Accuracy of lung ultrasound in diagnosing the etiology of respiratory symptoms in hospitalized children, a prospective cohort study

Accuracy of lung ultrasound in diagnosing the etiology of respiratory symptoms in hospitalized children, a prospective cohort study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Accuracy of lung ultrasound in diagnosing the etiology of respiratory symptoms in hospitalized children, a prospective cohort study Ali Ismail, Christian Sadaka, Dima Khreis, Ahmad Hijazi, Zeinab Sweid, and 13 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5461492/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 28 Nov, 2025 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Lung ultrasound (LUS) is an emerging radiation-free diagnostic tool with significant promise for identifying the causes of respiratory symptoms in hospitalized children. We conducted a prospective cohort study of 65 pediatric patients (September 2022–January 2023) to evaluate the accuracy of LUS in determining the etiology of acute respiratory illness compared to an expert panel’s final clinical diagnosis, and secondarily to chest X-ray (CXR). LUS examinations were performed within 24 hours of admission by an investigator blinded to CXR results, and the final diagnoses were established by a committee blinded to LUS findings. LUS demonstrated high diagnostic accuracy, correctly identifying the respiratory illness etiology in 92.3% of cases, with near-perfect agreement (κ = 0.86) with the final diagnosis. Sensitivity and specificity were high across etiologies, including bacterial pneumonia, viral infections, and asthma exacerbations. LUS also outperformed CXR in determining the cause of illness. Integrating LUS findings with clinical and laboratory data allowed effective differentiation between bacterial and viral infections, highlighting LUS as a safer and more accurate alternative to CXR. Our findings support the integration of LUS into pediatric respiratory care and warrant further research to validate its utility on a larger scale. Health sciences/Health care/Diagnosis Health sciences/Health care/Paediatrics Pediatric Lung Ultrasound Accuracy of LUS Bacterial and Viral Etiology Radiation-Free Diagnosis Ultrasound Algorithms in Pediatrics Sensitivity and Specificity Figures Figure 1 Figure 2 Figure 3 Figure 4 Background As global child mortality declines, lower respiratory tract infections (LRTIs) remain a leading cause of preventable death in young children, especially in low- and middle-income countries. 1 Childhood pneumonia mortality has fallen thanks to vaccines (against Streptococcus pneumoniae and Haemophilus influenzae type b) and better diagnostics, antibiotics, and supportive care. 1–3 However, disparities in healthcare access and overuse of antibiotics continue to exacerbate pediatric LRTI burdens, particularly in resource-limited regions. 4 Rising antimicrobial resistance tied to inappropriate antibiotic use underscores the need for accurate diagnosis of pediatric respiratory infections. 4 Rapidly identifying the etiology of pediatric LRTIs is crucial for guiding appropriate treatment. Common pathogens include respiratory viruses (such as Respiratory Syncytial Virus (RSV) and influenza) and bacteria like S. pneumoniae and H. influenzae. 5 Distinguishing between bacterial and viral pneumonia remains challenging with current methods. Invasive lung sampling (e.g., bronchoscopy) is rarely feasible in children, so clinicians rely on indirect tests (nasopharyngeal swabs, blood cultures, sputum) with limited diagnostic accuracy. 6 For example, a large pneumonia etiology study found 66% of pediatric cases were viral and only 8% bacterial (with 7% mixed infections), illustrating the predominance of viral causes. 7 Current guidelines advise against antibiotic use for presumed viral pneumonia absent evidence of bacterial co-infection, underscoring the importance of an accurate diagnosis. 8 Yet many children with viral LRTIs still receive unnecessary antibiotics, fuelling antimicrobial resistance. 9,10 Imaging is central in pediatric pneumonia diagnosis, but chest X-ray (CXR) has significant limitations. 11 CXR exposes children to ionizing radiation (raising lifetime cancer risk) and may not reliably distinguish pneumonia etiology or detect early lung changes. 12 Lung ultrasound (LUS) has emerged as a valuable alternative without these downsides. LUS is radiation-free, portable, and cost-effective; it allows bedside repeat exams to monitor illness or treatment response and is well-tolerated by children. 13 Over the past decade, LUS has been widely studied in children with pneumonia, demonstrating excellent diagnostic performance and safety. 14 Pediatricians are increasingly trained in LUS techniques; for example, the American Academy of Pediatrics and international critical care groups now advocate pediatric LUS use. 15 Growing evidence supports LUS as an effective tool for evaluating pediatric respiratory conditions (bronchiolitis, pneumonia, pleural effusions). 11,16,17 Multiple studies show LUS often outperforms CXR in diagnosing pneumonia. 13,16,17 A systematic review of 1,510 children reported LUS sensitivity ~95.5% and specificity ~98.2%, versus 86.8% and 95.3% for CXR. 13 Similarly, a meta-analysis found LUS ~96% sensitive and 93% specific for pediatric pneumonia, prompting calls for broader LUS use in pediatrics. 15 Additionally, LUS can track disease course; one case series demonstrated that serial LUS effectively monitored antibiotic response in children with community acquired pneumonia, underscoring its clinical utility. 14 LUS also shows promise in differentiating bacterial vs. viral LRTIs, which is critical for selecting appropriate therapy. Different studies found that pediatric bacterial pneumonias tend to appear on LUS as solitary, larger, often unilateral consolidations, whereas viral pneumonias show smaller, multifocal or bilateral lesions. 18,19 A prospective study of 85 children identified distinct LUS patterns for viral, bacterial, and even SARS-CoV-2 pneumonias, highlighting LUS’s potential to distinguish infection etiologies. 20 LUS may also help predict disease severity; certain ultrasound features correlate with complicated (e.g., effusive) vs. uncomplicated pneumonia in children. 21 More recently, a detailed prospective analysis of pediatric LRTIs showed specific LUS findings tied to bacterial, viral, or atypical causes, aiding targeted management. 22 Nonetheless, an earlier PICU study reported that LUS alone had lower accuracy in determining pneumonia etiology when sonographers were blinded to clinical information, emphasizing the need to interpret LUS alongside clinical and lab data for optimal accuracy. 23 Overall, evidence remains limited on how well integrating LUS with clinical assessment and laboratory results can definitively pinpoint the cause of respiratory symptoms in hospitalized children, highlighting a gap for further research. Methodology The Institutional Review Board (IRB) at the American University of Beirut (AUB) granted approval for this study. All methods were conducted in compliance with relevant guidelines and regulations. This single-center, prospective observational cohort study was carried out at the American University of Beirut Medical Center (AUBMC) in Lebanon between September 2022 and January 2023. In our study, the sample size was determined using a sample size calculator (https://wnarifin.github.io/ssc/sssnsp.html). We hypothesized that LUS accuracy would achieve a sensitivity of 95% and specificity of 95%, values consistent with those reported in the literature for diagnosing pneumonia with LUS. Given that the prevalence of LRTIs among hospitalized pediatric patients at our institution in the previous year was approximately 30%, and accounting for a 5% drop-out rate, a power of 90%, and an alpha value of 0.05, the estimated required sample size was 65 patients. Inclusion Criteria: Eligible patients were children aged 0 to 19 years who presented with respiratory symptoms , including any combination of cough, rhinorrhea, nasal congestion, shortness of breath, wheezing, tachypnea, or increased work of breathing — and were within 24 hours of admission to the Pediatric Ward or Pediatric Intensive Care Unit or had admission orders but were boarding in the Emergency Department (ED). To identify eligible patients, the investigative team—excluding the principal investigator—routinely accessed the EPIC electronic health record system. Following approval by the patient's primary physician, the study was explained to the parents or guardians, and written informed consent was obtained. Exclusion criteria Patients were excluded from the study if they had a pre-existing chronic lung disease, as these underlying abnormalities could significantly alter lung ultrasound interpretation, limiting the generalizability of findings to the healthy pediatric population. Patients were also excluded if the LUS was not performed within 24 hours of admission to ensure alignment with the initial clinical presentation minimizing temporal bias, if they were admitted under the PI’s care to avoid potential bias, or if the PI was not blinded to the CXR findings to maintain blinding and minimize interpretation bias. Data Collection In our study, patients were recruited after being admitted to the hospital, making it unfeasible to perform LUS before CXR to ensure blinding of the investigator. As part of their initial workup, patients underwent CXR in the ED upon admission. Patients were excluded if they were admitted under the PI’s care or if the PI was not blinded to the CXR findings — as shown in Figure 4. The PI was informed about the clinical presentation and laboratory findings of the patients but was not provided with CXR findings by the treating team. The PI performed the LUS and then submitted the data collection form to the research assistant. The PI could access EPIC only after completing the LUS and submitting the data collection form. Additionally, LUS images were stored in separate software, distinct from EPIC. The expert committee, responsible for the final clinical diagnosis, had no access to LUS images or reports (Supplementary Figure S1). Although the study algorithm (Figure 3) recommended repeating LUS within 48 hours in cases of diagnostic uncertainty, follow-up scans were not systematically performed or recorded as part of the study protocol. In most cases, a repeat LUS was not requested by the clinical team. As such, any follow-up LUS performed was used solely for clinical management purposes and was not included in the data collection. Lung Ultrasound Procedure The principal investigator, credentialed in lung ultrasound with over three years of experience, performed the LUS within 24 hours of the admission order. Ultrasound scans were performed using the portable CX50 ultrasound system (Philips, Bothell, WA, USA), equipped with a linear transducer (L: 12-3 MHz) and a convex transducer (C: 8-5 MHz). Due to the small body size of the participants, all participants were tested using the L: 12-3 MHz linear transducer at a 5 cm depth or less. Each lung was scanned in both longitudinal and transverse views, covering the midclavicular line anteriorly, the paraspinal line posteriorly, and the mid-axillary line (Figure 1). Each view included both the upper and lower zones, resulting in a total of 12 scanning zones. This protocol was based on the method described by Copetti and Catarossi. 24 LUS Findings A-lines: Horizontal artifacts appearing as a series of echogenic lines parallel and equidistant from the pleural line, indicating normal lung tissue (Figure 2A). B-lines: Vertical comet-tail artifacts originating from the pleural line. According to Lichtenstein’s classification, B-lines indicate fluid-filled interlobular septae surrounded by air, associated with various pathological conditions (Figure 2B). 25 Subpleural Consolidation (sub-centimetric): Localized beneath the pleura, typically less than 1 cm in size (Figure 2C). Consolidation: Refers to lung tissue becoming firm due to the accumulation of fluid or cellular material, typically measuring equal to or greater than 1 cm in size (Figure 2D and 2E). LUS Etiology Diagnosis: In our study, the initial classification of patients into infectious versus non-infectious etiologies was based on clinical judgment, as this distinction was essential for applying our diagnostic algorithm (Figure 3). Specifically, patients classified with infectious etiology were evaluated using the pathway outlined in Figure 3A, while those with a non-infectious etiology were assessed using Figure 3B to guide the interpretation of LUS findings. For instance, patients presenting with cough, chest pain, or shortness of breath along with fever and/or a positive respiratory viral test were considered to have an infectious etiology. In contrast, patients with a known history of asthma who exhibited acute respiratory deterioration, clinical wheezing, and improvement with bronchodilators—regardless of the presence of fever—were classified as having asthma exacerbations, which we considered a non-infectious etiology. Importantly, we did not rely on laboratory markers such as complete blood count, C-reactive protein (CRP), or absolute neutrophil counts (ANC) to differentiate between viral and bacterial infections. Instead, lung ultrasound findings were central to differentiating between viral and bacterial lower respiratory tract infections. In our study, we classified Mycoplasma pneumoniae under the bacterial infection category. Recent literature supports that Mycoplasma pneumoniae can present with LUS features similar to typical bacterial pneumonia, including lobar consolidations, dynamic air bronchograms, and pleural effusions. For instance, Buonsenso et al. reported that children with bacterial and atypical acute lower respiratory tract infections exhibited larger consolidations, often exceeding 1.5 cm, with dynamic air bronchograms and pleural effusions. These findings suggest that M. pneumoniae infections may share sonographic characteristics with bacterial pneumonia. 18 Infectious Etiology For suspected infectious etiology, further classification was based on the clinical stability. Clinically unstable patients were typically treated in the PICU and were more likely to receive antibiotics presumptively for bacterial pneumonia. For stable patients with suspected LRTIs, the first LUS assessment involved checking for the presence of B-lines (Figure 3). Absent B-lines In cases where B-lines were absent and no consolidations were detected, the diagnosis was consistent with upper respiratory tract infection (URTI). Unilateral B-lines : Unilateral, focal B-lines with a consolidation ≥1 cm, whether accompanied by pleural effusion and dynamic air bronchograms, pointed to bacterial pneumonia. When the air bronchograms were static or absent (lung hepatization), the cause was likely either bacterial pneumonia or atelectasis. 22,26–28 In cases of unilateral, focal B-lines without consolidation or with small consolidations (<1 cm), the condition was typically classified as either a mild viral infection or an early-stage bacterial infection. A follow-up LUS was recommended after 48 hours. If the follow-up showed progression to bilateral B-lines or improvement without antibiotic treatment, a viral etiology was presumed. Conversely, if the consolidation expanded to ≥1 cm, bacterial pneumonia was diagnosed. Clinically unstable patients or those diagnosed with bacterial, complicated, or superimposed bacterial pneumonia were evaluated for antibiotic therapy (Figure 3A). Bilateral B-lines : The presence of bilateral B-lines, with or without consolidation measuring less than 1 cm, was interpreted as indicative of viral changes, suggesting a diagnosis of either bronchiolitis for patients under 2 years old or viral bronchitis for those aged 2 and older. 19,22,29,30 While consolidation measuring 1 cm or larger typically indicates bacterial pneumonia, the simultaneous presence of bilateral B-lines raised the possibility of viral pneumonia. In these patients, clinical deterioration or complications such as pleural effusion or abscess formation, were suggestive of a bacterial pneumonia. Conversely, in the absence of these signs, a diagnosis of viral pneumonia was made, with a follow-up lung ultrasound recommended within 48 hours if deemed necessary. Should the repeat LUS show worsening, bacterial pneumonia was suspected. If there was no change or improvement without antibiotics, the diagnosis remained viral pneumonia (Figure 3A). Non-infectious etiology Patients with suspected non-infectious etiology were evaluated for conditions such as atelectasis, lung contusion, cardiogenic pulmonary edema, pulmonary embolism, chronic lung disease, or lung masses. However, our study specifically focused on individuals with LRTIs. Among these, we included patients who clinically presented with asthma exacerbations, whether triggered by infections or allergens, as suggested by their medical histories or clinical signs such as a sudden onset of shortness of breath and wheezing observed during physical examinations. In such instances, LUS findings could display a wide range of abnormalities, from normal results to various forms of consolidation, B-lines, or pleural effusion. 31 If the consolidation measured less than 1 cm, the etiology diagnosis was considered asthma exacerbation. This diagnosis was also applicable in cases of consolidation equal to or larger than 1 cm, with or without static air bronchograms; however, a repeat ultrasound was advised within 48 hours if clinically indicated. Should the repeat ultrasound show that the consolidation improved without antibiotic treatment, the diagnosis of asthma exacerbation was considered. Conversely, if the LUS findings showed worsening, it was interpreted as a possible superimposed infection. In patients with a clinical diagnosis of asthma exacerbation, a large or moderate effusion was considered indicative of bacterial infection and treated accordingly. For small pleural effusion a follow-up LUS within 48 hours was recommended. If the subsequent ultrasound demonstrated no change or improvement without antibiotics, the diagnosis of asthma exacerbation was supported. Conversely, if the pleural effusion worsened, a superimposed bacterial pneumonia was suspected and antibiotic therapy initiated (Figure 3B). Final Clinical Diagnosis: A committee of pediatric specialists, comprising a pulmonologist and a critical care physician, independently reviewed each patient’s complete clinical course post-discharge, remaining blinded to lung ultrasound findings. They determined the etiology—viral, bacterial, or asthma—through a comprehensive assessment that included the patient’s history, clinical presentation, physical examination, laboratory tests (such as complete blood count, C-reactive protein, viral panels via Polymerase Chain Reaction (PCR) or antigen detection, and available cultures, and chest X-ray findings. This rigorous process facilitated consensus on the final clinical diagnosis. Statistical Analysis Data were collected using a standardized form . Data management and analyses were done using IBM-SPSS software (version 28.0, Armonk, NY, USA) by a designated member of the research team who was not involved in data collection in order to minimize bias in data entry. Descriptive Statistics Demographic and other baseline characteristics were summarized using descriptive statistics. Continuous variables were presented as means and standard deviations (SD) if normally distributed or medians and interquartile ranges (IQR) for skewed distributions. Categorical variables were reported as frequencies and percentages. Diagnostic Accuracy Analysis The diagnostic accuracy of LUS in identifying the etiology of respiratory symptoms was assessed using the final clinical diagnosis as the reference gold standard and served as the primary outcome of the study. The secondary outcome was the comparison of LUS accuracy versus CXR accuracy, both against the final clinical diagnosis as the gold standard. This analysis was conducted both for overall and for specific etiological subgroups, including viral, bacterial, and asthma exacerbation. Sensitivity and specificity for each modality were calculated along with 95% confidence intervals to assess the precision of these estimates. The overall agreement between each diagnostic tool and the final clinical diagnosis was quantified using Cohen’s kappa coefficient (κ), with 95% confidence intervals provided to evaluate the consistency of the agreement. The interpretation of κ followed Cohen's guidelines: values ≤ 0 indicating no agreement, 0.01-0.20 slight agreement, 0.21-0.40 fair agreement, 0.41-0.60 moderate agreement, 0.61-0.80 substantial agreement, and 0.81-1.00 almost perfect agreement. 32 Comparative Analysis In our study, 11 out of the 65 patients included in the statistical analysis underwent only LUS because CXRs were not requested. To compare the accuracy of LUS and CXR, we employed the McNemar test specifically on the subset of patients (n=54) who underwent both diagnostic procedures. This test assessed whether the proportion of correct diagnoses differed significantly between the two methods. A two-tailed p-value of less than 0.05 was considered statistically significant. Results The study enrolled 99 patients who were admitted with respiratory symptoms between September 2022 and January 2023. Out of these, 65 patients met the inclusion criteria and were included in the final analysis (Figure 4). The median age of these patients was 2 years (IQR, 0.46-2 years), with 37 patients being males (57%) (Table 1). The mean duration for the lung ultrasound procedure was 8.4 minutes (SD = 2.5 minutes). Among the included patients, 62 were diagnosed with lower respiratory tract disease (LRTD), accounting for 95.4% of cases, while the remaining 3 patients (4.6%) had upper respiratory tract infections. Regarding the etiology of LRTD, 12 patients (19%) had bacterial infection (pneumonia), 43 (70%) had viral infections (bronchiolitis, bronchitis, viral pneumonia), and 7 (11%) were diagnosed with asthma exacerbations (Figure 4). Among the 43 patients with viral infections, 7 had viral pneumonia with a consolidation size greater than 1 cm (median diameter of 2 cm). Blood cultures were obtained for 47 patients, of whom only 4 had positive results: one patient had Klebsiella pneumoniae and E. coli ; one had Corynebacterium afermentans ; one had Acinetobacter variabilis ; and one had Streptococcus infantis and Staphylococcus haemolyticus . Viral testing (rapid antigen or PCR) was performed in 54 patients (83%), with respiratory syncytial virus (RSV) being the most commonly identified virus (12 patients), followed by rhinovirus (5), adenovirus (3), and influenza A (2). Of these, 13 patients were tested using a multiplex PCR assay, and none tested positive for Mycoplasma pneumoniae . A supplementary table (S2) presents selected laboratory parameters—including white blood cell count, CRP, ANC, and neutrophil-to-lymphocyte ratio (NLR)—among patients with final diagnoses of asthma, bacterial infection, and viral infection, to describe observable trends. CXR findings are summarized in Table 3. Primary Outcome The diagnostic accuracy of LUS compared to the final clinical diagnosis revealed that LUS had a sensitivity of 91.9% [95% CI: 85.2–98.7%] and a specificity of 100% [95% CI: 36.8–100.0%]. The agreement between LUS and the final clinical diagnosis, as measured by Cohen’s kappa, was 0.86 [95% CI: 0.77–0.94], indicating almost perfect agreement (Table 2). For patients with bacterial pneumonia, the sensitivity and specificity of LUS were 91.8% [95% CI: 61.6–99.8%] and 92.5% [95% CI: 61.6–99.8%], respectively. For patients with viral infections (bronchiolitis, bronchitis, viral pneumonia), the sensitivity and specificity were 90.7% [95% CI: 82.0–99.4%] and 95.5% [95% CI: 77.2–99.9%], respectively. For patients with asthma exacerbation, both sensitivity and specificity were 100.0% [95% CI: 65.2–100.0%] and 100.0% [95% CI: 95.0–100.0%], respectively (Table 2). Secondary Outcome Among patients who underwent both diagnostic procedures (n=54), the accuracy of diagnosis was significantly higher with LUS compared to CXR. The McNemar chi-squared statistic was χ² = 10.89, with a corresponding p-value = 0.001. The kappa coefficient for LUS was 0.84 [95% CI: 0.70–0.98], whereas CXR had a kappa of 0.47 [95% CI: 0.30–0.65]. The sensitivity and specificity of LUS were 90.2% [95% CI: 82.0–98.3%] and 100% [95% CI: 36.8–100.0%], respectively. In comparison, CXR had a sensitivity of 62.7% [95% CI: 49.5–76.0%] and a specificity of 100% [95% CI: 36.8–100.0%] (Tables 3 and 4). Subgroup Analysis In our study, we performed a subgroup analysis to evaluate the diagnostic accuracy of LUS in different age groups. For patients under 2 years of age (n=32), the total accuracy of LUS in diagnosing the etiology of respiratory symptoms was 90.6% (95% CI: 80.5, 100.0). The sensitivity for identifying bacterial infections was 66.7% (95% CI: 13.0, 100.0), while sensitivity for viral infections was high at 92.9% (95% CI: 83.0, 100.0). The specificity for bacterial infections was 93.1% (95% CI: 83.9, 100.0), but the specificity for viral infections was lower at 75% (95% CI: 33.0, 99.9). In contrast, for patients aged 2 years and older (n=33), LUS showed a higher total accuracy of 93.9% (95% CI: 85.0, 100.0). The sensitivity for bacterial infections was perfect at 100% (95% CI: 100.0, 100.0), while the sensitivity for viral infections was 86.7% (95% CI: 69.0, 100.0). The specificity for bacterial infections was 91.7% (95% CI: 80.6, 100.0), and for viral infections, it was 100% (95% CI: 84.7, 100.0). (Supplementary Table S3) Discussion This prospective study demonstrates point-of-care lung ultrasound can accurately identify the etiology of respiratory symptoms in hospitalized children when interpreted alongside clinical information. We observed an overall agreement of 92.3% between LUS-based diagnoses and the final expert clinical diagnoses, with a near-perfect kappa of 0.86. Notably, LUS maintained high diagnostic performance across different causes of illness: it showed strong agreement in detecting bacterial pneumonia and viral lower respiratory infections (each with sensitivities around 90% and substantial agreement, κ ~0.8) as well as asthma exacerbations (100% agreement with final diagnoses). We also noted a slight trend toward higher accuracy in older children (≥2 years) compared to infants, although LUS still performed well even in the under-2 age group. Collectively, these results position LUS as a reliable tool for pediatric respiratory diagnostics in the acute care setting. Our findings build on the growing evidence that LUS is an effective, child-friendly alternative to radiography for evaluating pediatric lung disease. Previous studies and meta-analyses have reported that LUS can diagnose pediatric pneumonia with high sensitivity and specificity (in the 93–98% range), often outperforming CXR. 11,15–17 However, those studies largely did not distinguish between bacterial and viral etiologies of pneumonia. Our study uniquely addresses this gap by demonstrating that integration of LUS results with clinical and laboratory findings can achieve accurate etiological differentiation at the bedside. This integrated approach is crucial, as an earlier pediatric ICU study showed that when sonographers were blinded to clinical context, LUS alone had markedly lower diagnostic accuracy (sensitivity dropping to ~56%) for identifying the cause of respiratory failure. 23 By contrast, our methodology—incorporating patient history, exam findings, and basic labs in the LUS interpretation—yielded near-perfect agreement with expert diagnoses, underscoring the importance of context-driven ultrasound interpretation. 18 Consistent with prior literature, we found distinct lung ultrasound patterns corresponding to different infection types, which aided our diagnostic accuracy. Studies have documented that bacterial pneumonias in children often appear on LUS as larger, solitary consolidations (frequently confined to one lung), whereas viral pneumonias tend to produce smaller, multifocal or bilateral lesions. 18,19 We observed similar trends in our cohort: for example, cases classified as bacterial infection typically had focal lobar consolidation on LUS, while viral cases showed diffuse B-lines or subpleural consolidations scattered in both lungs. A recent prospective study of pediatric pneumonias described comparably distinct LUS profiles for viral, typical bacterial, and even COVID-19 pneumonias, reinforcing the notion that ultrasound features can suggest the underlying pathogen. 20 In our data, LUS was highly sensitive in detecting viral bronchiolitis and viral pneumonia (over 90% sensitivity), which is on par with the 90.4% sensitivity reported by Caiulo et al. for bronchiolitis; notably, in that study LUS substantially outperformed CXR (73.1% sensitivity) for diagnosing bronchiolitis. 33 This high accuracy for viral illnesses is especially valuable, as it can help clinicians confidently identify viral infections and avoid unnecessary antibiotic use in these cases. 7,9,10 Our results also challenge a common assumption in lung ultrasound interpretation—that the presence of a large consolidation (≥1 cm) definitively indicates bacterial pneumonia. We identified several children with viral infections who had consolidations ≥1 cm on LUS (median ~2 cm); all of these patients recovered fully without antibiotics. This finding underscores that sizable consolidations can occur in viral pneumonia or severe bronchiolitis. Similarly, Berce et al. reported a median consolidation size of about 1.5 cm in confirmed viral pneumonias. 19 Buonsenso et al. also reported patients with viral pneumonia who had consolidation size of > 1cm. 18 Thus, consolidation size alone should not be viewed as a strict criterion for bacterial infection. Rather, the overall ultrasound pattern and clinical picture must be considered—some of these larger consolidations in viral infections may represent inflammatory or atelectatic changes rather than focal bacterial lobar pneumonia. In practice, for a stable patient with predominantly viral-appearing LUS findings (for instance, diffuse bilateral changes) but an unexpected large consolidation, a prudent approach would be watchful waiting with supportive care and a repeat LUS after 24–48 hours. Such a strategy, as we employed in our study, can ensure bacterial superinfection is not developing while sparing the child from immediate empiric antibiotics, thereby reducing unnecessary antibiotic exposure. Repeating LUS within 48 hours in our algorithm aligns with the findings of Buonsenso et al., who demonstrated that certain LUS patterns, both at diagnosis and 48 hours after treatment, were better predictors of antibiotic response in children with acute lower respiratory tract infections than clinical data and laboratory results. 14,21 In our subgroup analysis, the diagnostic performance of LUS varied notably by age group. Among children under 2 years of age, the sensitivity for identifying bacterial infections was 66.7% (95% CI: 13.0–100.0), and the specificity for viral infections was 75% (95% CI: 33.0–99.9). These modest values are likely influenced by the small sample size in this subgroup—only three patients had bacterial infections, and four were negative for viral infections—limiting the statistical reliability of the estimates. In contrast, for children aged 2 years and older, LUS showed excellent diagnostic accuracy, with both the sensitivity for bacterial infections and the specificity for viral infections reaching 100% (95% CI: 100.0–100.0 for bacterial infections and 95% CI: 84.7–100.0 for viral infections). These findings suggest that LUS may be more reliable in older children, potentially due to more distinguishable clinical and sonographic features. However, larger studies across all pediatric age groups are needed to validate these results and better define the role of LUS in differentiating between bacterial and viral lower respiratory tract infections. Beyond infections, this study demonstrates that LUS can contribute to the evaluation of non-infectious respiratory disorders like acute asthma exacerbations. We correctly identified all cases of asthma exacerbation using our LUS protocol in conjunction with clinical history, achieving perfect agreement (κ = 1) with the final diagnoses. This aligns with prior observations that children with asthma attacks can exhibit certain LUS abnormalities (such as bilateral comet-tail B-lines or small subpleural consolidations due to airway plugging), even though these ultrasound findings are not specific to asthma. 31 Without knowledge of the clinical context, such LUS findings could be misinterpreted as pneumonia or edema, which highlights a key point: LUS is most effective as an adjunct to, not a replacement for, clinical assessment. DeSanti et al. found that when LUS interpreters were blinded to clinical information, they correctly identified only 6 of 10 pediatric status asthmaticus cases based on ultrasound alone. 23 In our study, by factoring in the patients’ known history of asthma and presenting signs (wheezing responsive to bronchodilators), we avoided this pitfall. The implication is that while LUS may not have pathognomonic signs for asthma, it can still be very useful in the emergency setting to support an asthma diagnosis or exclude other causes of respiratory distress, especially when integrated with clinical judgement. Another important finding is that LUS outperformed chest X-ray in determining the cause of pediatric respiratory illness. We found LUS to be significantly more accurate than CXR (p < 0.001) for identifying the etiologic diagnosis, consistent with prior comparative studies. In our cohort, CXR results showed only moderate agreement with the final clinical diagnosis (κ ~0.47), considerably lower than the near-perfect concordance achieved with LUS. This result reinforces the growing consensus that lung ultrasound can replace or complement radiography for pediatric pneumonia diagnosis. 15,17 Given the cumulative risks of radiation in young patients, the superior performance of LUS that we and others have observed further supports adopting ultrasound as a first-line imaging modality in children with acute respiratory symptoms. Strengths Our study addresses a significant gap in the literature by demonstrating the high diagnostic accuracy of lung ultrasound when combined with clinical assessment, particularly in differentiating between bacterial and viral infections in patients with respiratory symptoms. A key strength of our research is its prospective cohort design, which minimized recall bias and established a clear temporal relationship between diagnostics and outcomes. 34 Additionally, lung ultrasounds were performed by a single, experienced investigator who was blinded to the chest X-ray results, thereby reducing both diagnostic and observer bias. 35 The implementation of standardized diagnostic algorithms facilitated a systematic and replicable integration of clinical, laboratory, and LUS findings, ultimately enhancing diagnostic accuracy. Limitations While our study provides valuable findings, it has several limitations. Firstly, it was conducted at a single tertiary care center, which may restrict the generalizability of our findings. 36 Furthermore, the relatively small sample size, predominantly consisting of patients with viral infections, could affect the calculations of sensitivity and specificity. 37 Furthermore, the lung ultrasounds were performed and interpreted by an investigator with advanced training in pediatric LUS; this level of expertise may not be available in all hospitals, and it underscores the need for adequate training programs if LUS is to be widely implemented. 38 In addition, the standard reference used in our study was the final clinical diagnosis established by two experts—a pediatric pulmonologist and a pediatric critical care physician—which may not reflect the gold standard for detecting the etiology of pulmonary diseases. Moreover, three patients were excluded from the study because the ultrasound machine was not accessible due to the lack of a dedicated ultrasound machine for the research study. Future directions Future research should involve larger, multicenter studies across diverse populations and clinical settings to validate our results and confirm the generalizability of LUS as a diagnostic modality in various healthcare environments. Further investigations are necessary to clarify the clinical relevance of sub-centimetric consolidations with or without air bronchograms observed on LUS. Determining whether these findings warrant antibiotic treatment or support a "watchful waiting" approach is crucial for guiding clinical decision-making and reducing unnecessary antibiotic use. Additionally, research is needed to redefine the size criteria of consolidations on lung ultrasound in cases of viral pneumonia. Our findings suggest that consolidations larger than 1 cm may still be associated with viral infections, challenging the traditional limitation to smaller consolidations. Clarifying these criteria could improve diagnostic accuracy and patient management strategies. Evaluations are also warranted to elucidate the LUS findings of atypical pathogens like Mycoplasma pneumoniae in children. Furthermore, larger, multicenter studies should systematically evaluate how age-related anatomical differences (e.g., chest wall thickness) impact LUS interpretation across age groups. Finally, the diagnostic algorithms (Figure 3) we utilized in this study for determining the etiology of infectious and non-infectious respiratory disorders showed promising results. Future studies should aim to validate the reliability and applicability of these algorithms in different clinical settings. By confirming their effectiveness, these diagnostic pathways could be integrated into clinical practice and incorporated into machine learning and artificial intelligence frameworks, thereby enhancing the utility of LUS in managing pediatric respiratory conditions. Conclusion This study demonstrates that LUS is a highly accurate and practical tool for diagnosing the etiology of respiratory symptoms in hospitalized children, outperforming chest X-ray in terms of diagnostic accuracy. By combining LUS findings with clinical and laboratory data, we were able to differentiate between bacterial and viral infections effectively, supporting more judicious use of antibiotics and improving antimicrobial stewardship. The use of LUS also eliminates the risks associated with radiation exposure from chest X-rays, making it a safer and more reliable imaging modality, especially for pediatric patients. Our findings highlight the potential of LUS as a first-line diagnostic tool in the management of pediatric respiratory diseases. Further studies, particularly multicenter trials with larger sample sizes, are needed to validate these results and refine diagnostic criteria for LUS, ultimately improving its implementation in clinical practice. Declarations Acknowledgments This study was funded by a seed grant from the American University of Beirut Medical Center (Medical Practice Plan - MPP Fund 320213). AI would like to acknowledge the training received under the Faculty Advancement Program (FAP) at the American University of Beirut. Special thanks to Dr. Mona Nabulsi for her mentorship and support in developing the proposal of this paper. The content is solely the responsibility of the authors. Author Contribution A.I., C.S., and N.A. contributed to the research question, proposal writing, and data collection. A.I. and D.K. contributed to the creation of algorithms. A.I. and Z.S. were responsible for data entry. A.I. and Z.M. carried out the statistical analysis. A.I., C.S., D.K., Z.S., F.T., A.M.H. and M.S. contributed to manuscript writing. A.I., D.K., H.R., J.A., R.S., S.K., A.M.H., and A.H.H. were involved in the literature review. N.G., J.M., S.M. and M.M. provided supervision and guidance. All authors reviewed and edited the draft and agreed to submit the manuscript for publication. Competing interest: The authors declare no competing interests. Data availability statement The datasets generated and/or analysed during the current study are not publicly available due to institutional requirements but are available from the corresponding author upon reasonable request. References Perin, J. et al. Global, regional, and national causes of under-5 mortality in 2000–19: an updated systematic analysis with implications for the Sustainable Development Goals. Lancet Child Adolesc. Health 6 , 106–115 (2022). Pavia, M., Bianco, A., Nobile, C. G. A., Marinelli, P. & Angelillo, I. F. Efficacy of Pneumococcal Vaccination in Children Younger Than 24 Months: A Meta-Analysis. Pediatrics 123 , e1103–e1110 (2009). Obonyo, C. O. & Lau, J. Efficacy of Haemophilus influenzae type b vaccination of children: a meta-analysis. Eur. J. Clin. Microbiol. Infect. Dis. 25 , 90–97 (2006). GBD 2016 Lower Respiratory Infections Collaborators. Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory infections in 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Infect. Dis. 18 , 1191–1210 (2018). US Burden of Disease Collaborators et al. The State of US Health, 1990-2016: Burden of Diseases, Injuries, and Risk Factors Among US States. JAMA 319 , 1444–1472 (2018). Gereige, R. S. & Laufer, P. M. Pneumonia. Pediatr. Rev. 34 , 438–456; quiz 455–456 (2013). Jain, S. et al. Community-Acquired Pneumonia Requiring Hospitalization among U.S. Children. N. Engl. J. Med. 372 , 835–845 (2015). Bradley, J. S. et al. The Management of Community-Acquired Pneumonia in Infants and Children Older Than 3 Months of Age: Clinical Practice Guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin. Infect. Dis. 53 , e25–e76 (2011). Pratt, M. T. G. et al. Prevalence of respiratory viruses in community-acquired pneumonia in children: a systematic review and meta-analysis. Lancet Child Adolesc. Health 6 , 555–570 (2022). Gerber, J. S. et al. Antibiotic Stewardship in Pediatrics. Pediatrics 147 , e2020040295 (2021). Jones, B. P. et al. Feasibility and Safety of Substituting Lung Ultrasonography for Chest Radiography When Diagnosing Pneumonia in Children. Chest 150 , 131–138 (2016). Kutanzi, K., Lumen, A., Koturbash, I. & Miousse, I. Pediatric Exposures to Ionizing Radiation: Carcinogenic Considerations. Int. J. Environ. Res. Public. Health 13 , 1057 (2016). Balk, D. S. et al. Lung ultrasound compared to chest X-ray for diagnosis of pediatric pneumonia: A meta-analysis. Pediatr. Pulmonol. 53 , 1130–1139 (2018). Buonsenso, D. et al. The Use of Lung Ultrasound to Monitor the Antibiotic Response of Community-Acquired Pneumonia in Children: A Preliminary Hypothesis. J. Ultrasound Med. 39 , 817–826 (2020). Pereda, M. A. et al. Lung Ultrasound for the Diagnosis of Pneumonia in Children: A Meta-analysis. Pediatrics 135 , 714–722 (2015). Yang, Y., Wu, Y. & Zhao, W. Comparison of lung ultrasound and chest radiography for detecting pneumonia in children: a systematic review and meta-analysis. Ital. J. Pediatr. 50 , 12 (2024). Heuvelings, C. C. et al. Chest ultrasound for the diagnosis of paediatric pulmonary diseases: a systematic review and meta-analysis of diagnostic test accuracy. Br. Med. Bull. 129 , 35–51 (2019). Buonsenso, D. et al. Role of lung ultrasound for the etiological diagnosis of acute lower respiratory tract infection (ALRTI) in children: a prospective study. J. Ultrasound 25 , 185–197 (2022). Berce, V., Tomazin, M., Gorenjak, M., Berce, T. & Lovrenčič, B. The Usefulness of Lung Ultrasound for the Aetiological Diagnosis of Community-Acquired Pneumonia in Children. Sci. Rep. 9 , 17957 (2019). Stoicescu, E. R. et al. Differentiating Viral from Bacterial Pneumonia in Children: The Diagnostic Role of Lung Ultrasound—A Prospective Observational Study. Diagnostics 14 , 480 (2024). Musolino, A. M. et al. Lung ultrasound features of children with complicated and noncomplicated community acquired pneumonia: A prospective study. Pediatr. Pulmonol. 54 , 1479–1486 (2019). Kharasch, S., Duggan, N. M., Cohen, A. R. & Shokoohi, H. Lung Ultrasound in Children with Respiratory Tract Infections: Viral, Bacterial or COVID-19? A Narrative Review. Open Access Emerg. Med. Volume 12 , 275–285 (2020). DeSanti, R. L. et al. Point-of-Care Lung Ultrasound to Diagnose the Etiology of Acute Respiratory Failure at Admission to the PICU*. Pediatr. Crit. Care Med. 22 , 722–732 (2021). Copetti, R. & Cattarossi, L. Ultrasound diagnosis of pneumonia in children. Radiol. Med. (Torino) 113 , 190–198 (2008). Lichtenstein, D. A. Lung ultrasound in the critically ill. Ann. Intensive Care 4 , 1 (2014). Lichtenstein, D., Mezière, G. & Seitz, J. The Dynamic Air Bronchogram. Chest 135 , 1421–1425 (2009). Omran, A. et al. Lung Ultrasound and Neutrophil Lymphocyte Ratio in Early Diagnosis and Differentiation between Viral and Bacterial Pneumonia in Young Children. Children 9 , 1457 (2022). Acosta, C. M., Belaunzarán, A. & Gonorazky, S. Accuracy of Transthoracic Lung Ultrasound for Diagnosing Anesthesia-induced Atelectasis in Children. Perioper. Med. International Liaison Committee on Lung Ultrasound (ILC-LUS) for the International Consensus Conference on Lung Ultrasound (ICC-LUS) et al. International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med. 38 , 577–591 (2012). Biagi, C. et al. Lung ultrasound for the diagnosis of pneumonia in children with acute bronchiolitis. BMC Pulm. Med. 18 , 191 (2018). Dankoff, S., Li, P., Shapiro, A. J., Varshney, T. & Dubrovsky, A. S. Point of care lung ultrasound of children with acute asthma exacerbations in the pediatric ED. Am. J. Emerg. Med. 35 , 615–622 (2017). McHugh, M. L. Interrater reliability: the kappa statistic. Biochem. Medica 276–282 (2012) doi:10.11613/BM.2012.031. Caiulo, V. A. et al. Lung ultrasound in bronchiolitis: comparison with chest X-ray. Eur. J. Pediatr. 170 , 1427–1433 (2011). Song, J. W. & Chung, K. C. Observational Studies: Cohort and Case-Control Studies. Plast. Reconstr. Surg. 126 , 2234 (2010). Delgado-Rodriguez, M. & Llorca, J. Bias. J. Epidemiol. Community Health 58 , 635 (2004). Bellomo, R., Warrillow, S. J. & Reade, M. C. Why we should be wary of single-center trials. Crit. Care Med. 37 , 3114 (2009). Chu, H. & Cole, S. R. Sample size calculation using exact methods in diagnostic test studies. J. Clin. Epidemiol. 60 , 1201–1202 (2007). Rutjes, A. W. S., Reitsma, J. B., Coomarasamy, A., Khan, K. S. & Bossuyt, P. M. M. Evaluation of diagnostic tests when there is no gold standard. A review of methods. Health Technol. Assess. Winch. Engl. 11 , iii, ix–51 (2007). Tables Tables 1 to 4 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files SupplementaryTablesandFigures.docx Tables1234.docx Cite Share Download PDF Status: Published Journal Publication published 28 Nov, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 19 May, 2025 Editor assigned by journal 19 May, 2025 Editor invited by journal 19 May, 2025 Reviews received at journal 06 May, 2025 Reviewers agreed at journal 03 May, 2025 Reviews received at journal 28 Apr, 2025 Reviewers agreed at journal 28 Apr, 2025 Reviewers invited by journal 28 Apr, 2025 Submission checks completed at journal 27 Apr, 2025 First submitted to journal 14 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5461492","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":449373150,"identity":"033c4b7e-b8b3-4d20-8733-63ee8d286697","order_by":0,"name":"Ali Ismail","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Ali","middleName":"","lastName":"Ismail","suffix":""},{"id":449373151,"identity":"ad71da3c-707e-47dc-ab7b-7e558c711e82","order_by":1,"name":"Christian Sadaka","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Sadaka","suffix":""},{"id":449373152,"identity":"edfa163e-55bf-48a0-b7b5-f77b4281037c","order_by":2,"name":"Dima Khreis","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Dima","middleName":"","lastName":"Khreis","suffix":""},{"id":449373153,"identity":"51c4f3f3-4ae5-4758-8b11-8f34455ea390","order_by":3,"name":"Ahmad Hijazi","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Ahmad","middleName":"","lastName":"Hijazi","suffix":""},{"id":449373154,"identity":"0a70ac97-338b-404b-94f9-9d4039e22abf","order_by":4,"name":"Zeinab Sweid","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Zeinab","middleName":"","lastName":"Sweid","suffix":""},{"id":449373155,"identity":"ce85c9ba-1334-47e4-86b4-04d19451a561","order_by":5,"name":"Faysal Tabbara","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Faysal","middleName":"","lastName":"Tabbara","suffix":""},{"id":449373156,"identity":"37524a4b-44c6-4b24-b7e2-2555a81c07d8","order_by":6,"name":"Maysaa Slika","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Maysaa","middleName":"","lastName":"Slika","suffix":""},{"id":449373157,"identity":"0d8df618-b8b3-425e-a261-9f3e9b94d53a","order_by":7,"name":"Hawraa Raoof","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Hawraa","middleName":"","lastName":"Raoof","suffix":""},{"id":449373158,"identity":"026dbe46-6738-42a9-8524-5f88ff1ce4f4","order_by":8,"name":"Nour Adra","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Nour","middleName":"","lastName":"Adra","suffix":""},{"id":449373159,"identity":"ce04dbc2-d669-4c1e-9f66-b5eebfbdac97","order_by":9,"name":"Jana Ayash","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Jana","middleName":"","lastName":"Ayash","suffix":""},{"id":449373160,"identity":"9a3ca769-ff04-4171-be77-38bda6cf295a","order_by":10,"name":"Rita Sebaaly","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Rita","middleName":"","lastName":"Sebaaly","suffix":""},{"id":449373161,"identity":"885925c7-18fb-4440-b30e-cdaf9bd9b397","order_by":11,"name":"Sana Kalaji","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Sana","middleName":"","lastName":"Kalaji","suffix":""},{"id":449373162,"identity":"2fe983d5-ccd3-4adb-aca3-f6c488fb50ba","order_by":12,"name":"Ali Hijazi","email":"","orcid":"","institution":"University of Balamand","correspondingAuthor":false,"prefix":"","firstName":"Ali","middleName":"","lastName":"Hijazi","suffix":""},{"id":449373163,"identity":"bd346e6d-79ea-4e70-837b-56b73aa21e1b","order_by":13,"name":"Jihane Moukhaiber","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Jihane","middleName":"","lastName":"Moukhaiber","suffix":""},{"id":449373164,"identity":"acdd751b-8b21-4be9-93a7-3b99ce4f3da1","order_by":14,"name":"Nisrine Ghalayini","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Nisrine","middleName":"","lastName":"Ghalayini","suffix":""},{"id":449373165,"identity":"2a496fa0-68aa-4b55-8e77-8a7d1306e15c","order_by":15,"name":"Ziyad R. Mahfoud","email":"","orcid":"","institution":"Weill Cornell Medicine","correspondingAuthor":false,"prefix":"","firstName":"Ziyad","middleName":"R.","lastName":"Mahfoud","suffix":""},{"id":449373166,"identity":"c7e03963-615a-4d27-94b5-b724324671fa","order_by":16,"name":"Marianne Nimah Majdalani","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArElEQVRIiWNgGAWjYHCChAMfgASQYcDAUHCAKC0PH86AazEgSgvjY2MekrTozm5Ok7apOZzHwN68TYLB4A5hLWZ3jqVJ5xw7XMzAc6wMqOUZEVpu5AC1sB1ObJDIMQNqOUyMlvxv0hb/gFrk3xCtJSHZmLENZAsP8VoSH/b2pRez8aQVWyQQ55eEhAM/vlnn8bMf3njjQwURIQYHbCAigQQNo2AUjIJRMArwAABlwjwrh5cyGgAAAABJRU5ErkJggg==","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":true,"prefix":"","firstName":"Marianne","middleName":"Nimah","lastName":"Majdalani","suffix":""},{"id":449373167,"identity":"d30aba96-d138-4dd2-a509-03aced456a97","order_by":17,"name":"Salman Mroueh","email":"","orcid":"","institution":"American University of Beirut Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Salman","middleName":"","lastName":"Mroueh","suffix":""}],"badges":[],"createdAt":"2024-11-15 15:23:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5461492/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5461492/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-26766-8","type":"published","date":"2025-11-28T15:58:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82042375,"identity":"f81563c8-3fb9-4cf4-9892-252271c68005","added_by":"auto","created_at":"2025-05-06 09:21:05","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":385705,"visible":true,"origin":"","legend":"\u003cp\u003eIllustration of pediatric lung ultrasound approaches.\u003c/p\u003e\n\u003cp\u003eIn Red: Anterior Approach (midclavicular lines)\u003c/p\u003e\n\u003cp\u003eIn Blue: Lateral Approach (midaxillary lines)\u003c/p\u003e\n\u003cp\u003eIn Black: Posterior Approach (vertical lines extending between the paraspinal and scapular lines).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePhotograph of the patient's chest taken with parental consent.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/952d106d8601139770b9a5ca.png"},{"id":82044162,"identity":"f8d2ef80-381d-469b-a178-3fc60447fe33","added_by":"auto","created_at":"2025-05-06 09:29:05","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":327069,"visible":true,"origin":"","legend":"\u003cp\u003eLung ultrasound findings. (A) Normal findings showing A-lines. (B) B-lines. (C) Sub-pleural Consolidation. (D) Consolidation with air-bronchogram. (E) Consolidation without air-bronchogram.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/cebfc9ea1144661728f6bf06.png"},{"id":82044161,"identity":"7f80800f-3285-406a-b8ea-2d840385033e","added_by":"auto","created_at":"2025-05-06 09:29:05","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":204401,"visible":true,"origin":"","legend":"\u003cp\u003eEtiology\u003cstrong\u003e \u003c/strong\u003eFlow diagram based on clinical and ultrasound findings. (A)\u003cstrong\u003e \u003c/strong\u003eInfectious Etiology Diagnosis. (B) Non-Infectious Etiology Diagnosis..\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/228abea8b0278f277dc9eafb.png"},{"id":82044160,"identity":"436c160b-5c79-4c85-bdfb-fe53535a099a","added_by":"auto","created_at":"2025-05-06 09:29:05","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":33518,"visible":true,"origin":"","legend":"\u003cp\u003eFlow Diagram of Study Participants.\u003c/p\u003e\n\u003cp\u003eLUS: lung ultrasound; CXR: Chest X-Ray.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/dacd43ba0f0bf494b6e4ebfc.png"},{"id":97178545,"identity":"777129e5-a466-40a9-a283-0568798222fb","added_by":"auto","created_at":"2025-12-01 16:10:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2003059,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/0d922bb1-1649-4426-8919-5ba77bbb3de1.pdf"},{"id":82042380,"identity":"e96193ff-0e37-4817-892a-93d306054d00","added_by":"auto","created_at":"2025-05-06 09:21:05","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1607454,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTablesandFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/14dbe5b235c24ddddebd80da.docx"},{"id":82042377,"identity":"003f3f55-43c5-4b66-964c-9345f6ff40f6","added_by":"auto","created_at":"2025-05-06 09:21:05","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":39051,"visible":true,"origin":"","legend":"","description":"","filename":"Tables1234.docx","url":"https://assets-eu.researchsquare.com/files/rs-5461492/v1/531f4b68f0cbb71161c19471.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Accuracy of lung ultrasound in diagnosing the etiology of respiratory symptoms in hospitalized children, a prospective cohort study","fulltext":[{"header":"Background","content":"\u003cp\u003eAs global child mortality declines, lower respiratory tract infections (LRTIs) remain a leading cause of preventable death in young children, especially in low- and middle-income countries.\u003csup\u003e1\u003c/sup\u003e Childhood pneumonia mortality has fallen thanks to vaccines (against Streptococcus pneumoniae and Haemophilus influenzae type b) and better diagnostics, antibiotics, and supportive care.\u003csup\u003e1\u0026ndash;3\u003c/sup\u003e However, disparities in healthcare access and overuse of antibiotics continue to exacerbate pediatric LRTI burdens, particularly in resource-limited regions.\u003csup\u003e4\u003c/sup\u003e Rising antimicrobial resistance tied to inappropriate antibiotic use underscores the need for accurate diagnosis of pediatric respiratory infections.\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eRapidly identifying the etiology of pediatric LRTIs is crucial for guiding appropriate treatment. Common pathogens include respiratory viruses (such as Respiratory Syncytial Virus (RSV) and influenza) and bacteria like S. pneumoniae and H. influenzae.\u003csup\u003e5\u003c/sup\u003e Distinguishing between bacterial and viral pneumonia remains challenging with current methods. Invasive lung sampling (e.g., bronchoscopy) is rarely feasible in children, so clinicians rely on indirect tests (nasopharyngeal swabs, blood cultures, sputum) with limited diagnostic accuracy.\u003csup\u003e6\u003c/sup\u003e For example, a large pneumonia etiology study found 66% of pediatric cases were viral and only 8% bacterial (with 7% mixed infections), illustrating the predominance of viral causes.\u003csup\u003e7\u003c/sup\u003e Current guidelines advise against antibiotic use for presumed viral pneumonia absent evidence of bacterial co-infection, underscoring the importance of an accurate diagnosis.\u003csup\u003e8\u003c/sup\u003e Yet many children with viral LRTIs still receive unnecessary antibiotics, fuelling antimicrobial resistance.\u003csup\u003e9,10\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eImaging is central in pediatric pneumonia diagnosis, but chest X-ray (CXR) has significant limitations.\u003csup\u003e11\u003c/sup\u003e CXR exposes children to ionizing radiation (raising lifetime cancer risk) and may not reliably distinguish pneumonia etiology or detect early lung changes.\u003csup\u003e12\u003c/sup\u003e Lung ultrasound (LUS) has emerged as a valuable alternative without these downsides. LUS is radiation-free, portable, and cost-effective; it allows bedside repeat exams to monitor illness or treatment response and is well-tolerated by children.\u003csup\u003e13\u003c/sup\u003e Over the past decade, LUS has been widely studied in children with pneumonia, demonstrating excellent diagnostic performance and safety.\u003csup\u003e14\u003c/sup\u003e Pediatricians are increasingly trained in LUS techniques; for example, the American Academy of Pediatrics and international critical care groups now advocate pediatric LUS use.\u003csup\u003e15\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eGrowing evidence supports LUS as an effective tool for evaluating pediatric respiratory conditions (bronchiolitis, pneumonia, pleural effusions).\u003csup\u003e11,16,17\u003c/sup\u003e Multiple studies show LUS often outperforms CXR in diagnosing pneumonia.\u003csup\u003e13,16,17\u003c/sup\u003e A systematic review of 1,510 children reported LUS sensitivity ~95.5% and specificity ~98.2%, versus 86.8% and 95.3% for CXR.\u003csup\u003e13\u003c/sup\u003e Similarly, a meta-analysis found LUS ~96% sensitive and 93% specific for pediatric pneumonia, prompting calls for broader LUS use in pediatrics.\u003csup\u003e15\u003c/sup\u003e Additionally, LUS can track disease course; one case series demonstrated that serial LUS effectively monitored antibiotic response in children with community acquired pneumonia, underscoring its clinical utility.\u003csup\u003e14\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eLUS also shows promise in differentiating bacterial vs. viral LRTIs, which is critical for selecting appropriate therapy. Different studies found that pediatric bacterial pneumonias tend to appear on LUS as solitary, larger, often unilateral consolidations, whereas viral pneumonias show smaller, multifocal or bilateral lesions.\u003csup\u003e18,19\u003c/sup\u003e A prospective study of 85 children identified distinct LUS patterns for viral, bacterial, and even SARS-CoV-2 pneumonias, highlighting LUS\u0026rsquo;s potential to distinguish infection etiologies.\u003csup\u003e20\u003c/sup\u003e LUS may also help predict disease severity; certain ultrasound features correlate with complicated (e.g., effusive) vs. uncomplicated pneumonia in children.\u003csup\u003e21\u003c/sup\u003e More recently, a detailed prospective analysis of pediatric LRTIs showed specific LUS findings tied to bacterial, viral, or atypical causes, aiding targeted management.\u003csup\u003e22\u003c/sup\u003e Nonetheless, an earlier PICU study reported that LUS alone had lower accuracy in determining pneumonia etiology when sonographers were blinded to clinical information, emphasizing the need to interpret LUS alongside clinical and lab data for optimal accuracy.\u003csup\u003e23\u003c/sup\u003e Overall, evidence remains limited on how well integrating LUS with clinical assessment and laboratory results can definitively pinpoint the cause of respiratory symptoms in hospitalized children, highlighting a gap for further research.\u003c/p\u003e"},{"header":"Methodology","content":"\u003cp\u003eThe Institutional Review Board (IRB) at the American University of Beirut (AUB) granted approval for this study. All methods were conducted in compliance with relevant guidelines and regulations. This single-center, prospective observational cohort study was carried out at the American University of Beirut Medical Center (AUBMC) in Lebanon between September 2022 and January 2023. In our study, the sample size was determined using a sample size calculator (https://wnarifin.github.io/ssc/sssnsp.html). We hypothesized that LUS accuracy would achieve a sensitivity of 95% and specificity of 95%, values consistent with those reported in the literature for diagnosing pneumonia with LUS. Given that the prevalence of LRTIs among hospitalized pediatric patients at our institution in the previous year was approximately 30%, and accounting for a 5% drop-out rate, a power of 90%, and an alpha value of 0.05, the estimated required sample size was 65 patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInclusion Criteria:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEligible patients were children aged 0 to 19 years who presented with respiratory symptoms\u0026nbsp;, including any combination of cough, rhinorrhea, nasal congestion, shortness of breath, wheezing, tachypnea, or increased work of breathing\u0026nbsp;\u0026mdash; and were within 24 hours of admission to the Pediatric Ward or Pediatric Intensive Care Unit or had admission orders but were boarding in the Emergency Department (ED). To identify eligible patients, the investigative team\u0026mdash;excluding the principal investigator\u0026mdash;routinely accessed the EPIC electronic health record system. \u0026nbsp;Following approval by the patient\u0026apos;s primary physician, the study was explained to the parents or guardians, and written informed consent was obtained.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExclusion criteria\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePatients were excluded from the study if they had a pre-existing chronic lung disease, as these underlying abnormalities could significantly alter lung ultrasound interpretation, limiting the generalizability of findings to the healthy pediatric population. Patients were also excluded if the LUS was not performed within 24 hours of admission to ensure alignment with the initial clinical presentation minimizing temporal bias, if they were admitted under the PI\u0026rsquo;s care to avoid potential bias, or if the PI was not blinded to the CXR findings to maintain blinding and minimize interpretation bias.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Collection\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our study, patients were recruited after being admitted to the hospital, making it unfeasible to perform LUS before CXR to ensure blinding of the investigator. As part of their initial workup, patients underwent CXR in the ED upon admission. Patients were excluded if they were admitted under the PI\u0026rsquo;s care or if the PI was not blinded to the CXR findings \u0026mdash; as shown in Figure 4. The PI was informed about the clinical presentation and laboratory findings of the patients but was not provided with CXR findings by the treating team. The PI performed the LUS and then submitted the data collection form to the research assistant. The PI could access EPIC only after completing the LUS and submitting the data collection form. Additionally, LUS images were stored in separate software, distinct from EPIC. The expert committee, responsible for the final clinical diagnosis, had no access to LUS images or reports (Supplementary Figure S1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAlthough the study algorithm (Figure 3) recommended repeating LUS within 48 hours in cases of diagnostic uncertainty, follow-up scans were not systematically performed or recorded as part of the study protocol. In most cases, a repeat LUS was not requested by the clinical team. As such, any follow-up LUS performed was used solely for clinical management purposes and was not included in the data collection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLung Ultrasound Procedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe principal investigator, credentialed in lung ultrasound with over three years of experience, performed the LUS within 24 hours of the admission order.\u0026nbsp;Ultrasound scans were performed using the portable CX50 ultrasound system (Philips, Bothell, WA, USA), equipped with a linear transducer (L: 12-3 MHz) and a convex transducer (C: 8-5 MHz). Due to the small body size of the participants, all participants were tested using the L: 12-3 MHz linear transducer at a 5 cm depth or less. Each lung was scanned in both longitudinal and transverse views, covering the midclavicular line anteriorly, the paraspinal line posteriorly, and the mid-axillary line (Figure 1). Each view included both the upper and lower zones, resulting in a total of 12 scanning zones. This protocol was based on the method described by Copetti and Catarossi.\u003csup\u003e24\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLUS Findings\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003eA-lines: Horizontal artifacts appearing as a series of echogenic lines parallel and equidistant from the pleural line, indicating normal lung tissue (Figure 2A).\u003c/li\u003e\n \u003cli\u003eB-lines: Vertical comet-tail artifacts originating from the pleural line. According to Lichtenstein\u0026rsquo;s classification, B-lines indicate fluid-filled interlobular septae surrounded by air, associated with various pathological conditions (Figure 2B).\u003csup\u003e25\u003c/sup\u003e\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eSubpleural Consolidation (sub-centimetric): Localized beneath the pleura, typically less than 1 cm in size (Figure 2C).\u003c/li\u003e\n \u003cli\u003eConsolidation: Refers to lung tissue becoming firm due to the accumulation of fluid or cellular material, typically measuring equal to or greater than 1 cm in size (Figure 2D and 2E).\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cstrong\u003eLUS Etiology Diagnosis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our study, the initial classification of patients into infectious versus non-infectious etiologies was based on clinical judgment, as this distinction was essential for applying our diagnostic algorithm (Figure 3). Specifically, patients classified with infectious etiology were evaluated using the pathway outlined in Figure 3A, while those with a non-infectious etiology were assessed using Figure 3B to guide the interpretation of LUS findings.\u003c/p\u003e\n\u003cp\u003eFor instance, patients presenting with cough, chest pain, or shortness of breath along with fever and/or a positive respiratory viral test were considered to have an infectious etiology. In contrast, patients with a known history of asthma who exhibited acute respiratory deterioration, clinical wheezing, and improvement with bronchodilators\u0026mdash;regardless of the presence of fever\u0026mdash;were classified as having asthma exacerbations, which we considered a non-infectious etiology.\u003c/p\u003e\n\u003cp\u003eImportantly, we did not rely on laboratory markers such as complete blood count, C-reactive protein (CRP), or absolute neutrophil counts (ANC) to differentiate between viral and bacterial infections. Instead, lung ultrasound findings were central to differentiating between viral and bacterial lower respiratory tract infections.\u0026nbsp;In our study, we classified \u003cem\u003eMycoplasma pneumoniae\u003c/em\u003e under the\u0026nbsp;bacterial infection\u0026nbsp;category. Recent literature supports that \u003cem\u003eMycoplasma pneumoniae\u003c/em\u003e can present with LUS features similar to typical bacterial pneumonia, including lobar consolidations, dynamic air bronchograms, and pleural effusions. For instance, Buonsenso et al. reported that children with bacterial and atypical acute lower respiratory tract infections exhibited larger consolidations, often exceeding 1.5 cm, with dynamic air bronchograms and pleural effusions. These findings suggest that \u003cem\u003eM. pneumoniae\u003c/em\u003e infections may share sonographic characteristics with bacterial pneumonia.\u003csup\u003e18\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eInfectious Etiology\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eFor suspected infectious etiology, further classification was based on the clinical stability. Clinically unstable patients were typically treated in the PICU and were more likely to receive antibiotics presumptively for bacterial pneumonia. For stable patients with suspected LRTIs, the first LUS assessment involved checking for the presence of B-lines (Figure 3).\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cu\u003eAbsent B-lines\u0026nbsp;\u003c/u\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eIn cases where B-lines were absent and no consolidations were detected, the diagnosis was consistent with upper respiratory tract infection (URTI).\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cu\u003eUnilateral B-lines\u003c/u\u003e:\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eUnilateral, focal B-lines with a consolidation \u0026ge;1 cm, whether accompanied by pleural effusion and dynamic air bronchograms, pointed to bacterial pneumonia. When the air bronchograms were static or absent (lung hepatization), the cause was likely either bacterial pneumonia or atelectasis.\u003csup\u003e22,26\u0026ndash;28\u003c/sup\u003e In cases of unilateral, focal B-lines without consolidation or with small consolidations (\u0026lt;1 cm), the condition was typically classified as either a mild viral infection or an early-stage bacterial infection. A follow-up LUS was recommended after 48 hours. If the follow-up showed progression to bilateral B-lines or improvement without antibiotic treatment, a viral etiology was presumed. Conversely, if the consolidation expanded to \u0026ge;1 cm, bacterial pneumonia was diagnosed. Clinically unstable patients or those diagnosed with bacterial, complicated, or superimposed bacterial pneumonia were evaluated for antibiotic therapy (Figure 3A).\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cu\u003eBilateral B-lines\u003c/u\u003e:\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe presence of bilateral B-lines, with or without consolidation measuring less than 1 cm, was interpreted as indicative of viral changes, suggesting a diagnosis of either bronchiolitis for patients under 2 years old or viral bronchitis for those aged 2 and older.\u003csup\u003e19,22,29,30\u003c/sup\u003e While consolidation measuring 1 cm or larger typically indicates bacterial pneumonia, the simultaneous presence of bilateral B-lines raised the possibility of viral pneumonia.\u0026nbsp;In these patients, clinical deterioration or complications such as pleural effusion or abscess formation, were suggestive of a bacterial pneumonia. Conversely, in the absence of these signs, a diagnosis of viral pneumonia was made, with a follow-up lung ultrasound recommended within 48 hours if deemed necessary. Should the repeat LUS show worsening, bacterial pneumonia was suspected. If there was no change or improvement without antibiotics, the diagnosis remained viral pneumonia (Figure 3A).\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eNon-infectious etiology\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003ePatients with suspected non-infectious etiology were evaluated for conditions such as atelectasis, lung contusion, cardiogenic pulmonary edema, pulmonary embolism, chronic lung disease, or lung masses. However, our study specifically focused on individuals with LRTIs. Among these, we included patients who clinically presented with asthma exacerbations, whether triggered by infections or allergens, as suggested by their medical histories or clinical signs such as a sudden onset of shortness of breath and wheezing observed during physical examinations. In such instances, LUS findings could display a wide range of abnormalities, from normal results to various forms of consolidation, B-lines, or pleural effusion.\u003csup\u003e31\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eIf the consolidation measured less than 1 cm, the etiology diagnosis was considered asthma exacerbation. This diagnosis was also applicable in cases of consolidation equal to or larger than 1 cm, with or without static air bronchograms; however, a repeat ultrasound was advised within 48 hours if clinically indicated. Should the repeat ultrasound show that the consolidation improved without antibiotic treatment, the diagnosis of asthma exacerbation was considered. Conversely, if the LUS findings showed worsening, it was interpreted as a possible superimposed infection.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn patients with a clinical diagnosis of asthma exacerbation, a large or moderate effusion was considered indicative of bacterial infection and treated accordingly. For small pleural effusion a follow-up LUS within 48 hours was recommended. If the subsequent ultrasound demonstrated no change or improvement without antibiotics, the diagnosis of asthma exacerbation was supported. Conversely, if the pleural effusion worsened, a superimposed bacterial pneumonia was suspected and antibiotic therapy initiated (Figure 3B).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFinal Clinical Diagnosis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA committee of pediatric specialists, comprising a pulmonologist and a critical care physician, independently reviewed each patient\u0026rsquo;s complete clinical course post-discharge, remaining blinded to lung ultrasound findings. They determined the etiology\u0026mdash;viral, bacterial, or asthma\u0026mdash;through a comprehensive assessment that included the patient\u0026rsquo;s history, clinical presentation, physical examination, laboratory tests (such as complete blood count, C-reactive protein, viral panels via Polymerase Chain Reaction (PCR) or antigen detection, and available cultures, and chest X-ray findings. This rigorous process facilitated consensus on the final clinical diagnosis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData were collected using a standardized form\u003cstrong\u003e.\u0026nbsp;\u003c/strong\u003eData management and analyses were done using IBM-SPSS software (version 28.0, Armonk, NY, USA)\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eby a designated member of the research team who was not involved in data collection in order to minimize bias in data entry.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDescriptive Statistics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDemographic and other baseline characteristics were summarized using descriptive statistics. Continuous variables were presented as means and standard deviations (SD) if normally distributed or medians and interquartile ranges (IQR) for skewed distributions. Categorical variables were reported as frequencies and percentages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDiagnostic Accuracy Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diagnostic accuracy of LUS in identifying the etiology of respiratory symptoms was assessed using the final clinical diagnosis as the reference gold standard and served as the primary outcome of the study. The secondary outcome was the comparison of LUS accuracy versus CXR accuracy, both against the final clinical diagnosis as the gold standard.\u0026nbsp;This analysis was conducted both for overall and for specific etiological subgroups, including viral, bacterial, and asthma exacerbation. Sensitivity and specificity for each modality were calculated along with 95% confidence intervals to assess the precision of these estimates. The overall agreement between each diagnostic tool and the final clinical diagnosis was quantified using Cohen\u0026rsquo;s kappa coefficient (\u0026kappa;), with 95% confidence intervals provided to evaluate the consistency of the agreement. The interpretation of \u0026kappa; followed Cohen\u0026apos;s guidelines: values \u0026le; 0 indicating no agreement, 0.01-0.20 slight agreement, 0.21-0.40 fair agreement, 0.41-0.60 moderate agreement, 0.61-0.80 substantial agreement, and 0.81-1.00 almost perfect agreement.\u003csup\u003e32\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComparative Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our study, 11 out of the 65 patients included in the statistical analysis underwent only LUS because CXRs were not requested. To compare the accuracy of LUS and CXR, we employed the McNemar test specifically on the subset of patients (n=54) who underwent both diagnostic procedures. This test assessed whether the proportion of correct diagnoses differed significantly between the two methods. A two-tailed p-value of less than 0.05 was considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe study enrolled 99 patients who were admitted with respiratory symptoms between September 2022 and January 2023. Out of these, 65 patients met the inclusion criteria and were included in the final analysis (Figure 4). The median age of these patients was 2 years (IQR, 0.46-2 years), with 37 patients being males (57%) (Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe mean duration for the lung ultrasound procedure was 8.4 minutes (SD = 2.5 minutes). Among the included patients, 62 were diagnosed with lower respiratory tract disease (LRTD), accounting for 95.4% of cases, while the remaining 3 patients (4.6%) had upper respiratory tract infections.\u003c/p\u003e\n\u003cp\u003eRegarding the etiology of LRTD, 12 patients (19%) had bacterial infection (pneumonia), 43 (70%) had viral infections (bronchiolitis, bronchitis, viral pneumonia), and 7 (11%) were diagnosed with asthma exacerbations (Figure 4). Among the 43 patients with viral infections, 7 had viral pneumonia with a consolidation size greater than 1 cm (median diameter of 2 cm).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBlood cultures were obtained for 47 patients, of whom only 4 had positive results: one patient had \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e and \u003cem\u003eE. coli\u003c/em\u003e; one had \u003cem\u003eCorynebacterium afermentans\u003c/em\u003e; one had \u003cem\u003eAcinetobacter variabilis\u003c/em\u003e; and one had \u003cem\u003eStreptococcus infantis\u003c/em\u003e and \u003cem\u003eStaphylococcus haemolyticus\u003c/em\u003e. Viral testing (rapid antigen or PCR) was performed in 54 patients (83%), with respiratory syncytial virus (RSV) being the most commonly identified virus (12 patients), followed by rhinovirus (5), adenovirus (3), and influenza A (2). Of these, 13 patients were tested using a multiplex PCR assay, and none tested positive for \u003cem\u003eMycoplasma pneumoniae\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eA supplementary table (S2) presents selected laboratory parameters\u0026mdash;including white blood cell count, CRP, ANC, and neutrophil-to-lymphocyte ratio (NLR)\u0026mdash;among patients with final diagnoses of asthma, bacterial infection, and viral infection, to describe observable trends. CXR findings are summarized in Table 3.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrimary Outcome\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diagnostic accuracy of LUS compared to the final clinical diagnosis revealed that LUS had a sensitivity of 91.9% [95% CI: 85.2\u0026ndash;98.7%] and a specificity of 100% [95% CI: 36.8\u0026ndash;100.0%]. The agreement between LUS and the final clinical diagnosis, as measured by Cohen\u0026rsquo;s kappa, was 0.86 [95% CI: 0.77\u0026ndash;0.94], indicating almost perfect agreement (Table 2).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor patients with bacterial pneumonia, the sensitivity and specificity of LUS were 91.8% [95% CI: 61.6\u0026ndash;99.8%] and 92.5% [95% CI: 61.6\u0026ndash;99.8%], respectively. For patients with viral infections (bronchiolitis, bronchitis, viral pneumonia), the sensitivity and specificity were 90.7% [95% CI: 82.0\u0026ndash;99.4%] and 95.5% [95% CI: 77.2\u0026ndash;99.9%], respectively. For patients with asthma exacerbation, both sensitivity and specificity were 100.0% [95% CI: 65.2\u0026ndash;100.0%] and 100.0% [95% CI: 95.0\u0026ndash;100.0%], respectively (Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary Outcome\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAmong patients who underwent both diagnostic procedures (n=54), the accuracy of diagnosis was significantly higher with LUS compared to CXR.\u0026nbsp;The McNemar chi-squared statistic was \u0026chi;\u0026sup2; = 10.89, with a corresponding p-value = 0.001.\u0026nbsp;The kappa coefficient for LUS was 0.84 [95% CI: 0.70\u0026ndash;0.98], whereas CXR had a kappa of 0.47 [95% CI: 0.30\u0026ndash;0.65]. The sensitivity and specificity of LUS were 90.2% [95% CI: 82.0\u0026ndash;98.3%] and 100% [95% CI: 36.8\u0026ndash;100.0%], respectively. In comparison, CXR had a sensitivity of 62.7% [95% CI: 49.5\u0026ndash;76.0%] and a specificity of 100% [95% CI: 36.8\u0026ndash;100.0%] (Tables 3 and 4).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSubgroup Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our study, we performed a subgroup analysis to evaluate the diagnostic accuracy of LUS in different age groups. For patients under 2 years of age (n=32), the total accuracy of LUS in diagnosing the etiology of respiratory symptoms was 90.6% (95% CI: 80.5, 100.0). The sensitivity for identifying bacterial infections was 66.7% (95% CI: 13.0, 100.0), while sensitivity for viral infections was high at 92.9% (95% CI: 83.0, 100.0). The specificity for bacterial infections was 93.1% (95% CI: 83.9, 100.0), but the specificity for viral infections was lower at 75% (95% CI: 33.0, 99.9).\u003c/p\u003e\n\u003cp\u003eIn contrast, for patients aged 2 years and older (n=33), LUS showed a higher total accuracy of 93.9% (95% CI: 85.0, 100.0). The sensitivity for bacterial infections was perfect at 100% (95% CI: 100.0, 100.0), while the sensitivity for viral infections was 86.7% (95% CI: 69.0, 100.0). The specificity for bacterial infections was 91.7% (95% CI: 80.6, 100.0), and for viral infections, it was 100% (95% CI: 84.7, 100.0).\u0026nbsp;(Supplementary Table S3)\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis prospective study demonstrates point-of-care lung ultrasound can accurately identify the etiology of respiratory symptoms in hospitalized children when interpreted alongside clinical information. We observed an overall agreement of 92.3% between LUS-based diagnoses and the final expert clinical diagnoses, with a near-perfect kappa of 0.86. Notably, LUS maintained high diagnostic performance across different causes of illness: it showed strong agreement in detecting bacterial pneumonia and viral lower respiratory infections (each with sensitivities around 90% and substantial agreement, \u0026kappa; ~0.8) as well as asthma exacerbations (100% agreement with final diagnoses). We also noted a slight trend toward higher accuracy in older children (\u0026ge;2 years) compared to infants, although LUS still performed well even in the under-2 age group. Collectively, these results position LUS as a reliable tool for pediatric respiratory diagnostics in the acute care setting.\u003c/p\u003e\n\u003cp\u003eOur findings build on the growing evidence that LUS is an effective, child-friendly alternative to radiography for evaluating pediatric lung disease. Previous studies and meta-analyses have reported that LUS can diagnose pediatric pneumonia with high sensitivity and specificity (in the 93\u0026ndash;98% range), often outperforming CXR.\u003csup\u003e11,15\u0026ndash;17\u003c/sup\u003e However, those studies largely did not distinguish between bacterial and viral etiologies of pneumonia. Our study uniquely addresses this gap by demonstrating that integration of LUS results with clinical and laboratory findings can achieve accurate etiological differentiation at the bedside. This integrated approach is crucial, as an earlier pediatric ICU study showed that when sonographers were blinded to clinical context, LUS alone had markedly lower diagnostic accuracy (sensitivity dropping to ~56%) for identifying the cause of respiratory failure.\u003csup\u003e23\u003c/sup\u003e By contrast, our methodology\u0026mdash;incorporating patient history, exam findings, and basic labs in the LUS interpretation\u0026mdash;yielded near-perfect agreement with expert diagnoses, underscoring the importance of context-driven ultrasound interpretation.\u003csup\u003e18\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eConsistent with prior literature, we found distinct lung ultrasound patterns corresponding to different infection types, which aided our diagnostic accuracy. Studies have documented that bacterial pneumonias in children often appear on LUS as larger, solitary consolidations (frequently confined to one lung), whereas viral pneumonias tend to produce smaller, multifocal or bilateral lesions.\u003csup\u003e18,19\u003c/sup\u003e We observed similar trends in our cohort: for example, cases classified as bacterial infection typically had focal lobar consolidation on LUS, while viral cases showed diffuse B-lines or subpleural consolidations scattered in both lungs. A recent prospective study of pediatric pneumonias described comparably distinct LUS profiles for viral, typical bacterial, and even COVID-19 pneumonias, reinforcing the notion that ultrasound features can suggest the underlying pathogen.\u003csup\u003e20\u003c/sup\u003e In our data, LUS was highly sensitive in detecting viral bronchiolitis and viral pneumonia (over 90% sensitivity), which is on par with the 90.4% sensitivity reported by Caiulo et al. for bronchiolitis; notably, in that study LUS substantially outperformed CXR (73.1% sensitivity) for diagnosing bronchiolitis.\u003csup\u003e33\u003c/sup\u003e This high accuracy for viral illnesses is especially valuable, as it can help clinicians confidently identify viral infections and avoid unnecessary antibiotic use in these cases.\u003csup\u003e7,9,10\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eOur results also challenge a common assumption in lung ultrasound interpretation\u0026mdash;that the presence of a large consolidation (\u0026ge;1\u0026nbsp;cm) definitively indicates bacterial pneumonia. We identified several children with viral infections who had consolidations \u0026ge;1\u0026nbsp;cm on LUS (median ~2\u0026nbsp;cm); all of these patients recovered fully without antibiotics. This finding underscores that sizable consolidations can occur in viral pneumonia or severe bronchiolitis. Similarly, Berce et al. reported a median consolidation size of about 1.5\u0026nbsp;cm in confirmed viral pneumonias.\u003csup\u003e19\u003c/sup\u003e Buonsenso et al. also reported patients with viral pneumonia who had consolidation size of \u0026gt; 1cm.\u003csup\u003e18\u003c/sup\u003e Thus, consolidation size alone should not be viewed as a strict criterion for bacterial infection. Rather, the overall ultrasound pattern and clinical picture must be considered\u0026mdash;some of these larger consolidations in viral infections may represent inflammatory or atelectatic changes rather than focal bacterial lobar pneumonia. In practice, for a stable patient with predominantly viral-appearing LUS findings (for instance, diffuse bilateral changes) but an unexpected large consolidation, a prudent approach would be watchful waiting with supportive care and a repeat LUS after 24\u0026ndash;48 hours. Such a strategy, as we employed in our study, can ensure bacterial superinfection is not developing while sparing the child from immediate empiric antibiotics, thereby reducing unnecessary antibiotic exposure. Repeating LUS within 48 hours in our algorithm aligns with the findings of Buonsenso et al., who demonstrated that certain LUS patterns, both at diagnosis and 48 hours after treatment, were better predictors of antibiotic response in children with acute lower respiratory tract infections than clinical data and laboratory results.\u003csup\u003e14,21\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eIn our subgroup analysis, the diagnostic performance of LUS varied notably by age group. Among children under 2 years of age, the sensitivity for identifying bacterial infections was 66.7% (95% CI: 13.0\u0026ndash;100.0), and the specificity for viral infections was 75% (95% CI: 33.0\u0026ndash;99.9). These modest values are likely influenced by the small sample size in this subgroup\u0026mdash;only three patients had bacterial infections, and four were negative for viral infections\u0026mdash;limiting the statistical reliability of the estimates. In contrast, for children aged 2 years and older, LUS showed excellent diagnostic accuracy, with both the sensitivity for bacterial infections and the specificity for viral infections reaching 100% (95% CI: 100.0\u0026ndash;100.0 for bacterial infections and 95% CI: 84.7\u0026ndash;100.0 for viral infections). These findings suggest that LUS may be more reliable in older children, potentially due to more distinguishable clinical and sonographic features. However, larger studies across all pediatric age groups are needed to validate these results and better define the role of LUS in differentiating between bacterial and viral lower respiratory tract infections.\u003c/p\u003e\n\u003cp\u003eBeyond infections, this study demonstrates that LUS can contribute to the evaluation of non-infectious respiratory disorders like acute asthma exacerbations. We correctly identified all cases of asthma exacerbation using our LUS protocol in conjunction with clinical history, achieving perfect agreement (\u0026kappa; = 1) with the final diagnoses. This aligns with prior observations that children with asthma attacks can exhibit certain LUS abnormalities (such as bilateral comet-tail B-lines or small subpleural consolidations due to airway plugging), even though these ultrasound findings are not specific to asthma.\u003csup\u003e31\u003c/sup\u003e Without knowledge of the clinical context, such LUS findings could be misinterpreted as pneumonia or edema, which highlights a key point: LUS is most effective as an adjunct to, not a replacement for, clinical assessment. DeSanti et al. found that when LUS interpreters were blinded to clinical information, they correctly identified only 6 of 10 pediatric status asthmaticus cases based on ultrasound alone.\u003csup\u003e23\u003c/sup\u003e In our study, by factoring in the patients\u0026rsquo; known history of asthma and presenting signs (wheezing responsive to bronchodilators), we avoided this pitfall. The implication is that while LUS may not have pathognomonic signs for asthma, it can still be very useful in the emergency setting to support an asthma diagnosis or exclude other causes of respiratory distress, especially when integrated with clinical judgement.\u003c/p\u003e\n\u003cp\u003eAnother important finding is that LUS outperformed chest X-ray in determining the cause of pediatric respiratory illness. We found LUS to be significantly more accurate than CXR (p \u0026lt; 0.001) for identifying the etiologic diagnosis, consistent with prior comparative studies. In our cohort, CXR results showed only moderate agreement with the final clinical diagnosis (\u0026kappa; ~0.47), considerably lower than the near-perfect concordance achieved with LUS. This result reinforces the growing consensus that lung ultrasound can replace or complement radiography for pediatric pneumonia diagnosis.\u003csup\u003e15,17\u003c/sup\u003e Given the cumulative risks of radiation in young patients, the superior performance of LUS that we and others have observed further supports adopting ultrasound as a first-line imaging modality in children with acute respiratory symptoms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStrengths\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur study addresses a significant gap in the literature by demonstrating the high diagnostic accuracy of lung ultrasound when combined with clinical assessment, particularly in differentiating between bacterial and viral infections in patients with respiratory symptoms. A key strength of our research is its prospective cohort design, which minimized recall bias and established a clear temporal relationship between diagnostics and outcomes.\u003csup\u003e34\u003c/sup\u003e Additionally, lung ultrasounds were performed by a single, experienced investigator who was blinded to the chest X-ray results, thereby reducing both diagnostic and observer bias.\u003csup\u003e35\u003c/sup\u003e The implementation of standardized diagnostic algorithms facilitated a systematic and replicable integration of clinical, laboratory, and LUS findings, ultimately enhancing diagnostic accuracy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhile our study provides valuable findings, it has several limitations. Firstly, it was conducted at a single tertiary care center, which may restrict the generalizability of our findings.\u003csup\u003e36\u003c/sup\u003e Furthermore, the relatively small sample size, predominantly consisting of patients with viral infections, could affect the calculations of sensitivity and specificity.\u003csup\u003e37\u003c/sup\u003e Furthermore, the lung ultrasounds were performed and interpreted by an investigator with advanced training in pediatric LUS; this level of expertise may not be available in all hospitals, and it underscores the need for adequate training programs if LUS is to be widely implemented.\u003csup\u003e38\u003c/sup\u003e In addition, the standard reference used in our study was the final clinical diagnosis established by two experts\u0026mdash;a pediatric pulmonologist and a pediatric critical care physician\u0026mdash;which may not reflect the gold standard for detecting the etiology of pulmonary diseases. Moreover, three patients were excluded from the study because the ultrasound machine was not accessible due to the lack of a dedicated ultrasound machine for the research study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFuture directions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFuture research should involve larger, multicenter studies across diverse populations and clinical settings to validate our results and confirm the generalizability of LUS as a diagnostic modality in various healthcare environments.\u003c/p\u003e\n\u003cp\u003eFurther investigations are necessary to clarify the clinical relevance of sub-centimetric consolidations with or without air bronchograms observed on LUS. Determining whether these findings warrant antibiotic treatment or support a \u0026quot;watchful waiting\u0026quot; approach is crucial for guiding clinical decision-making and reducing unnecessary antibiotic use.\u003c/p\u003e\n\u003cp\u003eAdditionally, research is needed to redefine the size criteria of consolidations on lung ultrasound in cases of viral pneumonia. Our findings suggest that consolidations larger than 1 cm may still be associated with viral infections, challenging the traditional limitation to smaller consolidations. Clarifying these criteria could improve diagnostic accuracy and patient management strategies. Evaluations are also warranted to elucidate the LUS findings of atypical pathogens like Mycoplasma pneumoniae in children. Furthermore, larger, multicenter studies should systematically evaluate how age-related anatomical differences (e.g., chest wall thickness) impact LUS interpretation across age groups.\u003c/p\u003e\n\u003cp\u003eFinally, the diagnostic algorithms (Figure 3) we utilized in this study for determining the etiology of infectious and non-infectious respiratory disorders showed promising results. Future studies should aim to validate the reliability and applicability of these algorithms in different clinical settings. By confirming their effectiveness, these diagnostic pathways could be integrated into clinical practice and incorporated into machine learning and artificial intelligence frameworks, thereby enhancing the utility of LUS in managing pediatric respiratory conditions.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study demonstrates that LUS is a highly accurate and practical tool for diagnosing the etiology of respiratory symptoms in hospitalized children, outperforming chest X-ray in terms of diagnostic accuracy. By combining LUS findings with clinical and laboratory data, we were able to differentiate between bacterial and viral infections effectively, supporting more judicious use of antibiotics and improving antimicrobial stewardship. The use of LUS also eliminates the risks associated with radiation exposure from chest X-rays, making it a safer and more reliable imaging modality, especially for pediatric patients. Our findings highlight the potential of LUS as a first-line diagnostic tool in the management of pediatric respiratory diseases. Further studies, particularly multicenter trials with larger sample sizes, are needed to validate these results and refine diagnostic criteria for LUS, ultimately improving its implementation in clinical practice.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by a seed grant from the American University of Beirut Medical Center (Medical Practice Plan - MPP Fund 320213). AI would like to acknowledge the training received under the Faculty Advancement Program (FAP) at the American University of Beirut. Special thanks to Dr. Mona Nabulsi for her mentorship and support in developing the proposal of this paper. The content is solely the responsibility of the authors.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contribution \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.I., C.S., and N.A. contributed to the research question, proposal writing, and data collection. A.I. and D.K. contributed to the creation of algorithms. A.I. and Z.S. were responsible for data entry. A.I. and Z.M. carried out the statistical analysis. A.I., C.S., D.K., Z.S., F.T., A.M.H. and M.S. contributed to manuscript writing. A.I., D.K., H.R., J.A., R.S., S.K., A.M.H., and A.H.H. were involved in the literature review. N.G., J.M., S.M. and M.M. provided supervision and guidance. All authors reviewed and edited the draft and agreed to submit the manuscript for publication.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interest:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are not publicly available due to institutional requirements but are available from the corresponding author upon reasonable request.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePerin, J. \u003cem\u003eet al.\u003c/em\u003e Global, regional, and national causes of under-5 mortality in 2000\u0026ndash;19: an updated systematic analysis with implications for the Sustainable Development Goals. \u003cem\u003eLancet Child Adolesc. Health\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 106\u0026ndash;115 (2022).\u003c/li\u003e\n\u003cli\u003ePavia, M., Bianco, A., Nobile, C. G. A., Marinelli, P. \u0026amp; Angelillo, I. F. Efficacy of Pneumococcal Vaccination in Children Younger Than 24 Months: A Meta-Analysis. \u003cem\u003ePediatrics\u003c/em\u003e \u003cstrong\u003e123\u003c/strong\u003e, e1103\u0026ndash;e1110 (2009).\u003c/li\u003e\n\u003cli\u003eObonyo, C. O. \u0026amp; Lau, J. Efficacy of Haemophilus influenzae type b vaccination of children: a meta-analysis. \u003cem\u003eEur. J. Clin. Microbiol. Infect. Dis.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 90\u0026ndash;97 (2006).\u003c/li\u003e\n\u003cli\u003eGBD 2016 Lower Respiratory Infections Collaborators. Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory infections in 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. \u003cem\u003eLancet Infect. Dis.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 1191\u0026ndash;1210 (2018).\u003c/li\u003e\n\u003cli\u003eUS Burden of Disease Collaborators \u003cem\u003eet al.\u003c/em\u003e The State of US Health, 1990-2016: Burden of Diseases, Injuries, and Risk Factors Among US States. \u003cem\u003eJAMA\u003c/em\u003e \u003cstrong\u003e319\u003c/strong\u003e, 1444\u0026ndash;1472 (2018).\u003c/li\u003e\n\u003cli\u003eGereige, R. S. \u0026amp; Laufer, P. M. Pneumonia. \u003cem\u003ePediatr. Rev.\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 438\u0026ndash;456; quiz 455\u0026ndash;456 (2013).\u003c/li\u003e\n\u003cli\u003eJain, S. \u003cem\u003eet al.\u003c/em\u003e Community-Acquired Pneumonia Requiring Hospitalization among U.S. Children. \u003cem\u003eN. Engl. J. Med.\u003c/em\u003e \u003cstrong\u003e372\u003c/strong\u003e, 835\u0026ndash;845 (2015).\u003c/li\u003e\n\u003cli\u003eBradley, J. S. \u003cem\u003eet al.\u003c/em\u003e The Management of Community-Acquired Pneumonia in Infants and Children Older Than 3 Months of Age: Clinical Practice Guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. \u003cem\u003eClin. Infect. Dis.\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, e25\u0026ndash;e76 (2011).\u003c/li\u003e\n\u003cli\u003ePratt, M. T. G. \u003cem\u003eet al.\u003c/em\u003e Prevalence of respiratory viruses in community-acquired pneumonia in children: a systematic review and meta-analysis. \u003cem\u003eLancet Child Adolesc. Health\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 555\u0026ndash;570 (2022).\u003c/li\u003e\n\u003cli\u003eGerber, J. S. \u003cem\u003eet al.\u003c/em\u003e Antibiotic Stewardship in Pediatrics. \u003cem\u003ePediatrics\u003c/em\u003e \u003cstrong\u003e147\u003c/strong\u003e, e2020040295 (2021).\u003c/li\u003e\n\u003cli\u003eJones, B. P. \u003cem\u003eet al.\u003c/em\u003e Feasibility and Safety of Substituting Lung Ultrasonography for Chest Radiography When Diagnosing Pneumonia in Children. \u003cem\u003eChest\u003c/em\u003e \u003cstrong\u003e150\u003c/strong\u003e, 131\u0026ndash;138 (2016).\u003c/li\u003e\n\u003cli\u003eKutanzi, K., Lumen, A., Koturbash, I. \u0026amp; Miousse, I. Pediatric Exposures to Ionizing Radiation: Carcinogenic Considerations. \u003cem\u003eInt. J. Environ. Res. Public. Health\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1057 (2016).\u003c/li\u003e\n\u003cli\u003eBalk, D. S. \u003cem\u003eet al.\u003c/em\u003e Lung ultrasound compared to chest X-ray for diagnosis of pediatric pneumonia: A meta-analysis. \u003cem\u003ePediatr. Pulmonol.\u003c/em\u003e \u003cstrong\u003e53\u003c/strong\u003e, 1130\u0026ndash;1139 (2018).\u003c/li\u003e\n\u003cli\u003eBuonsenso, D. \u003cem\u003eet al.\u003c/em\u003e The Use of Lung Ultrasound to Monitor the Antibiotic Response of Community-Acquired Pneumonia in Children: A Preliminary Hypothesis. \u003cem\u003eJ. Ultrasound Med.\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 817\u0026ndash;826 (2020).\u003c/li\u003e\n\u003cli\u003ePereda, M. A. \u003cem\u003eet al.\u003c/em\u003e Lung Ultrasound for the Diagnosis of Pneumonia in Children: A Meta-analysis. \u003cem\u003ePediatrics\u003c/em\u003e \u003cstrong\u003e135\u003c/strong\u003e, 714\u0026ndash;722 (2015).\u003c/li\u003e\n\u003cli\u003eYang, Y., Wu, Y. \u0026amp; Zhao, W. Comparison of lung ultrasound and chest radiography for detecting pneumonia in children: a systematic review and meta-analysis. \u003cem\u003eItal. J. Pediatr.\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 12 (2024).\u003c/li\u003e\n\u003cli\u003eHeuvelings, C. C. \u003cem\u003eet al.\u003c/em\u003e Chest ultrasound for the diagnosis of paediatric pulmonary diseases: a systematic review and meta-analysis of diagnostic test accuracy. \u003cem\u003eBr. Med. Bull.\u003c/em\u003e \u003cstrong\u003e129\u003c/strong\u003e, 35\u0026ndash;51 (2019).\u003c/li\u003e\n\u003cli\u003eBuonsenso, D. \u003cem\u003eet al.\u003c/em\u003e Role of lung ultrasound for the etiological diagnosis of acute lower respiratory tract infection (ALRTI) in children: a prospective study. \u003cem\u003eJ. Ultrasound\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, 185\u0026ndash;197 (2022).\u003c/li\u003e\n\u003cli\u003eBerce, V., Tomazin, M., Gorenjak, M., Berce, T. \u0026amp; Lovrenčič, B. The Usefulness of Lung Ultrasound for the Aetiological Diagnosis of Community-Acquired Pneumonia in Children. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 17957 (2019).\u003c/li\u003e\n\u003cli\u003eStoicescu, E. R. \u003cem\u003eet al.\u003c/em\u003e Differentiating Viral from Bacterial Pneumonia in Children: The Diagnostic Role of Lung Ultrasound\u0026mdash;A Prospective Observational Study. \u003cem\u003eDiagnostics\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 480 (2024).\u003c/li\u003e\n\u003cli\u003eMusolino, A. M. \u003cem\u003eet al.\u003c/em\u003e Lung ultrasound features of children with complicated and noncomplicated community acquired pneumonia: A prospective study. \u003cem\u003ePediatr. Pulmonol.\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 1479\u0026ndash;1486 (2019).\u003c/li\u003e\n\u003cli\u003eKharasch, S., Duggan, N. M., Cohen, A. R. \u0026amp; Shokoohi, H. Lung Ultrasound in Children with Respiratory Tract Infections: Viral, Bacterial or COVID-19? A Narrative Review. \u003cem\u003eOpen Access Emerg. Med.\u003c/em\u003e \u003cstrong\u003eVolume 12\u003c/strong\u003e, 275\u0026ndash;285 (2020).\u003c/li\u003e\n\u003cli\u003eDeSanti, R. L. \u003cem\u003eet al.\u003c/em\u003e Point-of-Care Lung Ultrasound to Diagnose the Etiology of Acute Respiratory Failure at Admission to the PICU*. \u003cem\u003ePediatr. Crit. Care Med.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 722\u0026ndash;732 (2021).\u003c/li\u003e\n\u003cli\u003eCopetti, R. \u0026amp; Cattarossi, L. Ultrasound diagnosis of pneumonia in children. \u003cem\u003eRadiol. Med. (Torino)\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 190\u0026ndash;198 (2008).\u003c/li\u003e\n\u003cli\u003eLichtenstein, D. A. Lung ultrasound in the critically ill. \u003cem\u003eAnn. Intensive Care\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1 (2014).\u003c/li\u003e\n\u003cli\u003eLichtenstein, D., Mezi\u0026egrave;re, G. \u0026amp; Seitz, J. The Dynamic Air Bronchogram. \u003cem\u003eChest\u003c/em\u003e \u003cstrong\u003e135\u003c/strong\u003e, 1421\u0026ndash;1425 (2009).\u003c/li\u003e\n\u003cli\u003eOmran, A. \u003cem\u003eet al.\u003c/em\u003e Lung Ultrasound and Neutrophil Lymphocyte Ratio in Early Diagnosis and Differentiation between Viral and Bacterial Pneumonia in Young Children. \u003cem\u003eChildren\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1457 (2022).\u003c/li\u003e\n\u003cli\u003eAcosta, C. M., Belaunzar\u0026aacute;n, A. \u0026amp; Gonorazky, S. Accuracy of Transthoracic Lung Ultrasound for Diagnosing Anesthesia-induced Atelectasis in Children. \u003cem\u003ePerioper. Med.\u003c/em\u003e\u003c/li\u003e\n\u003cli\u003eInternational Liaison Committee on Lung Ultrasound (ILC-LUS) for the International Consensus Conference on Lung Ultrasound (ICC-LUS) \u003cem\u003eet al.\u003c/em\u003e International evidence-based recommendations for point-of-care lung ultrasound. \u003cem\u003eIntensive Care Med.\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 577\u0026ndash;591 (2012).\u003c/li\u003e\n\u003cli\u003eBiagi, C. \u003cem\u003eet al.\u003c/em\u003e Lung ultrasound for the diagnosis of pneumonia in children with acute bronchiolitis. \u003cem\u003eBMC Pulm. Med.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 191 (2018).\u003c/li\u003e\n\u003cli\u003eDankoff, S., Li, P., Shapiro, A. J., Varshney, T. \u0026amp; Dubrovsky, A. S. Point of care lung ultrasound of children with acute asthma exacerbations in the pediatric ED. \u003cem\u003eAm. J. Emerg. Med.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 615\u0026ndash;622 (2017).\u003c/li\u003e\n\u003cli\u003eMcHugh, M. L. Interrater reliability: the kappa statistic. \u003cem\u003eBiochem. Medica\u003c/em\u003e 276\u0026ndash;282 (2012) doi:10.11613/BM.2012.031.\u003c/li\u003e\n\u003cli\u003eCaiulo, V. A. \u003cem\u003eet al.\u003c/em\u003e Lung ultrasound in bronchiolitis: comparison with chest X-ray. \u003cem\u003eEur. J. Pediatr.\u003c/em\u003e \u003cstrong\u003e170\u003c/strong\u003e, 1427\u0026ndash;1433 (2011).\u003c/li\u003e\n\u003cli\u003eSong, J. W. \u0026amp; Chung, K. C. Observational Studies: Cohort and Case-Control Studies. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 2234 (2010).\u003c/li\u003e\n\u003cli\u003eDelgado-Rodriguez, M. \u0026amp; Llorca, J. Bias. \u003cem\u003eJ. Epidemiol. Community Health\u003c/em\u003e \u003cstrong\u003e58\u003c/strong\u003e, 635 (2004).\u003c/li\u003e\n\u003cli\u003eBellomo, R., Warrillow, S. J. \u0026amp; Reade, M. C. Why we should be wary of single-center trials. \u003cem\u003eCrit. Care Med.\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 3114 (2009).\u003c/li\u003e\n\u003cli\u003eChu, H. \u0026amp; Cole, S. R. Sample size calculation using exact methods in diagnostic test studies. \u003cem\u003eJ. Clin. Epidemiol.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 1201\u0026ndash;1202 (2007).\u003c/li\u003e\n\u003cli\u003eRutjes, A. W. S., Reitsma, J. B., Coomarasamy, A., Khan, K. S. \u0026amp; Bossuyt, P. M. M. Evaluation of diagnostic tests when there is no gold standard. A review of methods. \u003cem\u003eHealth Technol. Assess. Winch. Engl.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, iii, ix\u0026ndash;51 (2007).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"Tables 1 to 4 are available in the Supplementary Files section."}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Pediatric Lung Ultrasound, Accuracy of LUS, Bacterial and Viral Etiology, Radiation-Free Diagnosis, Ultrasound Algorithms in Pediatrics, Sensitivity and Specificity","lastPublishedDoi":"10.21203/rs.3.rs-5461492/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5461492/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLung ultrasound (LUS) is an emerging radiation-free diagnostic tool with significant promise for identifying the causes of respiratory symptoms in hospitalized children. We conducted a prospective cohort study of 65 pediatric patients (September 2022\u0026ndash;January 2023) to evaluate the accuracy of LUS in determining the etiology of acute respiratory illness compared to an expert panel\u0026rsquo;s final clinical diagnosis, and secondarily to chest X-ray (CXR). LUS examinations were performed within 24 hours of admission by an investigator blinded to CXR results, and the final diagnoses were established by a committee blinded to LUS findings. LUS demonstrated high diagnostic accuracy, correctly identifying the respiratory illness etiology in 92.3% of cases, with near-perfect agreement (κ\u0026thinsp;=\u0026thinsp;0.86) with the final diagnosis. Sensitivity and specificity were high across etiologies, including bacterial pneumonia, viral infections, and asthma exacerbations. LUS also outperformed CXR in determining the cause of illness. Integrating LUS findings with clinical and laboratory data allowed effective differentiation between bacterial and viral infections, highlighting LUS as a safer and more accurate alternative to CXR. Our findings support the integration of LUS into pediatric respiratory care and warrant further research to validate its utility on a larger scale.\u003c/p\u003e","manuscriptTitle":"Accuracy of lung ultrasound in diagnosing the etiology of respiratory symptoms in hospitalized children, a prospective cohort study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-06 09:21:00","doi":"10.21203/rs.3.rs-5461492/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-05-20T03:52:17+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-05-20T03:35:11+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-05-20T03:34:58+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-06T09:32:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"6728648812992213664708815703232578039","date":"2025-05-04T02:41:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-29T00:55:51+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"260394808466190748110942477902283450311","date":"2025-04-29T00:42:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-04-28T21:18:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-27T05:45:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-04-14T19:42:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"cad6dcce-2c6c-47c1-ae11-6a1b342f1be8","owner":[],"postedDate":"May 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47818762,"name":"Health sciences/Health care/Diagnosis"},{"id":47818763,"name":"Health sciences/Health care/Paediatrics"}],"tags":[],"updatedAt":"2025-12-01T16:04:12+00:00","versionOfRecord":{"articleIdentity":"rs-5461492","link":"https://doi.org/10.1038/s41598-025-26766-8","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-11-28 15:58:22","publishedOnDateReadable":"November 28th, 2025"},"versionCreatedAt":"2025-05-06 09:21:00","video":"","vorDoi":"10.1038/s41598-025-26766-8","vorDoiUrl":"https://doi.org/10.1038/s41598-025-26766-8","workflowStages":[]},"version":"v1","identity":"rs-5461492","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5461492","identity":"rs-5461492","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}