Chapter 10
The diaphragm
Giovanni Ferrari1, Søren Helbo Skaarup2, Francesco Panero3 and John M. Wrightson4
1Pulmonary Medicine, Ospedale Mauriziano Umberto I, Torino, Italy. 2Dept of Pulmonary Medicine and Allergy, Aarhus University Hospital, Aarhus, Denmark. 3High Dependency Unit, San Giovanni Bosco Hospital, Torino, Italy. 4Oxford Centre for Respiratory Medicine, Churchill Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, UK.
Correspondence: Giovanni Ferrari, Pneumologia, Ospedale Mauriziano Umberto I, Largo Turati 62, 10128 Torino, Italy. E-mail: giovanniferrarister@gmail.com
This chapter covers US evaluation of diaphragm function, and describes the methods and US techniques used to measure movement and thickening of the diaphragm. Current validated techniques that assess diaphragm function are often invasive or expose the patient to ionising radiation. US is a noninvasive, easily repeatable bedside tool allowing direct visualisation of the muscle. There are a number of applications of DUS, from assessment and emergency medicine to respiratory medicine. DUS has become a useful tool in evaluating diaphragm dysfunction, suggesting its use to predict discontinuation from mechanical ventilation and to evaluate muscle atrophy during mechanical ventilation. It also has an important role in assessing respiratory effort, and in evaluating diaphragm paralysis and the mechanical consequences of pleural effusion. Thus, DUS is a fast, easy procedure that can identify diaphragm dysfunction, monitor muscle activity over time and provide useful information during invasive procedures such as thoracentesis.
Cite as: Ferrari G, Helbo Skaarup S, Panero F, et al. The diaphragm. In: Laursen CB, Rahman NM, Volpicelli G, eds. Thoracic Ultrasound (ERS Monograph). Sheffield, European Respiratory Society, 2018; pp. 129–147 [https://doi.org/10.1183/2312508X.10006917].
The diaphragm represents the main respiratory muscle but prior to the widespread use of US remained challenging to assess. This is of relevance given that diaphragm dysfunction is frequent in the critically ill, with patients developing a degree of muscular atrophy and weakness secondary to mechanical ventilation [1].
Ventilation-induced diaphragm dysfunction (VIDD) portends complications such as patient–ventilator asynchrony, weaning failure, pneumonia, prolonged intubation and a stay in the intensive care unit (ICU). Therefore, its recognition and grading is of significant importance. DUS has emerged as a valuable tool in such critically ill patients. By evaluating the movement and thickening of the diaphragm, clinically relevant information can be obtained for weaning trials, monitoring of respiratory efforts, investigating diaphragmatic paralysis and defining acute dyspnoea. Within a broader arena, DUS also allows assessment of suspected diaphragmatic paralysis, detection of structural abnormalities and prediction of probable pleural malignancy. A number of studies have demonstrated its easy learning, high reproducibility and good correlation with other validated measures.
This chapter provides a summary of the pathology affecting the diaphragm and proposes methods for assessing diaphragm function and structure using DUS. We highlight specific examples of DUS use, particularly where there is an evidence base.
DUS assessment in illness
Many disorders affect the structure and function of the diaphragm, including those affecting its innervation (central and peripheral nervous system and neuromuscular junction) and the muscular and tendinous diaphragm itself, and other disorders causing mechanical disruption of the diaphragm (including pleural and pulmonary causes) (table 1). Identification and quantification of the effects of these disorders on the diaphragm is of relevance to physicians from a wide range of specialities, including pulmonologists, intensivists, neurologists, spinal physicians, radiologists and surgeons.
Table 1. Disorders that affect the diaphragm and their site of effect
Site of effect | Disorder |
|---|---|
Neurological: brain and spinal cord | Arnold–Chiari malformation |
Multiple sclerosis | |
Motor neurone disease | |
Poliomyelitis | |
Quadriplegia | |
Spinal muscular atrophy | |
Stroke | |
Syringomyelia | |
Neurological: phrenic nerve | Charcot–Marie–Tooth disease |
Chronic inflammatory demyelinating polyneuropathy | |
Critical illness polyneuropathy | |
Idiopathic | |
Guillain–Barré syndrome | |
Neuralgic neuropathy | |
Tumour compression | |
Neuromuscular junction | Botulism |
Drugs/organophosphates | |
Lambert–Eaton syndrome | |
Myasthenia gravis | |
Muscle | Abdominal pathology (including ascites) |
Acid maltase deficiency | |
Disuse atrophy, ventilator-induced diaphragm dysfunction | |
Eventration | |
Hernia (e.g. Bochdalek, hiatus, Morgagni) | |
Glucocorticoids | |
Lung hyperinflation (COPD/asthma) | |
Lung tumour/other restrictive parenchymal disorder | |
Muscular dystrophies | |
Myositis (infectious, inflammatory, metabolic) | |
Pleural thickening/fluid | |
Tumour invasion |
Diaphragm functional assessment relied previously on cumbersome fluoroscopic or phrenic nerve stimulation studies, while anatomical assessment relied on cross-sectional imaging. Using US, both functional and structural assessment is possible in a multitude of settings [2–9]. These include: 1) the clinic room for assessment of basal opacity, where DUS enables differentiation of diaphragmatic paralysis, consolidation, subpulmonic effusion, diaphragmatic nodularity/thickening, diaphragmatic hernia and abdominal pathology, 2) the procedural room for assessment of diaphragm position prior to pleural puncture, where diaphragmatic inversion may predict a symptomatic benefit of pleural fluid drainage, 3) the emergency room for assessment of the diaphragm as part of rapid US assessment after trauma and of medical patients with acute dyspnoea, and 4) the ICU for assessment of basal opacities, where assessment of diaphragmatic function (particularly over time) is of relevance in weaning from mechanical ventilation.
DUS technique
The diaphragm is the most important muscle in breathing. Contraction of the diaphragm causes it to move in a craniocaudal direction, thereby increasing intrathoracic volume and decreasing intrathoracic pressure. This results in inspiratory airflow into the lungs. The intercostal and pectoral musculature also contributes to inspiration but to a lesser extent during tidal breathing. Sufficient diaphragmatic function is considered vital [5]. During exhalation, the diaphragm relaxes and, due to elastic recoil of the lungs, the diaphragm is passively drawn cranially to its resting position. In a forced exhalation manoeuvre, the diaphragm is relaxed, and contraction of internal intercostal and abdominal muscles leads to rapid exhalatory airflow.
Assessment of both diaphragm excursion and diaphragm thickening may be used as tools for functional assessment.
Evaluation of diaphragm excursion
M-mode
The time motion mode (M-mode) may be used to measure the excursion of the diaphragm. While there is no consensus on how the M-mode should be performed, most authors suggest the use of a curvilinear low-frequency transducer placed in the midclavicular line and angled in a cranial direction (figure 1). The M-mode line is placed at the posterior part of the diaphragm where there is maximal movement and excursion.
Figure 1. US evaluation of diaphragm movement with a curvilinear transducer in the sagittal plane in the midclavicular line (red line). The transducer is angled cranially and uses the liver as an acoustic window.
On the right side, the liver acts as an acoustic window, and the diaphragm is easily identified as a hyperechoic curved line abutting the liver (figure 2). Left hemidiaphragm imaging is more difficult due to the poor acoustic window of the spleen and the gas content of the stomach.
Figure 2. The right hemidiaphragm is identified as a curved hyperechoic structure posterior to the liver. The M-mode line (white line) is placed to measure maximal excursion (bottom panel).
Many studies describing the M-mode technique only assess the right hemidiaphragm [10, 11], while others note a lower success evaluating the left [12]. The probe should be placed in the subcostal area, between the midclavicular and the anterior axillary line, directing the US beam medially, dorsally and cranially [12]. The angle of incidence of the US beam should be perpendicular to the direction of the muscle movement [13]. As the patient inhales, the abdominal wall is pushed forward, potentially displacing the transducer, and this may result in an excursion measurement error [14]. It is therefore important to stabilise the transducer during the examination.
B-mode
Another method of quantifying diaphragm excursion is to use brightness mode (B-mode) to measure craniocaudal movement during breathing. This method is simple and can be used at both hemidiaphragms [15–17]. A curvilinear low-frequency transducer is placed perpendicularly in a lower intercostal space (ICS) between the mid- and posterior axillary line using the liver or spleen as an acoustic window. Other authors have proposed that the craniocaudal movement of the portal vein can also be measured as a surrogate for diaphragm excursion [18].
Other methods
New methods to assess diaphragm movement, such as speckle tracking, are under investigation but require further validation [19–21]. The rate of diaphragm excursion has also been studied [22].
Evaluation of diaphragm thickness and thickening
Diaphragm muscular thickness increases during contraction, and US evaluation of this provides further information about diaphragm function. A high-frequency linear transducer (7–12 MHz) is used to visualise the zone of apposition (where the diaphragm abuts the chest wall laterally), located a few centimetres deep [23]. The transducer can be placed across two ribs and then rotated obliquely to fit in the ICS for accurate sonographic identification of the diaphragm. Lower intercostal areas are examined in the anterior axillary line (figure 3). On inspiration, movement of the lung may prevent imaging of the diaphragm; when this occurs, the probe should be moved to a more inferior ICS.
Figure 3. US evaluation of diaphragm thickness and thickening during breathing. A high-frequency transducer is placed at the anterior axillary line (red line) identifying the diaphragm at the zone of apposition superficial to the liver or spleen.
During US, the diaphragm appears as a hypoechoic structure superior to the liver or spleen, flanked by two thin hyperechoic lines, which are the diaphragmatic pleura (superficially) and the peritoneum (deeper) (figure 4). The minimal thickness in the resting diaphragm is 12–15 mm [24, 25]. Accurate assessment of thickness requires that the US beam is perpendicular to the diaphragm, rather than oblique [26].
Figure 4. The diaphragm identified as a three-layered structure superficial to the liver or spleen. Hyperechoic layers of the diaphragmatic pleura (thoracic side) and the peritoneum (abdominal side) surround the hypoechoic diaphragm. Thickness is measured with the calliper function in maximal inhalation and exhalation (indicated by crosses and a dotted line). “1” indicates the diaphragmatic pleura.
An increase in diaphragm thickness during breathing is measured in either B-mode or M-mode using the calliper function in end-expiration (functional residual capacity (FRC)) and end-inspiration during deep breathing (total lung capacity (TLC)). At least three measurements should be recorded to minimise errors. Several derived parameters may be calculated, which relate thickness TLC to thickness at FRC [25, 26]. These include: 1) thickening ratio=tdiTLC/tdiFRC, where tdiTLC is the diaphragm thickness at TLC and tdiFRC is the diaphragm thickness at FRC, and 2) diaphragm thickening fraction (Dtf)=(tdiTLC−tdiFRC)/tdiFRC.
Standardisation and validity of DUS
Despite the previously discussed techniques of functional assessment using DUS, a clear consensus on the optimal technique has not been established. Normal values of diaphragm excursion and thickness and thickening fractions have, however, been established with good inter- and intra-observer variation. Both excursion and thickness have also been compared with non-US references, including spirometric measurements of exhaled air volumes.
Thickness
Measurement of diaphragm thickness by US demonstrates close correlation with post-mortem findings [27]. Normal values of diaphragm thickness were examined by CARRILLO-ESPER et al. [28] in end-expiration in 109 healthy men and women, finding a normal resting diaphragm of 18 mm (men) and 14 mm (women). No relationship was found between thickness and body mass index or thorax circumference [24, 28]. Diaphragm thickness should increase by at least 20% at maximal inspiration, with a minimal side-to-side variation [24]. Thickening during breathing is related to spirometric lung volumes [29]; however, during quiet breathing there is considerable variability, and up to one-third of healthy people show no increase in thickening during tidal breathing [25]. Intra- and inter-rater agreement is high [30], including during mechanical ventilation (although more so on the right than on the left) [31].
Excursion
Evaluation of excursion during breathing has mostly been undertaken using M-mode in a subcostal midclavicular view. There is a linear relationship between M-mode excursion and exhaled lung volume measured by spirometry [10, 11, 17]. A linear relationship has also been found using other US methods that measure craniocaudal movement of the diaphragm or the portal vein [18]. Movement is correlated to sex and weight [16]. Normal values for men and women have been proposed as 1.0 and 0.9 cm during quiet breathing, 1.8 and 1.6 cm during a sniff manoeuvre, and 4.7 and 3.6 cm during deep breathing [12]. Another study reported similar findings [32].
Examination is feasible and relatively fast, even by inexperienced operators [22]. Reproducibility is high but dependent on operator experience [22, 12]. Due to difficulty with sonographic access, reproducibility is lower for the left hemidiaphragm [17, 12].
When comparing DUS with chest radiography, US can detect poor diaphragm movement even with a normal radiographic diaphragm position, and, conversely, the finding of an elevated hemidiaphragm does not always correspond to decreased movement at US examination [33].
DUS can therefore be seen to provide rapid and reproducible information about diaphragm function that cannot be obtained as easily and safely by other diagnostic modalities. Despite the lack of consensus-based standardised methodology, it seems likely that DUS is set to become the primary tool for evaluation of diaphragm function.
Monitoring diaphragm thickness during mechanical ventilation
Mechanical ventilation is a lifesaving procedure, but prolonged mechanical ventilation (PMV) is associated with severe complications, such as tracheal injury, infection and ventilator-associated pneumonia. Moreover, PMV promotes diaphragm atrophy and contractile dysfunction [34], termed VIDD [35]. Although many patients on mechanical ventilation can be weaned early, difficulties in weaning are encountered in ∼30% of patients, with >40% of the time spent in the ICU being spent weaning from mechanical ventilation [36]. During mechanical ventilation, the ventilator provides part or all of the work of breathing for the patient, according to the ventilator modality employed. Several studies have shown that, during controlled mechanical ventilation (CMV), diaphragm atrophy has a rapid onset, while other muscles show no atrophy [37, 38].
DUS is a promising method to evaluate the diaphragm during mechanical ventilation. Experimental data suggest that a reduction in diaphragm thickness over time can indicate atrophy of the muscle itself. An early assessment of diaphragm dysfunction prior to extubation may be important to avoid the risk of extubation failure [39].
Assuming that mechanical ventilation has an unloading effect on respiratory muscles that leads to diaphragm atrophy and dysfunction, GROSU et al. [40] measured muscle thickness in seven intubated patients from the first day of intubation until extubation, tracheostomy or death. They found a daily average of 6% decrease in diaphragm thickness and that thinning of the diaphragm occurred within 48 h of intubation.
SCHEPENS et al. [41] studied right hemidiaphragm thickness in 53 patients ventilated for at least 72 h. The mean baseline value was 1.9 cm; compared with baseline, diaphragm thickness decreased >10% in 77% of the patients studied, remained stable in 19% and increased in 4%. The daily rate of thickness reduction was 10.9%. The main variable associated with diaphragm thinning was length of ventilation, while disease severity (assessed by the Simplified Acute Physiology Score II), steroid use, muscle relaxants or sepsis on admission were not associated with a decrease in thickness.
In a small study of eight patients receiving either pressure support ventilation (PSV) or assist-control ventilation (ACV), a decline in diaphragm thickness was observed only in patients receiving ACV, while subjects receiving PSV showed an increase in thickness [42]. Similarly, EL-MORSY et al. [43] studied diaphragm thickness in 67 mechanically ventilated patients, stratified by ventilatory modalities (CMV, ACV and spontaneous ventilation). CMV and ACV were associated with significantly higher rates of diaphragmatic atrophy than spontaneous modes. A daily decrease in thickness was observed in patients treated with CMV and ACV, while in spontaneous modalities a daily increase in thickness was observed [43].
ZAMBON et al. [44] measured daily atrophy rates in critically ill mechanically ventilated patients, categorising the subjects into four classes: 1) spontaneous breathing or continuous positive airway pressure, 2) PSV of 5–12 cmH2O (low PSV), 3) PSV of >12 cmH2O (high PSV) and 4) CMV. The authors reported, for the first time, that the degree of diaphragm atrophy was associated with different ventilation settings and observed a linear relationship between ventilation support and Dtf.
GOLIGHER et al. [45] described the evolution of diaphragm thickness over time in patients during mechanical ventilation. A total of 128 mechanically ventilated patients and 10 control patients were enrolled. Thickness and Dtf were measured daily. Diaphragm thickness remained unchanged in 47 subjects (44%), decreased by >10% in 47 subjects (44%) and increased by >10% in 13 subjects (12%). Changes in thickness occurred early. Thickness was stable over time in nonventilated control subjects and in patients following extubation [45].
To evaluate the impact of mechanical ventilation on the diaphragm, a recent study by GOLIGHER et al. [46] evaluated whether ventilator-induced changes in diaphragm thickness were associated with clinical outcomes, such as days of ventilation, re-intubation, tracheostomy or death. Both increases and decreases in diaphragm thickness and Dtf were assessed. The risk of PMV was increased with either a decrease or an increase in diaphragm thickness compared with baseline values [46]. Such findings suggest that atrophy is associated with PMV and also that insufficient muscle unloading (suggested by an increase in thickness) may be associated with diaphragmatic injury. The authors hypothesised that an intermediate Dtf could be associated with the shortest duration of ventilation. In a post-hoc analysis, the length of ventilation was lower in patients with an intermediate Dtf (mean value of 15–30%) over the first 3 days of ventilation. ICU length of stay and complications were lower in this group. Patients with Dtf values similar to normal subjects (i.e. 15–30%) during the first 3 days of ventilation had a shorter duration of mechanical ventilation, while subjects with lower or higher values of Dtf had a longer duration of mechanical ventilation and a higher risk of ventilation [46].
In conclusion, the findings of these studies suggest that changes in diaphragm thickness are common in mechanically ventilated patients, occur early and may be modulated by the intensity of respiratory muscle work performed by the patient [45]. The degree of atrophy may be associated with the duration of mechanical ventilation and not with other risk factors for muscle atrophy, such as sepsis.
Weaning from mechanical ventilation
Diaphragm dysfunction is frequent in critically ill patients either as a result of VIDD or due to direct effects of cardiac or abdominal surgery [47]. An early assessment of diaphragm dysfunction prior to weaning should be considered to avoid the risk of extubation failure. Previously, bedside assessment of diaphragm dysfunction was challenging, and the gold standard of measurement of transdiaphragmatic pressure (PTD) was usually limited to research studies [48].
The role of DUS in predicting weaning success has been investigated by studies measuring either excursion [39, 49, 50] or thickness and Dtf [50–52]. Both diaphragm excursion and Dtf measurements performed during a spontaneous breathing trial (SBT) in intubated/tracheotomised patients have shown utility in predicting weaning success.
JIANG et al. [49] assessed diaphragm excursion in mechanically ventilated patients by evaluating displacement of the liver or spleen. A threshold of 1.1 cm had a better performance in predicting weaning success compared with traditional parameters (maximal inspiratory pressure (MIP), rapid shallow breathing index (RSBI) and expiratory tidal volume) with a sensitivity and specificity of 84% and 83%, respectively, a positive predictive value (PPV) of 82%, a negative predictive value (NPV) of 86% and an accuracy of 84%.
KIM et al. [39] evaluated diaphragm dysfunction in 82 patients during a SBT by assessing vertical excursion or paradoxical movements of the diaphragm. A cut-off value of 1.0 cm was found, and patients with diaphragm dysfunction according to this criterion had longer weaning times and a higher frequency of re-intubation [39].
In a recent study, FARGHALY and HASAN [50] observed that a threshold value of diaphragmatic excursion of ≥10.5 mm was associated with successful extubation with a sensitivity and specificity of 87% and 71%, respectively.
Another recent study by SPADARO et al. [53] compared the role of a new index, the diaphragmatic-RSBI (D-RSBI), as opposed to RSBI, to predict weaning failure. Diaphragm displacement and D-RSBI performed better than traditional RSBI at predicting weaning outcome, and D-RSBI had the best diagnostic accuracy with a cut-off value of >1.3 breaths min−1 mm−1, giving a sensitivity of 94%, specificity of 65%, PPV of 57% and NPV of 96%.
Excursion as a weaning index should be assessed only during a SBT and not in patients receiving mechanical ventilation (which would give a false assessment of contractile activity). Moreover, variables such as abdominal or thoracic compliance, tidal volume, rib cage or abdominal muscle activity, and the presence of ascites may affect diaphragm motion and excursion assessment.
Several studies have evaluated the role of Dtf in predicting extubation success. DININO et al. [51] studied 63 patients weaned either by a SBT (27 patients) or a pressure support trial (36 patients). A Dtf cut-off value of 30% was associated with a PPV of 91% and NPV of 63% for successful extubation with a sensitivity and specificity of 88% and 71%, respectively. FERRARI et al. [52] evaluated Dtf in 46 tracheotomised patients during a SBT. A threshold Dtf of >36% was associated with a successful SBT with a sensitivity of 82%, specificity of 88%, PPV of 92% and NPV of 75%. Moreover, a significant difference between diaphragm thickness at TLC and residual volume was observed between patients who succeeded and those who failed the SBT. Finally, FARGHALY and HASAN [50] assessed Dtf percentage in 54 patients with a cut-of value of 34.2%, giving a sensitivity and specificity of 90% and 64%, respectively. In these three studies, the assessment of Dtf performed similarly to other weaning indexes, such as RBSI or MIP measurement.
Thus, according to these preliminary studies, the optimal threshold values for predicting weaning success are 1 cm for diaphragm excursion and 30–36% for Dtf. DUS provides important information for the weaning process. In particular, Dtf performs comparably to other weaning indexes, such as RSBI or MIP. Further randomised controlled studies are warranted to validate the role of DUS in the weaning process.
Respiratory effort monitoring
Validated methods of assessment of diaphragmatic function rely on direct measurement of patient-generated pressures: maximum inspiratory/expiratory pressure, airway pressure decrease at 100 ms after onset of inspiration (P0.1) and oesophageal pressure (Poes). PTD, calculated as the difference between Poes and gastric pressure (Pga), obtained from double-balloon catheters, together with twitch magnetic phrenic nerve stimulation, represent the gold standards [54]. Nevertheless, all of these suffer from limitations: they all need a discrete grade of collaboration from the patient, while the latter two are both invasive and potentially distressing [48].
DUS has been studied as a proxy of respiratory effort and work of breathing in healthy volunteers [17, 55] and in people suffering from neuromuscular diseases [56, 57]. DUS offers many advantages over other measures, being a noninvasive, repeatable bedside technique allowing direct visualisation of the diaphragm and discrimination of the dysfunctional side. Few studies have compared DUS with other methods in mechanically ventilated patients in terms of either movement of one hemidiaphragm or thickness change [31, 58–60].
LEROLLE et al. [59] compared the best diaphragmatic excursion on maximal inspiratory effort from FRC with both PTD and the derived Gilbert index (GI). GI evaluates the contribution of the diaphragm to respiratory pressure swing during quiet tidal breathing: a GI value of >0.30 is normal, while a GI value of ≤0 indicates severe diaphragmatic dysfunction [61]. The authors evaluated a cohort of 28 post-cardiac surgery patients requiring prolonged invasive mechanical ventilation, who were then at risk of diaphragmatic dysfunction [59]. All but one patient exhibited an abnormal PTD. The best diaphragmatic excursion and GI value correlated significantly (Spearman’s ρ=0.64; p=0.001), with lower values of excursion among those with severe muscle dysfunction (GI >0 versus GI ≤0, 30 versus 19 mm; p=0.007). Discrimination capability (assessed with a receiver operating characteristic curve) for the best diaphragmatic excursion with respect to a severe impairment (GI ≤0) was 93%, with an optimal cut-off point of 25 mm, showing a sensitivity and specificity of 100% and 85%, respectively. When applied to a different cohort of patients with an uncomplicated post-operative course, none of them showed a best diaphragmatic excursion of <25 mm.
VIVIER et al. [58] conducted a study on 12 patients undergoing noninvasive PSV as a weaning protocol after >48 h of invasive ventilation. They compared the Dtf of the right hemidiaphragm with both PTD and the transdiaphragmatic pressure–time product (PTPdi, an estimate of the work of breathing) at four different levels of pressure support (0, 5, 10 and 15 cmH2O, positive end-expiratory pressure of 5 cmH2O for all) during tidal ventilation. Dtf was calculated as the fractional thickening of the muscle at the apposition zone between end-inspiration and end-expiration. With increasing levels of pressure support (progressive muscular unloading), PTPdi and Dtf decreased similarly. Dtf was higher for spontaneously breathing patients (i.e. maximum workload) and lowest for the highest pressure support (i.e. minimum burden) (47.4% versus 16.3%; p<0.05). Dtf correlated well with PTPdi (Spearman’s ρ=0.74; p<0.001) [58]. This study suggests that Dtf is a useful indicator of the diaphragmatic contribution to the work of breathing.
More recently, GOLIGHER et al. [31] evaluated 66 patients receiving mechanical ventilation and nine healthy volunteers. Overall diaphragmatic function, estimated by Dtf, was considerably lower among the former than among the latter (p<0.0001), and was almost negligible in the subgroup undergoing neuromuscular blockage [31].
In 2011, UMBRELLO et al. [60] recruited 25 intubated patients after a major elective surgical intervention, who were deemed suitable for a weaning trial. Sedation and muscular blockage were withheld and PSV commenced. Dtf and the best excursion of the right hemidiaphragm, PTPdi and the respective oesophageal pressure–time product (PTPoes) for the Poes were collected. Unsurprisingly, they found that increasing levels of support were significantly associated with decreasing patient respiratory drive (P0.1) and effort. As in the study by VIVIER et al. [58], Dtf was significantly lower for rising levels of pressure support, being highest when breathing spontaneously (52.7% versus 13.0%; p<0.001). Dtf was significantly correlated with PTPoes (Pearson’s r=0.801; p<0.001) and PTPdi (Pearson’s r=0.701; p<0.001). In the same way, Dtf was positively and strongly associated with P0.1, PTPoes and PTPdi in a linear mixed model of regression analysis (p<0.001 for all). No correlation was found for the best diaphragmatic excursion and other measures.
To conclude, a few studies have investigated the role of DUS as an index of respiratory effort among critically ill patients. Some evidence supports the use of Dtf (rather than peak excursion) when compared with other measures of the work of breathing. Further studies should overcome the limit of small and highly selected cohorts, elucidate the differences between invasive and noninvasive ventilation (NIV), and narrow the wide variability of the data.
DUS assessment of diaphragmatic paralysis
An elevated hemidiaphragm on CXR may suggest diaphragmatic paralysis, but further studies are required to confirm palsy (rather than, for example, diaphragmatic eventration, pleural effusion or subdiaphragmatic pathology) [62]. Traditionally, fluoroscopic evaluation of the diaphragm during tidal breathing, deep breathing and a “sniff” test are undertaken. Such imaging is prone to error and is cumbersome, cannot be performed at the bedside and requires radiation exposure. Other investigation modalities include the measurement of PTD, phrenic nerve conduction studies, diaphragm needle electromyography and dynamic MRI. Similar evaluation is possible using DUS, which provides similar results to fluoroscopy with higher sensitivity and good inter- and intra-observer reproducibility [4, 12, 17, 24, 55, 63]. Sonographic assessment in suspected diaphragmatic paralysis may be achieved by studying both diaphragm inspiratory excursion and thickness.
Diaphragm inspiratory excursion
Both B-mode (two-dimensional) and M-mode may be used to evaluate diaphragm inspiratory excursion. Similar to fluoroscopy, sonographic features of diaphragm paralysis include: 1) an elevated hemidiaphragm, 2) decreased, absent or paradoxical motion during quiet respiration and 3) paradoxical motion under load (e.g. during a sniff manoeuvre) [62].
Figures 5 and 6 demonstrate M-mode analysis of a normally moving right hemidiaphragm (figure 5) with a poorly moving left hemidiaphragm in association with pleural fluid and thickening (figure 6).
