Ultrasound utilization in the diagnosis of diaphragm dysfunction compared to other modalities: A retrospective study

Introduction

Diaphragm is a dome shaped fibro-muscular assembly lies at the bottom of thoracic cavity and composed of an aponeurotic tendinous central ligament surrounded by peripheral muscle fibers. Diaphragm is divided into left and right hemi-diaphragms, each with different vascular and nerve supplies.[1-5] It is innervated by two phrenic nerves originating from cervical nerve roots C3 to C5. It is a physical barrier that separates the thoracic cavity from the abdominal cavity and functions primarily involuntarily with additional voluntary control when needed. It, therefore, is considered as principal muscle of ventilation. During inspiration, it contracts along with the accessory respiratory muscles resulting in expanding thoracic cavity, decreasing intrathoracic pressure and drawing air into the lungs. With relaxation of diaphragm, predominates elastic recoiling of lungs, causes exhalation. In addition, the diaphragm also assists in emesis, urination, and defecation by increasing intra-abdominal pressure and helps prevent gastroesophageal reflux by exerting external pressure at the esophageal hiatus.[6,7]

In young children, especially in neonates, the accessory respiratory muscles are often inadequate to compensate for a failing diaphragm. This leads to respiratory muscle fatigue with poor lung expansion and variable degree of potential atelectasis and ultimately result in respiratory failure. However, in adults and older children, the accessory respiratory muscles can often compensate for the paretic or paralyzed diaphragm.[8,9] Cardiac surgery done for underlying congenital heart diseases is the most common cause of abnormal diaphragmatic motion with a prevalence ranging from 0.3 to 12.8%.[10]

Based on symptoms like unexplained dyspnea, limitation to exercise and pain in shoulder, the primary utilized diagnostic tools include physical examination, laboratory investigation, and imaging modalities. Physical examination include diaphragm excursion, which involves percussion along the posterior chest to determine the displacement range of diaphragm during deep inspiration and deep expiration. Normal diaphragmatic excursion is 5–6 cm. Diminished diaphragmatic excursion is often associated with weakness of diaphragm or its paralysis.[7,11] Laboratory investigations of arterial/venous blood gases (BG’s) measure dissolved gases in and other properties of blood (pH, pCO₂, pO₂, base excess, O₂ saturation, etc.) and are most often performed on patients in critical care settings. BG’s are an indication of ventilation, gas exchange and acid-base status of blood, where blood is collected either from an arterial or venous blood supply.[12-16]

Chest radiographic X-ray provides an image, which allows physicians to perceive structure and morphology of diaphragm. It clearly demonstrates the elevation of diaphragm. In normal individuals, the left hemi-diaphragm is usually located one intercostal space lower than the right hemi-diaphragm. Slight elevation of right hemi-diaphragm is associated with the presence of liver under it. If a hemi-diaphragm is weak, then the normal negative intra-pleural pressure pulls the diaphragm cranially into the thoracic cavity. Consequently, the paralyzed diaphragm appears at a higher level. If the right side is paralyzed, the distance between the right and left diaphragm will be more than two intercostal spaces, and if the left side is paralyzed, both the hemi-diaphragms will appear on the same level [Figure 1].[17] In bilateral weakness, both hemi-diaphragms will appear at a higher level and might be missed on routine chest radiographs.[5]

Close

Figure 1

Chest radiograph showing: (a) Normal diaphragm appearance, (b and c) elevation of left and right hemi-diaphragm; respectively, following phrenic nerve paralysis in patients with post-cardiac surgery

Export to PPT

Diaphragm ultrasonography is a non-invasive, portable mode of imaging that eliminates the exposure to radiation and risk of transportation. It is widely used particularly in intensive care unit (ICU) where intense patient cooperation is not essential. Two ultrasound methods are characteristically used to assess diaphragmatic functions. First, the analysis of the dome excursion with M-mode approach which is well-tolerated test with a linear relationship between diaphragmatic movement and inspired volume and allows quantitative and qualitative assessment of diaphragmatic movement [Figure 2]. Second is the evaluation of diaphragmatic thickness and thickening during inspiration by analyzing the apposition zone.[1,18]

Close

Figure 2

M-mode ultrasound showing initial and follow-up ultrasound in a 4-month-old infant having post-cardiac surgery with left diaphragmatic paralysis. (a) Normal right hemi-diaphragm with inspiratory peak above the baseline. (b) First ultrasound shows a flat line indicative of absent motion of the left hemi-diaphragm. (c) Follow-up ultrasound shows some movement of the left hemi-diaphragm during inspiration. (d) Complete recovery of left diaphragmatic paralysis is seen with return of normal inspiratory peak

Export to PPT

The aim of this paper was to retrospectively study and to evaluate the accuracy of departmental ultrasound in assessment of diaphragm motion and to compare its results with other institutional parameter, namely, physical examination, laboratory investigation, and chest radiograph findings. Although, other institution parameters provide sufficient information to diagnose impaired diaphragm motion, ultrasonography is considered relatively sensitive tool. The final objective of this study was to support the diagnostic superiority and reliable of ultrasonography.

Materials and Methods

Results

From a total of 163 pediatric patients under the age of 14 years having clinical suspicion of diaphragmatic dysfunction, a total of 131 met clinical inclusion criteria for our study. The mean age was 1.6 (SD 2.6) years and there were 44.3% of male and 55.7% of female patients. A summary of demographic and baseline characteristics detailed in Table 1.

Table 1 Demographic and baseline characteristics of patients investigated for diaphragmatic dysfunction (n=131). Data shown are frequencies; n (%) or mean±SD

Table 2 summarizes the results of ultrasound, clinical suspicion, laboratory investigation and chest radiograph of right side hemi-diaphragm. There were 102 reports classified as normal by clinical diagnosis, out of which 45.1% of reports were correctly identified as normal by clinical suspicion, 9.8% were identified as paresis by clinical suspicion, and 45.1% were classified as paralysis. Out of 11 identified paralysis cases by clinical diagnosis, only one report (9.1%) was classified as paresis, whereas 10 (90.9%) reports were correctly classified as paralysis by clinical suspicion. Out of 18 clinically diagnosed paresis cases, 15 (83.3%) were identified as paralysis by clinical suspicion, whereas 3 (16.7%) reports were correctly identified as paresis. The results identified by clinical suspicion and clinical diagnosis were significantly different (P < 0.0001). Similar trend was seen for ultrasound and clinical diagnosis reports. The chest radiograph and laboratory investigation reports were not significantly different with clinical diagnosis reports.

Table 2 Results of comparing right hemi-diaphragm dysfunction diagnosis by clinical suspicion, laboratory investigation, ultrasound, and X-ray. Data shown are frequencies; n (%) and Fisher’s exact test calculated P-values

Table 3 summarizes results of ultrasound, clinical suspicion, laboratory investigation, and chest radiograph of the left side hemi-diaphragm. Out of 90 clinically diagnosed normal reports, 35 (38.9%) reports were correctly classified as normal by clinical suspicion, 84 (93.3%) by ultrasound, and only 26 (28.9%) by X-ray. 45 (50.0%) reports were identified as paralysis by clinical suspicion and X-ray, and only one (1.1%) by ultrasound. The results obtained by clinical suspicion and ultrasounds were significantly different from clinical diagnosis. The chest radiograph and laboratory investigation reports were not significantly different with clinical diagnosis.

Table 3 Results of comparing left hemi-diaphragm dysfunction diagnosis by clinical suspicion, laboratory investigation, ultrasound, and X-ray. Data shown are frequencies; n (%) and Fisher’s exact test calculated P-values

Table 4 summarizes the classification results of ultrasound, clinical suspicion, and X-ray in predicting clinical diagnosis on the right hemi-diaphragm. The ultrasound yielded 96.6% sensitivity and 90.3% PPVs, 99% NPV, and 97.1% specificity. The ultrasound results produced perfect classification for discriminating between clinical diagnosis results with accuracy of 96.95% and precision of 90.32%. The sensitivity and PPV for clinical suspicion were 100% and 34.1%, whereas, only 27.6% of sensitivity and 29.6% of PPV achieved by chest radiograph. The specificity and NPV were 45.1% and 100% for clinical suspicion, 81.4% and 79.8% for chest radiograph, respectively. The clinical suspicion and chest radiograph results produced poor classification for discriminating between clinical diagnosis results with accuracy of 57.25% and 69.47% and precision of 34.12% and 29.63%, respectively.

Table 4 Comparative results of the right hemi-diaphragm dysfunction diagnosis by clinical suspicion, ultrasound, and X-ray

Table 5 depicts the classification results of ultrasound, clinical suspicion, and chest radiograph in predicting clinical diagnosis on the left hemi-diaphragm. The ultrasound showed 90.2% sensitivity and 86.0% PPV, 95.5% NPV, and 93.3% specificity. Results showed little high variability in comparison to classification results obtained for the right hemi-diaphragm. The ultrasound results produced good classification for discriminating between clinical diagnosis results with accuracy of 92.37% and precision of 86.05%. The sensitivity and PPV for clinical suspicion were 95.1% and 41.5%, whereas 73.2% of sensitivity and 31.9% of PPV were achieved by chest radiograph. The specificity and NPV were 38.9% and 94.6% for clinical suspicion, and only 28.9% and 70.3% for chest radiograph, respectively. The clinical suspicion and chest radiograph results produced poor classification for discriminating between clinical diagnosis results with accuracy of 56.49% and 42.75% and precision of 41.49% and 31.91%, respectively.

Table 5 Comparative results of the left hemi-diaphragm dysfunction diagnosis by clinical suspicion, ultrasound, and X-ray

Figure 3 summarizes the accuracy results of clinical suspicion, ultrasound, and chest radiograph with clinical diagnosis performed on the left and right side hemi-diaphragm.

Close

Figure 3

Summary of accuracy results of clinical suspicion, ultrasound, and chest radiograph with clinical diagnosis performed on the left and right side hemi-diaphragm.

Export to PPT

Discussion

The term, diaphragmatic dysfunction includes eventration, paresis and paralysis.[6] Eventration is a permanent elevation of entire or part of the hemi-diaphragm caused by thinning.[6,7,26,27] Diaphragmatic paresis would be the partial loss of muscle strength to generate necessary pressure for adequate ventilation.[26,28] While paralysis means the total absence of this capacity. Diaphragmatic paralysis can arise from either weakness of the muscle itself or damage to its nerve supply. Depending on the severity of the paralysis and whether it is unilateral or bilateral, patients can have varied clinical manifestations such as dyspnea, intolerance to exercise, sleep disturbances, hypersomnia, and with a potential impact on survival.[4,5] A patient may be asymptomatic, often diagnosed during investigation of unexplained dyspnea or, occasionally, after the casual finding of a diaphragmatic elevation in an imaging radiograph performed for another purpose for unrelated aliment, while another may be ventilator dependent.[5,29]

Diaphragmatic dysfunction is still a matter of concern after cardiothoracic surgery, especially among young children and neonates. The prevalence of diaphragmatic dysfunction after cardiothoracic surgery in children varies from 0.3 to 12.8%.[10] Timely diagnosis of abnormal diaphragmatic motion is essential for patient in immediate post-operative period to minimize potential risk and to ascertain appropriate therapeutic approach. The diagnosis of diaphragm dysfunction can be done by laboratory investigation, diaphragmatic ultrasonography and chest radiograph.

Our retrospective study demonstrates the accuracy of ultrasonography as 96.95% on right side and 92.37% on the left side in diagnosing diaphragmatic dysfunction. Ultrasonography proved to be highly sensitive and specific diagnostic tool over chest radiograph and clinical suspicion approach. The second-best results were obtained by chest radiograph and prove to be more efficient in diagnosis of right hemi-diaphragm dysfunction. Slightly less accurate results were obtained by clinical suspicion as compared to chest radiograph.

Diaphragmatic ultrasound has some limitations. First, ultrasound systems have inherent resolution limits (usually 0.1 mm). In addition, the assessment of the left hemidiaphragm can be problematic. However, taking extra precautions during the diaphragmatic ultrasound examination (e.g., placing the patient in the supine position and rotating the transducer) can help overcome these limitations. Furthermore, because ultrasound is an operator-dependent examination, repeated training can improve accuracy. Moreover, although diaphragmatic ultrasound has been shown to have a steep learning curve when applied in healthy subjects, few studies have evaluated how to develop the appropriate skills. One study, involving a pediatric population, found that 4 h of hands-on diaphragmatic ultrasound training focused on the recognition of normal and abnormal diaphragmatic motion resulted in high concordance between the diaphragmatic ultrasound findings reported by a trainee and those reported by a pediatric intensivist.[30,31]

Conclusion

Diaphragmatic ultrasound appears to be a valid and reliable diagnostic tool for diagnosis of abnormal diaphragmatic motion. At present, the more emphasis should be given to use of ultrasonography due to its non-invasive and non-radiating characteristics, and convenient portability that eliminates risk of transportation. Thus, ultrasound assessment should be part of the diagnostic checklist, particularly in ICU where patient’s cooperation is not essential. Diaphragm ultrasound may act as an imaging tool guiding rehabilitation success in diaphragmatic dysfunction cases.

Authors Declaration Statements