The utility of chest x-ray and lung ultrasound in the management of infants and children presenting with severe pneumonia in low-and middle-income countries: A pragmatic scoping review

Background Chest x-ray (CXR) is commonly used (when available) to support clinical management decisions for child pneumonia and provide a reference standard for diagnosis in research studies. However, its diagnostic and technical limitations for both purposes are well recognised. Recent evidence suggests that lung ultrasound (LUS) may have diagnostic utility in pneumonia. This systematic scoping review of research on the utility of CXR and LUS in the management of severe childhood pneumonia aims to inform pragmatic guidelines for low- and middle-income countries (LMICs) and identify gaps in knowledge. Methods We included peer-reviewed studies published between 2000 and 2020 in infants and children aged from one month to nine years, presenting with severe pneumonia. CXR studies were limited to those from LMICs, while LUS studies included any geographic region. LUS and CXR articles were mapped into the following themes: indications, role in diagnosis, role in management, impact on outcomes, and practical considerations for LMIC settings. Results 85 articles met all eligibility criteria, including 27 CXR studies and 58 LUS studies. CXR studies were primarily observational and examined associations between radiographic abnormalities and pneumonia aetiology or outcomes. The most consistent finding was an association between CXR consolidation and risk of mortality. Difficulty obtaining quality CXR images and inter-reader variability in interpretation were commonly reported challenges. Research evaluating indications for CXR, role in management, and impact on patient outcomes was very limited. LUS studies primarily focused on diagnostic accuracy. LUS had higher sensitivity for identification of consolidation than CXR. There are gaps in knowledge regarding diagnostic criteria, as well as the practical utility of LUS in the diagnosis and management of pneumonia. Most LUS studies were conducted in HIC settings with experienced operators; however, small feasibility studies indicate that good inter-operator reliability may be achieved by training of novice clinicians in LMIC settings. Conclusions The available evidence does not support the routine use of CXR or LUS as essential tools in the diagnosis and initial management of severe pneumonia. Further evaluation is required to determine the clinical utility and feasibility of both imaging modalities in low-resource settings.

Background Chest x-ray (CXR) is commonly used (when available) to support clinical management decisions for child pneumonia and provide a reference standard for diagnosis in research studies.However, its diagnostic and technical limitations for both purposes are well recognised.Recent evidence suggests that lung ultrasound (LUS) may have diagnostic utility in pneumonia.This systematic scoping review of research on the utility of CXR and LUS in the management of severe childhood pneumonia aims to inform pragmatic guidelines for low-and middle-income countries (LMICs) and identify gaps in knowledge.
Methods We included peer-reviewed studies published between 2000 and 2020 in infants and children aged from one month to nine years, presenting with severe pneumonia.CXR studies were limited to those from LMICs, while LUS studies included any geographic region.LUS and CXR articles were mapped into the following themes: indications, role in diagnosis, role in management, impact on outcomes, and practical considerations for LMIC settings.
Results 85 articles met all eligibility criteria, including 27 CXR studies and 58 LUS studies.CXR studies were primarily observational and examined associations between radiographic abnormalities and pneumonia aetiology or outcomes.The most consistent finding was an association between CXR consolidation and risk of mortality.Difficulty obtaining quality CXR images and inter-reader variability in interpretation were commonly reported challenges.Research evaluating indications for CXR, role in management, and impact on patient outcomes was very limited.LUS studies primarily focused on diagnostic accuracy.LUS had higher sensitivity for identification of consolidation than CXR.There are gaps in knowledge regarding diagnostic criteria, as well as the practical utility of LUS in the diagnosis and management of pneumonia.Most LUS studies were conducted in HIC settings with experienced operators; however, small feasibility studies indicate that good inter-operator reliability may be achieved by training of novice clinicians in LMIC settings.

Conclusions
The available evidence does not support the routine use of CXR or LUS as essential tools in the diagnosis and initial management of severe pneumonia.Further evaluation is required to determine the clinical utility and feasibility of both imaging modalities in low-resource settings.
Pneumonia is a leading cause of death in under-five children living in under-resourced settings [1].Accurate and early diagnosis is important in reducing pneumonia-related mortality.Clinical features are non-specific, especially in young children, and there is no single test with high specificity or sensitivity for the diagnosis of pneumonia.Computed tomography (CT) of the chest may be considered a radiologic gold standard in diagnosing pneumonia, but cost, accessibility and radiation exposure make it unfeasible for routine use [2].

Overview of studies
Of 26 CXR studies with original data, 17 (65%) included children with WHO severe or very severe pneumonia (as per the 2005 definition) while the remaining had a broader inclusion criterion for pneumonia and severity.No studies evaluated children based on the WHO 2014 definition of severe pneumonia.Most CXR studies (n = 17, 65%) provided observational data about associations between CXR abnormality and clinical course, while one systematic review addressed predictors of pneumonia severity.
There were no studies evaluating indications for CXR or LUS.Included studies that did examine patient outcomes examined the association between imaging findings and patient outcomes rather than the impact of using the particular imaging modality.Further details of the nature and distribution of CXR and LUS studies can be found in Appendix S3 and S4 in the Online Supplementary Document.

Utility of chest x-ray in severe pneumonia
The 26 original CXR studies had heterogenous study aims and designs (Table 1).Most studies (80%) were conducted in LMICs.Only three studies were conducted solely in a low-income country (two in The Gambia and one in Ethiopia) and two were conducted over multiple sites which included low-income countries.Most CXR studies (92%) were based in emergency departments (EDs) or inpatient settings.One study was conducted in four peri-urban, primary health facility level study sites in Pakistan [18], while a large Gambian study [26] recruited children presenting to either a secondary care facility or hospital.Older children (5-9 years) were under-represented with only three studies including this age group [19,22,39].

Role of chest x-ray in diagnosis
Five studies [13,14,[16][17][18] explored the role of CXR in diagnosis.Clinical and radiographic definitions of pneumonia varied significantly across studies.The only standardised format for interpretation of CXRs was in the use of the WHO Primary Endpoint Pneumonia criteria [3], which was common in clinical studies.
Overall, severity as per WHO-defined pneumonia (2005) correlated poorly with radiographic pneumonia diagnosis [14,18,27].In a multi-centre study from India (n = 524), auscultation was found to be superior to CXR in distinguishing pneumonia from other respiratory disorders, using a final paediatrician diagnosis as a reference standard [14].Nizami et al. [18] evaluated the utility of performing CXR in children presenting with fever and cough in four peri-urban, primary level health facilities in Pakistan; approximately half of children with clinical upper respiratory tract infections had radiographic opacity or infiltration considered to be suggestive of infection.
A normal CXR was commonly reported (24%-72%) in children with WHO-defined severe pneumonia [20,27,30,32,36].The Pneumonia Etiology Research for Child Health (PERCH) study [27] evaluated the likelihood that a child with WHO-defined severe pneumonia ( 2005) and a normal CXR truly has pneumonia, based on the presence of two or more additional pneumonia signs, and reported that pneumonia was the "likely diagnosis" in 52% of these children.Infants (<12 months age) had the highest proportion of normal CXRs, possibly due to this age group having more cases of bronchiolitis.
Two studies [13,16], including a multi-centre study from South America (n = 2536), evaluated CXR in identifying bacterial pneumonia.Children with WHO severe pneumonia ( 2005) and Streptococcus pneumoniae isolated on blood culture or PCR were unlikely to have a normal CXR (negative predictive value (NPV) = 92%) [13].
Alveolar opacities and pleural effusion were associated with pneumococcal isolation in blood or pleural culture (P < .001).A study of 209 children in Brazil identified associations between a normal CXR and presumed viral pneumonia, as well as abnormal CXR and presumed bacterial pneumonia; however, assigning bacterial or viral aetiology on CXR alone would misclassify 30%-50% of cases [16].

Chest x-ray and outcomes
Eighteen studies evaluated associations between CXR abnormalities and severity of pneumonia (Table 2).
Five studies from countries in Africa and Asia comprising 20 754 children with severe pneumonia identified that radiographic dense consolidation at presentation was associated with a higher case fatality rate within one month of admission [24][25][26][27]32].In two of these studies, the strength of association decreased after multivariate analysis [24,25].The PERCH study found that radiographs of children who died were twice as likely to have WHO Primary Endpoint Consolidation (50% vs 24%) and less likely to have "other infiltrates" (18% vs 28%).
A systematic review (which also included studies from high-income countries (HICs)) evaluated predictors of pneumonia severity and found multi-lobar infiltrates and moderate or large pleural effusions to be predictive of mortality, along with several other factors such as younger age and hypoxaemia [23].A retrospective study [19] from Paraguay evaluated a multi-dimensional prognostic score for predicting pneumonia mortality; CXR abnormalities, particularly multi-lobar opacities, added value to the prediction score (odds ratio = 4.9; 95% confidence interval (CI) = 2.7-8.9).

Feasibility and practical considerations for LMIC settings
Six studies reported inter-observer reliability in CXR interpretation for pneumonia.Reliability was impacted by observer experience and the radiographic finding in question [30,35,[37][38][39][40] (n = 925, kappa value (ĸ) range = 0.15-0.7).Agreement was best for reporting lobar consolidation or primary endpoint pneumonia using the WHO standardised criteria for interpretation and became less reliable with more obscure findings such as "other infiltrates" [30,39].When CXR's with clear lobar pneumonia and pleural effusion were excluded, Bada et al. [37] identified poor concordance between junior doctors in interpreting other CXR abnormalities in children presenting with respiratory symptoms (ĸ = 0.2).An Ethiopian study found low concordance in interpretation of CXR for paediatric pneumonia among two senior radiologists when using a standardised checklist (ĸ = 0.45 for consolidation; ĸ = 0.65 for pleural effusion) [13].Patel et al. found an improvement in agreement and reduction in diagnosis of significant pathology following training in CXR interpretation using the WHO standardised criteria [34].Several studies from LMIC settings described significant issues with film quality, interpretation, and inter-observer variability, impacting the utility of CXR in these settings [14,20,27].Between 4%-20% of radiographs were determined to be uninterpretable in two recent, multi-site studies from hospital settings, despite technical assistance to optimise quality [20,27].In a study conducted at health facility level in Pakistan [18], 19% of radiographs were considered uninterpretable by the attending physician.Retrospective cohort study, N = 860 Multi lobar consolidation, pleural effusion, pneumothorax X Multi-lobar CXR involvement (OR = 4.9; 95% CI = 2.7-8.9;P < 0.0001), pleural empyema (OR = 2.6; 95% CI = 1.1-6.2;P < 0.05), pneumothorax (OR = 15; 95% CI = 2.9-76.6,P < 0.01) associated with mortality Awasthi, 2020 [20] Multi-site prospective surveillance study, N = 3214 Abnormal CXR X "Abnormal CXR" associated with fever, pallor, malnutrition, wheezing on auscultation CXR -chest x-ray, OR -odds ratio, PEP -primary endpoint pneumonia, PCV -pneumococcal vaccine, RR -risk ratio, CI -confidence interval, RCT -randomised controlled trial, WHO -World Health Organization, PPV -positive predictive value, NPV-negative predictive value *Primary endpoint consolidation and primary endpoint pneumonia are definitions developed by WHO for interpretation of CXR in pneumonia epidemiological studies [41].Primary endpoint consolidation is defined as a dense consolidation or confluent opacity that occupies a portion or whole of a lobe or the entire lung, that may or may not contain air bronchograms.Primary endpoint pneumonia refers to the presence primary endpoint consolidation or pleural effusion.www.jogh.org• doi: 10.7189/jogh.12.10013
The systematic reviews included a total of 32 studies, five of which were from LMICs (Peru [50,51], Egypt [56], India [57], and Turkey [59]).Most studies were conducted in emergency department or inpatient settings in HICs.Most studies used a reference standard for pneumonia which encompassed clinical and/or CXR features; however, computed tomography (CT) chest was the reference standard in a subset of patients in three studies [11,43,66].The pooled sensitivity of LUS for pneumonia diagnosis was consistently high in all systematic reviews (range = 93%-96.7%),while pooled specificity was slightly lower (range = 87%-96.1%).Compared with CXR, LUS had higher sensitivity in identifying lung consolidation and had a high negative predictive value for pneumonia [11,43,44,48].LUS was able to identify consolidation of less than 0.5-1 cm, which are often not distinguishable on CXR and may not represent true pneumonia.As a result, the specificity of LUS is reduced when comparing with CXR as a reference standard.using CT chest as a reference, was significantly lower than reported in other studies and similar to CXR (for consolidation, LUS sensitivity = 63% (95% CI = 0.34-0.85),specificity = 75% (95% CI = 0.58-0.87)).LUS had lower specificity compared with CXR for interstitial disease (P < .01),pleural effusion (P = 0.04), and consolidation (P = 0.26) in this patient population [62].Three systematic reviews [11,42,44] performed sub-analysis of LUS diagnostic accuracy based on operator experience and found that accuracy improved with experience, but that sensitivity and specificity were still moderate to high in novice sonographers following variable periods of training (pooled sensitivity = 80%-95%, pooled specificity = 91%-96%).There was limited reporting of diagnostic accuracy based on age.One systematic review [44] identified a slightly better sensitivity and specificity in studies with a mean age of participants aged over five years, compared with studies with a mean age of less than five years.On the other hand, Pereda et al. [11] reported better diagnostic accuracy in neonates than in older children.
All systematic reviews reported heterogeneity between studies as a consistent limitation, including in operator experience, ultrasound technique, equipment utilised, and sonographic criteria for pneumonia.

Role of lung ultrasound in diagnosis
The characteristics, aims, and findings of the five studies evaluating LUS in diagnosis of severe pneumonia are presented in Table 5.

Lung ultrasound and outcomes
No study directly assessed the impact of the use of LUS on patient outcomes.Eight studies evaluated the association between LUS abnormalities and outcomes, one of which [84] was from an LMIC (Turkey).One study [88] found an association between sonographic lung consolidation and lack of fever defervescence at 48 hours.Four observational studies (three from countries in Europe and one from Turkey) comprising 380 infants with bronchiolitis [84][85][86][87] found associations between LUS abnormalities and the need for respiratory support or hospital admission, but there was no clear benefit described over clinical assessment alone.
In 364 children with suspected community-acquired pneumonia in Italy and Taiwan, LUS was used to identify surgical complications [89][90][91].LUS was performed early in admission and identified signs of evolving complications of pneumonia before they were clinically apparent.Pleural effusion on initial ultrasound was associated with risk of complicated pneumonia, adverse outcome, or surgical intervention.Other LUS abnormalities were less consistently associated with outcomes.LUS performed better than both CXR and CT scan in identifying large and complicated effusions.

Feasibility and practical considerations for low-and middle-income country settings
Nine studies [50,62,[92][93][94][95][96][97]   Sustained levels of high agreement for diagnosis of PEP (κ>0.78),agreement was higher in children with danger signs.The training was resource intensive and expensive so may limit application in resource limited settings.This study provides criteria for PEP on LUS which requires further diagnostic validation to be widely implemented.Agreement for consolidation on LUS was moderate (0.55; 95% CI = 0.40-0.70),and on CXR was poor (0.36; 95% CI = 0.21-0.51).For interstitial disease, agreement was poor on LUS (0.32; 95% CI = 0.17-0.47),and substantial for CXR (0.63; 95% CI = 0.47-0.77).LUS performed by multiple sonographers and interpreted by multiple paediatric radiologists to improve applicability to clinical settings.

DISCUSSION
Although CXR is widely used for child pneumonia management, there is a lack of evidence to inform the optimal or pragmatic use of CXR for children presenting with possible pneumonia.Most existing research explored CXR as an independent tool in pneumonia diagnosis or evaluated association between CXR findings and patient outcomes, with significant limitations identified.Research evaluating LUS has focused mainly on diagnostic accuracy, and utility and feasibility are yet to be determined.Given the challenge of lack of a 'gold standard' for pneumonia diagnosis, ascertaining the precise diagnostic accuracy of CXR and LUS for pneumonia remains elusive.
While CXR had suboptimal performance as an independent tool in severe pneumonia, there was little high-quality evidence to support, refute, or guide the use of CXR as it is conventionally interpreted -with consideration to clinical and contextual variables such as age, comorbidities, duration of illness, and prior antibiotic treatment.Besides for diagnostic clarification, CXR is often used to investigate treatment failure and identify complications, comorbidities, or alternate diagnoses.However, there were no studies evaluating these indications for CXR.The lack of clear evidence is reflected in the substantial variation in national and international guidelines for use of radiographs in severe pneumonia in HIC and LMIC settings [6].
Despite many evidence gaps, there were a few consistent themes in the literature.Radiographic "dense" consolidation was associated with mortality in several studies from resource-limited settings.This has not been a convincing finding globally, and a recent systematic review including a large proportion of studies from HICs identified multi-lobar consolidation to be more predictive of adverse outcomes [23].The variability may reflect differences in local pneumonia-related epidemiology, vaccination status, patient co-morbidities, and availability of timely and appropriate treatment, and underscores the need for individualised interpretation and management of CXR abnormalities.Moreover, while positive CXR findings may be helpful in confirming a diagnosis, studies consistently identified a normal CXR in a small but significant proportion of children who were likely to have "true" severe pneumonia.Additionally, CXR was unable to reliably determine whether pneumonia is of bacterial or viral origin.There is wide inter-observer variability in the interpretation of CXR abnormalities, particularly with less experienced clinicians and more obscure findings, and the use of a standardised format for interpretation improved concordance.
LUS is superior to CXR in the detection of lung consolidation and pleural effusion, with high sensitivity consistently identified in several meta-analyses.Specificity was also high, but contingent upon the reference standard utilised for pneumonia, possibly making it lower in children with comorbidities.There is currently no consensus on the sonographic criteria for pneumonia or on an optimal scanning protocol.Small studies indicate LUS may have diagnostic capabilities in children with undifferentiated respiratory distress.Inter-operator reliability of LUS was influenced by operator experience, yet preliminary research on its feasibility in LMICs indicates that good diagnostic performance can be achieved following training of novice operators.Despite widespread availability of point of care ultrasound devices in HIC settings, LUS is not widely utilised as a first line investigation for pneumonia [98], particularly in settings where clinicians have no prior experience with its use.Barriers cited by clinicians include a lack of confidence in skills and challenge of maintaining expertise over time [67].While LUS is less costly to perform than CXR [65], the financial and personnel costs of training and ongoing support to maintain expertise in the longer term may be significant.

Strengths and limitations
Our scoping review was deliberately comprehensive in order to understand the range of relevant research.Therefore, we were able to obtain a clear picture of the strengths and limitations of the available research and identify significant gaps in knowledge.For both LUS and CXR studies, the lack of an optimal reference standard for pneumonia is a major limitation for evaluation.Moreover, both modalities lack a standardised ra-diological definition of pneumonia for clinical settings, hindering our ability to draw robust conclusions.By excluding CXR studies from HICs, we acknowledge there may have been data applicable to resource-limited settings which we could not include in our study.Similarly, we only evaluated recent literature from the last two decades, which may have excluded older studies relevant to informing and establishing the current practice of CXR.Any conclusions drawn regarding the practical utility of CXR and LUS is based on limited data as high-quality evidence was not available and for some outcomes there were no relevant studies.

Practical application
The limited diagnostic accuracy, combined with challenges of obtaining high quality images and need for expert interpretation do not support large scale programmatic use of CXR for pneumonia or specific recommendations for use in health facilities in resource-limited settings.CXR is unlikely to alter the initial case management of children with severe pneumonia.There is a paucity of research into the utility of CXR in specific contexts, such as exploring alternative diagnoses and investigating sudden deterioration or poor response to treatment.In such cases, CXR may identify complications such as pleural effusion, lung abscess, pneumothorax, pneumatoceles, or characteristic features of a specific aetiology such as tuberculosis.There is currently insufficient evidence to provide recommendations for the use of LUS in the case management of children with pneumonia.The use of LUS in the management of complicated pneumonia has already been established and may have a role in improving accuracy of a clinical diagnosis of pneumonia; further exploration of its utility and feasibility in low resource settings is needed.

CONCLUSIONS
Pneumonia is a heterogenous entity, encompassing a range of aetiologies with a clinical course dependant on host factors, clinical context, and access to timely and appropriate treatment.There is significant discrepancy between the evidence for utility of CXR and its widespread and ongoing use for pneumonia.High-quality, pragmatic research is needed to understand its role and perceived benefits, including the identification of alternate diagnoses or comorbidities in children presenting with severe pneumonia, investigation of treatment failure, or identification of complications.
LUS has the potential to improve diagnostic capabilities in low resource settings, with the benefit of greater sensitivity for pneumonia than CXR.However, there remain significant gaps in knowledge.High-quality, prospective studies are warranted in resource limited settings to explore LUS's role in improving the accuracy of clinical diagnosis as well as predicting and managing complications of pneumonia, and on the benefits and risks of this imaging tool.Although LUS findings in pneumonia are well described, consensus is required regarding scanning protocols and diagnostic criteria.Future feasibility studies should evaluate barriers to implementation and uptake of LUS by clinicians, with longitudinal follow up to understand the maintenance needs for both equipment and technical skills over time.
Children with co-morbidities which increase vulnerability to pneumonia-related mortality (such as malnutrition, HIV, and congenital heart disease) should be included in research of imaging tools to promote early and accurate diagnosis in this high-risk population.

of study Role of CXR in diagnosis
Assess the use of a modification of PIRO scale (used in adult pneumonia) and the association between modified PIRO score and mortality, stratified by risk.

Table 2 .
Association between chest x-ray abnormalities and outcome

Table 3 .
Systematic reviews evaluating diagnostic accuracy of lung ultrasound LMIC -low and middle-income country, CAP -community-acquired pneumonia, CT -computerized tomography, CXR -chest x-ray, y -years *Includes both paediatric and neonatal studies.†There is considerable overlap between systematic reviews of the studies included in each review.‡There were 5 original diagnostic accuracy studies from LMIC settings in total.

Table 4 .
Original lung ultrasound diagnostic accuracy studies

Study Study design Country Inclusion criteria Partici- pants (n) Reference standard for pneumonia Low-and middle-income country
One study which performed chest CT, CXR, and LUS, found high concordance between LUS and CT chest in identifying pneumonia (as determined by clinical course) when CXR was normal[66,71].LUS was superior to CXR and comparable to CT scan in its ability to identify and characterise pleural effusions[71].
Factors which may impact diagnostic accuracy of LUS include operator experience, presence of comorbidities, and patient age.LUS studies of diagnostic accuracy have primarily been conducted in children without significant comorbidities.A study from the USA evaluated children with suspected pneumonia and comorbidities such as malignancy, who required a CT chest as part of their clinical care[62].The accuracy of LUS CT -computerized tomography, CXR -chest x-ray, WHO -World Health Organization, AAP -American Academy of Pediatrics, mo -months, y -years *Final diagnosis -a diagnosis of bacterial pneumonia made by an experienced paediatrician or pulmonologist, blinded to LUS findings, based on clinical presentation, laboratory tests, CXR, and clinical course.

Table 4 .
continued out of five studies evaluating LUS in diagnosis of severe pneumonia were conducted inLMICs [50,[80][81][82].Two of these, one in Egypt and one in Turkey, evaluated the diagnostic utility of LUS in children presenting with undifferentiated respiratory symptoms, and found that LUS diagnosis of pneumonia, pleural effusion, and pneumothorax correlated well with a final physician diagnosis.A total of 208 children were enrolled in the studies and an experienced operator performed the bedside LUS[50,51].Another study from Turkey also reported ED physician diagnosis to have similar accuracy to LUS[84].A larger study of 832 children with re- spiratory symptoms in Peru found that LUS predicted pneumonia with greater accuracy than physicians did based on clinical signs and symptoms, CXR, pulse oximetry, and chest auscultation[80].LUS pneumonia correlated poorly with80,83].In a study of 378 children with respiratory symptoms conducted in Nepal and Peru[50], the proportion of children with consolidation on LUS was similar in children with WHO-defined pneumonia and children with respiratory symptoms not meeting WHO pneumonia criteria (23 vs 21%, P = 0.68).Consolidation occurred in 33% of children with WHO-defined severe pneumonia (n = 12).A study conducted in Slovenia[83]with 378 children correlated large and solitary consolidation on LUS with presumed bacterial pneumonia, and a pattern of small, multiple consolidations with viral pneumonia.The probable aetiology was assigned based on a combination of CXR, blood infection markers, blood cultures, and nasopharyngeal aspirate PCR.

Table 5 .
Lung ultrasound studies to diagnose cause of respiratory distress with high sensitivity and specificity for pneumonia (81.4% and 100%) using final clinical diagnosis as reference standard.Concordance between ED physician and LUS was high (kappa value 0.8).Similar high concordance was reported for diagnosis of acute bronchiolitis, asthma and croup, however, the method of LUS diagnosis of these conditions was not clear.Feasibility in this study was reported in time taken to perform LUS and make a diagnosis (mean time to LUS was 6.07 ± 3.92 min; time to sonographic diagnosis was shorter than to ED diagnosis: 29.32 ± 15.71 min (P < 0.001).

Table 5 .
continued for consolidation (range of kappa values = 0.635 to 0.78).The longer and more intensive training program resulted in sustained levels of high agreement for LUS findings over two-year follow-up; however, the program was resource intensive and costly[92].Two small studies conducted in HICs [96,97] compared feasibility of point-of-care LUS performed by experienced paediatric emergency physicians with CXR.The studies reported no missed pneumonia or increase in adverse events, shorter lengths of stay in the emergency department.and associated health care cost savings.A study conducted in the USA [62] assessed the reliability of LUS when performed by multiple operators and interpreted by multiple paediatric radiologists; inter-operator agreement for consolidation was lower than other studies (ĸ = 0.55; 95% CI = 0.40-0.70) .