An Observational Data Analytical Research on Pediatric Cardiomegaly as a Predictor of Structural or Functional Heart Diseases: Chest X-ray versus Echocardiography Comparison and Contemplation

Munesh Tomar; Fatima Afreen Ahmad; Maitri Chaudhuri; Vikas Agarwal

doi:10.4103/jiae.jiae_51_21

ORIGINAL RESEARCH

Year : 2022 | Volume : 6 | Issue : 1 | Page : 21-31

An Observational Data Analytical Research on Pediatric Cardiomegaly as a Predictor of Structural or Functional Heart Diseases: Chest X-ray versus Echocardiography Comparison and Contemplation

Munesh Tomar¹, Fatima Afreen Ahmad¹, Maitri Chaudhuri², Vikas Agarwal¹
¹ Department of Pediatric, LLRM Medical College, Meerut, Uttar Pradesh, India
² Department of Pediatric Cardiology, Manipal Hospital, Bengaluru, Karnataka, India

Date of Submission	24-Aug-2021
Date of Acceptance	21-Nov-2021
Date of Web Publication	23-Mar-2022

Correspondence Address:
Dr. Munesh Tomar
Department of Paediatrics, LLRM Medical College, Meerut - 250 004, Uttar Pradesh
India

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jiae.jiae_51_21

Abstract

Background: Chest X-ray (CXR) is a routine, noninvasive test advised in a plethora of pediatric non-cardiac conditions such as bronchopneumonia, respiratory distress, failure to thrive, suspected tuberculosis, and for preoperative anesthesia clearance. Finding of cardiomegaly (CM) on CXR is a common cause of referral to a pediatric cardiologist. The question lies in its significance: whether CM in CXR corroborates with cardiac disease (s), structural or functional. Materials and Methods: The data of 229 children (median age: 6 months), referred to pediatric cardiology unit for echocardiography with CXR depicting CM over a period of 2 years (June 2019 to July 2021), were retrospectively reviewed. Clinical and laboratory findings, CXR, and echocardiography were analyzed for all children to determine the positive predictive value of CM in CXR to detect cardiac disease. Echocardiography was taken as gold standard to reach the cardiac diagnosis. Results: True CM was noted in 85% (Group A), whereas 15% had false CM (Group B). Group A comprised structural heart defect in 71.7% and ventricular dysfunction in 13.3%. Less common causes were severe anemia, hypertensive heart failure, arrhythmogenic cardiomyopathy, diphtheritic cardiomyopathy, multisystem inflammatory syndrome in children, pericardial effusion, thiamine deficiency, and severe idiopathic pulmonary arterial hypertension. In Group B, the most common reasons for false diagnosis were expiratory film (n = 18), thymic shadow (n = 12), and chest deformity and mediastinal mass (n = 2 each). Conclusion: CM in CXR strongly correlates with cardiac involvement and has a high positive predictive value. Combining clinical, laboratory, and CXR interpretation enables the pediatrician to achieve a rapid functional diagnosis. Medical stabilization can be initiated with this knowledge pending the availability of echocardiography in resource-limited areas. Detailed evaluation at a pediatric cardiac center should be completed to reach a final diagnosis in all such patients. In the absence of congenital heart disease, known acquired causes leading to cardiac compromise should be actively looked for.

Keywords: Cardiomegaly, chest X-ray, dilated cardiomyopathy, multisystem inflammatory syndrome in children, pediatric echo, structural heart defect

How to cite this article:
Tomar M, Ahmad FA, Chaudhuri M, Agarwal V. An Observational Data Analytical Research on Pediatric Cardiomegaly as a Predictor of Structural or Functional Heart Diseases: Chest X-ray versus Echocardiography Comparison and Contemplation. J Indian Acad Echocardiogr Cardiovasc Imaging 2022;6:21-31

How to cite this URL:
Tomar M, Ahmad FA, Chaudhuri M, Agarwal V. An Observational Data Analytical Research on Pediatric Cardiomegaly as a Predictor of Structural or Functional Heart Diseases: Chest X-ray versus Echocardiography Comparison and Contemplation. J Indian Acad Echocardiogr Cardiovasc Imaging [serial online] 2022 [cited 2023 Sep 27];6:21-31. Available from: https://jiaecho.org/text.asp?2022/6/1/21/340638

Introduction

Joseph Perloff, the pioneer of clinical pediatric cardiology, considered chest X-ray (CXR) as one of the fundamental pillars in the tripod for diagnosing heart disease in childhood. It is also the most commonly ordered radiological test in pediatrics. The presence of CM alerts the physician. However, increased cardiothoracic ratio in rotated and/or expiratory films and underlying pulmonary-thoracic wall deformities often leads to false diagnosis. Pediatric electrocardiogram (ECG) often provides heterogeneous and inconclusive findings.

Our study was designed to test the validity of CM in CXR as a predictor of cardiac disease and its accuracy to reach an anatomical and physiological diagnosis. This was conceived remembering that pediatric echocardiography is often not available in resource-limited settings.

We reviewed our data of children with CM in CXR referred to pediatric cardiac unit for expert opinion.

Aims and objectives

The aims of our study were as follows:

To find the positive predictive value of cardiac enlargement (CE) and prediction of heart disease by CXR as compared to echocardiography. We aimed to find out whether or not CXR could be used as a screening and/or diagnostic tool for heart diseases in financially limited setups
Utilizing CXR in reaching accurate anatomical and functional diagnosis of cardiac anomaly, for example, increased or decreased pulmonary blood flow, specific chamber enlargement, pulmonary arterial hypertension, and/or venous hypertension
Combining clinical evaluation with laboratory investigations, metabolic screening, ECG, and ultimately echocardiography to reach a final diagnosis.

Materials and Methods

A retrospective study in pediatric patients of age group 2 month–18 years at a tertiary care hospital in Western Uttar Pradesh was undertaken from January 2019 to July 2021. During this period, 229 children with CXR diagnosis of CM were referred to pediatric cardiology unit for echocardiography and expert opinion. Clinical and CXR findings were examined first by a pediatric cardiologist before echocardiography. Next, echocardiography was performed in all by the same cardiologist. Additional investigations such as hemogram, renal and liver function test, thyroid function test, ECG, and metabolic screening were asked whenever indicated.

Exclusion criteria

Neonates were excluded from the study knowing the technical limitations in that age group. Children with known congenital heart disease or those on follow-up were also excluded.

Definition

CM on CXR has been defined by cardiothoracic index (CTI), calculated by dividing the widest transverse diameter of the heart by width of the thoracic cavity^[1] [Figure 1]. The value of CTI >0.55 in infants and >0.50 in children was taken as CM.^[2]

Figure 1: Good quality chest X-ray posteroanterior view of a 16-year-old male patient. The film is at the end of inspiration with adequate penetration, proper centering (marked by central line with equidistant medial ends of both clavicles). Black arrows mark the difference in the length of the right and left bronchi (right bronchus is shorter and more in line with trachea than left). There is situs solitus, levocardia, no cardiomegaly with normal pulmonary vasculature, and lung parenchyma. Cardiothoracic (CT) ratio is also measured for demonstration (CT ratio = R + L/D)

Click here to view

Echocardiogram was employed as gold standard for qualitative and quantitative assessment of CM and specific chamber enlargement.

Radiological algorithm

The CXRs available were interpreted as per this protocol:

Technical quality
Physiologic interpretations

Cardiac silhouette (cardiac size, chamber enlargement, and ventricular configuration)
Pulmonary vasculature

Pulmonary blood flow (PBF)
Pulmonary arterial hypertension (PAH)
Pulmonary venous hypertension (PVH).

Mediastinum: To look for widening of mediastinum.

Lung parenchyma and bony thorax.

Technical quality

View – Posteroanterior (PA) versus anteroposterior (AP): CXR in erect PA projection is considered the “gold standard.” There are conditions where PA projection is not possible and CXR is taken in AP projection, for example, small child or a sick patient, etc. The heart, being an anterior structure within the chest, is magnified by an AP projection. Magnification is exaggerated further by the shorter distance between the CXR source and the patient leading to more divergent beam to cover the same anatomical field. Majority of CXR in our study were taken in anteroposterior projection as the study population was very young (median age 6 months)
Position: A properly centered CXR allows correct interpretation of heart size, cardiac position, chamber enlargement, and lung volumes. In a well-centered CXR, the medial ends of the clavicle are equidistant from the midline [Figure 1]
Exposure: A CXR may be adequately exposed, overexposed, or underexposed. This seriously impacts the interpretation of lung vasculature and PVH.

The exposure is considered adequate if it shows the intervertebral disc spaces up to tracheal bifurcation (usually the first four-disc spaces) and not below it [Figure 1]
If vertebral bodies are visible clearly through the cardiac shadow, the CXR is “overexposed.” The lung parenchyma will appear darker, and the soft-tissue shadows such as lung infiltrates and pulmonary vascularity will be underestimated in an overexposed film [Figure 2]
On the other hand, in an “underexposed film,” one cannot differentiate between vertebral bodies and intervertebral discs. The lung parenchyma will appear whiter, with possible overestimation of the pulmonary vascularity [Figure 3].

Figure 2: Chest X-ray anteroposterior view of a 3-month-old infant with clinical features of congestive heart failure with mild cyanosis (SpO2: 80%). Chest X-ray showed cardiomegaly (cardiothoracic ratio: 70%), dilated pulmonary arteries, and increased pulmonary blood flow. Superior mediastinum is widened to the right side by the dilated superior vena cava (arrows). The film is overpenetrated and therefore, underestimating pulmonary vasculature. Echo showed nonobstructive total anomalous pulmonary venous connection to superior vena cava with dilated right atrium, right ventricle, and pulmonary arteries with pulmonary arterial hypertension

Click here to view

Figure 3: Chest X-ray anteroposterior view (underexposed film) of a 2-month-old infant with features of cardiac failure and echo diagnosis of large ventricular septal defect. Chest X-ray shows cardiomegaly (cardiothoracic ratio: 70%), left atrial enlargement, dilated pulmonary arteries, and features of increased pulmonary blood flow

Click here to view

Phase of respiration: Ideally, the film should be taken at the end of inspiration. It is often technically unachievable in small children.

If the film is taken at the end of inspiration, the hemidiaphragm crosses the sixth rib in the midclavicular line, eighth rib in the midaxillary line, and tenth rib posteriorly. The right diaphragm is slightly higher than left (1–2 cm) in situs solitus [Figure 1]
In the case of an expiratory film, cardiac shadow appears larger (impression of CM) and lung parenchyma appears hazy giving impression of increased pulmonary vascularity [Figure 4].

Figure 4: A 2-month-old infant with history of fever, respiratory distress, and positive antibody for severe acute respiratory syndrome coronavirus 2 (multisystem inflammatory syndrome in children). Chest X-ray (anteroposterior view), expiratory film, showing mild cardiomegaly with diffuse right lung infiltrates. On echo, there was mild left ventricular dysfunction, mild mitral regurgitation, and small pericardial effusion

Click here to view

Physiologic interpretations

CXR provides independent information on the cardiac physiology. Latter is more valuable than the anatomic information in our experience.

Cardiac silhouette

Description of the cardiac silhouette includes cardiac position (levocardia, mesocardia, and dextrocardia), cardiac size (CM or not), and its configuration. A knowledge of the normal structures forming the cardiac borders is needed to correctly interpret the enlargement of individual chambers or vessels.

The normal cardiac borders are formed by the following structures [Figure 5]:

Figure 5: Chest X-ray posteroanterior view marking the structures forming cardiac borders. Ao – aorta, IVC – inferior vena cava, LA – left atrium, LV – left ventricle, PA – pulmonary artery, RA – right atrium, RV – right ventricle

Click here to view

Right border from the above downward is formed by the superior vena cava and right atrium
Likewise, the left border is formed by the aortic knuckle, main pulmonary artery, left atrial appendage (LAA), and left ventricle (LV)
The medial two-third of the inferior border is formed by the right ventricle (RV) and lateral one-third by LV
In the lateral view, the anterior border is formed by RV.

Cardiac configuration

While certain cardiac configurations can point accurately to the specific cardiac conditions, deviations are very common indeed [Table 1] and [Figure 6],[Figure 7],[Figure 8].

Figure 6: Chest X-ray posteroanterior view of a 10-year-old child, a known case of restrictive ventricular septal defect, came with exertional palpitations and dyspnea. Chest X-ray showed cardiomegaly (cardiothoracic ratio: 62%), left ventricle type configuration. Echo showed restrictive doubly committed ventricular septal defect, aortic valve prolapse, and severe aortic regurgitation. Acquired severe aortic regurgitation due to aortic valve prolapse leads to features of cardiac failure and cardiomegaly on chest X-ray

Click here to view

Figure 7: Chest X-ray anteroposterior view of a 2-month-old infant, with features of congestive heart failure and cyanosis (SpO2: 60%). Chest X-ray showed cardiomegaly, right atrial enlargement, right ventricle type ventricular configuration, and decreased pulmonary blood flow. Echo showed tricuspid atresia, malposed great vessels, moderate mitral regurgitation, and severe pulmonary stenosis. The child was severely anemic (hemoglobin: 5 g/dl). In this child with complex complex congenital heart disease, pulmonary stenosis (decreased pulmonary blood flow), the cause of cardiomegaly was mitral regurgitation and severe anemia

Click here to view

Figure 8: Chest X-ray posteroanterior view of a 13-year-old girl with dyspnea. Chest X-ray showed cardiothoracic index of 0.55. Left atrium enlargement is obvious by double shadow in the right atrium (white arrow), widening of angle of carina, and left atrial appendage enlargement (orange arrow). Ventricular configuration is of left ventricle type. Echocardiography findings were rheumatic heart disease, severe mitral regurgitation, dilated left atrium and ventricle with pulmonary hypertension

Click here to view

Table 1: Characteristic features of specific cardiac chamber enlargement

Click here to view

Increased pulmonary blood flow

The following features in CXR point to increased pulmonary vascularity^[3],[4] [Figure 9],[Figure 10],[Figure 11]:

Figure 9: Chest X-ray posteroanterior view of an 18-year-old female showing severe cardiomegaly (cardiothoracic ratio: 82%), left ventricle type ventricular configuration, left atrium enlargement, dilated ascending aorta (arrow), dilated pulmonary arteries, and increased pulmonary blood flow. Echo showed patent ductus arteriosus with large left to right shunt, dilated left atrium/left ventricle. There was bicuspid aortic valve leading to aneurysmal dilatation of ascending aorta

Click here to view

Figure 10: Chest X-ray anteroposterior view of a 4-month-old infant with clinical features of congestive heart failure with cyanosis (SpO2: 74%). Chest X-ray showed cardiomegaly (CT ratio: 76%), left ventricle configuration, left atrium enlargement (note prominent left atrial appendage marked by orange arrow), and increased pulmonary blood flow. Dilated right pulmonary artery (black arrow) while the space normally occupied by main pulmonary artery is empty (yellow arrow) due to malposition of great vessels. Echo showed complete transposition of great vessels, nonrestrictive perimembranous ventricular septal defect, dilated left atrium, left ventricle, D-malposed aorta, and pulmonary arterial hypertension

Click here to view

Figure 11: Chest X-ray anteroposterior view of a 2-month-old infant, admitted in intensive care unit with pneumonia with SpO2 of 90%. Chest X-ray showed cardiomegaly (cardiothoracic ratio: 70%), left ventricle configuration, left atrium enlargement, narrow mediastinum, and increased pulmonary blood flow. Pulmonary artery origin and course are marked by arrows and showed higher origin. Also note large pneumonic patch in the right upper zone. Echo showed truncus arteriosus type I, large outlet ventricular septal defect, increased pulmonary blood flow, and hyperkinetic pulmonary arterial hypertension

Click here to view

Vascular shadows traced up to lateral one-third of lung field
Increased end-on vessels (>3 in one lung field and >5 in both lung fields)
End-on vessel diameter > accompanying end-on bronchus diameter.

Decreased pulmonary blood flow

The radiological features of decreased pulmonary vascularity are as follows [Figure 12] and [Figure 13]:

Figure 12: Chest X-ray anteroposterior view of an 11-month-old infant with clinical features of congestive heart failure with cyanosis (SpO2: 80%). Chest X-ray showed severe cardiomegaly (cardiothoracic ratio: 80%), right atrial enlargement, homogenous shadow in area normally occupied by main pulmonary artery, and decreased pulmonary blood flow. Echo showed severe Ebstein's anomaly of tricuspid valve, severe tricuspid regurgitation, right to left shunt across foramen ovale, and dilated infundibular chamber. Important to note sharp cardiac borders in the case of Ebstein's anomaly

Click here to view

Figure 13: Chest X-ray anteroposterior view of a 2-year-old child, with features of congestive heart failure and cyanosis (SpO2: 55%). Chest X-ray showed cardiomegaly, right atrial enlargement, broad base apex (single ventricle), and decreased pulmonary blood flow. Echo showed complex congenital heart disease (double inlet left ventricle), malposed great vessels, and severe pulmonary stenosis with large pericardial effusion. Obliteration of cardiophrenic angle (arrow) is a feature of pericardial effusion

Click here to view

Vascular shadows will be reduced and lung fields appear darker
Central pulmonary arteries are of small size
Vascularity can be traced only upto the middle one-third of the lung field.

Pulmonary arterial hypertension

Dilatation of main and branch pulmonary arteries
“Peripheral pruning”, as in the case of Eisenmenger syndrome: Pruning is defined as more than 50% loss of vessel diameter at branching points^[3]
Calcification in proximal pulmonary arteries: The presence of calcium suggests severe and long-standing hypertension.

Pulmonary venous hypertension

X-ray features of PVH reflect the degree and duration of the same.^[3] The typical findings according to progressive stages of PVH are described in [Table 2].

Table 2: Stages of pulmonary venous hypertension and CXR findings

Click here to view

Mediastinum

Thymic shadow

On a frontal CXR, the thymus appears as a prominent soft-tissue density in the superior mediastinum, with margins inseparable from the superior cardiac margins. The lateral view, however, confirms the opacity's location in the anterior mediastinum. The anterior lateral margin often has smooth undulations from the overlying ribs and the costal cartilages, the so-called “wave” or “ripple” sign. A small notch sometimes marks the inferior border between the thymus and the heart. The angular corner, usually of the right lobe flattened at the right minor fissure, gives the classical “sail sign” [Figure 14].^[5],[6]

Figure 14: Chest X-ray anteroposterior view of a 4-month-old infant referred for echo with diagnosis of cardiomegaly on chest X-ray. Chest X-ray showed enlarged thymus overlapping left heart border and was interpreted as cardiomegaly, with normal lung vasculature, and parenchyma. Notch (arrow) marks the demarcation of thymic and cardiac shadow. On echo, there was no structural defect, normal cardiac dimensions, and function. Prominent thymus was also noted on echocardiography

Click here to view

Echocardiography

Echocardiography was performed by using Philips Affiniti 50c machine. A segmental sequential approach was followed in all patients. Lesions were identified and tabulated as congenital heart defects, acquired heart disease, cardiomyopathies, and pericardial effusion. Hemodynamic effects including pressure gradients, pulmonary artery pressure, cardiac dimensions, chamber sizes, and ejection fraction were separately documented.

Twelve lead electrocardiogram

Twelve-lead ECG was done if history suggestive of arrhythmia or unexplained ventricular dysfunction and abnormal Doppler rhythm and rates were noted during echocardiography.

Hematological investigations

Hemogram, renal and thyroid function tests, karyotype, and metabolic screening for dilated cardiomyopathy were done whenever clinically indicated.

Results

Patient characteristics

Out of 229 patients studied, 147 (64%) were male and 82 (36%) were female, with a median age of 6 months.

The indications of CXR are summarized in [Table 3]. The most common indication was murmur with failure to thrive in 70 (30.5%) patients, followed by evaluation of recurrent pneumonia in 65 (28%) patients, respiratory distress in 39 (17%), clinical congestive heart failure in 22 (9.6%), cyanosis in 10 (4%), isolated poor weight gain in 6 (2.6%), fever in 5 (2.4%), trisomy 21 in 3 (1.3%), palpitation in 4 (1.7%), and chest pain in 2 (0.8%) patients.

Table 3: Indications for performing chest X-ray in the study population by the primary pediatrician

Click here to view

X-ray findings

One-hundred and ninety-five children (85%) had true CM, whereas 34 (15%) were wrongly reported (i.e., false CM) [Table 4]. Out of 195 patients with true CM, 13 had mild CM and 182 had moderate-to-severe CM (CT ratio >60%) which accounts for 6.6% and 93%, respectively.

Table 4: Cardiomegaly on Chest X-ray and its interpretation (n=229)

Click here to view

Ventricular configuration

LV type CM was diagnosed in 127 (65%) children and RV type configuration in 35 (18%) children. In 17% children, the ventricular configuration could not be commented on CXR.

Pulmonary vasculature

On detailed reporting of CXR, diagnosis of shunt lesions (increased PBF) was made in 110 (56.4%). However, PBF quantum (Qp) was underreported in 12 and overreported in 20 children when compared with echocardiography.

Decreased Qp was reported in 15 children (7.6%) on CXR compared to 22 children on echo. In 7 children with decreased Qp on echocardiography, CXR reporting erroneously commented as normal Qp. All children with decreased Qp in CXR were having RV type or undefined ventricle configuration.

Normal Qp was seen in 54 (27.6%) children on CXR compared to 30 children on echo.

Pulmonary venous hypertension

PVH was detected in 10 (5.1%) children on CXR although echo demonstrated 33 children with moderate-to-severe PVH. CXR of 20 children with PVH was interpreted as increased Qp with PAH and 2 were categorized as normal pulmonary vasculature. All children with PVH were sick, and echocardiography was performed inside intensive care unit.

[Figure 15] shows the interpretation of pulmonary vasculature (increased Qp, decreased Qp, normal Qp, and PVH) on CXR and comparison with echocardiography findings.

Figure 15: Comparison of pulmonary vascularity on chest X-ray and echocardiogram, PBF: Pulmonary blood flow, PVH: Pulmonary venous hypertension

Click here to view

False CM was attributed to prominent thymus in 12 patients, an expiratory/rotated film in 18 patients, and chest deformity (pectus excavatum) in 2 patients. Two patients had a mass in the superior mediastinum (mediastinal mass) [Table 4].

Echocardiographic findings

Echocardiographic findings were interpreted with clinical and CXR findings in mind. Structural heart disease, functional heart disease, or a pericardial effusion was detected by echocardiogram in all children with true CM [Table 5].

Table 5: Causes of true cardiomegaly (based upon laboratory findings and echocardiography) (n=195)

Click here to view

Structural heart disease (71.7%) was the main etiology leading to CM followed by cardiomyopathy (13.3%). Causes of ventricular dysfunction were dilated cardiomyopathy, arrhythmia, hypertensive cardiac failure, diphtheritic cardiomyopathy, and glycogen storage disorder in decreasing order [Figure 16] and [Figure 17]. Among structural defects, critical aortic stenosis (AS) or coarctation of aorta (CoA) (n = 5) and anomalous origin of left coronary artery from pulmonary artery (ALCAPA) (n = 2) had severe ventricular dysfunction. Third group (n = 29, 14.8%) was composed of miscellaneous causes such as hypothyroidism, thiamine deficiency, idiopathic PAH, multisystem inflammatory disorder of children, pericardial effusion, and anemia [Figure 18]. Seven children with severe anemia (hemoglobin level: 2–4 g/dl, mean: 3 g/dl) had cardiac chamber enlargement on echocardiography and presented with CM [Figure 19]. Interestingly, CM was precipitated by severe anemia in five children with tetralogy of Fallot (TOF).

Figure 16: An 11-year-old male child with features of congestive cardiac failure. Chest X-ray posteroanterior view showed severe cardiomegaly, left ventricle type configuration, and biatrial enlargement. Lung fields appear clear. On echo, there was severe biventricular dysfunction, moderate tricuspid and moderate mitral regurgitation, and pulmonary venous hypertension. Features of pulmonary venous hypertension are not very marked on chest X-ray

Click here to view

Figure 17: A 9-month-old infant with features of congestive heart failure, referred for echo with chest X-ray finding of severe cardiomegaly. On echo, there was severe left ventricular hypertrophy with dysfunction. The child was diagnosed to be having glycogen storage disorder (Pompe's disease)

Click here to view

Figure 18: An 8-month-old infant with multisystem inflammatory syndrome in children referred with diagnosis of mild cardiomegaly on chest X-ray. Clinically, the child was febrile, in respiratory distress, and was having pallor (Hb 5 g/dl). On echo, there was dilatation of cardiac chambers secondary to anemia

Click here to view

Figure 19: A 3-month-old infant with breathing difficulty referred to echocardiography in view of chest X-ray finding of cardiomegaly. Chest X-ray anteroposterior view showed moderate cardiomegaly, right atrial enlargement, and right ventricle type configuration. On echo, there was moderate tricuspid regurgitation with peak gradient 50 mmHg, dilated RA, and right ventricle. As per history, the baby was breastfed and the mother was taking polished rice which is deficient in thiamine. The baby was empirically treated as thiamine deficiency and responded to treatment

Click here to view

Diagnostic Accuracy

The positive predictive value (PPV) of CM in CXR in predicting major cardiac disease was found to be 85% in our study. This proved the high association of cardiac disease when CM appeared in CXR. This was an important clinical lesson.

Discussion

Cardiac anomalies result in significant changes in CXR in a number of situations. It is possible to recognize important anatomical and physiological clues and also triage the severity of illness if a CXR is interpreted carefully with clinical findings in mind. These inferences can have a meaningful impact on clinical decision-making in a variety of situations. Before the development of echocardiography and other imaging modalities, CXR played an integral role in diagnosis and periodic evaluation of the patients with heart diseases. With advancement in diagnostic techniques, the role and ability to accurately interpret CXRs have unfortunately declined. However, this simple, accessible, and inexpensive investigation still holds significant independent value as a primary investigation tool in the evaluation of pediatric cardiac diseases. Advanced, expensive imaging modalities are not needed in all.

Satou et al. prospectively evaluated 95 children (median age: 5 years) to determine the usefulness of heart size on CXR in predicting CE.^[7] The presence of CE by echocardiography was used as the gold standard, and CXR was done to compare echo and CXR diagnosis of CE. Their study showed that the sensitivity of CXR to identify CE was 58.8% (95% confidence interval (CI) 32.9, 81.6], with a PPV of 62.5% (35.4, 84.8). The specificity was 92.3% (84.0, 97.1), with a negative predictive value of 91.1% (82.6, 96.4). These data suggest that the assessment of CE on CXR to predict CE by echocardiography has a relatively high specificity and negative predictive value but a low sensitivity and PPV. The limitations of CXR as a diagnostic test should be understood by clinicians using the test when screening children for cardiac diseases.

In a study in 1999, Birkebaek et al. studied reproducibility and the accuracy of a pediatric cardiologist's assessment of CXR with respect to the presence or absence of heart defects in 98 children with an asymptomatic heart murmur and found the mean sensitivity 0.3, mean PPV 0.4, mean specificity 0.86, and the mean predictive value of a negative test 0.80.^[8] The study concluded that the use of CXR could not be recommended for initial evaluation of an asymptomatic child with a heart murmur. A study by Jane L McKee in adult patients showed that sensitivity and specificity of CXR to identify CM as 40% and 91% respectively. However, this study was done in adults presenting with non-ST elevation myocardial infarction, hence a different result in infants and children would not be surprising.^[9]

In this study, we retrospectively evaluated the diagnostic accuracy of frontal CXR with CM of 229 children referred to pediatric cardiology unit. All children underwent detailed cardiac evaluation, history, examination, interpretation of CXR by a pediatric cardiologist, and echocardiography. As per referral records, indications of ordering a CXR in the majority of children were clinically significant symptomatic stage (90%) while 10% were referred with incidental detection of CM on CXR as part of preoperative anesthesia clearance.

True CM was found in 85%, whereas false CM was noted in15% of the children.

In children with true CM, the most common cause was structural heart defect (71.7%), followed by ventricular dysfunction (13.3%) and miscellaneous disorders in 14.8% of the children. Common structural defects were shunt lesions (ventricular septal defect, atrial septal defect, patent ductus arteriosus, and atrioventricular septal defect in decreasing order) and cyanotic heart defects (total anomalous pulmonary venous connection, transposition of great arteries, truncus arteriosus, and single ventricle) followed by critical obstructive lesions (AS and CoA with ventricular dysfunction). CM is not a classical feature of TOF, but in our study, 5 children with TOF had CM on CXR and the culprit was severe anemia. Causes of ventricular dysfunction in our series were dilated cardiomyopathy, glycogen storage disorder, arrhythmogenic cardiomyopathy, diphtheria, hypertension, ALCAPA, and critical left-sided obstructive lesions. Miscellaneous causes of CM were anemia, thiamine deficiency, hypothyroidism, primary PAH with severe PAH and tricuspid regurgitation, pericardial effusion, multisystem inflammatory syndrome in children, and hypertensive heart failure.

On detailed reporting of CXR with CM, we were able to reach a functional diagnosis in majority (80%) of the children. In children with false CM, thymic shadow, expiratory film, or pectus excavatum as a contributing factor was easily picked up.

Our study supports that CXR, a noninvasive, cost-effective, and easily available investigation, should be considered as part of initial evaluation in children. If interpreted properly, a functional diagnosis can be reached in majority of the cases. A PPV of 85% indicates that it can be used as a good screening test to screen for cardiac diseases in outpatient clinics. CXR also helps to assess the severity of illness. Hence, pediatric echocardiography, a comparatively costly modality with less availability of expertise in a developing country like India can be prioritized .

Limitations of our study

We included only those patients who had both CXR and echocardiography performed as part of their clinical care so that we could simultaneously compare these two imaging modalities head-to-head. As part of research protocol in government institute, we ensured that there was no financial burden on the family.

Furthermore, we relied on referral from a treating pediatrician suspecting CM in CXR. There is a definite possibility that some true cardiac patients could not reach the pediatric cardiac center due to this referral bias. This can be reduced only with growing awareness and reorientation of primary pediatrician in interpreting CXR.

Finally, in this study, only the children with CXR diagnosis of CM were included. Hence, it was not possible to estimate sensitivity or negative predictive value.

Conclusion

CXR has a variable sensitivity but high specificity in determining CM and reaching a functional cardiac diagnosis. Thus, CM on CXR needs a detailed workup in the form of detailed head-to-toe clinical examination. Leaning to interpret CXR is a fascinating lesson. CXR can also be used as a screening tool to rule out cardiac disease in resource-limited settings. Initial medical therapy can be started considering the clinical scenario and CXR findings together. Echocardiography remains the gold standard but should be used only if initial investigations are abnormal and trained workforce is available.

It is the responsibility of an ordering physician to interpret the CXR with clinical findings in mind and then order further diagnostic tests such as ECG and echocardiography as needed.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patient (s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initial s will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

Munesh Tomar is an editorial board member of the Journal of The Indian Academy of Echocardiography & Cardiovascular Imaging. The article was subject to the journal's standard procedures, with peer review handled independently of this editor and their research groups. There are no other conflicts of interest.

References

1.	Ovitt TW. The chest roentgenogram. In: Moss AJ, Adams FH, editors. Heart Disease in Infants, Children, and Adolescents. Baltimore: Williams and Wilkins; 1995. p. 182-3.
2.	Abdulla, X-Ray Ra-id, Douglas M. Luxenberg. “Chapter 2 Cardiac Interpretation of Pediatric Chest.” (2017). Ra-id Abdulla (ed.), Heart Diseases in Children: A Pediatrician's Guide, DOI 10.1007/978-1-4419-7994-0_2, © Springer Science+Business Media, LLC 2011.
3.	Elliot LP. Pulmonary vascularity on the plain chest radiograph. Cardiol Clin 1983;1:545-64.
4.	Hegde M, Vijayalakshmi IB. Role of radiography in congenital heart diseases In: Vijayalakshmi IB, Rao S, Chugh R, editors. A comprehensive Approach to Congenital Heart Diseases. 1^st ed. New Delhi: Jaypee Brothers Medical Publishers (P) Ltd.; 2013. p. 190-202. DOI: 10.5005/jp/books/12075_13.
5.	Kuhn JP, Slovis TL, O Haller J. Caffey's Pediatric Diagnostic Imaging. 10^th ed., Vol. 1. St Louis: Mosby; 2004. p. 1175-84.
6.	Singleton EB, Wagner ML, Dutton RV. Radiologic Atlas of Pulmonary Abnormalities in Children. Philadelphia, PA: WB Saunders; 1971. p. 13-23.
7.	Satou GM, Lacro RV, Chung T, Gauvreau K, Jenkins KK. Heart size on chest X-ray as a predictor of cardiac enlargement by echocardiography in children. Pediatr Cardiol 2001;22:218-22.
8.	Birkebaek NH, Hansen LK, Elle B, Andersen PE, Friis M, Egeblad M, et al. Chest roentgenogram in the evaluation of heart defects in asymptomatic infants and children with a cardiac murmur: Reproducibility and accuracy. Pediatrics 1999;103:E15.
9.	McKee JL, Ferrier K. Is cardiomegaly on chest radiograph representative of true cardiomegaly: A cross-sectional observational study comparing cardiac size on chest radiograph to that on echocardiography. N Z Med J 2017;130:57-63.