Comparison of Chest Ultrasound to Chest Radiography in screening for Pediatric Tuberculosis

doi:10.21203/rs.3.rs-2994488/v1

Download PDF

Research Article

Comparison of Chest Ultrasound to Chest Radiography in screening for Pediatric Tuberculosis

https://doi.org/10.21203/rs.3.rs-2994488/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

There is a lack of robust data on the utility of chest ultrasound scans (CUS) for triage and screening for pediatric TB. Therefore, we set out to compare CUS findings to chest radiography (CXR), which is the recommended imaging modality in children with presumptive PTB.

Methods

Eighty children up to 14 years of age with presumptive TB had a CUS and CXR performed by two separate radiologists, who looked for the presence of consolidation, lymphadenopathy and pleural effusion on both modalities. These were compared using Fisher’s exact test for independence to determine whether there was a significant difference in the findings between the two modalities. Cohen's kappa coefficient was used to calculate the inter-reader agreement between CXR and CUS. The following interpretations of the results were used: less than zero, no agreement; 0 to 0.20 slight agreement, 0.21 to 0.40 fair agreement, 0.41 to 0.60 moderate agreement, 0.61 to 0.80, substantial agreement and 0.81 to 1, almost perfect agreement. STATA version 15 was used for analysis.

Results

The proportion of children with abnormal findings (consolidation, lymphadenopathy and pleural effusions) on CUS was 65% (52/80) and 81.3% (65/80) on CXR. 62.5% (33/52) of those with abnormal findings on CUS and 51.3% (33/65) on CXR were likely to have TB. The overall inter-reader agreement for these characteristics was moderate (κ-0.42).

Conclusion

CUS detected more abnormalities in children with suspected TB. Overall, the findings were comparable to those of CXR, except for lymphadenopathy. Ultrasound is a promising screening and triage tool for treating pediatric TB. However, there is a need to compare the CUS with the gold standard.

Chest ultrasound

Chest radiograph

Comparison

Screening

Pediatric Tuberculosis

Perihilar or mediastinal lymphadenopathy is the most characteristic finding of childhood TB on chest radiography (CXR) [1–3]. CXR, is a two-dimensional grey-scale modality in which lymph nodes are difficult to discern from overlapping vascular and mediastinal structures (1). Subsequently, there is a poor intra-and inter-reader agreement for lymphadenopathy on CXR (k = 0.00-0.40) [4, 5].

Other common CXR findings described in children with PTB are consolidation and pleural effusion, but these are also observed in other respiratory infections [6]. However, the available literature reported that pleural effusions and consolidation are more common in children with microbiologically confirmed PTB [7].

CXR is the imaging modality of choice for screening and triaging pulmonary TB [8]. This is despite the exposure of children to ionizing radiation and its low availability in resource-limited settings. There is also a scarcity of equipment and skilled personnel to run and interpret the images. In contrast, CUS devices are relatively inexpensive, portable, and do not emit ionizing radiation. Sometimes, ultrasonography may not require radiological staff [9]. This makes them attractive for practical reasons in low-resource settings. In addition, CUS is a bedside test that can be performed in less than 8 minutes and is easily reproducible [6].

However, there is a lack of robust data on the use of CUS [7, 9–12]. The purpose of this study, therefore, was to compare CUS and CXR findings in children with presumptive PTB. This may potentially result in the consideration of CUS as a modality for screening and triaging for pediatric TB.

Study Design and Settings

This cross-sectional study was conducted at Mulago National Referral Hospital in Kampala, Uganda, East Africa from December 2020 to May 2021. On average, 12 children were diagnosed with TB monthly at the hospital [13]. These children were diagnosed based on clinical signs and symptoms, with a review of a chest radiograph if available, and Xpert MTB/RIF Ultra testing if a sputum sample could be obtained according to the national guidelines [14].

Study Participants

The study participants were children 0–14 years with a cough of any duration with HIV or a cough for > 2 weeks for children without HIV, weight loss or failure to thrive over the last three months, persistent fever for > 2 weeks not responding to any treatments or having been in contact with someone with confirmed Pulmonary Tuberculosis (PTB).

Children were excluded from the study if they were already on TB treatment or TB prophylaxis for > 72 hours. They were also not eligible if their parents did not consent for those under eight years or if they did not assent for those older than eight years. Participants were enrolled using simple random sampling in pre-selected wards.

Radiological tests

All study participants had a CUS and CXR performed by radiologists with a combined experience of about 32 years. Two study radiologists interpreted the images. If there was a discrepancy, they sought consensus with the principal investigator (PI), who was also a radiologist with at least 15 years of experience acting as an arbitrator. The radiologist who performed the chest ultrasound scan were blinded to the chest radiograph findings.

Chest Ultrasound Scan

A portable, low-cost, greyscale ultrasound machine (Edan, Model U60) with a bandwidth of 5.0 to 10.0 MHz linear probes was used to scan the chest of the children in this study. The radiologists were responsible for the methodological scanning of the children. The older children were scanned either sitting or supine plus in left and right lateral decubitus positions to examine the lungs and pleura, while the younger children were scanned on the laps of the caregiver. The chest was divided into 14 regions: the left and right, anterior, lateral, posterior, superior, and inferior regions, and the two apices. All regions were scanned in longitudinal and transverse (intercostal) planes. Mediastinal ultrasound was performed through the suprasternal notch. The child was placed in the supine position with the neck slightly extended to improve access to the suprasternal notch, and transverse and oblique views were obtained. All the cine clips were saved.

Chest Radiograph

To ensure the highest achievable quality of CXR reports, the posterior anterior and lateral CXRs were independently reviewed by two radiologists reporting on a standardized record sheet for pleural effusions, lymphadenopathy, and consolidations, plus the radiological diagnosis of the likelihood of TB.

In cases of disagreement on the final diagnosis, a third radiologist was consulted, and the final diagnosis was based on a consensus opinion. All the images were digitally achieved.

Sample Size

To compare CUS with CXR in children with presumptive TB, the following formula for the sample size for diagnostic accuracy was used:

Z_{α/2 =} 1.96, standard normal value corresponding to 5% level of significance

S_N = 0.76, Assuming a sensitivity of Chest Ultrasound in the detection of TB of 76%.

d = 0.1, precision

P = 0.89, Prevalence of pulmonary TB among children under 15 (Wobudeya et al., 2015)

we get a sample size of N = 80 participants.

A sample size of 80 participants was considered in this study.

Data Analysis

The study characteristics were summarized using mean and standard deviation (SD) for normally distributed numerical data and median with interquartile range (IQR) for non-normally distributed numerical data. Categorical data are summarized using frequencies and proportions.

The presence of consolidation, lymph nodes, and pleural effusion on both modalities (CXR and CUS) was compared using Fisher’s exact test for independence to determine if there was a significant difference in the findings between the two tests. Cohen's kappa coefficient was used to calculate the inter-reader agreement between CXR and CUS. The following interpretations of the results were used: less than zero, no agreement; 0 to 0.20 slight, 0.21 to 0.40 fair, 0.41 to 0.60 moderate, 0.61 to 0.80, substantial and 0.81 to 1, almost perfect agreement. STATA version 15 was used for analysis.

Ethical Considerations

Ethical approval

to conduct this study was granted by the Makerere University School of Medicine Research Ethics Committee (REC No. 2020 − 138). All methods were performed in accordance with the relevant guidelines and regulations or declaration of Helsinki. Informed consent was obtained from their parents or legal guardians given that they were all under 16 years. All investigations used in the study were part of the standard of care for a long time for different purposes, but not for the diagnosis of PTB.

Characteristics of the Study Participants

The characteristics of the participants are listed in Table 1. Eighty children were subjected to CUS and CXR examination, with 52.5% (42/80) being male. Their average age was 42.9 months (1-144). The majority of the children (50.0%, 40/80)) were aged between one and five years, with 21.3% (17/80) being younger. In contrast, 15.0% (12/80) and 13.7% (11/80) were aged 5–10 years and 10–15 years, respectively.

Table 1

Characteristics of the study participants
Characteristic	Total Participants (n = 80)
Sex, n (%)
Male	42 (52.5)
Female	38 (47.5)
Mean Age in months	42.89
Age Ranges, n (%)
< 1 year	17 (21.3)
1-<5 years	40 (50.0)
5years -<10 years	12 (15.0)
10 years-<15 years	11 (13.7)

Comparison of Chest Ultrasound Findings and Chest Radiography Findings

The comparison of Chest ultrasound and Chest radiography findings is in Table 2. A total of 65% (52/80) of the children had abnormal findings (consolidation, pleural effusion, or lymph nodes) on CUS, while 81.3% (65/80) of the children had abnormal CXR findings. Among those with abnormal findings, 62.5% (n = 50) and 51.3% (n = 41) were likely to have TB on CUS and CXR, respectively.

Consolidation was found in 38.8% (31/80) of children on CUS. Among those with consolidation on CUS, 93.5% (29/31) were likely to have TB. Consolidation on CXR was observed in 33.75% (n = 27) of the children, all of whom were likely to have TB. There was no significant difference between consolidation findings on either CUS or CXR (P = 0.491).

Pleural effusion was detected in 20.0% (16/80) of CUS scans. All those who were found to have pleural effusions were likely to have TB. Pleural effusions were found in 13.8% of the CXRs. Almost all of these CXRs were likely to have TB ((90.9% (10/11)). There was no significant difference between the findings of pleural effusion on CUS and CXR (P = 0.407).

A total of 6.25% (n = 5) of children had lymphadenopathy on CUS. All of these patients were likely to have TB. However, 33.8% (27/80) of the children had lymphadenopathy on CXR, and only 14.8% (4/27) were likely to have TB. The differences in the findings of lymphadenopathy between CUS and CXR were significant (P < 0.001).

Table 2

Comparing Chest Ultrasound scan and Chest Radiography Features
	CXR Findings		CUS Findings
Features	TB Likely, n (%)	TB Unlikely, n (%)	TB Likely, n (%)	TB Unlikely, n (%)	Fishers exact test (P-value)
Consolidation					0.491
Yes	27 (81.8)	0 (0.0)	29 (59.2)	2 (6.5)
No	6 (18.2)	47 (100.0)	20 (40.8)	29 (93.5)
Pleural Effusion					0.407
Yes	10 (30.3)	1 (2.1)	16 (32.7)	0 (0.0)
No	23 (69.7)	46 (97.9)	33 (67.3)	31 (100.0)
Lymph adenopathy
Yes	4 (1.7)	23 (25.3)	5 (10.2)	0 (0.0)	< 0.001
No	1 (3.3)	52 (49	44 (89.8)	31 (100.0)

Inter-reader agreement on chest ultrasound scan and chest radiograph

Table 3 shows the inter-reader agreement for CUS and CXR. The inter-reader agreement for consolidation and pleural effusion on the CUS and CXR was moderate (κ-0.41 & κ-0.43, respectively), while that of lymphadenopathy was slight (κ-0.16). The overall inter-reader agreement for the likelihood of TB was moderate (κ-0.42) for both CUS and CXR.

Table 3

Inter-reader agreement of Chest Ultrasound scan and Chest Radiography
Imaging Finding	Proportions on CXR	Proportions on CUS	Degree of Agreement	Kappa kohen statistics	P-Value
Consolidation	27/80	31/80	58/80	0.407	0.0001
Pleural Effusion	11/80	16/80	67/80	0.425	< 0.001
Lymph adenopathy	27/80	5/80	56/80	0.162	0.012
Overall (TB Likely)	33/80	49/80	56/80	0.423	< 0.001

The purpose of this study was to compare CUS findings with those of CXR in children with presumptive PTB. More children had an abnormal finding on CXR than on CUS, but the reverse was true for those with a likelihood of TB. There was no significant difference in finding consolidations and pleural effusions between CUS and CXR, unlike lymphadenopathy. When the reading of these features was compared between CUS and CXR, there was a moderate inter-reader agreement for consolidations and pleural effusions, but only a slight inter-reader agreement for lymphadenopathy.

Slightly more pleural effusions were detected on CUS (20%) than on CXR (13.75%), but the difference was not statistically significant (P = 4.07). This is comparable to the literature that reported CUS to be superior in detecting pleural effusion compared to CXR [6, 15, 16]. Pleural effusion is more likely to be associated with TB on both CXR and CUS, although pleural effusion can occur in children with pneumonia caused by other bacteria (pneumococcus and Staphylococcus) or viruses [6]. However, these children are usually younger and acutely unwell compared to children with pleural effusion due to TB [17]. PTB is a probable diagnosis when pleural effusion is present, especially if there is consolidation.

The presence of consolidation was associated with the likelihood of TB on both the CUS and CXR. Consolidation is commonly seen in children with pulmonary TB. However, it has also been observed in children with bacterial infections and less commonly, viral pneumonia. “Co-infection” was also common, where children with underlying pulmonary TB present co-infected with viral or bacterial pneumonia [17]. We encourage radiologists and clinicians in the presence of consolidation to look for other features, especially pleural effusion.

Less lymphadenopathy was detected on CUS than on CXR. This contrasts with the available literature, which reported more lymphadenopathy in CUS than in CXR [6, 10]. Lymphadenopathy with or without parenchymal abnormalities is the radiological hallmark of primary TB in children. This is often more prevalent in children aged < 3 years than in those between 4-and 15 years [18]. Since lymphadenopathy, especially in the mediastinal region, is likely to be the only feature of CXR, CUS efforts should be made to look for it in presumptive TB patients, especially those less than 3 years of age.

This study has key implications for clinical practice, especially in the management of children with suspected TB, using ultrasonography as a screening tool. CUS can detect some features suggestive of TB in children, and it has the potential to be used as an initial modality before children are exposed to chest x-rays with ionizing radiation. CUS has become even more useful in rural and remote healthcare settings where x-ray equipment is not available. Thus, we encourage the screening of suspected children in these settings initially with CUS.

This study compared the findings of CUS to CXR. Ideally, we should have compared CUS with Computer Tomography (CT) and Magnetic Resonance and Imaging (MRI), which are the gold standard for imaging. However, Chest CT scan has very high radiation dose making it not a first line imaging modality in children who have radiosensitive tissues. MRI scans are few and very expensive in the setting where the study took place. Despite this, however, the study does provide useful indicators that CUS can be used as an initial screening tool for children suspected to have TB and aid in identifying which ones require further imaging and management. Moving forward there is need for comparing the additional benefit of using CUS to diagnose TB in children earlier.

CUS detected more abnormalities among children that were likely to have TB. Its findings were comparable to CXR in finding other features except for lymphadenopathy. Ultrasound is a promising screening and triage tool for treating pediatric TB. However, there is a need to compare the CUS with the gold standard.

Acknowledgement

We would like to acknowledge the study participants and their caregivers for allowing to participate in this study and in turn contribute to science. We are grateful for your participation.

Ethics approval

Ethical approval to conduct this study was granted by the School of Medicine Research Ethics Committee at Makerere University (REC No. 2020-138).

Consent to participate

Informed consent was obtained from their parents or legal guardians given that they were all under 16 years.

Conflicts of interest

The authors declare no competing interests exist.

Funding

There was no funding secured for this work, all the authors donated their time, to complete it.

Author’s contribution

All the authors read and approved the final manuscript. The following authors participated in developing the idea into a concept (GE, CO, FA, AGM and MK), data collection and analysis (MO, CO, GE, SB, MK, DIN), drafting the manuscript (GE, CO, MO, FA, DIN, SB, AGM and MK), reviewing, critically appraising, and approving the final version of the manuscript (GE, CO, MO, FA, DIN, SB, AGM and MK)

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

Consent for publication

Not applicable.

García-Basteiro AL, López‐Varela E, Augusto OJ, et al. Radiological findings in young children investigated for tuberculosis in Mozambique. PLoS ONE. 2015;10(5):e0127323.
Richter-Joubert L, Andronikou S, Workman L, Zar HJ. Assessment of airway compression on chest radiographs in children with pulmonary tuberculosis. Pediatr Radiol. 2017;47(10):1283–91. https://doi.org/10.1007/s00247‐017‐3887‐9.
Schaaf HS, Marais BJ, Whitelaw A, Hesseling AC, Eley B, Hussey GD et al. Culture-confirmed childhood tuberculosis in Cape Town, South Africa: a review of 596 cases. BMC Infect Dis. 2007; 7: 140. doi: 10.1186/1471-2334-7-140 PMID: 18047651.
du Toit G, Swingler G, Iloni K. Observer variation in detecting lymphadenopathy on chest radiography. Int J Tuberc Lung Dis. 2002;6:814–7.
Swingler GH, du Toit G, Andronikou S, van der Merwe L, Zar HJ. Diagnostic accuracy of chest radiography in detecting mediastinal lymphadenopathy in suspected pulmonary tuberculosis. Arch Dis Child. 2005;90:1153–6.
Heuvelings CC, Bélard S, Andronikou S, Lederman H, Moodley H, Grobusch MP, Zar HJ. Chest ultrasound compared to chest X-ray for pediatric pulmonary tuberculosis. Pediatr Pulmonol. 2019 Dec;54(12):1914–20. 10.1002/ppul.24500. Epub 2019 Sep 1. PMID: 31475477; PMCID: PMC6899616.
World Health Organization. (‎2016)‎. Chest radiography in tuberculosis detection: summary of current WHO recommendations and guidance on programmatic approaches. World Health Organization https://apps.who.int/iris/handle/10665/252424.
Bigio J, Kohli M, Klinton JS, MacLean E, Gore G, Small PM, et al. Diagnostic accuracy of point-of-care ultrasound for pulmonary tuberculosis: A systematic review. PLoS ONE. 2021;16(5):e0251236. https://doi.org/10.1371/journal. pone.0251236.
Bosch-Marcet J, Serres‐Creixams X, Zuasnabar‐Cotro A, Codina‐Puig X, Catala‐Puigbo M, Simon‐Riazuelo L. Comparison of ultrasound with plain radiography and CT for the detection of mediastinal lymphadenopathy in children with tuberculosis. Pediatr Radiol. 2004;34:895–900.
Bosch-Marcet J, Serres‐Créixams X, Borrás‐Pérez V, Coll‐Sibina MT, Guitet‐Juliá M, Coll‐Rosell E. Value of sonography for follow‐up of mediastinal lymphadenopathy in children with tuberculosis. J Clin Ultrasound. 2007;35:118–24.
Moseme T, Andronikou S. Through the eye of the suprasternal notch: point-of‐care sonography for tuberculous mediastinal lymphadenopathy in children. Pediatr Radiol. 2014;44:681–4.
Heuvelings CC, Bélard S, Andronikou S, Jamieson-Luff N, Grobusch MP, Zar HJ. Chest ultrasound findings in children with suspected pulmonary tuberculosis. Pediatr Pulmonol. 2019;54:463–70. https://doi.org/10.1002/ppul.24230.
Wobudeya E, Lukoye D, Lubega IR, Mugabe F, Sekadde M, Musoke P. Epidemiology of tuberculosis in children in Kampala district, Uganda, 2009–2010; a retrospective cross-sectional study. BMC Public Health 2015 Sep 25;15:967. doi: 10.1186/s12889-015-2312-2. PMID: 26407719; PMCID: PMC4582927.
Ministry of Health. National Tuberculosis and Leprosy Program. Management of tuberculosis in children: a health worker guide. Kampala, Uganda: Uganda Ministry of Health; 2015.
Soni NJ, Franco R, Velez MI, et al. Ultrasound in the diagnosis and management of pleural effusions: ultrasound and pleural effusions. J Hosp Med. 2015;10(12):811–6. https://doi.org/10.1002/jhm.2434.
Lichtenstein D, Goldstein I, Mourgeon E, Cluzel P, Grenier P. Comparative diagnostic performances of auscultation, chest radiography, and lung ultrasonography in acute respiratory distress syndrome. Anesthesiology. 2004;100(1):9–15.
Diagnostic CXR. Atlas for Tuberculosis in Children. A Guide to Chest X-ray Interpretation, 2nd ed.; International Union against Tuberculosis and Lung Disease (The Union): Paris, France, 2022. Available online: https://theunion.org/technical-publications/diagnostic-cxr-atlas-for-tuberculosis-in-children (accessed on 16 April 2022).
Burroughs M, Beitel A, Kawamura A, et al. Clinical presentation of TB in culture-positive children. Pediatr Infect Dis. 1999;18:440–6.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comparison of Chest Ultrasound to Chest Radiography in screening for Pediatric Tuberculosis

Status:

Version 1

Abstract

Background

Methods

Study Design and Settings

Study Participants

Radiological tests

Chest Ultrasound Scan

Chest Radiograph

Sample Size

Data Analysis

Ethical Considerations

Results

Characteristics of the Study Participants

Comparison of Chest Ultrasound Findings and Chest Radiography Findings

Inter-reader agreement on chest ultrasound scan and chest radiograph

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1