Diagnostic Ability of Chest Ultrasound in Selective Paediatric Pneumonia Alternative to CT scan: A single-center Comparative Observational Study

Introduction: Chest CT is the gold standard method of the diagnostic evaluation of patients with pneumonia. Lung sonography has been lately explored as an alternative modality to decrease radiation hazards. This study aimed to evaluate the diagnostic ability of lung ultrasonography in detecting paediatric pneumonia at presentation and follow up and comparing findings with chest CT scan. Methods: A cross-sectional observational study was performed at a paediatric hospital from August 2019 to April 2021. We studied 106 children (ages from 45 days to 14 years) referred by the paediatrician with clinical data of pneumonia. All children underwent CT chest examination; 90 showed positive, while 16 showed no pneumonia. Ultrasonography was performed on all patients within 24 hours after CT examination. Results: We found that lung ultrasound showed 100% specificity, 82.2 %, sensitivity in the diagnosis of paediatric pneumonia, accuracy was 84.9%, positive predictive value was 100% and negative predictive value was 50%. Their sensitivity and specificity are 100% in complicated pneumonia by parapneumonic effusion, empyema, and abscess formation. Conclusions: Chest ultrasound is a simple technique that can be performed in everyday practice, with high specificity and sensitivity compared to a chest CT scan in diagnosis and follow up of pneumonia in the pediatric age group Abstract


Introduction
Pneumonia in children remains a condition that is challenging to diagnose accurately because the presenting signs and symptoms are nonspecific, might be subtle, and vary, depending on the patient's age, responsible pathogen, and severity of the infection. 1,2 Chest radiography (CXR) is considered the test of choice for diagnosing pneumonia in children. Nevertheless, it is not 100% sensitive nor specific, and variation exists in intra-and inter-observer agreement among radiologists. [3][4][5][6] A chest CT scan is usually considered the ideal gold stander for pneumonia patients with a non-conclusive CXR. However, it is not recommended for routine use because of its high radiation risk, high cost, and not associated improvement in outcomes. 5 Original Article DOI: 103126/JNPS.V4113

LUS versus CT in paediatric pneumonia
Lung ultrasonography (LUS) has been lately explored as an alternative modality to decrease radiation hazards. On lung, ultrasound pneumonia appears as a hypoechoic consolidated area of varying size and shape, with irregular borders. The echotexture can appear homogeneous or inhomogeneous. 7 Lung sliding is reduced or absent. 5 Pneumonia's most common sonographic feature is the air bronchogram, characterized by lens-shaped internal echoes within the hypodense area or echogenic lines and corresponds to air inclusions or air-filled bronchioles and bronchi. Dynamic air bronchograms can be observed. This finding rules out atelectasis. 6 Fluid bronchograms are characterized by anechoic or hypoechoic tubular structures with hyperechoic walls, without perfusion signs inside at color Doppler examination. [5][6][7] Pleural effusion is easily detected in the US and appears as an anechoic area in the pleural space. 8 In paediatric patients, as in adults, lung ultrasound demonstrated a diagnostic accuracy higher or not inferior to CXR. 6,9 This study aimed to evaluate the diagnostic ability of LUS in the detection of paediatric pneumonia at presentation and follow-up and comparing findings with chest CT scan.

Methods
This was a single center, cross-sectional observational study performed at a Paediatric Hospital from October 2019 to April 2021. The study protocol was approved by the local institutional Ethics Committee (no 118/2019). We obtained express informed written consent from parents of all eligible patients accepted to participate in this study following the Declaration of Helsinki. Children with clinical signs and symptoms of pneumonia consisting of fever, cough, and shortness of breath, tachypnea, rib retraction and grunting as well as decreased air entry, fine crepitation and bronchial breathing at the auscultatory examination, in which the suspicion of pneumonia met the WHO criteria 10 (Clinically defined as age-specific tachypnea and chest indrawing) for diagnosis of pneumonia, were included in the study. We excluded children with congenital heart disease or other pulmonary background pathology, cerebellar palsy, thoracic wall malformations, thoracic trauma, obese patient, and bronchopneumonia seen at CT examination. We used a Voulson E6 GE ultrasound machine. Children underwent lung ultrasound using one of two probes, depending on the child's age and the thickness of the subcutaneous adipose tissue: a linear probe, with a frequency of 6 -9 MHz; a convex probe (3)(4)(5). The ultrasound examination was performed by a senior radiologist (With more than five years of experience) in a time interval maximum 1-day post-CT study with a time of examination averaged 5 -8 minutes. We needed no special preparation, no sedation, or fluid restrictions during the ultrasound examination.
We examined the patient in supine decubitus position and sitting position for the old child while the infant reviewed by both supine positions on a table or hold from parents for the irritable child and prone position. We divided each hemithorax into three areas to cover the whole lung: the anterior area delimited by parasternal and anterior axillary lines, the lateral space between the anterior and posterior axillary lines, and the rear area delimited by the paravertebral and posterior axillary lines. 11 We made a focused approach for a specific region. Each part was scanned in the longitudinal and oblique plane, mediallateral and up-down. We examined the anterior and lateral areas of the chest while the infants were in supine decubitus. The posterior region was examined in prone decubitus in infants, while we used the sitting position to scan the rear wall in older patients. A noncontrast chest CT study was performed on 80 patients (other patients take contrast according to senior in charge opinion) using Brilliance Philips 64 slice machine. A radiologist reported the CT findings blindly to the results of the US. On lung ultrasound, consolidation, seen as "hepatization"-liver-like images or parenchymal images-with air or liquid bronchogram and anfractuous edges were considered ultrasound diagnostic parameters for pneumonia. In our study, the presence of a bronchogram inside consolidation was considered mandatory for pneumonia. The finding of dynamic air bronchogram (air bronchograms can have intrinsic dynamic centrifugal movements due to breathing) on LUS attests bronchial patency and rules out atelectasis. Detecting more than three B lines (which appear as vertical hyperechoic lines that arise from the pleural line) in one region between two ribs is considered evidence of interstitial lung syndrome. When we identified lung consolidation, we took the largest dimension longitudinal, transverse and sagittal axis and recorded the anatomical location for follow-up visits. An abscess was appreciated as a well-defined intrapulmonary rounded or oval hypoechoic lesion with variable thickness outer margins and may show internal fluid and air. Pulmonary necrosis was seen as a focal hypoechoic area within the consolidated lung, similar to CT. Lung sliding was appreciated as a horizontal movement of the pleural line in synchrony with the respiratory cycle, indicating a sliding movement of the visceral pleura against the parietal pleura. The movement disappears and cannot be detected with the LUS in a patient with pneumonia. We assessed the volume of pleural effusion in the supine position, and the transducer was aligned perpendicular to the dorsolateral chest wall measurement taken at inspiration by measuring the maximum distance between visceral and parietal pleura in millimeter multiplying by twenty. For pleural effusion diagnosis, anechoic accumulations were characteristic for uncomplicated pleural effusions or associated with septae or floating echos in complicated pleural effusions. On chest CT scan, parenchymal consolidation was defined Original Article LUS versus CT in paediatric pneumonia as air-space density with air bronchograms. The abscess was described as an intra-pulmonary cavity containing fluid and air, taking peripheral enhancement after IV contrast. Pulmonary necrosis was defined as an area of decreased density within a consolidated lung that shows no enhancement relative to the adjacent parenchyma. Pleural effusion is a free fluid density defined as loculated if the collection had a lobulated or lenticular shape with a convex border. The data was analyzed using Statistical Package for Social Sciences (SPSS) version 25. The data presented as mean, standard deviation, and ranges. Frequencies and percentages give categorical data. A Chi-square test assessed the association between U/S finding results and specific information. A level of P -value less than 0.05 was considered significant.

Results
One hundred six patients with signs, symptoms, and pneumonia-specific clinical presentation were evaluated with ages 45 days to 14 years with a mean of 6.5 years and standard deviation (SD) of ± 4.11 years. The highest proportion of study patients was > 5 years (47.2%). Regarding gender proportion, males were higher than females (66% versus 34%) with a male to female ratio of 1.94:1. Regarding residency proportion, rural was higher than urban area (51% versus 49%) while regarding immunization status, the higher proportion was partial immunization (54.7%) ( Table 1).  Figure 1).   By CT -scan, empyema was diagnosed in 5.7%, while by LUS, it was diagnosed in 3.8% of patients (Table 2) ( Figure 4).  Patients with complicated pneumonia could have more than one pleural effusion, abscess, and empyema. Six patients have empyema on CT associated with pleural effusion; two are missed by ultrasound. Twelve cases have cavitary lesions by CT scan, six associated with pleural effusion, two cavities missed by ultrasound examination. The study results revealed high sensitivity and specificity of 100% for LUS to detect complicated pneumonia, compared to CT (Table 3).  The distribution of study patients by follow-up is shown in (Table 4). In this study, six cases of consolidative lesion missed follow-up; three of them had pleural effusion. During the follow-up, 71% of cases of pneumonia and 83.3% of effusion cases were entirely resolved.
In this study, all patients with pneumonia aged < 2 years were not diagnosed by LUS with a significant association between LUS finding results and age (P = 0.001). There were no significant associations (P ≥ 0.05) between LUS finding results and all other characteristics (Table 5).

Discussion
The total number of study patients enrolled in the current study was 106. All of them were children who presented with signs and symptoms of pneumonia. Regarding general data of patient distribution, the current study shows results that are slightly different from other studies. 12 In the present study, 71% of cases of pneumonia and 83.3% of effusion cases were entirely resolved. This finding was comparable to the study by Saraya et al, 12 which showed almost complete resolution of the pneumonic hepatization in 53.8%. In comparison, associated pleural effusion was resolved in 94.1% of treated cases. In the current study, all patients with pneumonia aged < 2 years were not diagnosed by LUS with a significant association between LUS finding results and age (P = 0.001). We attributed this to the small number of patients aged less than two years (8 cases) included in our study; all have noncomplicated central pneumonia not reaching the pleural surface. There were no significant associations (P ≥ 0.05) between LUS finding results and all other characteristics, including sex, residency, immunization status, and clinical presentation. A few limitations that exist for this study must be acknowledged. The number of patients is still limited for broader applicability, and a more extensive, multicenter study would provide more generalizable results. Another limitation is that we did not assess the diagnostic ability of lung ultrasound for lung necrosis due to a lack of cases.

Conclusions
Chest ultrasound is a simple tool that can be performed in everyday practice, with high specificity and sensitivity compared to chest CT scan in diagnosis and follow-up of complicated pneumonia in the pediatric age group. Therefore, ultrasound needs to be encouraged not just as a valid diagnostic alternative but as a necessary ethical choice.