absence of clearly documented cuff leak test results and sought to address this through electronic medical record (EMR) enhancement. METHODS:We conducted a single-center quality improvement project with multiple Plan-Do-Study-Act cycles to improve cuff leak test documentation. Interventions included test standardization and the addition of a dedicated EMR flowsheet row for documentation. Outcomes were compared before and after flowsheet row implementation using run charts. The primary outcome was cuff leak test documentation, with UAO requiring treatment and postextubation dexamethasone as balancing measures. RESULTS:Following flowsheet row implementation, overall cuff leak documentation increased from 78% to 91% (P < .001), respiratory therapist (RT) documentation increased from 60% to 86% (P < .001), and inclusion in a progress or extubation note increased from 47% to 83% (P < .001). More subjects had a documented cuff leak before extubation (74% vs 63%, P < .001), and fewer received dexamethasone before extubation (28% vs 34%, P = .03). Run charts demonstrated a statistically significant shift in overall documentation from 78% to 92% following flowsheet row implementation, RT documentation increased from 61% to 91% before declining to 83%, while provider documentation increased from 45% to 79%, followed by a second shift to 89%. No changes were observed in UAO requiring treatment or dexamethasone administration. CONCLUSIONS:Addition of a dedicated EMR flowsheet row significantly improved overall cuff leak documentation, RT documentation, and provider documentation. It was also associated with reduced pre-extubation dexamethasone exposure with no change in UAO requiring treatment.
Prone positioning (PP) has evolved from an experimental maneuver into a cornerstone therapy for moderate-to-severe ARDS, supported by robust evidence of survival benefit. Its physiological rationale includes reducing dorsoventral mechanical heterogeneity, promoting dependent alveolar recruitment, improving ventilation-perfusion (V˙/Q˙) matching, and attenuating ventilator-induced lung injury (VILI). In preclinical models, PP also homogenizes the pleural pressure response to spontaneous breathing effort, thereby attenuating effort-dependent lung injury, a mechanism particularly relevant for awake PP. Despite these well-characterized mechanisms, the clinical response to PP is highly individualized, and a substantial proportion of patients do not show meaningful improvement in oxygenation. Furthermore, optimal PEEP during PP cannot be reliably inferred from conventional global monitoring parameters or from supine PEEP settings. Electrical impedance tomography (EIT) is a radiation-free, continuous bedside imaging tool that enables real-time spatial assessment of regional lung ventilation, perfusion, and alveolar recruitment. Applied during PP, EIT can directly visualize the redistribution of tidal ventilation toward dorsal lung regions, identify persistent collapse or overdistention, guide individualized PEEP titration, and quantify V˙/Q˙ matching, all of which are inaccessible from global metrics alone. This narrative review examines the physiological mechanisms of PP, the temporal dynamics of ventilation and perfusion responses, and the clinical heterogeneity of mechanical response. It evaluates the current evidence base for EIT as a tool to guide individualized ventilator management during PP in ARDS, distinguishing established findings from physiologically derived rationale and areas requiring future clinical validation. Clinical implications and future directions, including awake proning, serial sessions, and PP during extracorporeal membrane oxygenation, are also discussed.
BACKGROUND:Administration of inhaled bronchodilators during mechanical ventilation is technically challenging, and data on the use of long-acting agents in this setting remain limited. METHODS:In this prospective study, 23 hemodynamically stable COPD subjects receiving mechanical ventilation and considered ready for weaning were enrolled. Tiotropium/olodaterol (2.5/2.5 µg, 2 puffs) was delivered via an inline adapter. Outcomes included peak inspiratory pressure (PIP), plateau pressure (Pplat), mean airway pressure (P¯aw), respiratory system resistance (RRS), tidal volumes (VTi and VTe), and the SpO2/FIO2. RESULTS:The median age was 76 years (interquartile range, 48-90), with 82.6% male subjects and 69.6% classified as GOLD stage 3 to 4. Compared with baseline, statistically significant changes occurred in PIP (P < .001), Pplat (P = .041), RRS (P = .004), and ventilator-measured VTi (P = .02) and VTe (P = .006). Variations in P¯aw (P = .067) and SpO2/FIO2 (P = .14) were not statistically significant. In exploratory subgroup analyses, greater changes in tidal volumes were observed in male subjects (VTi and VTe), subjects younger than 65 years (VTi), and those with a tracheostomy (VTe). No hemodynamic instability or treatment-related discontinuations were observed. CONCLUSIONS:Inline delivery of tiotropium/olodaterol during mechanical ventilation was feasible and was associated with changes in respiratory mechanics in COPD subjects. Given the observational design and absence of a control group, these findings should be interpreted as physiologic observations rather than evidence of treatment efficacy.
BACKGROUND:Portable home ventilators operating in pressure-targeted passive leak mode are widely prescribed for patients with neuromuscular disease (NMD). In this passive ventilation configuration, delivered tidal volume (VTi) is determined not by clinician-set parameters alone but by device-specific pressurization algorithms. Inter-device VTi variability has been demonstrated in cross-sectional bench studies; however, whether firmware updates alter these characteristics longitudinally-and whether such alterations are clinically important-has not been systematically examined. NMD is a progressive condition in which respiratory system compliance deteriorates over time, further amplifying the clinical relevance of device pressurization characteristics. METHODS:We performed bench testing of portable home ventilators in 3 measurement cohorts: 2018 (no. = 7 devices), 2020 (no. = 3 devices), and 2021 (no. = 4 devices). All cohorts used an identical protocol: single-limb passive leak circuit (Intersurgical), 1.8-m corrugated tubing, SmartLung test lung (IMT Analytics, Buchs, Switzerland), inspiratory positive airway pressure (IPAP) 15 cm H2O/expiratory positive airway pressure 5 cm H2O, TI1.0 s, breathing frequency 10 breaths/min, rise time fastest, and Body Temperature, Pressure, Saturated (BTPS) correction. Three test conditions simulated NMD mechanics (C30/R5: compliance 30 mL/cm H2O, resistance 5 cm H2O/L/s), high airway resistance (C30/R50), and low compliance (C20/R5). Data were logged continuously for each condition for 60 min at a 10-Hz output interval (CITREX H4 analyzer; IMT Analytics), yielding ∼590 complete breath cycles per condition. Ensemble-averaged VTi, pressurization indices at 300 and 500 ms (P300 ms, P500 ms), and inspiratory flow at 80% of the set TI(F80%TI) were derived from all breath cycles. RESULTS:Under C30/R5 in the 2018 cohort, inter-device VTi ranged from 308 to 424 mL (range 116 mL; 32.7% of the cohort mean). Devices were classified into 2 flow profile categories: high late-TI flow (VOCSN, F80%TI = 11.3 L/min; Vivo 60, 9.8 L/min) and early flow decay (Stellar 100 [VELIA], Trilogy 100, PB560, Astral 150, A40; F80%TI 3.0-7.2 L/min). Longitudinal comparison revealed clinically important VTi reduction following firmware updates: VOCSN decreased from 424 ± 4.8 mL (v4.01.06R, 2018) to 269 ± 4.5 mL (v4.13.00R, 2021), a reduction of 155 mL (-37%); Trilogy EVO decreased from 396 ± 2.0 mL (v1.01, 2020) to 275 ± 1.2 mL (v1.04.06, 2021), a reduction of 121 mL (-31%); and Vivo 45LS decreased from 515 ± 2.6 mL (early lot, 2020) to 316 ± 3.0 mL (revised firmware, 2021), a reduction of 199 mL (-39%). In each case, P300 ms was preserved across firmware versions, whereas F80%TI fell sharply (VOCSN: 11.3 to 1.4 L/min; Trilogy EVO: 8.3 to 2.0 L/min; Vivo 45LS: 21.5 to -0.5 L/min), consistent with compromised pressure maintenance fidelity during the latter phase of inspiration. Under C20/R5, VTi converged to 114-134 mL across all devices and cohorts regardless of firmware version. CONCLUSIONS:Inter-device VTi variability of up to 32.7% under identical settings indicates that device selection directly influences ventilation adequacy in passive NMD ventilation. Firmware updates produced a clinically important 31-39% reduction in VTi without any modification of clinician-set parameters, with altered pressure maintenance fidelity in the latter phase of inspiration demonstrated as the primary mechanism. These changes reflect proprietary algorithm modifications that were implemented, typically not detailed in standard firmware release notes available to clinicians. In patients with low respiratory system compliance, device substitution is unlikely to resolve the VTi deficit; IPAP adjustment represents a primary strategy within the existing circuit. Clinicians managing ventilator-dependent NMD patients should be aware that the VTi actually delivered may not correspond to expectations based on set parameters and should verify VTi delivery after firmware updates. Knowledge of device-specific pressurization characteristics should inform ventilator selection commensurate with disease severity and progression.
BACKGROUND:Mechanical power integrates the energy delivered to the respiratory system during mechanical ventilation but may not reflect the mechanical stress applied to lung tissue. Metrics focused on the elastic component of ventilatory energy may provide a more physiologically relevant estimate of parenchymal stress. We evaluated elastic static power and elastic static power normalized to predicted body weight (PBW) and their association with ARDS severity and mortalityMethods:We conducted a single-center retrospective cohort study in an ICU, analyzed using Bayesian modeling. Adult subjects (N = 137) receiving volume-controlled invasive mechanical ventilation for COVID-19 pneumonia were included, encompassing individuals without ARDS and with mild, moderate, or severe ARDS. Elastic static power and PBW-normalized elastic static power were derived from ventilatory parameters recorded within the first 24 h after initiation of mechanical ventilation. Bayesian regression models were used to evaluate associations between these variables and ARDS severity, ICU mortality, and 28-day mortality. RESULTS:Posterior distributions for elastic load metrics favored greater ARDS severity and increased mortality. Elastic static power favored moderate-to-severe ARDS (relative risk 1.32, 95% credible interval [CrI] 1.10-1.57), ICU mortality (hazard ratio [HR] 1.36, 95% CrI 1.16-1.60), and 28-day mortality (HR: 1.31, 95% CrI: 1.17-1.48). Similar patterns were observed after normalization to PBW. CONCLUSIONS:Posterior distributions for elastic ventilatory energy, in both its nonnormalized and PBW-normalized forms, consistently favored greater ARDS severity and mortality. These findings support the use of elastic load metrics as physiologically relevant indicators of ventilatory stress.
BACKGROUND:Identifying predictors of noninvasive respiratory support failure is crucial to avoid delays in intubation. This study assessed, in a homogeneous group of patients with SARS-CoV-2-related acute hypoxemic respiratory failure (AHRF), the association between outcome and early measured expired tidal volume normalized to predicted body weight (VTe/PBW), a surrogate of lung-distending pressure. METHODS:This single-center retrospective study was conducted at a Swiss tertiary-care university hospital. Data were collected from medical records of patients admitted to the ICU between March 2020 and April 2022 for SARS-CoV-2-related AHRF requiring noninvasive respiratory support, with VTe recorded during an early positive pressure session using an oronasal mask and a double-limb circuit. VTe/PBW and other potential factors associated with outcome were compared between patients with favorable and unfavorable (intubation and/or ICU death) outcomes using Fisher's exact and Mann-Whitney tests. The optimal VTe/PBW cutoff associated with unfavorable outcome was identified with the Youden index. Univariate and multivariate analyses were performed. RESULTS:Eighty subjects were included, 38 (47.5%) with unfavorable outcome. Median VTe/PBW was higher in the unfavorable outcome group (9.8 [8.6-11.5] vs 8.5 [7.6-10.1] mL/kg PBW, P = .02). In univariate analysis, VTe/PBW was associated with outcome (odds ratio [OR] = 1.27, 95% CI = 1.06-1.57), as were age, gender, Sequential Organ Failure Assessment, SAPS II, SpO2/FiO2 ratio at admission, and VOX index. The area under the receiver operating characteristic curve for the correlation between VTe/PBW and unfavorable outcome was 0.66. Based on the Youden index, the VTe/PBW cutoff most strongly associated with the outcome was 8.5 mL/kg PBW. Nevertheless, because differences across VTe/PBW cutoffs were small, this threshold should be considered hypothesis-generating. CONCLUSIONS:VTe/PBW during an early positive-pressure ventilation session was associated with worsening outcomes in subjects with SARS-CoV-2-related AHRF. The VTe/PBW cutoff most strongly associated with unfavorable outcome was 8.5 mL/kg PBW.
BACKGROUND:Noninvasive respiratory support (NRS) is frequently trialed in pediatric ARDS despite limited evidence supporting its efficacy. We aim to describe patient, disease, and NRS characteristics associated with NRS failure in ARDS and investigate outcomes associated with early versus delayed intubation. METHODS:This is a retrospective cohort study. Patients meeting pediatric ARDS criteria (2023 PALICC-2 definition) initially supported by NRS were included. Eligible patients were divided into 2 groups: NRS success (NRS use without subsequent endotracheal intubation) versus NRS failure (NRS use with subsequent endotracheal intubation). Subject characteristics, disease severity, and NRS usage patterns were evaluated. NRS failure rate, PICU and hospital stay, and mortality are reported. The NRS failure group was further stratified by early (≤8 h) versus late (>8 h) intubation. Chi squared, Fisher exact, and Wilcoxon rank sum tests were performed as appropriate. RESULTS:123 subjects were included (79 NRS success and 44 NRS failure). The initial median SpO2/FIO2 ratio for the entire cohort was 160 (mild/moderate ARDS). Initial SpO2/FIO2 ratios did not differ between the groups (P = .12). Subjects in the NRS failure group were more likely to have genetic comorbidities (P = .009), higher PRISM III scores (P < .001), and ARDS etiologies of sepsis and aspiration. The NRS failure rate was 35.8%, and mortality was low at 4.9%. PICU (P < .001) and hospital (P < .001) stay were significantly longer in the failure group. Within the failure group, the median time to intubation was 8 h. Subjects intubated before 8 h had fewer ventilator days (P = .052) and significantly shorter PICU (P = .03) and hospital (P = .033) stay compared with subjects intubated after 8 h. CONCLUSIONS:NRS failure was associated with longer PICU and hospital stay. Early intubation in the failure group was associated with decreased PICU and hospital stay.
BACKGROUND:High-flow nasal cannula (HFNC) liberation strategies in pediatric intensive care often transition patients directly from HFNC to room air. At high altitude, oxygen requirements may persist after flow-dependent respiratory support is no longer needed. We evaluated the association between a protocolized, respiratory therapist-driven HFNC liberation strategy and time to successful liberation from high-flow support in children with acute respiratory failure. METHODS:We conducted a retrospective cohort study using prospectively collected LARed Network registry data from a tertiary PICU in Bogotá, Colombia (2,600 m altitude). In September 2022, the unit implemented an HFNC liberation strategy designed to separate persistent oxygen requirement from ongoing need for high-flow support. The primary analysis included the first HFNC episode per PICU admission from March 2018 to March 2025. The primary outcome was time to successful HFNC liberation. Kaplan-Meier curves and multivariable Cox regression were used; gamma log-link models were used for sensitivity analyses, including a prespecified bronchiolitis subgroup. RESULTS:We included 1,086 PICU admissions, 627 before and 459 after implementation. Median HFNC duration decreased from 59.8 h (interquartile ranges [IQR] 34.7-91.4) to 53.1 h (IQR 32.3-80.5, P = .01). Kaplan-Meier analysis showed earlier HFNC liberation after implementation (log-rank P < .01). Implementation was associated with earlier liberation in adjusted Cox regression (hazard ratio 1.17, 95% CI 1.03-1.34, P = .02). PICU stay decreased from 4.7 days (IQR 3.2-6.9) to 4.0 days (IQR 2.9-5.8, P < .01). In bronchiolitis, implementation was associated with an 18.2% relative reduction in HFNC duration (95% CI 7.7-27.5%, P < .01). CONCLUSIONS:In a high-altitude PICU, a respiratory therapist-driven HFNC liberation strategy was associated with earlier liberation from high-flow support. A strategy that preserves oxygen delivery while testing tolerance of minimal flow may be useful in high-altitude settings.
BACKGROUND:Almost 2.5 million neonatal deaths occur annually, disproportionately affecting low- and middle-income countries. Respiratory support devices remain largely inaccessible in these regions because of prohibitive costs and infrastructure requirements. The WHO-style bubble CPAP provides accessible continuous positive pressure but cannot deliver bi-level support when CPAP alone is insufficient. We hypothesized that a low-cost bubble noninvasive ventilatory (Bubble-NIV) system, meant to augment the WHO-style bubble CPAP, would perform non-inferiorly to the ventilator-driven RAM-NIV in delivering positive pressure across variable nasal interface conditions in bench testing. METHODS:We developed a low-cost Bubble-NIV system (∼$50) with a servo motor-driven lever to drive bi-level pressures via variable distal tubing water submersion depths. Performance was compared with ventilator-driven RAM-NIV in a static in vitro model across 3 nasal cannula fits (80%, 70%, 60% nostril occlusion). Primary outcome was mean airway pressure measured in the nasopharynx. RESULTS:Bubble-NIV maintained bi-level pressures and delivered pressures closer to target (mean airway pressure, mean goal: 15 cm H2O) than ventilator-driven RAM-NIV across all nasal interface scenarios: 15.1 versus 13.4 cm H2O at 80% occlusion (P < .001), 12.6 versus 7.4 cm H2O at 70% occlusion (P < .001), and 5.6 versus 1.8 cm H2O at 60% occlusion (P < .001). The performance advantage increased with worsening leak (1.1-fold to 3.1-fold). Optimal Bubble-NIV pressures occurred with flows set within 2 L/min of minimal bubbling. Oscillatory waveforms of Bubble-NIV were greater than RAM-NIV. CONCLUSIONS:A low cost Bubble-NIV system demonstrated superior nasal leak tolerance compared with ventilator-driven RAM-NIV in bench testing. The low-cost, nonproprietary design addresses accessibility barriers to delivering bi-level noninvasive respiratory support for neonates in resource-limited settings.
COPD remains a leading cause of global morbidity and mortality, and a substantial minority of patients remain disabled by hyperinflation, chronic bronchitis, and exacerbations despite optimized inhaled pharmacotherapy and pulmonary rehabilitation. Over the past two decades, bronchoscopic interventions have matured from clinical trial investigation into guideline-endorsed treatments for patients with refractory symptoms despite optimal medical treatment. This state-of-the-art review synthesizes the mechanistic rationale, evidence-based clinical trial data, patient-selection criteria, key controversies, and future research needs for bronchoscopic lung volume reduction (BLVR) and airway-directed therapies. In rigorously selected patients with severe emphysema, hyperinflation, and absence of interlobar collateral ventilation, endobronchial valves deliver clinically meaningful, durable improvements in FEV1, 6-min walk distance, dyspnea, and quality of life with moderate-to-high certainty and carry the highest guideline rating (GOLD Evidence A). Pneumothorax, occurring in roughly 18-34% of valve recipients, is the principal early complication, necessitating several days of in-patient monitoring that add complexity and cost to care. Endobronchial coils and thermal vapor ablation offer fissure-independent but more modest benefit; airway scaffolds are an emerging, tissue-sparing option that may extend candidacy to patients ineligible for existing bronchoscopic approaches; targeted lung denervation may stabilize lung function and symptoms but did not reduce exacerbations in its pivotal trial; and bronchial rheoplasty and metered cryospray are emerging options for the chronic bronchitis phenotype. The field is now advancing toward expanding candidacy for endobronchial valve therapy through collateral ventilation conversion, individualized complication prediction, and phenotype-matched airway therapies. Multidisciplinary, imaging-driven patient selection that addresses the key features of a patient's symptoms while avoiding significant comorbidity remains the cornerstone of successful outcomes.
BACKGROUND:Patient performance on extubation readiness tests (ERTs) informs the decision to extubate after cardiac surgery. ERT pass criteria at our center include age-based breathing frequencies and weight-adjusted tidal volumes (VT). Evidence validating these criteria in children after cardiac surgery is limited. We hypothesized that higher breathing frequency and/or lower VT during the final ERT prior to extubation, even if meeting pass criteria, would be associated with re-intubation within 48 h. METHODS:This was an observational, retrospective cohort study involving children <18 years of age receiving invasive mechanical ventilation for >24 h after cardiac surgery for Society of Thoracic Surgery (STAT) Mortality category ≥3 from May 2022 to April 2024. The primary outcome was extubation failure. The secondary outcome was the interval between the first passed ERT and extubation. Factors associated with re-intubation were explored using univariate and multivariable analyses. RESULTS:Two hundred sixty-eight subjects met inclusion criteria with a median (interquartile range) age of 58.5 (6.1-210.2) days and weight of 4.2 (3.3-7.5) kg. Thirty (11.2%) subjects experienced extubation failure. These subjects had a higher breathing frequency during the ERT (42 [36-52] vs 36 [28-44] breaths/min, P = .01). On univariate analysis, subjects who required re-intubation were younger, had longer courses of invasive mechanical ventilation, and more frequently received preoperative invasive mechanical ventilation, delayed sternal closure, or were in a higher STAT category. Multivariable analysis revealed no association between re-intubation and breathing frequency (adjusted odds ratio [aOR] = 1.02, 95% CI: 0.99-1.06), VT (aOR = 1.0, 95% CI: 0.78-1.25), or the Rapid Shallow Breathing Index (RSBI; expressed as breathing frequency/VT) (aOR = 1.11, 95% CI: 0.96-1.27). The interval between the first passed ERT and extubation did not differ between groups (1.4 [0.5-3.6] vs 1.2 [0.4-2.4] days, P = .22). CONCLUSIONS:There was no association between the final ERT breathing frequency, VT, or RSBI and re-intubation within 48 h.
BACKGROUND:Bubble humidification is often utilized with low-flow nasal cannulas despite a lack of evidence to support its use. The practice was discontinued in our children's hospital based on recommendations from the American Association for Respiratory Care clinical practice guideline for management of pediatric patients with oxygen in the acute care setting and other available literature. The aim of this study was to evaluate staff feedback post de-implementation of the practice. METHODS:Following staff and stakeholder communication, bubble humidification for low-flow oxygen therapy was discontinued in May 2024. Respiratory and nursing staff were invited to complete a post-de-implementation survey in August 2024. Domains included respondent demographics, clinical practice guideline familiarity, receipt of communication, agreement with the practice change, and patient concerns. Responses were collected electronically. RESULTS:Responses from 132 staff were analyzed. The majority were female (105, 79%), nurses (81, 61%), and had 5 years or less of work experience (49, 37%). Eighty-six (65%) reported familiarity with the guideline. Overall, 59 (45%) were indifferent to the practice change and 46 (35%) disagreed. More than half (75, 57%) were concerned about patients. Nurses were significantly younger (≤25 years, P = .052; 26-34 years, P = .005) and less experienced (0-5 years, P = .009) than respiratory therapists. There were no significant differences between disciplines for patient concern. Nasal irritation/dryness was the most common theme for the reason for disagreement and patient concern followed by epistaxis. CONCLUSIONS:Few nurses and respiratory therapists agreed with discontinuing bubble humidification despite clinical practice guideline recommendations, and many were concerned about patients. Prospective, randomized clinical trials for acute and chronic oxygen delivery are warranted.
BACKGROUND:Positive expiratory pressure (PEP) devices facilitate airway clearance, but the determinants of pressure transmission during therapy are not fully understood. METHODS:In this bench study, a flow-dependent PEP device was tested using tracheostomy tubes with internal diameters of 7, 8, and 9 mm under different oxygen flows (0-15 L/min), tidal volumes (260-886 mL), and breathing frequencies (10-32 breaths/min). Tracheal and airway pressures were measured. RESULTS:A total of 486 observations were analyzed. Mean maximum tracheal pressure was 11 ± 7 cm H2O for 7 mm and 8 mm tubes and 10 ± 6 cm H2O for 9 mm tubes. Oxygen flow was the main determinant of maximum tracheal pressure (β = 3.19, P < .001), while tidal volume (β = 2.75, P < .001) and breathing frequency (β = 4.44, P < .001) were also independently associated with higher pressures. Compared with the 7-mm cannula, maximum tracheal pressure was significantly lower only with the 9-mm cannula. Minimum tracheal pressure was independently associated with oxygen flow (β = 1.28, P < .001), tidal volume (β = -2.56, P < .001), and breathing frequency (β = -3.00, P < .001), with significant interaction effects between ventilatory variables. Maximum airway-opening pressure-maximum tracheal pressure gradient was negative in 354 of 486 observations and became progressively less negative with increasing cannula diameter. Oxygen flow was associated with a higher gradient (β = 0.16, P < .001), whereas tidal volume (β = -0.59, P < .001) and breathing frequency (β = -0.87, P < .001) were associated with a lower gradient. CONCLUSIONS:Tracheal pressure during a flow-dependent PEP therapy was influenced by oxygen flow, ventilatory pattern, and airway diameter. These findings highlight the dynamic behavior of a flow-dependent PEP device and the importance of patient-specific respiratory mechanics during therapy.
