Kyomuhangi-CHS-Masters.Pdf

MAKERERE UNIVERSITY

COLLEGE OF HEALTH SCIENCES

DEPARTMENT OF RADIOLOGY

ACCURACY OF CHEST ULTRASOUND IN DIAGNOSING PNEUMONIA IN PEDIATRIC PATIENTS AT MULAGO NATIONAL REFERRAL HOSPITAL, KAMPALA, UGANDA.

PRINCIPAL INVESTIGATOR:

DR KYOMUHANGI AGNES, MBChB, MUK

SUPERVISORS:

1. DR BUGEZA SAM

MBChB(MUK), MMED (Rad).

2. DR EREM GEOFFREY

MBChB(MUST), MMED(Rad)

3. DR MWOROZI EDISON ARWANIRE

MBChB(MUK), MMED (SENIOR CONSULTANT, Pead).

A DISSERTATION SUBMITTED TO SCHOOL OF GRADUATE STUDIES IN PARTIAL FULLFILLMENT OF THE REQUIREMENT FOR AWARD OF THE DEGREE OF MASTERSAggie OF MEDICINE IN RADIOLOGY AT MAKERERE UNIVERSITY. [Date]

AUGUST 2019 i

DECLARATION

I Kyomuhangi Agnes, hereby declare that the work presented in this dissertation has not been presented for any other degree in this university.

Signed…………………………………. ………………………………….

DR. KYOMUHANGI AGNES Date

This dissertation has been submitted for examination with approval of the following supervisors;

Signed…………………………………. ………………………………….

DR. BUGEZA SAMUEL Date

MBChB, MMed Rad

Specialist Radiologist / lecturer, College of Health Sciences, Makerere University.

Signed……………………………….... …………………………………..

DR. EREM GEOFFREY Date

MBChB, MMed Rad

Specialist Radiologist / lecturer, College of Health Sciences, Makerere University.

Signed…………………………………. …………………………………….

DR. MWOROZI EDISON ARWANIRE Date

MBChB, MMed Pead

Consultant Pediatrician Mulago National Referral Hospital / Senior lecturer, College of Health Sciences, Makerere University.

DEDICATION

To my family, for being a constant source of inspiration, I am eternally grateful for their love, unwavering encouragement and all round support during the course of my masters programme.

iii

ACKNOWLEDGEMENTS

The development and completion of this course/work was first of all made possible, by the Almighty God who has been faithful providing me with grace, mercy and strength.

The funding to do this study was made possible by Uganda Cancer Institute (UCI-AfDB) scholarship which sponsored me throughout my masters programme and this study.

Special thanks go to my supervisors; Dr. Bugeza Samuel, Dr. Erem Geoffrey and Dr. Mworozi Edison Arwanire for their guidance and advise they gave me throughout the preparation of this dissertation.

I would also like to thank my mother, Ms. Bekinga Monica Kyomuhangi, my guardian, Mr. John Diisi and my siblings for their sympathetic ear, wise counsel, support and for providing happy distraction to rest my mind outside this course and study.

I further extend my special thanks to Dr. Muyinda Zeridah and Dr. Ameda Faith for their support and help during this study.

My sincere thanks also go to my classmates for their encouragement and help during this study and dissertation preparation.

Special thanks to the members of the department of radiology, Mulago hospital, mostly Mr. Bwangu and Mr. Senoga who assisted me during this study.

I thank the pediatric team at Acute Care Unit at Mulago National Referral Hospital for their co- operation that enabled me carry out my research smoothly.

Finally, huge and special thanks to the study participants and their parents / guardians for participating in this study and providing the information that made it a success.

Table of Contents ABBREVIATIONS AND ACRONYMS...... ix OPERATIONAL DEFINITIONS...... x ABSTRACT ...... 12 INTRODUCTION ...... 14 1.1 PROBLEM STATEMENT ...... 15 1.2. JUSTIFICATION...... 15 1.3. RESEARCH QUESTIONS...... 16 1.4. OBJECTIVES OF THE STUDY...... 16 1.4.1. GENERAL OBJECTIVE...... 16 1.4.2. SPECIFIC OBJECTIVES...... 16 2.0. LITERATURE REVIEW ...... 18 2.1. Burden of pneumonia...... 18 2.2. Aetiology of pneumonia...... 18 2.3. Pathogenesis of pneumonia...... 19 2.4. Risk factors of pneumonia...... 19 2.5. Clinical diagnosis of pneumonia...... 19 2.6. Role of plain chest radiography in diagnosis of pneumonia...... 20 2.7. Role of chest computed tomography in diagnosis of pneumonia...... 21 2.8. Role of chest ultrasound in diagnosing pneumonia...... 21 3.0 METHODOLOGY ...... 25 3.1. STUDY DESIGN...... 25 3.2. STUDY SETTING...... 25 3.3. STUDY POPULATION...... 25 3.3.1. Target population ...... 25 3.3.2. Accessible population ...... 25 3.3.3. Eligible population ...... 26 3.3.4. Inclusion criteria ...... 26 3.3.5. Exclusion criteria ...... 26 3.4. Sample size determination...... 26 3.4.1. Sampling strategy...... 26

3.5. DATA COLLECTION TECHNIQUE...... 27 3.5.1. Patient demographics...... 27 3.5.2. The chest X rays...... 27 3.5.3. Chest ultrasounds...... 27 3.6. DATA COLLECTION TOOLS...... 31 3.6.1 Independent variables ...... 31 3.6.2. Dependent variables ...... 31 3.7. Quality control...... 32 3.8. Data management...... 33 3.8.1. Data analysis...... 33 3.9. Ethical considerations...... 33 3.10. Limitations of the study...... 34 4.0 RESULTS...... 35 4.1. Participants’ characteristics ...... 36 4.1.1. Demographic characteristics ...... 36 4.1.2 Clinical features ...... 37 4.1.3 Radiological diagnostic yield of pneumonia by both Chest ultrasound and Chest X-ray ...... 38 4.2 Diagnostic accuracy of chest ultrasonography in children with suspected pneumonia at Acute Care Unit ...... 40 4.2.1 Accuracy of chest ultrasonography with final clinical diagnosis as a standard reference in the detection of pneumonia...... 40 4.2.2 Comparison of chest ultrasonography with chest radiography...... 41 4.3 The correlation between the clinical and radiological findings of chest ultrasonography and chest radiography ...... 43 4.4 Images ...... 45 5.0 DISCUSSION ...... 48 5.1 Introduction ...... 48 5.2 Demographic profile ...... 48 5.3 Clinical Profile ...... 48 5.4 Radiological findings ...... 49 5.5 Accuracy of chest ultrasonography in diagnosing pneumonia...... 51 5.5 Limitations...... 53

6.0. Conclusion and recommendation...... 54 6.1 Conclusions...... 54 6.2 Recommendations...... 54 REFERENCES...... 55 APPENDIX 1 ...... 59 DATA COLLECTION TOOL...... 59 APPENDIX 2A ...... 63 CONSENT FORM – ENGLISH VERSION ...... 63 CONSENT FORM – LUGANDA VERSION...... 66 APPENDIX 2B ...... 69 ASSENT FORM -ENGLISH VERSION ...... 69 ASSENT FORM-LUGANDA VERSION ...... 71 APPENDIX 3 ...... 73 WORK PLAN...... 73 APPENDIX 4 ...... 74 BUGDET FOR STUDY ...... 74

vii

LIST OF FIGURES AND TABLES.

Figure 1: Conceptual Framework 17

Figure 2 and 3: Illustrations 29

Figure 4: Flow chart 30

Figure 5: Study profile 35

Figure 6-10: Images from this study 45-47

Table 1: Baseline demographic Characteristics 36

Table 2: Clinical features 37

Table 3a and 3b: Radiological diagnostic yield 38-39

Table 4a and 4b: Comparing chest ultrasound with final diagnosis 40

Table 5a and 5b: Comparing chest ultrasound with chest X-ray 41

Table 6a: Correlation between clinical findings and chest ultrasound 43

Table 6b: Correlation between clinical findings and chest X-ray 44

viii

Abbreviations and acronyms. ACU- Acute Care Unit.

ALR - Anterior lower region.

AUR- Anterior upper region.

CAP - Community acquired pneumonia.

CT - Computed tomography

CUS - Chest ultrasound.

CXR- Chest X-ray.

LLR - Lateral lower region.

LUR- Lateral upper region.

MNRH – Mulago National Referral Hospital.

PLR- Posterior lower region.

PUR- Posterior upper region.

WHO - World Health Organization.

OPERATIONAL DEFINITIONS. 1) Pediatric patient: any patient below 18 years. However, in this study, it’s the age from 2months to 12 years. 2) Community acquired pneumonia (CAP): an acute infectious pulmonary disease developing outside the hospital or in the first 48–72 hrs. of hospital stay. In this study, it is referred to as pneumonia. 3) Chest ultrasound: it is a noninvasive diagnostic exam that uses sound waves to acquire images of organs within the chest cavity. 4) Lung consolidation (Lobar pneumonia): this is when air is replaced by fluid/pus/mucus in the alveolar sacs. In this study, it was also called alveolar process (on CXR). 5) Interstitialprocess (Broncho/lobular pneumonia): this presence of peribronchial inflammation leading to patchy or reticular infiltrates. This was called B-lines on chest ultrasound. 6) Pneumonia on chest ultrasound is the presence of lung consolidation (hypoechogenic area of varying size and shape with poorly defined borders), with air or fluid bronchograms, a vascular tree-shaped pattern or presence of confluent B-lines, with/without pleural effusion. 7) Air bronchograms: these are linear or dot-like echogenic structures which are often seen within a consolidation. They represent air-filled bronchi and at times demonstrate intrinsic movement on real time ultrasound which is termed as dynamic air bronchogram and this rules out atelectasis. 8) Fluid bronchograms; these are linear or branching hypoechoic structures which are often seen within a consolidation, they represent bronchi filled with fluid/ pus/mucus and do not have visible walls. 9) A lines: these are horizontal, regularly spaced hyper-echogenic lines representing reverberations of the pleural line. These are motionless and are artifacts of repetition representing normal lung. 10) B lines; these are hyperechoic vertical narrow based lines arising from the pleural line to the deep edge of the ultrasound screen without fading, they obliterate the A-lines where they cross and correlate with thickened interlobular septae or ground-glass appearance when they are more than 3 or confluent, in the same view.

11) Shred sign (fractal sign): this is a static sonographic sign observed at the edges of an intralobular consolidation. 12) Tissue-like sign: this is a static sonographic sign observed in non-intralobular consolidation. 13) Pleural effusion: this is dependent collection within the pleural cavity. 14) Plain chest radiography: A projection radiograph of the chest used to diagnose chest conditions like pneumonia. 15) Standard reference (gold standard): this is an accepted test (combined tests) that is assumed to be able to determine the true disease state of a patient regardless of positive or negative test findings or sensitivities or specificities of other diagnostic tests used. 16) The final diagnosis based on overall clinical findings (diagnosis on discharge) and CXR findings was used as the standard reference, in this study.

ABSTRACT Background.

Pneumonia is a major health threat worldwide and a leading infectious cause of death in children under 5 years causing more deaths than AIDS, malaria, and measles. The current guidelines suggest that the diagnosis of pneumonia in pediatric patients can mainly be made clinically using WHO criteria for diagnosing pneumonia with chest X-ray (CXR) reserved for severe or complicated cases of pneumonia. However, it has relatively low accuracy, is associated with delays in diagnosis and exposes children to ionizing radiation. Some studies have shown that chest ultrasonography (CUS) is accurate at diagnosing community acquired pneumonia (CAP) in children. It is also cheaper, more readily available, safer than chest radiography and a good tool for easy bedside use and follow up of patients. However, very few studies have been done in children and none in our settings.

Objective of the study. The aim of this study was to establish the accuracy of chest ultrasound in diagnosing pneumonia in pediatric patients at Mulago National Referral Hospital.

Method and materials. This was a cross sectional study conducted at acute Care Unit(ACU), MNRH. Children aged 2 months to 12 years admitted with a clinical suspicion of pneumonia were prospectively enrolled for the study and underwent both chest X-ray and chest ultrasound (performed by the principal investigator blinded to CXR findings). Both investigations were performed within 24 hrs. of admission. Chest sonography was carried out using a high frequency (7-12 MHz) linear probe. Different CUS findings including; B-lines, lung consolidation, CXR findings; alveolar process, interstitial process and/or pleural effusion were assessed. The final discharge diagnosis was based on history and physical examination and CXR findings. The final diagnosis was based on overall clinical findings and CXR findings and was used as a reference standard compared with CUS to determine the accuracy of chest ultrasound in diagnosing pneumonia. The bivariate and multivariate analysis were done to determine the association between radiological and clinical findings.

Results. Of the 280 patients enrolled, 252 patients had complete data for analysis. The mean age was 21.4 months with more male participants 131(52%).CUS was positive in 164(64.7%) cases; 149(59.0%) consolidation and 62(24.6%) B-lines; 115(46.0%) had at least one patterns; 48(19.0%)

12 had both; 29(12.0%) pleural effusion. CXR was positive in 95(37.7%) cases; 82(32.5%) alveolar process; 18(7.1%); 90(35.7%) had at least one pattern; 5(2.0%) had both patterns; 9(3.6%) pleural effusion. 205(81.3%) had a final clinical diagnosis of pneumonia.

CUS sensitivity was 72%(95% CI [65-78]), specificity 67%(95% CI [52-81]), PPV 91%(95% CI [85-95]), NPV 35%(95% CI [25-46]), likelihood ratios, 2.2 (95% CI [1.44-3.37]) for positive and 0.42(95% CI [0.31-0.56], for negative, ROC 0.7(95% CI [0.62-0.77]). When compared to CXR findings alone, CUS sensitivity was 96% [95% CI, 90-99], specificity of 54% [95% CI, 46-62], PPV of 56% [95% CI, 48-64], NPV of 96% [95% CI, 89-99].

Hypoxia (SP 02 <93%) showed a strong association with both CUS and CXR, tachypnea with only CUS and inability to feed with only CXR, with p-values <0.05. The odds of diagnosing pneumonia by CUS in children with cough was 3.9; [95% CI; 1.19-9.62; p-0.022], and those with with hypoxia 1.9; [95% CI; 1.05-3.33; p-0.035]. The odds of diagnosing pneumonia by CXR in children with hypoxia 1.9 [95% CI; 1.07-3.26; p-0.028].

Conclusion

CUS was found to have a high sensitivity and therefore be used as a screening tool or add- on tool to CXR to diagnose pneumonia especially when CXR is negative. Hypoxia and tachypnea are the main predictors of pneumonia radiologically using CUS and CXR in children with clinical suspicion of pneumonia

1.0.INTRODUCTION Pneumonia is a major health threat worldwide and causes more deaths in children than AIDS, malaria, and measles. The World Health Organization predicts that pneumonia has a universal annual incidence of almost one million mortality cases in children, comprising approximately 15% of all deaths of individuals under 5 years of age.(1) Despite the introduction of the pneumococcal conjugate vaccine, community-acquired pneumonia is still considered as a common cause of morbidity among children aged ≤5 years in developed countries, where the incidence reaches from 10–40 cases per 1,000 subjects.(2) Currently, the diagnosis of pneumonia is made based on history and physical examination of presence of cough, and/or difficult breathing, fast breathing with/without chest in drawing, with/without fever, reserving the use of CXR only in severe or complicated cases(3), and is the current practice in Mulago National Referral Hospital. Despite these indications, CXR is also requested for mild and moderate cases because of the poor reliability of the history and physical examination and is based on the fact that the radiological investigations are entirely harmless.(4, 5) In addition to this, CXR examination is not always informative: in 2–5% of patients the plain chest X-ray picture is atypical and may lead to considerable overlap in differentiating pneumonia from effusions and masses since they may all appear as homogenous opacity on chest x-ray(6), can be negative in early stages of the disease, misses consolidations <1cm(7), has a low sensitivity and relatively low accuracy, interpretation is subjective and follow up using it leads to more radiation exposure. In such cases CT is required, although this leads to additional exposure of children to even larger amounts of radiation, yet children have a greater risk of getting radiation induced cancers(8, 9). Although, air was originally considered an enemy of ultrasound and the lung considered an organ not amenable to ultrasonography examination because of increased acoustic impedance of air, chest ultrasound can now be used as a good alternative to other radiological methods for pneumonia diagnosis. In 1986, a new method of evaluating CAP by the use of lung ultrasonography to demonstrate sonographic air bronchograms within lung consolidations by B. Weinberg(10)was discovered and numerous subsequent studies have since been done and have shown and proved that it is an accurate and reliable tool in the diagnosis of pneumonia(2, 11-13), with the possibility of a follow-up until the complete resolution of lung injury without exposure

14 to ionizing radiation. It can be repeated at any time and can be done in different positions which is very important in management of pediatric patients. However, most of the studies have been done in adults(14) with few studies done in children(15). Ultrasound characteristics considering the age pulmonary morphology have not been determined(16), the dynamics of ultrasound image of resolving inflammatory infiltrate has not been monitored, ultrasound examination algorithm has not been worked out and there is lack of both data and knowledge especially about the efficiency of CUS in diagnosis of CAP in our setting. The aim of this study was to bridge this knowledge gap and to offer evidence of CUS as a reliable, easily repeatable and noninvasive means in the diagnosis and follow up of CAP and to determine its accuracy in diagnosing CAP in children.

1.1 PROBLEM STATEMENT Pneumonia is the leading infectious cause of death in children in Africa and south Asia, according to the WHO and Uganda alone(1, 17) but confirmation of clinically suspected diagnosis is still problematic.

The current routine diagnosis of CAP is based on history, physical examination, blood tests and at times chest radiography which has a relatively low accuracy in children(6, 18) and even lower in the critically ill, exposes children to ionizing radiation, causes delays in processing and diagnostic dilemmas and a challenge to position children.

Chest CT which has a high accuracy to diagnose pneumonia is expensive(19), causes delays, exposes children to greater ionizing radiation and is not readily available and yet accurate and timely diagnosis is very important to improve the prognosis and avoid sequela of complications of pneumonia in children.

Lung ultrasound has been demonstrated to show a high reliability and accuracy in detection of CAP without exposure to ionizing radiation thus safe yet, there is no data available about the use of CUS in the diagnosis of CAP or been adopted in Ugandan settings.

1.2. JUSTIFICATION. CUS has been found to be reliable and accurate in diagnosing CAP in many studies using CT as a gold standard but these have mainly been done in adults with fewer studies in

15 children(Angelika, R. Copetti et al).(20, 21). Some studies have also been done using CXR as a gold standard (Anne-Sophie Claes et al)(22), .

It is also readily available, cheaper, faster and safer and provides a safer choice for follow-up until the complete resolution of lung injury and can be repeated at any time and done at the bedside.

This however, has not been demonstrated or adopted in the Ugandan clinical settings and there is no data available about the use of CUS in the diagnosis of CAP in Uganda.

Findings of this study were to help clinicians judge or opt for a safer radiological method/tool to use in diagnosing CAP among pediatric patients.

1.3. RESEARCH QUESTIONS. 1. What is the sensitivity and specificity of chest ultrasound in diagnosing pneumonia in pediatric patients at MNRH? 2. What is the correlation between the clinical and radiological findings in pediatric patients admitted with pneumonia at MNRH?

1.4. OBJECTIVES OF THE STUDY. 1.4.1. GENERAL OBJECTIVE. To determine the accuracy of chest ultrasound in diagnosing pediatric patients admitted with suspected clinical pneumonia at MNRH.

1.4.2. SPECIFIC OBJECTIVES. 1. To determine the sensitivity and specificity of chest ultrasound in diagnosing pediatric patients admitted with suspected clinical pneumonia at MNRH. 2. To determine the correlation between the clinical and radiological findings in pediatric patients admitted with suspected clinical pneumonia at MNRH.

Figure1: Conceptual framework.

2.0. LITERATURE REVIEW 2.1. Burden of pneumonia. Community-acquired pneumonia is an acute infectious pulmonary disease of different etiology (mainly bacterial) developing outside the hospital or in the first 48–72 hrs. of hospital stay, in a previously normal person. Pneumonia is the single largest infectious cause of death in children worldwide. Pneumonia killed 920 136 children under the age of 5 in 2015, accounting for 16% of all deaths of children under five years old and is most prevalent in South Asia and sub-Saharan Africa(1). The African Region has, in general, the highest burden of global child mortality. Although it comprises about 20% of the world’s population of children aged less than 5 years, it has about 45% of global under-5 deaths and 50% of worldwide deaths from pneumonia in this age group(23).

2.2. Aetiology of pneumonia. Bacteria, viruses, and fungi infections can cause pneumonia. Bacteria are the most common aetiological pathogens of CAP and have traditionally been divided into two groups designated ‘typical’ and ‘atypical(24). The most common typical pathogens include Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, group A streptococci and Moraxella catarrhalis. Atypical pathogens include bacteria that cannot be seen with a Gram's stain and cannot be cultured in regular sputum or blood culture medium. The most common atypical pathogens include Legionella spp., Mycoplasma pneumoniae and Chlamydophila pneumoniae. More than 100 micro-organisms have been implicated in community-acquired pneumonia, but most cases are caused by Streptococcus pneumoniae(25), being identified in 30–50% of pneumonia cases.(26-28). The second most common organism isolated in most studies is H. influenzae type b (Hib; 10–30% of cases), followed by S. aureus and K. pneumoniae.

Viruses are another well-recognized etiology of CAP(29),but because of the lack of rapid standardized tests for diagnosis, their real magnitude as aetiological agents of CAP has not been well defined. The role of influenza in the aetiology of CAP may also be increasing as a result of the emergence of the 2009 H1N1 influenza A virus pandemic. Respiratory syncytial virus is the leading viral cause of viral pneumonia being identified in 15–40% of pneumonia or bronchiolitis

18 cases admitted to hospital in children in developing countries, followed by influenza A and B, parainfluenza, human metapneumovirus and adenovirus(30-32). Primary respiratory infection by this virus increases the risk of secondary bacterial pneumonia and viral or bacterial coinfection is a common finding in young children with pneumonia in developing countries (approximately 20– 30% of episodes)(33). During the 2009 H1N1 influenza A virus pandemic, a combined viral and bacterial aetiology of CAP was detected in ∼30% of hospitalized patients with influenza CAP(34).

2.3. Pathogenesis of pneumonia Pneumonia is an acute infection of the lung parenchyma. Inhalation of infectious particles is probably the most important pathogenetic mechanism in the development of community-acquired pneumonia. Once bacteria reach the tracheobronchial tree, defects in local pulmonary defenses can make infection more likely such as; cough reflex can be impaired by poor nutrition, mucociliary transport depressed by dehydration and second hand smoking, prior infection with influenza virus, and chronic airway disease. Inflammatory cells drawn to infected areas of the pulmonary tree, release proteolytic enzymes, altering the bronchial epithelium and ciliary clearance mechanisms and stimulating the production of excess mucus(35).The signs and symptoms of acute pneumonia develop over hours to days

2.4. Risk factors of pneumonia. The reason that CAP is so common relates to the very high prevalence of specific risk factors for this infection in patients worldwide(36). Most affected children in epidemiologic studies of pneumonia are less than 5 years of age. Disease incidence among this age group is higher than at any other age, including adults over 65 years(37). Among older children, disease burden is substantially decreased and continues to decline through young adulthood. Other risk factors include poor nutritional status, low housing quality (e.g., overcrowding, environmental exposures), secondhand smoke exposure, comorbidities such as asthma, congenital heart disease and preceding upper respiratory tract infection(38, 39).

2.5. Clinical diagnosis of pneumonia.

The clinical diagnosis of CAP is made based on symptoms and signs which include fever, cough, tachypnea, grunting, retractions, hypoxia, abdominal pain, decreased breath sounds, and crackles or rales(40-43). The World Health Organization (WHO) criteria are widely used in the developing world and are intended to identify young children with high suspicion of clinical pneumonia in

19 resource-limited settings in which radiography and laboratory testing are not available(44). According to the revised criteria (WHO 2014), children less than 5 years of age with cough plus fast breathing and/or chest indrawing are said to have pneumonia which requires home therapy with oral amoxicillin and severe pneumonia if they have any general danger signs (not able to drink, persistent vomiting, convulsions, lethargic or unconscious, stridor in a calm child or severe malnutrition) which requires referral and injectable therapy(45).

2.6. Role of plain chest radiography in diagnosis of pneumonia. Chest radiography is routinely used to confirm the diagnosis of CAP in the developed world and is based on the presence of acute infiltrates on a CXR in a child with symptoms of an acute respiratory illness, as based on the WHO criteria for standardized interpretation of CXR for the diagnosis of CAP. However, CXR should not supplant clinical judgment, especially in the outpatient setting since the radiological definition of acute infiltrates is also caused by other acute lower respiratory illnesses (e.g., acute asthma exacerbated by viral upper respiratory illness).

So it should be reserved for those children in whom the diagnosis of CAP is in question, or for those more critically ill children who require hospitalization. In such cases, plain CXR may reveal complications requiring intervention or prolonged antibiotic therapy (e.g., effusion, abscess) or findings suggestive of certain etiologies (e.g., pneumatoceles in staphylococcal pneumonia)(46). WHO has developed criteria to standardize chest radiograph interpretation for the diagnosis of pneumonia in children(47). These standardized definitions have also been applied in the clinical settings. The finding of alveolar infiltrate had substantial interrater agreement (κ = 0.58–0.73 across studies) (48-50). In contrast, the finding of interstitial infiltrate had poor inter-observer reliability. These studies highlight the importance of clarifying the presence or absence of alveolar infiltrates in particular when interpreting chest radiographs.

However, several studies point to chest radiography’s low sensitivity for the diagnosis of pneumonia as proved in one of the studies done by (Jared T.et al) who reviewed a random sample of 105 patients with a diagnosis of CAP whose admission and subsequent chest radiographs were interpreted without knowledge of the clinical data. They noted twenty-one percent (22/105) of patients with a clinical diagnosis of CAP had negative chest radiographs at presentation and fifty- five percent of patients with initially negative chest radiographs who had follow-up studies developed an infiltrate within 48 hours(51). In addition to this, image quality of CXR is lower

20 when patients are in the decubitus position and when portable devices are used; it implies patient mobilization, radiation exposure, and high inter-observer variability(19).

2.7. Role of chest computed tomography in diagnosis of pneumonia. Chest CT has greater sensitivity than does chest radiography for the diagnosis of pneumonia,(52) but its use is reserved for more complex cases and when therapeutic failure occurs. Its main drawbacks are the need to transfer the patient to the radiology unit, increased exposure to radiation, and high cost.

2.8. Role of chest ultrasound in diagnosing pneumonia. Although it is not considered in current clinical practice guidelines, bedside chest ultrasonography for pathologic pleuropulmonary conditions has attracted interest in recent years, because it is portable and fast, does not use radiation, is easily reproducible, and allows real-time scanning(53). This was first described by (Weinberg et al) in 1986 who described a new method of evaluating CAP by the use of lung ultrasonography (LUS). They sonographically evaluated 30 pediatric patients in a comparative study with clinical and radiographic evidence of pneumonia. A new sonographic pattern of pulmonary consolidation associated with air-filled bronchi in children was presented. The consolidated area of the lung was described as usually hypoechoic, poorly defined, and wedge-shaped. The air-filled bronchi produce linear, high-amplitude branching echoes that converge toward the lung root. Posterior acoustic shadowing and reverberation artifacts are seen accompanying the proximal large bronchi. Twenty-eight patients were successfully examined with sonography. The cases were divided into three groups: (1) consolidated lung without pleural effusion (19 patients); (2) consolidated lung with small pleural effusion (five patients); and (3) consolidated lung with partial compression atelectasis due to large pleural effusion (four patients) (10). More studies using LUS in the diagnosis of CAP have done seen.

Although transthoracic USS has been used to assess a variety of pulmonary conditions in several studies, only recently has it been considered a valid method for the assessment of interstitial pulmonary conditions in patients with connective tissue diseases using HRCT (High resolution computed tomography) at the concurrent gold standard(54, 55). The ultrasound assessment of ILD (Interstitial lung disease) is determined by the presence and quantification of B-lines, which consist of tails generated by the reflection of the ultrasound beam from thickened sub-pleural interlobar septa at the lung surface interface. In study of USS in evaluation of interstitial pneumonia by V.

Lo Giudice et al, 55 patients referred to ultrasonography units for evaluation of probable viral or viral-like infections of the respiratory tract were found to have the signs of interstitial lung disease described by Lichtenstein revealed “comet-tail” artifacts in the anterolateral lung fields in 31 (56.36%) patients and mixed patterns consisting in increased density associated with ring-down artifacts in 24 (46.64%). Pleural involvement was also observed in 34 cases (61.82%)(56).

In a study to evaluate lung ultrasound in the diagnosis of CAP at emergency department, (Stefano et al,) studied 49 patients with suspected pneumonia and results were as follows, pneumonia was confirmed in 32 cases (65.3%). In this group they had 31 (96.9%) positive lung US and 24 (75%) positive CXR. In 8 (25%) cases, lung US was positive with a negative CXR. In this group, CT scan was used to confirm the CUS results. In one case, US was negative and CXR positive. Follow- up turned out to be always consistent with the diagnosis (12).

In a prospective multicenter study done by (Angelika et al) in 14 European centers to define the accuracy of lung ultrasound (LUS) in the diagnosis of community-acquired pneumonia (CAP), three hundred sixty-two (362) patients with suspected CAP were enrolled. Diagnosis was based on history, clinical examination, laboratory testing, and LUS were performed as well as the reference test, which was a radiograph in two planes or a low-dose CT scan in case of inconclusive or negative radiographic but positive LUS findings. The results were as follows; CAP was confirmed in 229 patients (63.3%). LUS revealed a sensitivity of 93.4% (95% CI, 89.2%-96.3%), specificity of 97.7% (95% CI, 93.4%-99.6%), and likelihood ratios (LRs) of 40.5 (95% CI, 13.2- 123.9) for positive and 0.07 (95% CI, 0.04-0.11) for negative results. A combination of auscultation and LUS increased the positive LR to 42.9 (95% CI, 10.8-170.0) and decreased the negative LR to 0.04 (95% CI, 0.02-0.09). They found 97.6% (205 of 211) of patients with CAP showed breath-dependent motion of infiltrates, 86.7% (183 of 211) an air bronchogram, 76.5% (156 of 204) blurred margins, and 54.4% (105 of 193) a basal pleural effusion. During follow-up, median C-reactive protein levels decreased from 137 mg/dL to 6.3 mg/dL at days 13 to 16 as did signs of CAP; median area of lesions decreased from 15.3 cm2 to 0.2 cm2 and pleural effusion from 50 mL to 0 mL (20).

In a study done to compare the diagnostic accuracy of ultrasound and chest X-ray (CXR) in children with suspected pneumonia by (R. Copetti et al), Seventy-nine (79) children aged from 6 months to 16 years with clinical signs suggestive of pneumonia underwent lung ultrasound and

22 chest radiography and results were as follows. Lung ultrasound was positive for the diagnosis of pneumonia in 60 patients, whereas CXR was positive in 53. In four patients with negative CXR and positive ultrasound findings, pneumonia was confirmed by chest computed tomography (CT) (performed for recurrent pneumonia in the same location). In the other three patients with negative CXR and positive ultrasound findings, the clinical course was consistent with pneumonia(21).

CUS is also a good, portable and noninvasive diagnostic tool for follow up in patients with CAP as proved by (Reissig A. et al) in 2007 in a prospective study to identify sonographic features associated with pneumonia at admission and during the course of the disease under treatment. Thirty patients (30) with CXR-proven pneumonia underwent transthoracic sonography (TTS) on day 0, between days 1 and 3, 4 and 7, 8 and 14, 15 and 21, and after day 21. TTS was assessed according to: number, location, shape, echogenicity, echotexture, echo structure, breath-depending movement, size of pneumonic lesions, broncho-aerogram, fluid bronchogram, superficial fluid alveologram, necrotic areas, vascularity and incidence of local and/or basal pleural effusion and findings were as follows, thirty-three pneumonic infiltrates were eligible for analysis in 30 patients. In 57% (17/30), the likely pathogenic microorganism was identified. Pneumonia was recognized as a hypoechoic area of varying size (mean size between 33.7 × 9.38 and 91.2 × 45.3 mm) and shape, with irregular and blurred margins along with a nonhomogeneous echotexture. The most characteristic feature was a positive bronchoaerogram (32/33). Sixty-one percent (20/33) revealed basal and 9% (3/33) local effusion. During follow-up, lesions decreased in size or disappeared (30/33) or decreased in number (4/9). The bronchoaerogram became less pronounced (13/32), basal pleural effusion either diminished (7/20) or dissipated (7/20), as did localized effusion (3/3). In 30 cases, the course of pneumonia was comparable using X-ray and TS(13).

Between 2013 and 2014 in a monocentric prospective study to evaluate the performance of ultrasound in detecting lung consolidation in children suspected of pneumonia, in comparison to the current gold standard, chest X-rays, (Anne-Sophie Claes et al,) studied 143 children (mean age 3 years; limits between 8 days and 14 years) referred for chest X-ray for suspected pneumonia. Each child was examined by chest ultrasound by an examiner blinded to the chest X-ray. And results were as follows; ultrasound detected at least one area of consolidation in 44 out of 45 patients with positive X-rays. Of the 59 areas of consolidation on X-ray, ultrasound identified 54. In the 8 patients with negative X-ray, ultrasound revealed 17 areas of consolidation. The mean size

23 of consolidations visible only on ultrasound was 9.4 mm; for consolidations visible on both techniques the mean size was 26 mm (p < 0.0001). The sensitivity and specificity of ultrasound were calculated at 98% and 92%. PPV and NPV were 85% and 99%, respectively(22).

In a study to determine the accuracy of point-of-care ultrasonography for the diagnosis of pneumonia in children and young adults by a group of clinicians, in 2013 by Shah VP et al, test performance characteristics for the ability of ultrasonography to diagnose pneumonia were determined using chest radiography as a reference standard. Subgroup analysis was performed in patients having lung consolidation exceeding 1 cm with sonographic air bronchograms detected on ultrasonography; specificity and positive likelihood ratio (LR) were calculated to account for lung consolidation of 1 cm or less with sonographic air bronchograms undetectable by chest radiography. Results were as follows, two hundred patients were studied (median age, 3 years; interquartile range, 1-8 years); 56.0% were male, and the prevalence of pneumonia by chest radiography was 18.0%. Ultrasonography had an overall sensitivity of 86% (95% CI, 71%-94%), specificity of 89% (95% CI, 83%-93%), positive LR of 7.8 (95% CI, 5.0-12.4), and negative LR of 0.2 (95% CI, 0.1-0.4) for diagnosing pneumonia by visualizing lung consolidation with sonographic air bronchograms. In subgroup analysis of 187 patients having lung consolidation exceeding 1 cm, ultrasonography had a sensitivity of 86% (95% CI, 71%-94%), specificity of 97% (95% CI, 93%-99%), positive LR of 28.2 (95% CI, 11.8-67.6) and negative LR of 0.1 (95% CI, 0.1-0.3) for diagnosing pneumonia. (57)

Studies have also been done to compare LUS with CT in assessing consolidations in patients with respiratory symptoms. In 2015, (Peiman Nazerian et al) carried out a prospective study on an emergency department population complaining of respiratory symptoms of unexplained origin. 285 patients who had a chest CT scan performed for clinical reasons were consecutively recruited. LUS was targeted to evaluate lung consolidations with the morphologic characteristics of pneumonia, and then compared to CT. Results were as follows CT was positive for at least one consolidation in 87 patients. LUS was feasible in all patients and in 81 showed at least one consolidation, with a good inter-observer agreement (k = 0.83), sensitivity 82.8% (95% CI 73.2%- 90%) and specificity 95.5% (95% CI 91.5%-97.9%). Sensitivity raised to 91.7% (95% CI 61.5%- 98.6%) and specificity to 97.4% (95% CI 86.5%-99.6%) in patients complaining of pleuritic chest pain. In a subgroup of 190 patients who underwent also chest radiography (CXR), the sensitivity

24 of LUS (81.4%, 95% CI 70.7%-89.7%) was significantly superior to CXR (64.3%, 95% CI 51.9%- 75.4%) (P<.05), whereas specificity remained similar (94.2%, 95% CI 88.4%-97.6% vs. 90%, 95% CI 83.2%-94.7%) (58)

3.0 METHODOLOGY 3.1. STUDY DESIGN Across-sectional study.

3.2. STUDY SETTING The study was conducted at Mulago National Referral and teaching Hospital (MNRH)for Makerere University in Kampala, the capital city of Uganda. The hospital serves Kampala city, the suburbs and referrals from all over the country as well as the Great lakes region and South Sudan. It is one of the centers that offers tertiary public medical services in the country. The radiology department in the in the hospital has an ultrasound unit with more than ten heavy-duty ultrasound machines and two portable machines. The department is currently fragmented into three stations due to the on-going renovations taking place in MNRH.

The sites include; Old Mulago (department of Radiology, Department of surgery and Department of pediatrics), Obstetrics and Gynecology Department (Kawempe), Department of of medicine (Kiruddu hospital).

The study was conducted at Acute Care Unit(ACU)in upper Mulago hospital. The ACU is an emergency department that receives all non –specific emergency pediatric cases including those suspected to have community acquired pneumonia from around Kampala and referrals from other centers. These after triage and stabilization are admitted on the general pediatric wards where they continue their treatment until discharge.

According to ACU data, an average of about 5 children are admitted with a clinical suspicion of pneumonia.

3.3. STUDY POPULATION. 3.3.1. Target population All children admitted through ACU during the study period.

3.3.2. Accessible population All children admitted through the ACU with a diagnosis of suspected of pneumonia/severe pneumonia, during the study period. 25

3.3.3. Eligible population Children aged from 2 months to 12 years admitted with suspected clinical diagnosis of pneumonia or severe pneumonia during the study period.

3.3.4. Inclusion criteria Children who were admitted with suspected pneumonia or severe pneumonia at ACU during the study period that fulfilled the eligibility criteria and had assent forms or whose parents gave an informed consent to participate in the study.

3.3.5. Exclusion criteria 1. Children who had chest tubes or catheters in situ.

2. Patients with non-diagnostic images.

3.4. Sample size determination. Sample size was calculated using the Kish Leslie (1965) formula for estimating population proportions.

푍2 푃(1−푃) Sample size n= 푑2

Z is the score for 95% confidence interval and it is 1.96.

P is the expected proportion of cases with positive CUS findings. This is 0.76(76%) as derived from a study done by (R. Copetti et al) to compare the diagnostic accuracy of ultrasound and chest X-ray (CXR) in children with suspected pneumonia. He reported an incidence of 60/79 (75.95%).

1-p is 0.24. d is the tolerable error. It is 5%(0.05) in this study.

Sample size n=1.96x1.96x0.76x0.24 /0.05x0.05= 280.

3.4.1. Sampling strategy All children suspected to have or admitted with pneumonia at ACU and meet the inclusion criteria were consecutively recruited in the study after their parents or guardian gave informed consent on their behalf and ascent consent for those above 7 years of age.

3.5. DATA COLLECTION TECHNIQUE. 3.5.1. Patient demographics. Every patient who was clinically diagnosed and admitted with pneumonia by a pediatrician at ACU was identified. Relevant medical history was taken to help with the inclusion criteria. The parent or guardian of the patient were told about the study in a language he or she understood. Consent to include the patient in the study was sought. Relevant socio-demographic information of the patient was obtained and recorded in data collection tool.

3.5.2. The chest X rays. In the current practice at MNRH, not all cases admitted with pneumonia get CXRs done, they are reserved for complicated cases (those that do not improve on medical treatment after 72 hrs.).

During this study, all the patients recruited in the study had CXRs done on them from department of radiology at Mulago National referral hospital using an analogue machine. The processing and development of the films was also manual and interpreted by an independent radiologist blinded to the CUS findings. The diagnosis of pneumonia was made in accordance with the (WHO) criteria for the standardized interpretation of pediatric chest radiographs (30) by an independent radiologist before or after the chest ultrasound.

3.5.3. Chest ultrasounds.

Chest ultrasound is not routinely ordered for by the pediatricians and was carried out independently.

The CUS was done using a SIUI (Shantou Institute of Ultrasound Instruments), Apogee 1000, model 2015, portable ultrasound machine with a high frequency and resolution, 7–12 MHz, linear array transducer.

The CUS was carried out by the principal investigator with guidance of a consultant radiologist who was supervising this study and were both blinded to the CXR findings.

The chest wall was divided into 6 anatomical regions. The anterior region which is between the sternal border and the anterior axillary line, the lateral region which lies between the anterior axillary line and the posterior axillary line then the posterior region which lies between the vertebral column and the posterior axillary line. Each region was further divided into upper and lower region. The anterior and lateral regions were examined in supine position with the shoulders

27 fully straightened on the sides, the posterior regions in a sitting up or lateral decubitus positions, or on their parents’ shoulders or while breastfeeding (for the younger babies), to minimize anxiety. The lung was visualized through intercostal window with the probe placed perpendicular, oblique and parallel to the ribs in all the 12 regions, as described by Copetti R et al(21).

Additional views through the subcostal and intercostal acoustic windows of the liver and the spleen were used to study the lung base and pleural effusion. Diaphragmatic motion was evaluated using intercostal and subxiphoid approaches. Images through the sternal were used to examine the superior anterior mediastinum. (59).Color Doppler ultrasound was used to evaluate the vascularity of lung lesions in some cases.

ILLUSTRATIONS.

A B Fig 2; A) A diagram showing the main anatomical divisions for probe positioning. B) Sonogram to illustrate the appearance of normal lung parenchyma on chest ultrasound (BAT SIGN). (Courtesy of the internet).

The diagnosis of CAP was made basing on the presence of lung consolidation (hypoechogenic area of varying size and shape with poorly defined borders), air or fluid bronchograms, superficial fluid alveologram or B-lines (>=3) and/or the presence of pleural effusion as defined by Reissig A et al (10, 13).

a) b) c) Fig 3; Chest sonograms showing abnormal lung parenchymal lung; a) Shred Sign, b) Fluid bronchograms, c) B-lines. (Courtesy of the internet).

Both investigations were done in the same setting (i.e. one after the other), CUS at ACU, using a portable machine and the patients were taken to the department of radiology for CXR. Chest x-ray reports were done daily and given to patients to help in their management by the clinicians.

The final diagnosis which was reached by the clinicians while patients were on wards, after laboratory tests in some cases or before discharge of the patients, is usually recorded in the discharge books on the wards. We did a follow up where we noted these for all patients in this

29 study. The final diagnosis was combined with the CXR findings and was used as the standard reference in this study.

All the findings and the images were recorded on both soft and hard forms in order to evaluate the accuracy of the method.

Figure 4: Flow chart

3.6. DATA COLLECTION TOOLS. A preformed and standardized mixed data collection tool was used to collect relevant information about the patient.

3.6.1 Independent variables These included record of; age, sex, next of kin, contact, and address of the patients. Immunization status. Medical history of chronic illnesses like asthma, congenital heart disease, tuberculosis, malnutrition were taken.

3.6.2. Dependent variables 3.6.2.1 Clinical symptoms and signs of pneumonia. These included;

i. Cough ii. Fever iii. Fast breathing iv. Chest wall indrawing v. Intercostal recessions vi. Cyanosis vii. Stridor viii. Altered mentation ix. Inability to feed.

3.6.2.2. Characteristics of the Chest ultrasounds. These included;

i. Location of lung injury. (Right, Left or both sides) ii. Air bronchograms (dynamic, adynamic) iii. Fluid bronchograms iv. Presence of shred sign or tissue-like sign. v. B lines. vi. Pleural effusion. vii. Vascular pattern. viii. Other findings like masses, abscesses etc.

3.6.2.3. Characteristics of the chest x-rays. These included;

i. Site of the opacities. ii. Alveolar lung disease (consolidation). iii. Interstitial lung disease (infiltrates). iv. Presence of pleural effusion. v. Other findings like lymph nodes, masses, cavities etc.

3.7. Quality control. A pilot study to test the data collection tool was done prior to starting the study.

Research assistants were trained by the principle investigator on the standard operational procedures for data collection and management prior to the start of the study. The Principal investigator supervised the research assistants.

The CUS was done by the principal investigator on all patients. The CXRs were done by a trained radiographer. The principal investigator was blinded to the radiographs until the CUSs were carried out. Data was checked for completeness before the participants were discharged. The principal investigator had weekly meetings with the research assistants.

Completed data collection tools were kept under lock and key with access limited to the investigating team. Proper archiving of the materials used to obtain and manage the data was done to ensure security and accountability of the data. Data backup was done regularly and electronically and data was kept securely with limited access to the data base via encrypted passwords.

A record of Informed consent from patients/parents and caregivers of children that met the inclusion criteria was obtained to confirm their decision to participate in the study. Consenting parents or caregivers signed or placed a thumb print on consent forms translated in either English or Luganda. Children who were 7 years and above signed assent forms (English or Luganda), as well. An interpreter was availed to aid in communication in case the parent / guardian understood another language.

The data was backed up in a portable hard disc to ensure data safety.

3.8. Data management. Data obtained from the completed data collection tool was cleaned and entered to Epidata software. Double entry was done to reduce errors. Data entry was done by principle investigator and trained research assistant. Regular data cleaning and proper archiving of all material used in data management was done to ensure transparency.

3.8.1. Data analysis. The collected data from the study collection tool was double entered using Epidata version 3.1 and with range and consistency checks done. Data was transferred to STATA version 14.0 (STATA Corporation, College Station, Texas, USA) for analysis and all p values of <0.05 were considered to indicate statistical significance.

The data set was checked for missing values. Analysis was restricted to all available non missing data for the classical analysis and restricted each model to the non-missing values for the variables in the model. Categorical data was displayed as frequencies, median, means, standard deviations, minimum or maximum and percentages.

In order to determine the accuracy of chest ultrasound, its findings were compared to final diagnosis and CXR findings in a two by two table. The accuracy was presented as a percentage with sensitivity, specificity, negative and positive predictive value, positive and negative likelihood ratios and a corresponding 95% confidence interval. Bivariate and multivariate analysis were done using logistic regression, variables were considered significant if their p-values were <0.05 and p-values <0.1 were borderline.

3.9. Ethical considerations. Informed consent were obtained from the participants.

Appropriate ethical approval was sought from the Departments of Radiology and Pediatrics, School of Medicine Research and Ethics Committee (SOMREC) of Makerere University and Mulago hospital ethical and research committees.

Principal investigator’s roles were.

• Selecting the cases.

• Carrying out the chest ultrasounds.

• Writing radiology (CUS) reports.

3.10. Limitations of the study.  Presence of subcutaneous emphysema and chest tubes might limit the use of CUS.  Presence of skeletal deformity might limit the exploration of the chest regions.  Poor visualization of deep lesions.

3.11. Dissemination of results.

Results were presented to the Makerere university annual Research Dissemination Conference. They were also submitted to the School of Medicine and Research and Ethics Committee (SOMREC), Makerere university library and both departments of radiology and pediatrics of college of health sciences. A compiled paper will be submitted for publishing to a peer reviewed journal.

4.0 RESULTS. A total of 280 patients aged from 2 months to 12 years, admitted with a clinical suspicion of pneumonia were consecutively enrolled between January 2018 and June 2018, from Acute Care Unit(ACU), in upper Mulago hospital. Of these, 252 had all data for analysis. 205(81.3%) patients had a final diagnosis of pneumonia and 47(18.7%) did not have pneumonia at discharge. (fig 2).

Figure 5: Study profile. 280 children met the inclusion criteria 28 Excluded due to; Incomplete investigations=11 Incomplete data=06 Poor image quality =11 252 children had all data for analysis

Final clinical diagnosis CUS CXR

Pos Pos Neg Pos Neg Neg n=163 n=98 n=157 n= 205 n=47 (64.7%) n=89 (37.7%) (62.3%) (81.3%) (18.7%) (35.3%)

Consolidation, n=149 Alveolar process, n=82 (32.5%) (59.1%) Interstitial process, n=18 (7.1%) B-Lines, n= 62 (24.6%) Had both, n=5 (2.0%) Had both, n=48 (19.1%) Had at least one, n=90 (35.7%)

Had at least one, n=115 Pleural effusion, n= 9(3.6%) (45.6%)

Pleural effusion, n=29(12%)

4.1. Participants’ characteristics 4.1.1. Demographic characteristics Table 1: Social demographics of the participants admitted with clinical suspicion of pneumonia at Acute Care Unit(ACU)

Characteristics Clinical suspicion Final diagnosis Final diagnosis p-value pneumonia pneumonia no pneumonia

(n=252) (n=205) (n=47) n (%) n (%) n (%) Age (months) [Mean], sd 21.4(24.2) 20.1(23.8) 26.7(25.6) 0.0944

<6 mo. 57 (22.6) 49(24.0) 8(17.0) 0.264 6 - 11 63 (25.0) 51 (24.8) 12(25.5) 12 - 23 53 (21.0) 46 (22.4) 7 (15.0) 24 - 59 54 (21.4) 42 (20.5) 12 (25.5) ≥60 25 (10.0) 17 (8.3) 8 (17.0)

Sex Female 121 (48.0) 103 (50.2) 18 (38.3) 0.139 Male 131 (52.0) 102 (49.8) 29 (61.7)

Immunization Yes 243(96.0) 197(96.0) 46(98.0) No 9 (4.0) 8 (4.0) 1 (2.0) 0.554

Chronic illnesses Absent 213(84.5) 171(83.4) 42(89.4) 0.262 CHD (known) 16(6.4) 12(5.9) 4(8.5) SCD 12(4.8) 12(5.9) 0 Others 11(4.3) 10(4.8) 1 (2.1)

CHD-Congenital heart disease SCD-Sickle cell disease The mean age of the patients enrolled was 21.4(Sd±24.2) months, with the biggest proportion, 63(25%) of the participants being between 6-11 months. However, there was essentially equal distribution among the age ranges below 59 months which was 90% of the all those screened.

The male participants 131(52%) were slightly higher than the females 121(48%). Most of the patients were immunized 243(96%) and did not have chronic illness 213(84.5%).

The baseline characteristics of the patients with final diagnosis of pneumonia did not differ much from those without pneumonia, with all the p-values >0.05.

4.1.2 Clinical features Table 2: Clinical features of patients admitted with suspected pneumonia at Acute Care Unit(ACU)

Variables Clinical Final diagnosis Final diagnosis P- suspicion pneumonia no pneumonia value pneumonia n (%) n (%) n (%) Cough 234 (93.0) 192 (94.0) 42 (89.4) 0.302 Temp(C), mean, sd 38.2(0.91) 38.2(0.94) 38.3(0.78) >37.5 202 (80.2) 163(79.5) 39 (83.0) 0.436 36.5-37.5 43(17.1) 35(17.1) 8(17.0) <36.5 7(2.8) 7(3.4) 0.0 Oxygen saturation, mean, sd 92.2(4.3) 91.9(4.4) 93.6(3.5) <93% 118(47.0) 104(51.0) 14(30.0) 0.009 ≥93% 134(53.2) 101(49.3) 33(70.2) Respiratory Rate, mean, sd 54 (7.8) 54 (8.1) 52 (6.5) 0.0960 Intercostal recessions 219 (87.0) 174 (85.0) 45 (96.0) 0.046 Chest wall indrawing 242 (96.0) 198 (97.0) 44 (94.0) 0.347 Cyanosis 7 (3.0) 7 (3.0) 0.0 0.199 Stridor 5 (2.0) 5 (2.4) 0.0 0.279 Altered mentation 6 (2.4) 6 (3.0) 0.0 0.235 Inability to feed 17(6.8) 17 (8.3) 0.0 0.041

Majority of the patients 234 (93%) had a history of cough, were febrile with 202(80.2%) having a temp>37.5 C, had intercostal recessions 219(87%) and chest wall indrawing 242(96%).

The clinical presentation of children with a final diagnosis pneumonia had a strong association with those without pneumonia at discharge in regards to oxygen saturation, intercostal recession and inability to feed with p-values <0.05.

Children without the pneumonia at discharge had a greater percentage of intercostal recessions 96% as compared to those who were found to have pneumonia at 85%. (Table 2).

4.1.3 Radiological diagnostic yield of pneumonia by both Chest ultrasound and Chest X-ray Table 3 a: Radiological features of patients admitted with clinical suspicion of pneumonia at Acute Care Unit

Radiological findings Clinical Final Final P- suspicion diagnosis diagnosis value pneumonia pneumonia no pneumonia

n (%) n (%) n (%) Abnormal features by 164(64.7%) 147 (72.0%) 16(34.0%) 0.000 CUS Consolidation 149(59.0%) 134(65.4%) 15(32.0%) 0.000 B-Lines 62(24.6%) 58(28.3%) 4(9.0%) 0.005 At least one 115(46.0%) 102(50.0%) 13(28.0%) 0.000 Both 48(19.0%) 45(22.0%) 3(6.4%) Pleural effusion 29(12.0%) 27(13.2%) 2(4.3%) 0.084

Abnormal features by 95 (37.7%) 94 (45.9%) 1 (2.1%) 0.000 CXR Alveolar Process 82 (32.5%) 81(39.5%) 1 (2.1%) 0.000 Interstitial process 18 (7.1%) 18 (8.8%) 0 0.035 At least one 90(35.7%) 89 (43.4%) 1(2.1%) Both 5(2.0%) 5(2.4%) 0 0.000 Pleural effusion 9(3.6%) 9(4.4%) 0 0.144 Cardiomegaly 46 (18.3%) 42 (20.5%) 4 (9.0%) 0.055

Of the 252, admitted with clinical suspicion of pneumonia, 164(64.7%) were found to have radiological features of pneumonia by CUS and 95(37.7%), of the same patients, were identified by CXR. This yield was found to be statistically significant with a p-value <0.0001.

Air space disease was the commonest radiological pattern with 149(59%) being demonstrated by CUS and 82(32.5%) demonstrated by CXR. CUS was able to demonstrate at least one of the patterns in 115(46%) of the patients and 90(35.7%) by CXR.

CUS demonstrated pleural effusion in 29(12.0%) of the patients compared to 9(3.6%) by CXR. 46(18.3%) were found to have cardiomegaly on CXR compared to 16(6.4%) who had known congenital heart disease.

Table 3 b: Distribution of pneumonia by hemithorax

Investigation Clinical Final diagnosis Final diagnosis P-value suspicion pneumonia no pneumonia pneumonia n (%) n (%) n (%) CUS (RT LUNG) Consolidation 123(81.5%) 110(81.0%) 13(87.0%) 0.584 B-lines 42(17.0%) 38(19.0%) 4(9.0%) 0.096 CXR (RT LUNG) Alveolar Process 77(89.5%) 76(89.4%) 1(100%) 0.731 Interstitial process 14(74.0%) 14(74.0%) 0 -

CUS (LT LUNG) Consolidation 86(57.0%) 80(59.0%) 6(40.0%) 0.162 B-lines 40(16.0%) 39(19.0%) 1(2.13%) 0.004 CXR (LT LUNG) Alveolar Process 24(28.0%) 24(28.2%) 0 0.531 Interstitial process 9(47.4%) 9(47.4%) 0 -

Most of the radiological pneumonia (both radiological patterns, but mainly airspace) was in the right hemithorax.

The demonstration of B-lines in the left hemithorax by CUS in patients with a final diagnosis of pneumonia showed a strong association with those without pneumonia at discharge with a p-value of 0.004.

Those with interstitial process on CXR in both hemithoraces were found to have pneumonia at discharge.

4.2 Diagnostic accuracy of chest ultrasonography in children with suspected pneumonia at Acute Care Unit 4.2.1 Accuracy of chest ultrasonography with final clinical diagnosis as a standard reference in the detection of pneumonia. Table 4 a: 2 by 2 table Final diagnosis Chest ultrasound Positive Negative Total Positive 148 (72%) 15 163 Negative 58 31 (67%) 89 Total 206 46 252

Of the 206 with final diagnosis of pneumonia, 148(72%) were diagnosed by CUS and of the 46 without pneumonia at discharge, CUS demonstrated pneumonia in 15(33%).

Table 4b: Diagnostic accuracy of chest ultrasonography with final clinical diagnosis as a standard reference in the detection of pneumonia Variable Sensitivity Specificity PPV NPV LR + LR- ROC % [95% Cl] %[95% Cl] %[95% Cl] %[95% Cl] %[95% Cl] %[95% Cl] %[95% CI] Overall 72[65-78] 67[52-81] 91[85-95] 35[25-46] 2.2[1.4-3.4] 0.4[0.3-0.6] 0.7[0.6-0.8] Age spec [months] <6 82[69-91] 57[18-90] 93[81-99] 31[9-61] 1.9[0.8-4.6] 0.3[0.1-0.8] 0.7[0.5-0.9] 6 - 11 69[54-81] 67[35-90] 90[76-97] 33[16-55] 2.1[0.9-4.7] 0.5[0.3-0.8] 0.7[0.5-0.8] 12 - 23 63[48-77] 71[29-96] 94[79-99] 23[8-45] 2.2[0.7-7.3] 0.5[0.3-1.0] 0.7[0.5-0.9] 24 - 59 69[53-82] 75[43-95] 91[75-98] 41[21-64] 2.8[1.0-7.5] 0.4[0.2-0.7] 0.7[0.6-0.9] ≥60 82[57-96] 63[25-92] 82[57-96] 63[25-92] 2.2[0.9-5.5] 0.3[0.1-1.0] 0.7[0.5-0.9]

Sex Spec Female 75[66-83] 53[28-77] 91[83-96] 26[13-43] 1.6[1.0-2.7] 0.5[0.3-0.8] 0.6[0.5-2.7] Male 69[59-78] 76[57-90] 91[86-96] 41[28-55] 2.8[1.5-5.5] 0.4[0.3-0.6] 0.7[0.6-0.8]

The above results yielded an accuracy of 71.0% with a sensitivity of 72% [95% CI, 65-78], specificity of 67% [52-81], PPV 91% [95% CI 85-95], NPV 35% [95% CI 25-46] with likelihood ratios (LRs) 2.1[95% CI, 0.8-3.3] for positives and 0.4[95% CI, 0.3-0.6] for negatives.

4.2.2 Comparison of chest ultrasonography with chest radiography. Table 5 a: 2 by 2 table

Chest x-ray Chest ultrasound Positive Negative Total Positive 91 (96%) 72 163 Negative 4 85 (54%) 89 Total 95 157 252 Of the 95 patients with pneumonia on CXR, CUS diagnosed 91(96.0%) and of the 157 without pneumonia on CXR, CUS identified 72(46%) with pneumonia.

Table 5 b: Diagnostic accuracy of chest ultrasonography using chest radiography as a standard reference in the detection pneumonia.

Variables Sensitivity Specificity PPV NPV LR+ LR- ROC % [95% CI] % [95% CI] %[95% CI] %[95% CI] %[95% CI] %[95% CI] %[95% CI] Overall 96[90-99] 54[46-62] 56[48-64] 96[89-99] 2.1[1.8-2.5] 0.1[0.0-0.2] 0.8[0.7-0.8] Age spec [months] <6 100[86-100] 39 [23-58] 56[39-70] 100 [75-100] 1.7[1.3-2.2] 0.00 0.7[0.6-0.8] 6-11 82[57-96] 46 [31-61] 36[21-53] 88 [68-97] 1.5[1.1-2.1] 0.4[0.1-1.1] 0.6[0.5-0.8] 12-23 100[83-100] 67[48-82] 68[48-82] 100[85-100] 3.0[1.9-4.9] 0.00 0.8[0.8-0.9] 24-59 96[78-100] 68[49-83] 69[50-84] 96[77-100] 3.0[1.8-5.0] 0.1[0.0-0.4] 0.8[0.7-0.9] >=60 100[72-100] 57[29-82] 65[38-86] 100[63-100] 2.3[1.3-4.3] 0.00 -

Sex spec Female 98[90-100] 50[38-62] 61[49-71] 97[85-100] 2.0[1.5-2.5] 0.0[0.0-0.3] 0.7[0.7-0.8] Male 93[81-99] 57[46-68] 51[39-62] 94[85-99] 2.2[1.7-2.8] 0.1 [0.0-0.4] 0.8[0.7-0.8]

The above yielded an accuracy of 69.8% with a sensitivity of 96% [95% CI, 90-99], specificity of 54% [95% CI, 46-62], PPV of 56%[95% CI, 48-64], NPV of 96%[95% CI, 89-99].

The sensitivity of CUS was good across all age groups except 6-11 months with 82%(95% CI, 57-96]. The sensitivity between both sexes was almost the same. The specificity better between 12-59 months with values above 68% [ 95% CI, 49-83].

The specificity of CUS among children between 12-23 months and 24-59 months was slightly higher with 67% [95% CI, 48-82] and 68% [95% CI, 49-83], respectively. And it was a better diagnostic tool among the same age groups with ROC of >0.82.

4.3 The correlation between the clinical and radiological findings of chest ultrasonography and chest radiography Table 6 a: The correlation between clinical findings and chest ultrasonography

Variables Clinical Final Final Bivariate Multivariate suspicion diagnosis diagnosis Unadjusted Adjusted pneumonia Pneumonia No OR (95% CI) p- OR (95% CI) p- pneumonia value value n n (%) n (%) Cough 234 155(66.0) 79(34.0) 2.5(0.93,6.46) 0.069 3.9(1.19,9.62) 0.022 Temp(oC) 202 129(64.0) 73(36.0) 0.6(0.11,2.96) 0.510 0.4(0.07,2.15) 0.271 SpO2 (<93) 118 87(74.0) 31(26.0) 2.1(1.26,3.65) 0.005 1.9(1.05,3.33) 0.035 Tachypnea 223 150(67.3)) 73(32.7) 2.5(1.16,5.54) 0.020 2.2(0.95,4.94) 0.067 Intercostal 219 143(65.3) 76(34.7) 1.2(0.58,2.59) 0.600 0.9(0.40,2.05) 0.811 recession Chest wall 242 159 (66.0) 83(34.0) 2.9(0.79,10.47) 0.110 2.7(0.69,10.50) 0.151 indrawing Cyanosis 7 6(86.0) 1(14.0) 3.4(0.40,28.39) 0.265 3.6(0.11,119.7) 0.472 Stridor 5 4(80.0) 1(20.0) 2.2(0.24,20.11) 0.480 0.6(0.02,17.71) 0.794 Altered mentation 6 5(83.0) 1(17.0) 2.8(0.32,24.22) 0.353 0.6(0.01,25.53) 0.757 Inability to feed 17 14 (82.0) 3 (18.0) 2.7(0.75,9.64) 0.128 1.6(0.31,8.39) 0.572 SP 02-Oxygen saturation

In bivariate analysis, only oxygen saturation <93% and tachypnea were strongly associated with CUS, with p-values <0.005. Cough was just borderline with a p-value <0.1.

Table 6 b: The correlation between clinical findings with chest radiography.

Variables Clinical Final diagnosis Final diagnosis Bivariate Multivariate suspicion Pneumonia No pneumonia Unadjusted Adjusted pneumonia OR(95% CI) p- OR (95% CI) p- value value n n (%) n (%) Cough 234 90(38.5) 144(61.5) 1.6 (0.56,4.71) 0.371 1.9(0.63,6.11) 0.246 Temp(C) 202n 75(37.1) 127(62.9) 1.2 (0.23,5.78) 0.868 0.8(0.14,4.96) 0.840 SpO2(<93) 118 55(46.6) 63(53.4) 2.1(1.22,3.44) 0.006 1.9(1.07,3.26) 0.028 Tachypnea 223 87(39.0) 136(64.0) 1.7(0.71,3.96) 0.236 1.4(0.58,3.55) 0.438 Intercostal 219 80(36.5) 139(63.5) 0.7 (0.33,1.45) 0.326 0.5(0.22,1.11) 0.083 recession Chest wall 242 93(38.4) 149(61.6) 2.5(0.52,12.01) 0.254 2.7(0.51,13.96) 0.247 indrawing Cyanosis 7 5 (71.4) 2(28.6) 4.3(0.82,22.65) 0.085 4.1(0.23,74.03) 0.335 Stridor 5 3(60.0) 2(40.0) 2.5(0.41,15.40) 0.315 0.6(0.03,11.57) 0.723 Altered 6 4(66.7) 2(33.3) 3.4(0.61,19.0) 0.162 0.6(0.02,16.10) 0.755 mentation Inability to feed 11 11(64.7) 6(35.3) 3.3(1.18,9.23) 0.023 2.0(0.53,7.82) 0.297

SP 02-Oxygen saturation

In bivariate analysis, low oxygen saturation and inability to feed showed a strong associated with CXR with p-values<0.05. Cyanosis was borderline with a p-value 0.085.

In multivariate analysis, the odds of diagnosing pneumonia with CXR were 1.9 times higher in patients with low oxygen saturation as compared to those with normal saturation. OR=1.9[95% CI; 1.07-3.26, p=0.028]

4.4 Images

A) B) Figure 6: (A case with normal radiological findings): A 4-year-old male admitted with pneumonia. A; Shows normal chest sonography findings (in the left anterior upper region). B; Normal AP chest x-ray findings.

A B)

Figure 7: (A case with abnormal CUS and normal CXR): A 5-year-old old admitted with pneumonia. A; A shred sign seen in the left posterior lower region (arrow). B; Normal lung fields with cardiomegaly.

A) B)

Figure 8: (A case of interstitial disease). A 3 months old baby admitted with severe pneumonia. A; Sonogram of the right anterior upper region showing a white lung. B; An AP chest radiograph showing bilateral ground glass opacification of both lung fields.

A) B)

C) D)

Figure 9: (A case of left lung consolidation and its different features ). An 18-months old admitted with pneumonia. A; A sonogram of the left anterior upper region showing a shred sign with dot-like echogenicities representing air bronchograms (arrows). B) Sonogram of the left lateral lower region showing, the diapragm(echogenic curve), spleen, hypoechoic collection, arrow head(pleural effusion) and part of the consolidated lung containing sonolucent areas, stars, which have no flow on colour doppler in (C), signifying necrosis(necrotising pnuemonia). D) AP radiograph showing a homogenous opacity in the left middle and lower lung zones forming a silhoutte sign with the left cardiac border, obscuring the left costophrenic angle.

A) B)

Figure 10: A 14-months old admitted with pneumonia. A) A left Para cardiac wedge shaped shred sign with air broncograms (echogenic foci). B) M-Mode showing dynamic motion of the air bronchograms. C) An AP radiograph showing a radio-opacity forming a positive silhouette sign with the left cardiac border.

5.0 DISCUSSION 5.1 Introduction Pneumonia is the leading infectious cause of illness and death in children younger than 5 years old in sub-Sahara Africa and Asia, claiming a life of a child every 20 seconds and accounts for 16 percent of all deaths of children under age 5 worldwide (1). Despite the clinical guidelines/protocols that have been put in place to timely diagnose and manage it, its diagnostic gold standard remains elusive among this age group.

This study was conducted to determine the accuracy of CUS in diagnosing pneumonia in pediatric patients and to correlate the clinical findings and radiological findings among these patients admitted with suspicion of pneumonia at Acute Care Unit (ACU) at Mulago National Referral Hospital.

5.2 Demographic profile Most patients 131(52%) were male and 121(48%) were female, showing a slight predilection for males. The highest proportion, 90%, were below 59 months of age (limits from 2 months to 12 years) with an overall mean age of 21.4(Sd±24.2) months, this was in keeping with the findings of WHO that stated pneumonia as a major infectious cause of health threat worldwide and a common cause of death in children especially those <5 yrs. of age(1).This could be because the immune systems of children <5 yrs. of age aren’t fully develop and while most healthy children can fight the infection with their natural defenses, children whose immune systems are compromised are at higher risk of developing pneumonia(1).

In this study, 243(96%) of the patients were fully immunized or had immunization up to date and 213(84.5%) of the patients did not have known chronic illness. This is important because a child's immune system maybe weakened by comorbidities like HIV, CHD, SCD, asthma and malnutrition or undernourishment, especially in infants who are not exclusively breast-fed predisposing her/him to infections.

However, these baseline demographic characteristics of the patients did not differ much between those with a final diagnosis of pneumonia and without pneumonia, with all the p-values >0.05.

5.3 Clinical Profile Majority of the patients 234 (93.0%) had a history of cough, were febrile with 202(80.2%) having a temp >37.5 C, had intercostal recessions 219(87.0%) and chest wall indrawing 242(96.0%).

Based on the 2014, WHO criteria for diagnosing pneumonia(45), these findings show that most of the enrolled patients had pneumonia.

The other predictors for severe pneumonia were not common with stridor 5(2.0%), cyanosis 7(3.0%), altered mentation 6(2.4%) and inability to feed 17(6.8%) were not common.

The clinical presentation of children with a final diagnosis pneumonia had a strong association with those without pneumonia in regards to oxygen saturation, intercostal recessions and inability to feed with p-values <0.05.

Patients without pneumonia at discharge had a greater percentage 45(96.0%) of intercostal recessions as compared to those with pneumonia at 198(85.0%), this could be because a big proportion 39(83.0%), without pneumonia were febrile at presentation compared to those with pneumonia and children who are febrile are also tachypneic which is usually associated with intercostal recessions.

5.4 Radiological findings Of the 252 patients admitted with clinical suspicion of pneumonia, CUS showed pulmonary abnormalities consistent with pneumonia in 163 (64.7%)and CXR was positive for pneumonia in 95 (37.7%), of the same patients. This yield was found to be statistically significant with a p-value <0.0001. These findings were similar to studies done by Copetti et al (21), Parlamento et al(12), Anne-Sophie et al(22).

Nonoccurrence of positive CUS findings with negative CXR findings could have occurred due to a number of factors; consolidations in the retro-cardiac or sub diaphragmatic which may be invisible on posterior-anterior chest radiographs but may be visible on lateral views which were not done in this study; the ability of CUS to detect very small (subcentimetre) consolidations which are not visible on CXR, as described by Shah et al(57); Jared et al(51) also found that CXR had a low sensitivity in early stages of infection and couldn’t demonstrate infiltrates until about 48 hrs. of the disease manifesting; misclassification of liver or spleen as lower lobe consolidation as stated by Anne-Sophie et al(22), in there study to determine the performance of ultrasound in detecting lung consolidations in pediatric patients with pneumonia; and high variability of interpretation of images.

The negative CUS findings with positive CXR findings could have occurred due to; consolidations not reaching the pleural surface often located in the perihilar as highlighted by Luri D et al(60)who found 7 patients with perihilar consolidations on CXR and zero cases on CUS, in his study to evaluate the sensitivity of CUS Vs CXR in detecting lung consolidations and pleural effusion; or could be due to consolidations located in regions hard to reach with CUS such as retro-scapular, supraclavicular or axillary regions.

Air space disease (consolidations/alveolar process) was a more common radiological pattern than interstitial disease on both CUS and CXR. CUS identified a higher percentage of patients with consolidations, 149 (59.0%)as compared to CXR which identified 82(32.5%). This was similar to findings by Copetti et al(21), Boursiani et al(61)and Caiulo el at(62) who reported a higher percentage of consolidations detected by CUS as compared to CXR in their studies which were comparing the two investigations.

CUS demonstrated 62(24.6%)of the interstitial disease cases and CXR 18 (7.1%) These findings differed from a study done by Boursiani et al(61) who found that CXR was better at detecting interstitial pattern than CUS which was better at detecting alveolar disease with CXR having a sensitivity of 95.5% and CUS at 92.4%.

This could be because B-lines are a nonspecific feature of interstitial lung disease that cannot reliably distinguish transudative and exudative causes of interstitial edema or infective from non- infective inflammatory process(63).

Most of the radiological pneumonia, in this study, was found in the right lung fields, air space disease more than interstitial process. This is attributed to the anatomical orientation of the right bronchus and consolidations in retro cardiac regions(63).

CUS demonstrated pleural effusion in 29(12.0%) of the patients compared to 9(3.6%) by CXR. This is because, CXR can only detect the presence of pleural effusion in patients in the orthostatic position only if the volume of the effusion is at least 200 mL and the sensitivity of this method further decreases in the supine position, whereas ultrasound can detect effusions as small as 20 mL (64).

These findings are comparable to findings by Luri et al(60) who demonstrated pleural effusion in 15 patients using CUS compared to CXR that showed only 8 patients, in a study to evaluate

50 usefulness of lung ultrasound in children suspected to have pneumonia and Urbankowski et al(65)found that pleural effusion was demonstrated by CUS in 54.3% of patients, while radiographic signs of pleural effusion were found in only 12.1% of patients.

46(18.3%) of the patients in this study were found to have radiographic features of cardiomegaly with a p-value of 0.055. Cardiac echo was recommended in these patients, in this study, though follow up was not done. This is because children with heart disease are at a greater risk of developing respiratory tract infections. These findings were in close keeping with a study done by W.E. Sadol et al. who found that the prevalence of CHD in children with pneumonia was high and suggested evaluation of children with pneumonia for children with CHD(66).

5.5 Accuracy of chest ultrasonography in diagnosing pneumonia. Using final diagnosis as standard reference;

Of the 206 with final diagnosis of pneumonia, 148(72%) were diagnosed by CUS and of the 46 without pneumonia, CUS demonstrated pneumonia in 15(33%). This yielded a sensitivity of 72% [95% CI, 65-78], specificity of 67% [52-81], PPV 91% [95% CI 85-95], NPV 35% [95% CI 25- 46] with likelihood ratios (LRs) 2.1[95% CI, 0.8-3.3] for positives and 0.4[95% CI, 0.3-0.6] for negatives.

The result in this study were however lower than those in the study done by Lorio G et al(67) who found that the sensitivity of lung ultrasound was 96.5% (95% CI [82.2%–99.9%]), specificity of 95.6% (95% CI [78.0%–99.9%]), positive likelihood ratio of 22.2 (95% CI [3.2–151.2]), and negative likelihood ratio of 0.04(95% CI[0.01–0.25])for diagnosing pneumonia, when he compared discharge diagnosis (made on the basis of history and physical examination, laboratory and instrumental tests, including Chest radiography)

It high lights a percentage 15(33.0%) of false positives read by CUS which further emphasizes its high sensitivity that enables it to identity other pulmonary abnormalities that are not necessarily pneumonia such as non-infective processes(63).

In a meta-analysis of CUS for diagnosis of CAP in children by Maria A et al(15), they found in a subgroup analysis (5 studies), that when CXR was used as a standard reference alone, the sensitivity of CUS was similar to that when both clinical criteria and CXR were used to diagnose CAP, but specificity decreased to 84%, from 93%.

When CUS was compared with chest radiography;

Of the 95 patients diagnosed with pneumonia by CXR, CUS diagnosed 91(96.0%) and of the 157 without pneumonia on CXR, CUS identified 72(46%) with pneumonia. This yielded a sensitivity of 96%(95% CI, 90-99), specificity of 54% (95% CI, 46-62), PPV of 56% (95% CI, 48-64), NPV of 96% (95% CI, 89-99), as seen in (table 5 a & b).

The high sensitivity by CUS in these results, on top of the reasons stated above, could be attributed to the small body habitus and chests in children which enable easy identification of consolidations(which are usually sub-pleuric) though these may also be due to non-infiltrative processes like atelectasis(obstructive and compressive) which present like consolidations (particularly in bronchiolitis or asthma)(15). Similarly, smaller consolidations (sub centimeter) which are detected by CUS but missed by CXR lead to a false-positive reading by CUS. This was highlighted by Shah et al(57), who found a sensitivity of 86% and a specificity of 89% when using CXR as a reference standard, with a high level of discordance between CXR and CUS when consolidation size was ≤1 cm. When including only consolidation size >1 cm, specificity was 97%. This would likely explain the lower specificity when CXR alone is used as the reference standard, though the specificity in this study was much lower. Similar findings were also presented by Esposito et al(2) who showed a very high diagnostic performance of CUS (as compared to CXR) in children with CAP suspicion, with sensitivity, specificity, PPV and NPV of 97.9%, 94.5%, 94.0% and 98.1%, respectively and in a meta-analysis of 8 studies using CXR as a standard reference by Maria A et al(15), showed high sensitivity and specificity of CUS in childhood CAP (95% and 93% respectively).

Ultrasonography of the left lower chest can be challenging, with the combination of spleen and air in the stomach being mistaken for lung consolidation and sonographic air bronchograms. In addition, the thymus has a sonographic appearance with hyper echogenic foci that can be mistaken for lung consolidation(57). However, no errors were made in mistaking the thymus for lung consolidation. Recognition of these potential pitfalls was included in the pilot study.

5.6 Correlation between clinical findings and radiological findings

In bivariate analysis with CUS, only oxygen saturation <93% and tachypnea were strongly associated with CUS, with p-values <0.05. Cough was just borderline with a p-value <0.1.

In multivariate analysis, the odds of having a diagnosis of pneumonia on chest ultrasound were 3.9 times higher among children with cough as compared to those without cough and this was statistically significant with a p-value of 0.022, OR=3.9; [95% CI; 1.19-9.62; P=0.022] while the odds of having a diagnosis of pneumonia on chest ultrasound were 1.9 times higher in children with low oxygen saturation as compared to those with normal oxygen saturation. OR=1.9; [95% CI; 1.05-3.33; p=0.035]. This was comparable to findings by Chao, J.H. et al(68) who found a positive correlation between positive CUS and Sat O2 ≤92%, with p-value of 0.001, in his study to determine accuracy of CUS in diagnosing pneumonia in children with bronchiolitis.

In bivariate analysis with CXR, low oxygen saturation<93% and inability to feed showed a strong associated with CXR with p-values<0.05. Cyanosis was borderline with a p-value 0.085. In multivariate analysis, the odds of diagnosing pneumonia with CXR were 1.9 times higher in patients with low oxygen saturation as compared to those with normal saturation, OR=1.9[95% CI; 1.07-3.26, p=0.028].

These findings were in close keeping with the findings made by Mark I et al(69), in a study to associate clinical and history findings with radiographic findings who found that history of chest pain, focal rales, duration of fever, and oximetry levels at triage were significant predictors of pneumonia while the presence of tachypnea, retractions, and grunting were not associated with pneumonia. Hypoxia (oxygen saturation ≤92%) was the strongest predictor of pneumonia (odds ratio: 3.6 [95% CI, 2.0–6.8), though this was slightly lower in our study. Lynch et al,(43)also found that history of fever, tachypnea, retractions, grunting, rales and decreased breath sounds were associated with radiographic pneumonia.

5.5 Limitations. The final clinical diagnosis which was used as a reference standard in this study is not the gold standard for diagnosing pneumonia.

Computed tomography, CT, of the chest is the best and accurate gold standard for diagnosing pneumonia, particularly for very small lung consolidations, pleural effusions and centrally located. This was however not practical for us to obtain chest CT in children enrolled in this study and is not our standard of care here in MNRH.

CXRs were assessed by only one radiologist who was blinded to the clinical findings. Therefore, the results of imaging studies were observer-dependent and inter-observer bias could not be evaluated.

Most of the clinical findings were copied from the medical records/files which were made by the medical personnel (medical students and doctors) at ACU. This could explain the possible false positives of those diagnosed with pneumonia clinically. There was no standardization of the gargets used to take the vitals of the patients enrolled in the study.

6.0. Conclusion and recommendation. 6.1 Conclusions.  Chest ultrasound shows high sensitivity in detection of pneumonia.  The commonest radiological signs of children admitted with pneumonia are consolidations, appearing more on the right than the left hemithoraces.  Hypoxia and tachypnea are the main predictors of pneumonia radiologically using CUS and CXR in children with clinical suspicion of pneumonia.

6.2 Recommendations.  Chest ultrasound should be included in screening of children with clinical suspicion of pneumonia.  Chest ultrasound should be considered as an alternative or add-on diagnostic tool to the conventional chest radiography to diagnose pneumonia in children especially for follow up purposes.  Clinicians / pediatricians can be trained to diagnose childhood pneumonia in different clinical settings (particularly emergency departments), in rural or resource limited settings where radiography is not readily available to enhance patient management.  A case control study on a bigger scale can be done in the future to compliment these findings.

REFERENCES. 1. World Health Organization . Pneumonia. Geneva SAM, 2015]. (Fact Sheet No 331). Available from: www.who.int/mediacentre/factsheets/fs331/en/. 2. Esposito S, Papa SS, Borzani I, Pinzani R, Giannitto C, Consonni D, et al. Performance of lung ultrasonography in children with community-acquired pneumonia. Italian Journal of Pediatrics. 2014;40:37-. 3. Harris M, Clark J, Coote N, Fletcher P, Harnden A, McKean M, et al. British Thoracic Society guidelines for the management of community acquired pneumonia in children: update 2011. Thorax. 2011;66(Suppl 2):ii1. 4. Shah S, Bachur R, Kim D, Neuman MI. Lack of Predictive Value of Tachypnea in the Diagnosis of Pneumonia in Children. The Pediatric Infectious Disease Journal. 2010;29(5):406-9. 5. Little MP. Risks associated with ionizing radiationEnvironmental pollution and health. British Medical Bulletin. 2003;68(1):259-75. 6. 2 Samsygina GA Infektsii respiratornogo trakta u detey rannego vozrasta [Respiratory infections in young children] Moscow: Pul’s М; 2013; 260 p 7. Raoof S, Feigin D, Sung A, Raoof S, Irugulpati L, Rosenow EC, III. Interpretation of Plain Chest Roentgenogram. CHEST.141(2):545-58. 8. Allisy-Roberts P, Williams JR, Farr RF. Farr's physics for medical imaging. Edinburgh: Saunders ; Elsevier; 2008. 9. Sorantin E, Weissensteiner S, Hasenburger G, Riccabona M. CT in children–dose protection and general considerations when planning a CT in a child. Eur J Radiol. 2013;82. 10. Weinberg B, Diakoumakis EE, Kass EG, Seife B, Zvi ZB. The air bronchogram: sonographic demonstration. American Journal of Roentgenology. 1986;147(3):593-5. 11. Cortellaro F, Colombo S, Coen D, Duca PG. Lung ultrasound is an accurate diagnostic tool for the diagnosis of pneumonia in the emergency department. Emerg Med J. 2012;29. 12. Parlamento S, Copetti R, Di Bartolomeo S. Evaluation of lung ultrasound for the diagnosis of pneumonia in the ED. The American Journal of Emergency Medicine.27(4):379-84. 13. Reissig A, Kroegel C. Sonographic Diagnosis and Follow-Up of Pneumonia: A Prospective Study. Respiration. 2007;74(5):537-47. 14. Chavez MA, Shams N, Ellington LE, Naithani N, Gilman RH, Steinhoff MC, et al. Lung ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta-analysis. Respiratory Research. 2014;15(1):50. 15. Pereda MA, Chavez MA, Hooper-Miele CC, Gilman RH, Steinhoff MC, Ellington LE, et al. Lung Ultrasound for the Diagnosis of Pneumonia in Children: A Meta-analysis. Pediatrics. 2015;135(4):714. 16. Dianova TI, Safonov DV. Ultrasound Monitoring and Age Sonographic Characteristics of Community-Acquired Pneumonia in Children2015. 113-9 p. 17. Walker CLF, Rudan I, Liu L, Nair H, Theodoratou E, Bhutta ZA, et al. Global burden of childhood pneumonia and diarrhoea. The Lancet.381(9875):1405-16. 18. Prayle A, Atkinson M, Smyth A. Pneumonia in the developed world. Paediatric Respiratory Reviews.12(1):60-9. 19. Albaum MN, Hill LC, Murphy M, Li YH, Fuhrman CR, Britton CA, et al. Interobserver reliability of the chest radiograph in community-acquired pneumonia. PORT Investigators. Chest. 1996;110(2):343-50. 20. Reissig A, Copetti R, Mathis G, Mempel C, Schuler A, Zechner P, et al. Lung Ultrasound in the Diagnosis and Follow-up of Community-Acquired Pneumonia. CHEST.142(4):965-72.

21. Copetti R, Cattarossi L. Ultrasound diagnosis of pneumonia in children. La radiologia medica. 2008;113(2):190-8. 22. Claes A-S, Clapuyt P, Menten R, Michoux N, Dumitriu D. Performance of chest ultrasound in pediatric pneumonia. European Journal of Radiology. 2017;88(Supplement C):82-7. 23. 2008]. WpppdUNPDAfheuouaoA. World population prospects. population database United Nations Population Division. 2008(population database ). 24. Bartlett JG. Is activity against "atypical" pathogens necessary in the treatment protocols for community-acquired pneumonia? Issues with combination therapy. Clin Infect Dis. 2008;47 Suppl 3:S232-6. 25. Loeb M. Community-acquired pneumonia. BMJ Clinical Evidence. 2010;2010:1503. 26. Shann F. Etiology of severe pneumonia in children in developing countries. Pediatric infectious disease. 1986;5(2):247-52. 27. Adegbola RA, Falade AG, Sam BE, Aidoo M, Baldeh I, Hazlett D, et al. The etiology of pneumonia in malnourished and well-nourished Gambian children. Pediatr Infect Dis J. 1994;13(11):975-82. 28. Falade AG, Mulholland EK, Adegbola RA, Greenwood BM. Bacterial isolates from blood and lung aspirate cultures in Gambian children with lobar pneumonia. Annals of tropical paediatrics. 1997;17(4):315-9. 29. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007;44 Suppl 2:S27- 72. 30. Weber MW, Mulholland EK, Greenwood BM. Respiratory syncytial virus infection in tropical and developing countries. Tropical medicine & international health : TM & IH. 1998;3(4):268-80. 31. Simoes EA. Respiratory syncytial virus infection. Lancet (London, England). 1999;354(9181):847-52. 32. Stensballe LG, Devasundaram JK, Simoes EA. Respiratory syncytial virus epidemics: the ups and downs of a seasonal virus. Pediatr Infect Dis J. 2003;22(2 Suppl):S21-32. 33. Forgie IM, O'Neill KP, Lloyd-Evans N, Leinonen M, Campbell H, Whittle HC, et al. Etiology of acute lower respiratory tract infections in Gambian children: I. Acute lower respiratory tract infections in infants presenting at the hospital. Pediatr Infect Dis J. 1991;10(1):33-41. 34. Bacterial coinfections in lung tissue specimens from fatal cases of 2009 pandemic influenza A (H1N1) - United States, May-August 2009. MMWR Morbidity and mortality weekly report. 2009;58(38):1071-4. 35. van der Poll T, Opal SM. Pathogenesis, treatment, and prevention of pneumococcal pneumonia. The Lancet. 2009;374(9700):1543-56. 36. Welte T, Torres A, Nathwani D. Clinical and economic burden of community-acquired pneumonia among adults in Europe. Thorax. 2012;67(1):71-9. 37. Jokinen C, Heiskanen L, Juvonen H, Kallinen S, Karkola K, Korppi M, et al. Incidence of community-acquired pneumonia in the population of four municipalities in eastern Finland. American journal of epidemiology. 1993;137(9):977-88. 38. Grant CC, Emery D, Milne T, Coster G, Forrest CB, Wall CR, et al. Risk factors for community-acquired pneumonia in pre-school-aged children. Journal of paediatrics and child health. 2012;48(5):402-12.

39. Heiskanen-Kosma T, Korppi M, Jokinen C, Heinonen K. Risk factors for community- acquired pneumonia in children: a population-based case-control study. Scandinavian journal of infectious diseases. 1997;29(3):281-5. 40. Redd SC, Vreuls R, Metsing M, Mohobane PH, Patrick E, Moteetee M. Clinical signs of pneumonia in children attending a hospital outpatient department in Lesotho. Bulletin of the World Health Organization. 1994;72(1):113-8. 41. Mathews B, Shah S, Cleveland RH, Lee EY, Bachur RG, Neuman MI. Clinical predictors of pneumonia among children with wheezing. Pediatrics. 2009;124(1):e29-36. 42. Murphy CG, van de Pol AC, Harper MB, Bachur RG. Clinical predictors of occult pneumonia in the febrile child. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine. 2007;14(3):243-9. 43. Lynch T, Platt R, Gouin S, Larson C, Patenaude Y. Can we predict which children with clinically suspected pneumonia will have the presence of focal infiltrates on chest radiographs? Pediatrics. 2004;113(3 Pt 1):e186-9. 44. Organization WH, Development DoCaAHa, PBoHCfCGftMoCIwLR, , Geneva, Organization WH. Guidelines for the Management of Common Illnesses with Limited Resources , 2005Department of Child and Adolescent Health and Development, Pocket Book of Hospital Care for Children. 2005. 45. WHO. Revised WHO Classification and Treatment of Pneumonia in Children at Health Facilities: Evidence Summaries. GENEVA, WHO. 2014. 46. Williams DJ, Shah SS. Community-Acquired Pneumonia in the Conjugate Vaccine Era. Journal of the Pediatric Infectious Diseases Society. 2012;1(4):314-28. 47. Cherian T, Mulholland EK, Carlin JB, Ostensen H, Amin R, de Campo M, et al. Standardized interpretation of paediatric chest radiographs for the diagnosis of pneumonia in epidemiological studies. Bulletin of the World Health Organization. 2005;83(5):353-9. 48. Neuman MI, Lee EY, Bixby S, Diperna S, Hellinger J, Markowitz R, et al. Variability in the interpretation of chest radiographs for the diagnosis of pneumonia in children. Journal of hospital medicine. 2012;7(4):294-8. 49. Ben Shimol S, Dagan R, Givon-Lavi N, Tal A, Aviram M, Bar-Ziv J, et al. Evaluation of the World Health Organization criteria for chest radiographs for pneumonia diagnosis in children. Eur J Pediatr. 2012;171(2):369-74. 50. Xavier-Souza G, Vilas-Boas AL, Fontoura M-SH, Araújo-Neto CA, Andrade SCS, Cardoso M-RA, et al. The inter-observer variation of chest radiograph reading in acute lower respiratory tract infection among children. Pediatric Pulmonology. 2013;48(5):464-9. 51. Hagaman JT, Panos RJ, Rouan GW, Shipley RT. Admission Chest Radiograph Lacks Sensitivity in the Diagnosis of Community-Acquired Pneumonia. The American Journal of the Medical Sciences.337(4):236-40. 52. Brenner DJ, Hall EJ. Computed tomography--an increasing source of radiation exposure. N Engl J Med. 2007;357(22):2277-84. 53. Sartori S, Tombesi P. Emerging roles for transthoracic ultrasonography in pulmonary diseases. World journal of radiology. 2010;2(6):203-14. 54. Soldati G, Copetti R, Sher S. Sonographic Interstitial Syndrome. Journal of Ultrasound in Medicine. 2009;28(2):163-74. 55. Hasan A, Makhlouf H. B-lines: Transthoracic chest ultrasound signs useful in assessment of interstitial lung diseases. Annals of Thoracic Medicine. 2014;9(2):99-103.

56. Lo Giudice V, Bruni A, Corcioni E, Corcioni B. Ultrasound in the evaluation of interstitial pneumonia. Journal of Ultrasound. 2008;11(1):30-8. 57. Shah VP, Tunik MG, Tsung JW. Prospective evaluation of point-of-care ultrasonography for the diagnosis of pneumonia in children and young adults. JAMA Pediatr. 2013;167(2):119-25. 58. Nazerian P, Volpicelli G, Vanni S, Gigli C, Betti L, Bartolucci M, et al. Accuracy of lung ultrasound for the diagnosis of consolidations when compared to chest computed tomography. The American Journal of Emergency Medicine.33(5):620-5. 59. S. FYGETAKI1 IT, S. Stefanaki2, S. Antonopoulos2, E. Sfakianaki2; 1Heraklion, GR/GR, 2Heraklion/GR Pediatric chest: From x-ray to ultrasound. A pictorial review of 173 patients". (ECR 2014 / C-0451). 60. Iuri D, De Candia A, Bazzocchi M. Evaluation of the lung in children with suspected pneumonia: usefulness of ultrasonography. La radiologia medica. 2009;114(2):321-30. 61. Boursiani C, Tsolia M, Koumanidou C, Malagari A, Vakaki M, Karapostolakis G, et al. Lung Ultrasound as First-Line Examination for the Diagnosis of Community-Acquired Pneumonia in Children. Pediatric Emergency Care. 2017;33(1):62-6. 62. Caiulo VA, Gargani L, Caiulo S, Fisicaro A, Moramarco F, Latini G, et al. Lung ultrasound characteristics of community-acquired pneumonia in hospitalized children. Pediatr Pulmonol. 2013;48(3):280-7. 63. Stadler JAM, Andronikou S, Zar HJ. Lung ultrasound for the diagnosis of community- acquired pneumonia in children. Pediatric Radiology. 2017;47(11):1412-9. 64. Prina E, Torres A, Carvalho CR. Lung ultrasound in the evaluation of pleural effusion. Jornal brasileiro de pneumologia : publicacao oficial da Sociedade Brasileira de Pneumologia e Tisilogia. 2014;40(1):1-5. 65. Urbankowska E, Krenke K, Drobczyński Ł, Korczyński P, Urbankowski T, Krawiec M, et al. Lung ultrasound in the diagnosis and monitoring of community acquired pneumonia in children. Respiratory Medicine. 2015;109(9):1207-12. 66. Sadoh WE, Osarogiagbon WO. Underlying congenital heart disease in Nigerian children with pneumonia. Afr Health Sci. 2013;13(3):607-12. 67. Iorio G, Capasso M, De Luca G, Prisco S, Mancusi C, Laganà B, et al. Lung ultrasound in the diagnosis of pneumonia in children: proposal for a new diagnostic algorithm. PeerJ. 2015;3:e1374. 68. Chao JH, Lin RC, Marneni S, Pandya S, Alhajri S, Sinert R. Predictors of airspace disease on chest X-ray in emergency department patients with clinical bronchiolitis: a systematic review and meta-analysis. AcadEmerg Med. 2016;23. 69. Neuman MI, Monuteaux MC, Scully KJ, Bachur RG. Prediction of Pneumonia in a Pediatric Emergency Department. Pediatrics. 2011;128(2):246.

APPENDIX 1 DATA COLLECTION TOOL. A STUDY TO DETERMINE THE ACCURACY OF CHEST ULTRASOUND IN PEDIATRIC PATIENTS ADMITTED WITH PNEUMONIA AT MULAGO NATIONAL REFERRAL HOSPITAL, KAMPALA UGANDA.

CASE IDENTIFICATION CODE______

A. PATIENT DEMOGRAPHICS

1)Age |__||__|/|__||__| mm/yrs.

2)Sex of the patient F/M

3)Home address(District) ______

4)Patient’s contact (NOK of patient)______

5)Occupation of the parent or care giver______

B. PAST MEDICAL HISTORY. a) Immunization □Yes □ No b) Chronic illness______

C. CLINICAL FINDINGS 1) Fever(temp) |__||__|◦C 2) Cough: □ Present □ Absent 3) SP 02: ______4) Respiratory rate: |__||__||__| per minute 5) Intercostal recessions: □ Present □ Absent 6) Chest-wall indrawing: □ Present □ Absent 7) Cyanosis: □Present □ Absent 8) Stridor: □Present □Absent 9) Altered mentation: □Yes □ No 10) Inability to feed. □Yes □ No

D. CHEST RADIOGRAPH CHARACTERISTICS. Zone of lung Alveolar Interstitial Pleural Others injury. process process effusion RIGHT LUNG Upper zone Middle zone Lower zone LEFT LUNG Upper zone Middle zone Lower zone BOTH LUNGS

E. CHEST ULTRASOUND CHARACTERISTICS. Location of lung Lung consolidation Air fluid Pleural effusion Vascular B- lines Others injury (Shred/ bronchogram bronchogram (characteristics) pattern tissue-like sign) (Dynamic/ adynamic)

RIGHT LUNG AUR ALR LUR LLR PUR PLR LEFT LUNG AUR ALR LUR LLR PUR PLR BOTH LUNGS

Superior mediastinum Posterior mediastinum Pericardium

APPENDIX 2A CONSENT FORM – ENGLISH VERSION THE ACCURACY OF CHEST ULTRASOUND IN DIAGNOSING PEDIATRIC PATIENTS ADMITTED WITH PNEUMONIA AT MULAGO NATIONAL REFERRAL, KAMPALA, UGANDA.

Study ID: |__||__||__||__|

Principal Investigator: I am (representing) Dr. Kyomuhangi Agnes, a post graduate student at the Department of Radiology, College of Health Sciences Makerere University.

I would like to request you to allow your child to participate in a research study. Participation in this study is voluntary. If you accept to be part of this study, you will be requested to sign at the end of this page.

Background of the study.

Community-acquired pneumonia is an acute infectious pulmonary disease developing outside the hospital or in the first 48–72 hrs. of hospital stay.

It is a major health threat worldwide and is the leading cause of deaths in children under 5 yrs. The clinical diagnosis is made based on history findings of cough, and/or difficult breathing, fast breathing with/without chest in drawing, with/without fever and auscultatory findings.

Chest radiography is recommended or reserved to diagnose complicated cases of pneumonia but it has low accuracy. However, chest ultrasound has been found to be cheaper, more accurate, reliable and a radiation free tool in the diagnosis of pneumonia.

Purpose of the Research

The aim of this study is to determine the accuracy of chest ultrasound in diagnosing pneumonia in pediatric patients admitted with pneumonia at Mulago national referral hospital, Kampala, Uganda.

To achieve this, the study will enroll participants who are admitted with pneumonia and fit the inclusion criteria and these will undergo both ultrasound and chest radiography investigations which will be done within 24-hour s of each other. This study shall be on for a period of 6 months.

Study procedures

Before carrying out the investigations, I will ask you for information about yourself and your child including age, home address, occupation which I will record in a pre-coded form that I have here with me. Then your child will be taken for chest X-ray which will be done while yourchild is standing or lying down. Then chest ultrasound. Both investigations will be done from the radiology department of Mulago Hospital.

Benefits of the study:

You will not incur any extra cost when your child participates in this study. Both investigations will be paid for by the principal investigator and chest x-ray and ultrasound reports given to he/she for proper management by the clinicians.

The participation of your child in this study will aid me in collecting information that can be used to determine and improve the protocol used to diagnose pneumonia in children

Risks: The dose of radiation that will be used for Chest X-ray acquisition will be low and thus not detrimental. Chest ultrasound is very safe.

Statement of Confidentiality: Your child will be given a unique identification number and your child’s name or any other information that may reveal his/her identity will not be included in the report. All information and test results obtained will be locked away and only accessed by the principal investigator and research assistant, thus all will be kept confidential.

Costs for Participation: You will not incur any extra cost when you participate in this study.

Contact Information for Questions or Concerns

In case you have special concerns about this study, you can contact the chairperson of Makerere University school of medicine ethical review board; Prof. Ponsiano Ocama on 0772421190

In case you still have further question regarding this study you may contact us through the principal investigator: Dr. Kyomuhangi Agnes; Mobile: 0754121610; E-mail: [email protected]/[email protected].

Signature and Consent/Permission to be in the Research

I have read and understood the above information. I have had the opportunity to ask questions about it and any questions that I have asked have been answered to my satisfaction. I consent voluntarily for my child to participate in this study.

Name of participant’s parent/guardian signature/thumb print Date

………………………… ………….. ………… …………………

Name of witness Signature of witness Date

………………………… ………….. ………… …………………

Name of investigator signature Date

…………………… …………………… ………………

CONSENT FORM – LUGANDA VERSION. EKIWANDIIKO KYOKUKKIRIZA OKWETABA MU KUNOONYEREZA

OBUTUUFU BWOKUWULILIZA EKIFUBA MU KUKEBERA ABAANA ABALWADDE ABAWEEBWA EBITANDA NEKIFUBA KYAMAWUGWE MU DWALIRO EKKULU E MULAGO, KAMPALA, UGANDA

Namba yokunoonyereza: …

Akulira okunoonyereza: Nze (nkiikirira) musawo Kyomuhangi Agnes, omuziyi wa ddiguli eyokubili mu kitongole kya ‘Radiology’, mu tendekero lyebyobulamu eya setendekero ya Makerere.

Twagala okukusaba okukkiriza omwanawo okwetaba mu kunoonyereza. Okwetaba mu kunoonyereza kuno kwa kyeyagalire. Bwokkiriza, okwetabamu ojja kusabibwa okuteeka omukono kunkomelero yekiwandiiko kino.

Enyanjula

Ekifuba kyamawugwe bulwadde obwomutawaana obusiigibwa obukulira wabweru wedwaliro oba wakati wesaawa 48–72 ezokubeera mu dwaliro.

Kyantiisa nyo eli ebyobulamu mu nsi yonna era kyekisinga okuviirako okufa mu baana abali wansi wemyaka etaano (5). Okukebera kwedwaliro kukolebwa okusinziira ku byafaayo ebyazuulibwa ku kifuba, ne oba okussa okuzibu, okussiza okumukumu, okubeera oba obutabeera namusujja.

Okuwuliriza ekifuba kulagibwa oba kukolebwa mu kukebera ensonga zekifuba ezikalubye naye kulina obutuufu butono. Wabula, ekifaananyi kyekifuba kizuulidwa okuba ekyalayisi, kilungi, kyesigika, era tekilina maanyi gaabulabe agekyuma mu kukebera ekifuba.

Omugaso gwokunoonyereza

Okunoonyereza kuno kugenderera okuzuula ekitundu kyokukaanya wakati wokukebera ekifuba wamu Nokukikuba ekifaananyi mu kukebera obulwadde bwamawugwe obumanyidwanga ‘Pneumonia’ mu balwadde abato abaweeledwa ebitanda kulwa ‘pneumonia’ ku dwaliro ekkulu e Mulago, Kampala Uganda.

Okufuna kino, okunoonyereza kujja kuteekamu abantu abaweledwa ebitanda n’obulwadde bwamawugwe obumanyidwanga ‘Pneumonia’ era nga bagwa mu ttuluba lyabo abateekebwa mu kunoonyereza era bano bajja kuyita mu kukeberwa kwokulabibwa ekifuba nokukubwa ekifaananyi ebijja okukolebwa mu saawa 24 oluvanyuma lwokuweebwa ekitanda. Okunoonyereza kuno kujja kumala ebbanga lyamyezi mukaaga.

Emitendera gyokunoonyereza

Nga tetunnaba kukebera, nja kukubuuza obubaka obukukwatako (omwanawo) omuli emyaka, gyobeera, omulimu byenaawandiika mu kiwandiiko ekiliko enamba kyennina wano. Olwo omwanawo ajja kutwalibwa okukubwa ekifaananyi kyekifuba ekijja okukolebwa ngomwanawo ayimilidde oba nga agalamidde. Oluvanyuma ekifuba kijja kukeberwa. Okwekebejja kwombi kujja kukolebwa ku kitongole kya ‘Radiology’ ekyedwaliro ly’e Mulago.

Emiganyuro gyokunoonyereza

Tojja kusasaanya sente ndala zonna nga wetaba mu kunoonyereza kuno. Akulira okunoonyereza kuno, ajja kusasulila ebifaananyi byona era ebizuulidwa bijja kuweebwa biyambe omusawo wo mukkuwa obujjajjabi.

Okwetabamu kwomwanawo mu kunoonyereza kuno kujja kunyamba mu kukungaanya obubaka obuyinza okukozesebwa okuzuula wamu nokutumbula enkola ekozesebwa okukebela obulwadde bwamawugwe obumanyidwanga ‘Pneumonia’ mu baana.

Akatyabaga;

Tewaliiwo katyabaga gyoli mu kunoonyereza kuno. Amassanyalaze gekifaananyi kyekifuba mattini mukunonyereza kuno wamu nokuwuliliza ekifuba byonna tebilina buzibu.

Olunyiliri lwokukuuma ebyama Omwanawo ajja kuweebwa namba eyenjawulo era erinya lyomwanawo oba obubaka obulala bwonna obuyinza okumulabisa tebujja kuteekebwa mu alipoota. Obubaka bwonna wamu nebinaava mu kukebera ebinaafunibwa bijja kusibibwa era bilabibweko akulira okunoonyereza yekka wamu nayamba ku kunoonyereza, olwo byonna bijja kukuumibwa mu kyaama.

Kulwebibuuzo wamu nensonga endala

Bwoba olina ensonga ezenjawulo oba ebyokwemulugunya ebikwata ku kunoonyereza kuno, oyinza okutuukilira sentebe wakakiiko akakwasisa empisa mu kunoonyereza akayitibwa ‘School of Medicine Research Ethics Committee’ sabakenkufu Ponsiano Ocama ku ssimu 0772421190.

Bwoba okyalina ebibuuzo ebilala ebikwata ku kunoonyereza kuno oyinza okututuukilira ngoyita mu akulira okunoonyereza musawo Kyomuhangi Agnes; essimu 0754121610, omutimbagano [email protected].

Omukono wamu nokukkiriza / olukusa okubeera mu kunoonyereza

Nsabidwa kulwomwana wange okwetaba mu kunoonyereza kuno. Nsomye era nentegeera obubaka obwo waggulu. Nfunye omukisa okubuuza ebibuuzo ebikukwatako era ebibuuzo byonna byanukudwa kulwokumatila kwange.

Nkikiriza nga neyagalidde kulwomwana wange okwetaba mu kunoonyereza kuno.

Erinya lyomuzadde weyetabyemu/omukuza Omukono/ ekinkumu Enaku zomwezi

……………………………………………. ……………………….. ………………

Erinya lyomujulizi Omukono gwomujulizi Enaku zomwezi

…………………………………………… ………………………...... ………………

Erinya lyanoonyereza Omukono Enaku zomwezi

APPENDIX 2 B ASSENT FORM -ENGLISH VERSION ACCURACY OF CHEST ULTRASOUND IN DIAGNOSING PEDIATRIC PATIENTS ADMITTED WITH PNEUMONIA AT MULAGO NATIONAL REFERRAL, KAMPALA, UGANDA Study ID: ……………… Introduction I am Dr. Kyomuhangi Agnes, a post graduate student at the Department of Radiology, Makerere University College of Health Sciences, P.O. Box 7072 Kampala Uganda. Telephone number is +256-754121610 Email address: [email protected] I am requesting you to participate in a research study. Participation is voluntary. If you accept to be part of this study, you will be asked to sign at the end of this page. Background of the study. Community-acquired pneumonia is an acute infectious pulmonary disease developing outside the hospital or in the first 48–72 hrs. of hospital stay. It is a major health threat worldwide and is the leading cause of deaths in children under 5 yrs. The clinical diagnosis is made based on history findings of cough, and/or difficult breathing, fast breathing with/without chest in drawing, with/without fever and auscultatory findings. Chest radiography is recommended or reserved to diagnose complicated cases of pneumonia but it has low accuracy. However, chest ultrasound has been found to be cheaper, more accurate, reliable and a radiation free tool in the diagnosis of pneumonia. Purpose of the Study The aim of this study is to determine the accuracy of chest ultrasound in diagnosing pneumonia in pediatric patients admitted with pneumonia at Mulago national referral hospital, Kampala, Uganda. Study procedure If you volunteer to participate in this study, you will be asked some questions about yourself then, you will undergo both ultrasound and chest radiography investigations. The chest X-ray will be taken while you are standing or lying down. Then chest ultrasound will be done while you are lying down, on your side or while sitting up. Both investigations will be done from the radiology department of Mulago Hospital, within 24 hrs. after admission. Respect and Confidentiality

The information collected will only be used for research purposes and your name will not be used in reference to the information collected. Every form will have a study number that will be used and all the information will be securely stored under lock and key. Benefits There will not be any monetary benefit or compensation to you and you will not incur any expenses when you participate in this study. Both investigations will be paid for by the principal investigator and chest x-ray and ultrasound reports given to he/she for proper management by the clinicians. The findings from this study will be used to determine and improve the protocol used to diagnose pneumonia in children. Risk / Discomfort; The dose of radiation that will be used for Chest X-ray acquisition will be low and thus not detrimental. Chest ultrasound is very safe. Your rights as a study participant Participation in this study is voluntary and you are free to withdraw from the study at any point. This will not in any way affect the care that you’re receiving from the hospital. Contacts for further information. For issues concerned with your right as a study participant, contact the chairman of the School of Medicine Research and Ethics Committee Professor Ponsiano Ocama on telephone number 0772421190. For any further information about this study, feel free to contact me Dr. Kyomuhangi Agnes– the principal investigator on 0754121610. Statement of Consent I have fully understood the purpose and nature of this study; the risks and potential benefits and it is my choice to participate and that withdrawal from the study will not affect my treatment in the clinic. I hereby voluntarily choose to participate in the study as signed below: …………………………………………….. ……………………… Name/signature or thumbprint of child Date ……………………………………………….. ……………………… Name/signature of Interviewer Date ……………………………………………….. ………………… Name/signature of witness Date

ASSENT FORM-LUGANDA VERSION EKIWANDIIKO KYABAANA EKYOKJUKKIRIZA OKWETABA MU KUNOONYEREZA OMUTWE GWOKUNOONYEREZA: OBUTUUFU BWOKUWULILIZA EKIFUBA MU KUKEBERA ABAANA ABALWADDE ABAWEEBWA EBITANDA NEKIFUBA KYAMAWUGWE MU DWALIRO EKKULU E MULAGO, KAMPALA, UGANDA Namba yokunoonyereza ……………… Enyanjula Nze (…………….) /Musawo Kyomuhangi Agnes, omuyizi weddaala elyokubili mu kitundu kya ‘Radiology’ mu tendekero lyebyobulamu elimanyidwa nga ‘Makerere University College of Health Sciences’, akasanduuko ka posita namba 7072 Kampala Uganda. Enamba yessimu eri +256-754121610. Omutimbagano gwa yintaneti: [email protected] Nkusaba okwetaba mu kunoonyereza. Okwetabamu kwa kyeyagalire. Bwokkiriza okubeera mu kunoonyereza, ojja kusabibwa okuteeka omukono ku nkomelero yolupapula luno. Ebyafaayo byokunoonyereza. Ekifuba kyamawugwe bulwadde obwomutawaana obusiigibwa obukulira wabweru wedwaliro oba wakati wesaawa 48–72 ezokubeera mu dwaliro. Kyantiisa nyo eli ebyobulamu mu nsi yonna era kyekisinga okuviirako okufa mu baana abali wansi wemyaka etaano (5). Okukebera kwedwaliro kukolebwa okusinziira ku byafaayo ebyazuulibwa ku kifuba, ne oba okussa okuzibu, okussiza okumukumu, okubeera oba obutabeera namusujja. Okuwuliriza ekifuba kulagibwa oba kukolebwa mu kukebera ensonga zekifuba ezikalubye naye kulina obutuufu butono. Wabula, ekifaananyi kyekifuba kizuulidwa okuba ekyalayisi, kilungi, kyesigika, era tekilina maanyi gaabulabe agekyuma mu kukebera ekifuba. Omugaso gwokunoonyereza Ekigendererwa kyokunoonyereza kuno kwekuzuula obutuufu bwekifaananyi kyekifuba mu kekebela obulwadde bwamawugwe mu baana abalwadde abaweeledwa ekitanda nekifuba mu dwaliro ekkulu e Mulago, Kampala, Uganda. Emitendera gyokunoonyereza Bweweyama okwetaba mu kunoonyereza kuno, ojja kubuuzibwayo ebibuuzo ebikukwatako, olwo ojja kuyita mu kukebelebwa wamu nokukubwa ekifaananyi kyekifuba. Ekifaananyi kyekifuba kijja kukubibwa ng’otudde oba ngogalamidde, Olwo, okuwuliliza ekifuba kujja kukolebwa ngogalamidde oba ngotudde. Okwekebejja kwombi kujja kukolebwa mu kitundu kya ‘Radiology’ ekyedwaliro ly’e Mulago mu saawa 24 oluvanyuma lwokuweebwa ekitanda. Ekitiibwa nokukuuma ebyama

Obubaka obukungaanyizibwa bujja kukozesebwa kulwa mirimu gya kunoonyereza kwokka era erinyalyo telijja kukozesebwa nga lyekuusa ku bubaka obukungaanyizidwa. Buli kiwandiiko kijja kubeerako namba yokunoonyereza ejja okukozesebwa era obubaka bwonna bujja kukuumibwa bulungi nekkufulu nekisumuluzo. Emiganyuro Tewajja kubaawo kuganyurwa kwa nsimbi oba okuddizibwa gyoli era tojja kusaasanya nsimbi zonna mu kunoonyereza kuno. Akulira okunoonyereza kuno, ajja kusasulila ebifaananyi byona era ebizuulidwa bijja kuweebwa biyambe omusawo wo mukkuwa obujjajjabi. Ebizuulidwa okuva mu kunoonyereza bijja kukozesebwa okuzuula wamu nokulongoosa enkola ekozesebwa okukebera obulwadde bwekifuba kyamawugwe ekiyitibwa ‘Pneumonia’ mu baana. Akatyabaga/Ebitali bilungi Tewaliiwo katyabaga gyoli mu kunoonyereza kuno. Amassanyalaze gekifaananyi kyekifuba mattini mukunonyereza kuno wamu nokuwuliliza ekifuba byonna tebilina buzibu. Eddembelyo ngeyetabye mu kunoonyereza Okwetaba mu kunoonyereza kuno kwa kyeyagalire era oli waddembe okuvaamu akadde konna. Kino tekijja mungeri yonna kukosa nzijanjabayo gyofuna okuva mu dwaliro Endagiliro kulwobubaka obusingako. Kulwensonga ezekuusa ku ddembelyo ngeyetaba mu kunoonyereza, tuukilira sentebe wakakiiko akakwasaganya ebyempisa mu kunoonyereza akayitibwa ‘School of Medicine Research and Ethics Committee’ sabakenkufu Ponsiano Ocama ku namba yessimu 0772421190. Kulwobubaka obusingawo obukwata ku kunoonyereza kuno, beera waddembe okutuukilira nze musawo Kyomuhangi Agnes– akulira okunoonyereza ku ssimu 0754121610. Olunyilili lwokukkiriza Ntegedde bulungi omugaso wamu nekikula kyokunoonyereza kuno; akatyabaga wamu nemiganyuro era kusalawo kwange okwetabamu era nti okuvaamu kwange tekijja kukosa bujanjabi bwange okuva mu kilinika. Neyagalira okulondawo okwetaba mu kunoonyerza nga bwekiteekedwako omukono wammanga: …………………………………………….. ……………………… Erinya/omukono oba ekinkumu kyomwana Ennaku zomwezi ……………………………………………….. ……………………… Erinya/omukono gwabuuza Enaku zomwezi ……………………………………………….. ……………………….. Erinya/omukono gwomuujulizi Enaku zomwezi

APPENDIX 3 WORK PLAN. YEAR 2017 ACTIVITY Jan Feb March April May June July Aug Sept Oct Nov Dec Proposal writting X X X X

Presentation to X department ETHICAL X X APPROVAL 2018

Jan Feb March April May June July Aug Sept Oct Nov Dec

Pilot testing of X study stool Data collection X X X X X X X

Data Analysis X X

Report writting X X X X X

2019

Jan Feb March April May June

Presentation of X X the dissertation.

APPENDIX 4 BUDGET FOR STUDY UNIT COST (UG SH) ITEM QUANTITY TOTAL COST Printing and photocopying 1,000,000 1,000,000 1,000,000 Research assistants 2 1,500,000 3,000,000 (300,000 per month) X-rays 300 films 10000 3,000,000 Sonopaper 50 rolls 70000 3,500,000 Sonogel 50,000 per jerry can 4 200,000 Stationary 1,000,000 1,000,000 1,000,000 Statistician 1 800,000 800,000 Data storage 500,000 500,000 500,000 (HD , CDs) Pulse oximeter 500000 1 500,000 Total 13,500,000/=