Our study demonstrates a correlation between ultrasound measurement of diaphragm excursion and TV in healthy volunteers, but the large variability in the data in ICU patients indicates a less obvious association. This precludes a reliable estimation of the absolute value of TV from diaphragm excursion measurement in the clinical setting.

The difference in TV-diaphragm excursion ratio between ICU patients and healthy volunteers may be explained by a smaller distribution of TV in ICU patients. Indeed, an additional (sensitivity) analysis of the TV-diaphragm excursion relationship in healthy volunteers when including only breaths in the same TV range as ICU patients (TV ≤ 1250 mL) indicated that the TV-diaphragm excursion ratio is comparable to ICU patients [231 (209–255) mL/cm]. In addition, altered respiratory mechanics and potential effects of PEEP on diaphragm efficiency [13] have likely played a role. An earlier study showed that the application of PEEP resulted in caudal displacement of the diaphragm and decreased the diaphragm contractile efficiency. Possibly, the decrease in TV-diaphragm excursion ratio in ICU patients compared to healthy volunteers is explained by PEEP.

Our results contrast with earlier studies that observed a fair correlation between TV and diaphragm excursion [9, 10]. However, these studies were performed in non-clinical settings, and used simple linear regression analysis without accounting for multiple measurements per participant. Furthermore, their larger TV-diaphragm excursion ratios (555 and 625 mL/cm, respectively) may be explained by recruitment of accessory muscles, since participants in earlier studies were instructed to inhale up to total lung capacity.

There are limitations of our study to acknowledge. First, we did not quantify accessory muscle use, although the association between diaphragm excursion and TV is affected by these muscles. Occult recruitment of accessory muscles may, therefore, have distorted the TV-diaphragm excursion ratio especially in ICU patients. However, we evaluated the potential of diaphragm excursion as bedside tool to monitor TV. Simultaneous evaluation of accessory muscle recruitment might have improved the understanding of the association between diaphragm excursion and TV, but would also complicate its clinical applicability. Second, the average time on invasive mechanical ventilation in the studied ICU patients was rather short. We recognize that the association between diaphragm excursion and TV may differ in patients with prolonged invasive mechanical ventilation due to diaphragm muscle dysfunction [14]. However, the targeted population to use diaphragm excursion as proxy for TV would concern non-intubated patients rather than those with prolonged invasive mechanical ventilation. Third, we excluded patients with high BMI due to difficulty of imaging the diaphragm, and also patients with exacerbation of obstructive lung disease due to flattening of their diaphragm resulting from pulmonary hyperinflation. This may affect the generalizability of our results as these are common comorbidities in the ICU population. Fourth, images from multiple breaths were obtained once in each participant. The use of a single ultrasonographer may imply that if this method were to be translated to clinical practice more variability from different observers may be introduced. However, the reproducibility of diaphragm excursion measurements via ultrasound has already been substantiated in a large study [8]. Consequently, we reasoned that imaging performed by multiple observers was deemed unnecessary in this study. Finally, TV was derived from the flow tracings but under different gas conditions (body temperature, pressure, water vapor saturated in ventilated patients and ambient temperature and pressure in healthy subjects); this will not affect the primary conclusion and between-subject variability but may result in a slightly higher absolute ratio (mL/cm) for healthy volunteers compared to ventilated patients.

Adequate diaphragm imaging is pivotal to establish the TV-diaphragm excursion ratio in the critical care setting. M-mode ultrasound measures unidimensional diaphragm movement and requires diaphragm motion perfectly aligned with the M-mode line. Even then, commonly employed one-dimensional measures of diaphragm excursion cannot capture the complete diaphragm motion, which is multidimensional. Our study emphasizes the complexity of the resultant relationship between a single measurement of diaphragm excursion and TV. Hence, a one-dimensional measure is unsuitable to determine absolute values or a safe cutoff for TV. Advanced techniques such as speckle tracking may have superior performance by quantifying diaphragm motion in multiple dimensions [15, 16]. Prior studies were often hampered by the application of inspiratory pressure support during ultrasound measurements. However, it should be stressed that such measurements of excursion should be performed in patients without inspiratory ventilator support [12, 17] to reliably reflect the patient’s own contribution to generating TV, such as done in our study. The relationship between diaphragm excursion as measured with speckle tracking and TV and its possible role in predicting the need for intubation in non-intubated patients with acute respiratory failure requires further study.