Serial Diaphragm Ultrasound Evaluation During Inspiratory Muscle Training in Prolonged Mechanical Ventilation

Prolonged mechanical ventilation (PMV) is associated with substantial morbidity, mortality, and healthcare resource utilization. In Taiwan, approximately 170,000 patients receive mechanical ventilation annually, of whom an estimated 13%-16% remain ventilator-dependent for more than 21 days and 3.7%-5.6% require ventilatory support for more than 63 days. Diaphragm dysfunction is an important physiological contributor to difficult ventilator liberation. Prolonged mechanical unloading, systemic inflammation, critical illness, and disuse may result in diaphragmatic atrophy, structural injury, and impaired contractility.

Inspiratory muscle training (IMT) is used during ventilator rehabilitation to improve inspiratory muscle strength and endurance. IMT may be delivered using resistive-loading or threshold-loading devices. Threshold-loading devices provide a relatively constant, flow-independent inspiratory resistance and permit training intensity to be prescribed according to maximal inspiratory pressure (Pimax). Similar to peripheral skeletal muscle training, respiratory muscle adaptation is influenced by the magnitude and progression of the applied workload. However, the appropriate training intensity for patients with PMV and diaphragm dysfunction remains uncertain. Higher-intensity training may provide a stronger physiological stimulus but may be less well tolerated, whereas lower-intensity training may be easier to perform but may produce a smaller training response.

The effects of IMT have traditionally been evaluated using global respiratory measurements such as Pimax, tidal volume, respiratory rate, minute ventilation, and the rapid shallow breathing index. Although these measurements provide clinically relevant information, most are effort-dependent and do not directly assess diaphragm-specific mechanical recovery. Point-of-care diaphragm ultrasound provides a noninvasive bedside method for serial evaluation of diaphragmatic function. Diaphragm excursion reflects craniocaudal diaphragmatic displacement during inspiration, while diaphragm thickening fraction reflects inspiratory muscle contraction within the zone of apposition.

This prospective, single-center, randomized controlled trial compared high-intensity and low-intensity inspiratory muscle training (IMT) in adults with prolonged mechanical ventilation and inspiratory muscle weakness who had been transferred from an intensive care unit to a specialized respiratory care center. Eligible participants were randomly assigned to receive an individualized IMT workload based on baseline maximal inspiratory pressure (Pimax). The high-intensity group trained at 50% of baseline Pimax, and the low-intensity group trained at 20% of baseline Pimax.

IMT was administered using a Dofin® adjustable-resistance threshold inspiratory muscle trainer (GaleMed Corporation, Taiwan). During each training session, the device was connected directly to the endotracheal or tracheostomy tube by a respiratory therapist after temporary disconnection from the mechanical ventilator. Each cycle included 6-10 inspiratory efforts, and five cycles were performed daily. Rest periods of 1-3 minutes were provided between cycles to reduce acute respiratory muscle fatigue. Pimax was reassessed twice weekly, and the training resistance was adjusted according to the updated measurement.

Participants were continuously observed during IMT for signs of respiratory distress or hemodynamic instability. Training was discontinued and mechanical ventilation was resumed when clinically significant oxygen desaturation, tachycardia, arrhythmia, diaphoresis, pallor, cyanosis, altered mental status, severe dyspnea, or other evidence of intolerance occurred. When a participant could not tolerate the prescribed high-intensity workload, resistance was reduced in 10% increments. A session was terminated when intolerance persisted after a reduction of more than 20% from the prescribed workload. Training continued for up to 2 weeks or until successful liberation from mechanical ventilation.

Baseline demographic and clinical characteristics were prospectively recorded. Respiratory mechanics, including respiratory rate, tidal volume, minute ventilation, rapid shallow breathing index, and Pimax, were assessed twice weekly during the intervention period. Mechanical ventilation duration and ventilator days after IMT initiation were also recorded.

Point-of-care diaphragm ultrasound was performed at baseline and on days 4, 7, and 10 after IMT initiation. During ultrasound assessment, ventilatory support was changed to continuous positive airway pressure mode to standardize the measurement condition. Right diaphragmatic excursion was measured using M-mode ultrasonography with a convex-array transducer positioned near the intersection of the right costal margin and midclavicular line. Diaphragm thickness was measured using a high-frequency linear-array transducer placed over the right zone of apposition near the anterior axillary line. Thickening fraction was calculated from the difference between inspiratory and expiratory diaphragm thickness relative to expiratory thickness.

Serial changes in respiratory mechanics and diaphragm ultrasound measurements were compared between the high-intensity and low-intensity inspiratory muscle training groups using statistical methods appropriate for repeated measurements. Ventilator liberation was evaluated using time-to-event methods. Kaplan-Meier analysis was used to compare the cumulative probability of successful ventilator liberation, with participants who did not achieve liberation censored at 28 days. Group differences were assessed using the log-rank test, and Cox proportional hazards regression was used to evaluate the association between inspiratory muscle training intensity and time to successful ventilator liberation.

The study was approved by the Institutional Review Board of Chang Gung Medical Foundation, Taiwan. Written informed consent was obtained from each participant's legally authorized representative before enrollment. Study procedures were conducted in accordance with the Declaration of Helsinki.

Inclusion criteria

• Willing and able to provide written informed consent.
• Adults aged >20 years.
• Patients requiring mechanical ventilation for >21 days and admitted to the Respiratory Care Center (RCC).
• Hemodynamically stable and not receiving vasopressor therapy.
• Clinically stable and considered by the attending physician to be ready to begin weaning from mechanical ventilation. Ventilator settings must meet all of the following criteria:

(1) Fraction of inspired oxygen (FiO₂) < 0.60. (2) Positive end-expiratory pressure (PEEP) ≤10 cmH₂O. (3) Able to trigger spontaneous breaths on the ventilator. (4) Respiratory rate <25 breaths/min. (5) Respiratory muscle weakness confirmed by maximal inspiratory pressure (MIP), defined as MIP <30 cmH₂O.

Exclusion criteria

• Hemodynamic instability, including cardiac arrhythmia, decompensated heart failure, or acute coronary syndrome.
• Active hemoptysis.
• Presence of a chest tube.
• Use of home mechanical ventilation.
• Uncontrolled pain or dyspnea.
• Patients determined to be ventilator-dependent and requiring long-term mechanical ventilation.
• Patients or their legally authorized representatives who are unwilling or unable to provide written informed consent.