Hindawi

BioMed Research International

Volume 2019, Article ID 6202405, 8 pages
https://doi.org/10.1155/2019/6202405

Research Article
Radiologic Diagnosis and Hospitalization among
Children with Severe Community Acquired Pneumonia:
A Prospective Cohort Study

Meiron Hassen,1,2 Alemayehu Toma,3 Mulugeta Tesfay,4 Eyoel Degafu,5 Solomon Bekele,6
Freshwork Ayalew,7 Abel Gedefaw,8 and Birkneh Tilahun Tadesse 2
1
Adare General Hospital, Pediatrics Unit, Hawassa, Ethiopia
2
Hawassa University, College of Medicine and Health Sciences, Department of Pediatrics and Child health, Hawassa,
P.O. Box 1560, Ethiopia
3
Hawassa University, College of Medicine and Health Sciences, School of Pharmacy, Department of Pharmacology and Toxicology,
Hawassa, P.O. Box 1560, Ethiopia
4
Bete-Abrham Primary Hospital, Department of Radiology, Hawassa, P.O. Box 851, Ethiopia
5
Alatyon General Hospital, Department of Radiology, Hawassa, P.O. Box 1267, Ethiopia
6
Hawassa University, College of Medicine and Health Sciences, Department of Radiology, Hawassa, P.O. Box 1560, Ethiopia
7
Hawassa University, College of Medicine and Health Sciences, School of Laboratory Medicine, Hawassa, P.O. Box 1560, Ethiopia
8
Hawassa University, College of Medicine and Health Sciences, Department of Gynecology and Obstetrics, Hawassa,
P.O. Box 1560, Ethiopia

Correspondence should be addressed to Birkneh Tilahun Tadesse; [email protected]

Received 15 November 2018; Revised 13 December 2018; Accepted 20 December 2018; Published 9 January 2019

Academic Editor: Jonathan Muraskas

Copyright © 2019 Meiron Hassen et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objectives. This study was designed to assess the role of chest radiography for the diagnosis of pneumonia and assess the association
of clinical characteristics with radiologic findings and predictors of hospitalization among children with severe community acquired
pneumonia. Methods. A prospective study was conducted on 122 children between ages of 3 month and 14 years admitted to pediatric
emergency unit with diagnosis of severe pneumonia from September 1st to November 30th , 2017. Eligible children were subjected
to chest radiography which was read by two senior radiologists independently (R1 and R2). Disagreements between R1 and R2 were
resolved by a third senior radiologist (R3). Level of agreement between radiologists was assessed using Cohen’s kappa coefficient.
Clinical and laboratory parameters which could explain the variability in the duration of hospital stay were assessed using a linear
regression mode. Independent predictors were assessed using multiple linear regression. Results. The median age of the cohort was
10.0 months (interquartile range (IQR): 6.75–24.0); 76 (62.3%) were male. Nearly half, 63 (51.6%) did not have radiologic evidence
of pneumonia. There was low level of agreement between R1 and R2 in reporting consolidation (kappa=0.435, p-value≤0.001),
haziness (kappa=0.375, p-value≤0.001), and infiltration (kappa=0.267, p-value=0.008). Children with higher recorded temperature
were more likely to have radiologic abnormalities suggesting pneumonia (p-value=0.033). The median duration of hospitalization
was 3 days (IQR: 1-4 days); 118 (96.7%) were discharged with improvement. Height-for-age z-score (Coef.=0.203, R2 =0.041, p-
value=0.027); and hemoglobin level (Coef.=-0.249, R2 =0.062, p-value=0.006) explained 4.1% and 6.2% of the variability in the
duration of hospital stay, respectively. Conclusion. Radiologic evidence of pneumonia was absent in half of the children with severe
pneumonia. There was low agreement between senior radiologists in reporting chest radiographic findings, potentially necessitating
harmonization activities to uniformly implement the WHO guidelines in reading chest radiographs.

1. Background The global annual incidence of pneumonia is 150 to 156

million cases, accounting for approximately 10-20 million
Worldwide, pneumonia was responsible for 15% of childhood hospitalizations [1]. In Ethiopia, acute respiratory tract infec-
deaths in 2016, with highest incidence in developing countries [1]. tions, particularly pneumonia, are the leading causes of child
2 BioMed Research International

mortality and morbidity. Pneumonia accounts for 18% of Chest radiographic films are usually read by radiologists.
childhood deaths in Ethiopia [2]. Typically, cough, pleuritic However, there are no data to assess the validity of the
chest pain, fever, fatigue, and loss of appetite are presenting readings by the radiologists and the role of CXR in treating
symptoms of pneumonia. Children and the elderly with children with pneumonia. In the current study three clinical
pneumonia have varying presenting features, which include questions were posed: (1) What is the level of agreement
headache, nausea, abdominal pain, and absence of one or between two radiologists in reading CXR films of children
more of the prototypical symptoms [3, 4]. with severe pneumonia? (2) What is the magnitude and
The diagnosis of pneumonia is based mainly on clinical predictors of radiologic pneumonia in children? (3) What are
parameters including respiratory symptoms and signs, which the predictors of duration of hospitalization among children
incur no cost and are sensitive, but are also nonspecific and with severe pneumonia in Ethiopia?
could lead to unnecessary prescription of antibiotics and
drug resistance. According to World Health Organization 2. Methods and Materials
(WHO) guidelines, severe pneumonia is diagnosed when
there is cough and fast breathing (defined based on respi- 2.1. Study Setting, Design, and Research Questions. The study
ratory rate for age) with one or two of the severity signs was conducted in Hawassa University Specialized Compre-
including chest in-drawing (in infants <2 months of age), hensive Hospital (HUCSH), which is found in the southern
grunting, cyanosis, inability to feed, lethargy, or convulsion region of Ethiopia. HUSCH is the first referral hospital
[5]. Chest X-ray (CXR) has been the main stay modality in established in the region serving as a teaching hospital for the
the investigation of chest infection since its invention in the College of Medicine and Health Science of Hawassa Univer-
late 19th century, despite the advances in imaging modalities sity, with a catchment population of 10-12 million. It receives
[6]. The WHO recommends CXR for all patients clinically about 43,384 patients annually. The hospital gives free service
diagnosed with severe pneumonia at tertiary centers [7– for patients with high priority pediatric conditions, including
11]. Chest radiograph improves the diagnosis of pediatric severe pneumonia.
community acquired pneumonia to a certain degree and A prospective short cohort study was conducted from
may prevent overtreatment with antibiotics [12]. However, it September 1st to November 30th , 2017. Children with a
is important to understand that interpretation is subject to clinical diagnosis of severe community acquired pneumonia
perceptual and cognitive limitation and errors [13]. were followed from admission until discharge. All pediatric
The interpretation of radiographs is difficult in young patients between ages 3 months and 14 years, who presented
children and is affected by the radiographer’s experience and to the pediatric emergency department and were diagnosed
the amount of clinical information available. Additionally, with severe community acquired pneumonia as per WHO
chest radiography cannot reliably distinguish between viral guidelines, were included. All patients fulfilling inclusion
and bacterial pneumonia and is often unable to detect early criteria during the study period were enrolled consecutively.
changes of pneumonia [6, 11, 14]. Other drawbacks of chest
There were three primary objectives of the study, which
radiography include exposure to ionizing radiation, cost,
included (1) assessing the pattern of CXR abnormalities
the time and space used, and the need to wait from the
among children with a clinical diagnosis of severe community
radiograph and to see the clinician again. Moreover, chest
acquired pneumonia; (2) assessing the level of agreement
radiography has a low sensitivity compared to advanced
among senior radiologists (MD + Radiology) on the WHO
imaging tools like lung ultrasound, computed tomography,
classification of CXR findings among children with severe
and magnetic resonance imaging [14–16].
community acquired pneumonia; and (3) assessing the deter-
Timely diagnosis and management of severe pneumonia,
minants of duration of hospitalization of children with severe
and shorter hospital stay is crucial to decrease childhood
morbidity and mortality in resource limited settings. Pro- community acquired pneumonia.
longed hospitalization could be associated with unfavorable
outcomes and is also a cost burden to the public [17–21]. 2.2. Exclusion Criteria. Children with congestive heart fail-
Several preventable risk factors have been reported to be ure, hospitalization within 14 days of a prior hospitalization
associated with prolonged hospitalization among children episode or onset of pneumonia within 4 days of hospi-
with severe community acquired pneumonia. Failing to talization, those diagnosed with tuberculosis, and children
exclusively breastfeed for the first 6 months, inappropriate with foreign body aspiration or aspiration pneumonia were
complimentary feeding, anemia, malnutrition, exposure to excluded from participating in the study.
parental smoking, inadequate antibiotic use, lack of aware-
ness of parents, overuse of nonsteroidal anti-inflammatory 2.3. Diagnosis and Management of Severe Community
drugs, and indoor air pollution contribute to prolonging Acquired Pneumonia. In this study, the diagnosis of severe
duration of hospitalization [22, 23]. History of full vaccina- community acquired pneumonia was considered if the
tion on the other hand decreased length of stay in the hospital children presented with cough and/or with tachypnea defined
[24]. by respiratory rate count of more than cut off for age (i.e.,
Ethiopia is one of the high burden countries for childhood >/= 50 for those between ages of 3 months and 1 year; >/= 40
morbidity and mortality associated with severe pneumonia. for children between the ages of 1-5 years) and the presence
It is common practice for clinicians working in hospitals of one of intercostal and/or subcostal recessions, chest
to routinely request CXR for the diagnosis of pneumonia. in-drawing, grunting, cyanosis, or altered mental status [5].
BioMed Research International 3

Table 1: #Form 2. CXR Assessment Checklist. Radiologist: Please complete the following table for each CXR film.

SN Parameter Response Comment

1 Patient ID ——————
(1) R1 ◻
2 Radiologist Code (2) R2 ◻
(3) R3 ◻
3 Chest X ray code ——————
(1) Adequate ◻
4 Film Quality (2) Sup optimal ◻
(3) Not interpretable ◻
(1) Yes ◻ If yes,
5 Consolidation
(2) No ◻ location——————
(1) Yes ◻ If yes,
6 Infiltration
(2) No ◻ location——————
(1) Yes ◻ If yes,
7 Haziness
(2) No ◻ location——————
(1) Yes ◻ If yes,
8 Pleural effusion
(2) No ◻ location——————
(1) Yes ◻ If yes,
9 Atelectasis
(2) No ◻ location——————
(1) Yes ◻ If yes,
10 Fibrosis
(2) No ◻ location——————
(1) Yes ◻
11 Pleural thickening
(2) No ◻
(1) Yes ◻
12 Hyperinflation
(2) No ◻
13 Index

Treatment of severe community acquired pneumonia guidelines (Table 1). Findings and conclusion in which both
involves respiratory support—including oxygen administra- radiologists agreed were taken as final and did not require
tion via a nasal catheter or prong, antipyretics if fever reading by a third senior radiologist, R3 who was used as a
>38.5∘ Celsius and antibiotics—intravenous crystalline peni- tie breaker for the study. Chest X-ray films with discordant
cillin in four divided doses until the child can take oral readings in the initial evaluation were arbitrated by R3. The
antibiotics. Additionally, any environmental, nutritional, or finding which was reported by two radiologists (R1 and R3 or
socioeconomic risk factors are carefully assessed and modifi- R2 and R3) was considered as a final CXR finding.
able factors are addressed.
2.5. Data Collection Procedures. Every patient admitted to
2.4. CXR Diagnosis of Pneumonia. Based on the WHO guide- the emergency unit was registered in our study registration
lines [8], the standard criteria for the radiologic diagnosis of form by trained data collector nurses. At enrolment, through
pneumonia in a two view plain chest radiograph are defined medical chart review and use of structured questionnaires,
as follows: we collected dependent variables like demographics (age,
sex, and address) of the child; clinical presentation (cough,
Endpoint consolidation: a dense or fluffy opacity that fever, difficulty breathing, grunting, cyanosis, convulsions,
occupies the whole lobe of the lung or the entire lung inability to feed, and change in level of consciousness),
often containing air bronchogram. previous history of similar problem, immunodeficiency (HIV,
Nonend point consolidation: linear and patchy densi- malnutrition, and diabetes mellitus), physical examination
ties in a lacy pattern involving both lungs, featuring (vital signs, nutritional status, intercostal or subcostal retrac-
peribronchial thickening and multiple areas of atelec- tion, nasal flaring, chest in-drawing, wheezing, crepitations,
tasis. bronchial breath sounds, cyanosis, signs of rickets, and
Pleural effusion: presence of fluid in the lateral pleural mental status), and laboratory results (white blood cell count,
space evidenced by obliterating the costophrenic absolute neutrophil count, hemoglobin, platelet count, C-
angle between the lung and chest wall. reactive protein, erythrocyte sedimentation rate, and blood
culture). The children were followed during their stay in the
Initially, CXR was independently read by two senior radi- hospital for evaluation of oxygen requirement, antibiotics
ologists (MD + 3-Year Radiology Specialty)—referred to as therapy, feeding, persistence of fever, tachypnea, and duration
R1 and R2—following a checklist adopted from the WHO of hospital stay and treatment outcome.
4 BioMed Research International

Nutritional assessment was done using the WHO recom- Antibiotic treatment at admission consisted of crystalline
mended growth curves [25]. Z-scores were used to classify penicillin (56, 45.9%), ceftriaxone (64, 52.5%), and a combi-
nutritional status. nation of crystalline penicillin and ceftriaxone (2, 1.6%).
Questionnaires were completed by trained data collection
nurses. The study nurses accompanied patients to the radiol- 3.2. Radiologic Evidence of Clinically Diagnosed Severe Pneu-
ogy department for CXR, and films were read in sequence by monia. Over half (63, 51.6%) of the children diagnosed with
R2, followed by R1. R3 read all the films with disagreements severe pneumonia did not have radiologic evidence of pneu-
at the end of data collection. Each radiologist was asked monia according to WHO CXR criteria. The only difference
to complete the checklist of comments and findings. The in clinical presentation between children with abnormal CXR
readings by R1 and R2 of the CXR films were used by the and those with normal CXR was a higher body temperature
treating clinicians for patient care. We ensured that the study at admission among those with radiologically confirmed
procedures did not delay patient treatment in any way. pneumonia (p=0.033). Although not statistically significant,
white cell count and hemoglobin levels were higher in
2.6. Data Analysis. Descriptive statistics were analyzed and children with normal CXR, and platelet counts were higher
presented as frequency (percentage), mean (standard devi- in those with abnormal CXR findings (Table 2).
ation), and median (interquartile range). The association
of sociodemographic, clinical, and laboratory findings with 3.3. Factors Associated with Duration of Hospitalization. The
CXR findings or CXR diagnosis of pneumonia was assessed majority 118 (96.7%) were discharged home, whereas 3 (2.5%)
using Fisher’s exact test for categorical variables and Mann- absconded against medical advice and 1 (0.8%) died. The child
Whitney U test. Nonparametric tests were used as we had a that died was 8 months old who presented with fast breathing,
small sample size and the data were not normally distributed. cough, fever, and grunting of 3 days’ duration. The median
Agreement between the two radiologists (R1 and R2) was length of hospitalization was 3 days (IQR 2.75–4) days.
assessed using the Cohen’s kappa coefficient. On bivariate analysis, height-for-age z-score (p=0.027) and
Predictors of duration of hospital stay were assessed using hemoglobin level (p=0.006) were associated with prolonged
a linear regression model. Multiple linear regression was used hospital stay (Table 3). However, on multiple linear regression
to assess the independent predictors which could signifi- modelling, hemoglobin was the only independent variable
cantly explain the variability in the duration of hospital stay associated with prolonged hospitalization (p=0.007; Table 4).’
among children with severe pneumonia. Variables which on
bivariate analysis showed significance at a level of p<0.2 were
3.4. Agreement between Radiologists in Interpreting CXR Find-
included in the multiple linear regression model. P-values of
ings of Children with Severe Pneumonia. Radiologic diagnos-
<0.05 were considered as being statistically significant in the
tic readings by R1 and R2 were not significantly different in
multiple linear regression model. SPSS for windows version
interpretation of atelectasis, fibrosis, pleural thickening, and
22 was used to analyze the data.
hyperinflation in children with severe pneumonia. Comment
on film quality as adequate for reading was 74.6% by R1 and
2.7. Ethical Consideration. Ethical approval was obtained 69.7% by R2 (kappa=0.634 and p-value≤0.001). Presence of
from the Ethical Review Board of Hawassa University College consolidation and haziness was reported in 22.1% and 9.8%
of Medicine and Health Sciences and permission for data (kappa=0.435, p-value≤0.001); 9.8% and 2.5% (kappa=0.0375,
collection was obtained from HUCSH- Manager and Head p-value=0.001), respectively, by R1 and R2 (Table 5). Overall,
of Department of Pediatrics. After brief explanation of the 43 (35.2%) CXR films required arbitration by R3.
purpose of the study, informed consent was obtained from
parents/guardians/or care givers of eligible children. Confi-
dentiality was assured by excluding participant names and
4. Discussion
other identifiers from the analytic database. Participants had The present study evaluated the role of CXR for the diagnosis
the right not to participate or withdraw from the study at of severe pneumonia by radiologists in a cohort of chil-
any point. Patients judged unstable to undergo CXR were dren hospitalized at HUCSH, Hawassa, Southern Ethiopia.
excluded from the study. Evidence based definitive diagnosis assists clinicians to
rationalize treatment options, so as to prevent prolonged
3. Results hospitalization and overburdening of healthcare services and
minimize cost to the public health sector.
3.1. Sociodemographic and Baseline Characteristics. A total In this study, the only clinical finding that discriminated
of 122 children with severe pneumonia were followed from significantly between children with radiographically con-
admission to discharge, loss to follow-up, or death. The firmed pneumonia and those with normal CXR was degree
median age of the cohort was 10.0 months IQR (6.75-24.0 of fever. Higher body temperature has been associated with
months); 76 (62.3%) were male. The majority of the children, delay in achieving clinical stability, although pattern of fever
77 (63.1%) were from the Southern Nations, Nationalities, was not associated with radiologic findings [26].
Peoples’ Region; 26 (21.3%) and 19(15.6%) were from the The WHO has developed a clinical case definition of
neighboring Oromia region and Hawassa city, respectively. pneumonia for health care workers who work in settings with
Twenty-two (18%) children had been hospitalized previously. limited diagnostic capabilities [5]. Over the ensuing three
BioMed Research International 5

Table 2: Demographic and clinical characteristics of children with severe pneumonia in Hawassa University Comprehensive Specialized
Hospital.

Radiologic evidence of pneumonia

Baseline data P-value
Present (n=59) Absent (n=63)
Age (months), Median (IQR) 10 (8-24) 10 (5-24) 0.550∗
Sex, Male (%) 40 (67.8) 36 (37.1) 0.264†
Duration of illness (days), Median (IQR) 3 (1-4) 3 (2-5) 0.269∗
Fever, Yes (%) 47 (46.1) 55 (53.9) 0.329†
Grunting, Yes (%) 26 (44.1) 32 (50.8) 0.474†
Vomiting, Yes (%) 22 (37.3) 24 (37.1) -
Previous admission, Yes (%) 12 (20.3) 10 (15.9) 0.639†
Indoor cooking, Yes (%) 35 (59.3) 31 (49.2) 0.281†
Wheezing, Yes (%) 57(96.6) 54 (85.7) 0.055†
Crepitation, Yes (%) 50 (84.7) 56 (88.9) 0.595†
Mentation, Yes (%) 58 (98.3) 61 (96.8) -
Respiratory rate, Median (IQR) 68 (60-76) 64 (54-70) 0.060∗
Pulse rate, Median (IQR) 148 (140-160) 146 (136-152) 0.076∗
Temperature (∘ C), Median (IQR) 38.1 (37.8-38.7) 37.8 (37.2-38.5) 0.033∗
SaPO2, Median (IQR) 85.0 (78.0-89,0) 85.5 (82-81.25) 0.187∗
White Blood Cell count, Median (IQR) 9,900 (7625-13,425) 10,000 (7,000-13,300) 0.998∗
Platelet count, Median (IQR) 352,000 (231,000-417,750) 317,000 (248,000-412,000) 0.724∗
Hemoglobin, Median (IQR) 11.2 (9.65-12.59) 11.8 (10.6-12.6) 0.224∗
Weight-for-Age Z-score, Median (IQR) -0.85 (-1.73-0.50) -0.15 (-1.11-0.71) 0.067∗
Height-for-Age Z-score, Median (IQR) -0.61 (-2.01-0.76) -0.25 (-1.29-1.06) 0.144∗
Weight-for-Height Z-score, Median (IQR) -0.46 (-1.93-0.89) -0.29 (-1.30-0.52) 0.495∗
BAZ, Median (IQR) -0.5 (-1.93-0.80) -0.43 (-1.55-0.63) 0.443∗
∗Mann-Whitney U test was used to compare the medians between groups; †Fisher’s exact test was used for categorical variables; IQR, Interquartile Range;
SaPO2, Saturation of Oxygen; BAZ, Body Mass Index-for-Age Z-score

decades, this definition has served as the primary means of from Brazil reported a concordance of 86.7% (kappa=0.683)
identifying pneumonia among children in low and middle with one radiologist reporting consolidation in 32.9% of
income countries. However, studies examining the specificity CXRs whilst the other radiologist reported consolidation
of this definition for the presence of radiographic pneumonia in 28.3% [31]. These findings are much better than the
report discordant readings by radiologists. Only about 44% of findings of the current study which found a statistically sig-
more than 7,800 children meeting criteria for WHO-defined nificant lack of concordance between radiologists for consol-
severe pneumonia in The Gambia had radiographic findings idation (kappa=0.435), infiltration (kappa=0.267), haziness
[27, 28]. Consistent with these, the current study found that (kappa=0.375), and effusion (kappa=0.658). The significant
radiologic evidence was found in 51.6% of children who were difference in interpretation of CXR between radiologists in
clinically diagnosed with WHO-defined severe pneumonia. our study may have been due to differences in training and
Discordant interpretation of CXR by radiologists may be experience between the specialists.
due to variability in the use of the WHO CXR scoring system Proper diagnosis and treatment of pneumonia in children
by health professionals or variation in the interpretation of are associated with less hospitalization and lower cost to
chest radiographs by radiologists. Use of lung ultrasound families and the healthcare setting. The median duration
has been shown to supplement CXR in the confirmation of hospitalization in our study was 3 days (IQR: 1–4 days)
of clinical pneumonia and improves diagnostic yield of which is similar to a report from a high income setting in
pneumonia in resource limited settings [29]. Results of our the Netherlands, where hospitalization ranged from 3 to 6
study suggest that there is a need for improved interpretation days [32]. Age, deranged vital signs, chest in-drawing, and
of CXRs by radiologists, and this would be anticipated to be infiltrates on radiography were the strongest predictors of
even more necessary for clinicians that lack formal radiologic severity of pneumonia in children [33, 34]. Malnutrition is
training. associated with longer hospital stay due to the poor immune
Our findings on agreement between radiologists on status, which could lead to other complications [35, 36]. Our
most CXR findings were much lower than results reported findings showed that hemoglobin level was an independent
by studies from western settings. For instance, a Spanish predictor of duration of hospitalization. However, the current
study found a concordance of 93.1% between radiologists study did not reveal any relationship between nutritional
for CXR interpretation (kappa=0.8) [30]. Another study status, physical findings, and duration of hospitalization.
6 BioMed Research International

Table 3: Bivariate linear regression model of predictors of duration of hospitalization among children with severe pneumonia, Hawassa,
Ethiopia.

Variable Coefficient R2 P-value

Age in months -0.123 0.015 0.177
Sex, Male 0.022 0.000 0.810
Duration of illness in days 0.070 0.005 0.443
Presence of fever -0.084 0.007 0.355
Presence of vomiting -0.116 0.013 0.204
Presence of grunting 0.070 0.005 0.440
Previous admission 0.060 0.004 0.515
Indoor cooking -0.004 0.000 0.969
Breast feeding 0.071 0.005 0.435
Pulse rate (count per minute) 0.115 0.013 0.207
Respiratory rate (count per minute) -0.048 0.002 0.599
Temperature in Celsius 0.117 0.014 0.198
Saturation of Oxygen (SaPO2) 0.031 0.001 0.772
Presence of cyanosis -0.047 0.002 0.611
Finding of crepitation crepitation -0.029 0.001 0.755
Finding of wheezing -0.016 0.000 0.862
Level of consciousness 0.123 0.015 0.178
White blood cell count (cells/mm3 ) -0.140 0.020 0.126
Platelet count (cells/mm3 ) -0.141 0.020 0.122
Hemoglobin (gm/dL) -0.249 0.062 0.006
Height-for-age Z score 0.203 0.041 0.027
Weight-for-height Z score -0.049 0.002 0.605
Weight-for-age Z score 0.108 0.012 0.241
Body Mass Index-for-Age Z-score -0.035 0.001 0.703
CXR positive (based on WHO criteria) 0.002 0.000 0.979

Table 4: Results of multiple linear regression analysis of predictors of duration of hospitalization among children with severe pneumonia,
Hawassa, Ethiopia.

Duration of hospitalization≠
Variable
Beta T p-value
Age in months 0.051 0.544 0.588
Temperature in Celsius 0.098 1.059 0.292
Loss of consciousness -0.052 -0.587 0.558
White blood cell counts per mL -0.129 -1.420 0.159
Platelet counts per mL -0.103 -1.128 0.262
Hemoglobin, g/dL -0.252 -2.727 0.007
Height for ages z-score 0.176 1.862 0.065
=Adjusted
̸ R2 = 0.081; p-value = 0.022.

The study has several strengths including the prospective In conclusion, this study revealed that the level of
design and the high burden setting. However, the results agreement between the trained radiologists in interpreting
could have been more representative if more than one center CXR findings of children with a clinical diagnosis of severe
were included, especially in an effort to address district pneumonia was alarmingly low, particularly for commonly
and primary hospitals which might have different patient expected findings in cases of pneumonia. There is a low
populations. Due to the low mortality rate, we did not sensitivity of CXR for the diagnosis of childhood pneumonia.
have sufficient power to evaluate the association between Interventions such as training, preparation of guidelines
chest radiographic evidence of pneumonia classification and and other measures to build capacity for competence in
mortality as an outcome in multivariable analysis. interpretation of pediatric CXRs would be anticipated to
BioMed Research International 7

Table 5: Agreement between senior radiologists in reading chest x-ray films of children with severe pneumonia, Hawassa, Ethiopia.

R1 R2 Concordance‡ R3
Finding Kappa∗ p-value†
N (%) N (%) N (%) Total n/ N, %
Film quality, %Adequate 91 (74.6) 85 (69.7) 103 (85.5) 28/43, 65.1 0.634 < 0.001
Consolidation present 27 (22.1) 12 (9.8) 101 (84.2) 13/41, 31.7 0.435 < 0.001
Infiltration present 20 (16.4) 13 (10.7) 99 (82.5) 6/41, 14.6 0.267 0.008
Haziness observed 12 (9.8) 3 (2.5) 111 (92.5) 2/41, 4.9 0.375 0.001
Effusion seen 3 (2.5) 3 (2.5) 118 (98.3) 0/41, 0 0.658 0.001
Atelectasis detected 0 0 120 (100) - - -
Fibrosis present 0 0 120 (100) - - -
Pleural thickening seen 0 0 120 (100) - - -
Hyperinflation observed 1 (0.8) 0 120 (100) - - -
WHO Diagnosis of
58 (47.5) 30 (24.6) 86 (71.7) 21/41, 51.2 0.395 ≤ 0.001
abnormal CXR
†Fisher’s exact test; CXR, chest X-ray; ∗agreement between R1 and R2; ‡concordance between R1 and R2.

improve radiologic diagnosis at our facility. Hemoglobin lev- from a point prevalence survey, 2011 to 2012,” Eurosurveillance,
els were found to be an independent predictor of prolonged vol. 23, no. 32, 2018.
hospitalization in our study, which suggests a possible role for [2] EDHS, “Ethiopian Demographic and Health Survey,” https://
hemoglobin analysis to predict outcome in our setting. dhsprogramcom/pubs/pdf/FR328/FR328pdf, 2016.
[3] S. N. Grief and J. K. Loza, “Guidelines for the evaluation
and treatment of pneumonia,” Primary Care: Clinics in Office
Data Availability Practice, vol. 45, no. 3, pp. 485–503, 2018.
The data used to support the findings of this study are [4] M. Maheshwari and S. Maheshwari, “Clinico-radiological pro-
available from the corresponding author upon request. file and outcome of novel H1N1-infected patients during 2009 to
2014 pandemic at tertiary referral hospital in Rajasthan,” Journal
of the Association of Physicians of India, vol. 63, no. MAY, pp.
Conflicts of Interest 42–45, 2015.
[5] WHO, Integrated Managment of Childhood Illnesses, WHO,
The authors have no conflicts of interest to declare. Geneva, Switzerland, 2014, http://www.who.int/maternal child
adolescent/documents/IMCI chartbooklet/en/.
Authors’ Contributions [6] H. C. Magree, F. M. Russell, R. Sa’aga et al., “Chest X-ray-
confirmed pneumonia in children in Fiji,” Bulletin of the World
Meiron Hassen conceived the study, drafted the proposal, Health Organization, vol. 83, no. 6, pp. 427–433, 2005.
and was involved in all implementation stages of the [7] WHO, Standardization of Interpretation of Chest Radiographs
project. Alemayehu Toma, Freshwork Ayalew, Abel Gedefaw, for the Diagnosis of Pneumonia in Children, World Health
and Birkneh Tilahun Tadesse reviewed the proposal and Organization Pneumonia Vaccine Trial Investigators Group,
were involved in all implementation stages of the project Geneva, Switzerland, 2001, WHO document WHO/V&B/01.35.
and write-up. All authors reviewed the proposal and the [8] T. Cherian, E. K. Mulholland, J. B. Carlin et al., “Standardized
final manuscript. All authors approved final version of the interpretation of paediatric chest radiographs for the diagnosis
manuscript. Mulugeta Tesfay and Eyoel Degafu are senior of pneumonia in epidemiological studies,” Bulletin of the World
radiologists and were involved in reading the CXR films. Health Organization, vol. 83, no. 5, pp. 353–359, 2005.
[9] N. Fancourt, M. D. Knoll, B. Barger-Kamate et al., “Standardized
interpretation of chest radiographs in cases of pediatric pneu-
Acknowledgments monia from the PERCH study,” Clinical Infectious Diseases, vol.
64, pp. S253–S261, 2017.
We would like to thank Hawassa University for supporting [10] N. Mahomed, N. Fancourt, J. de Campo et al., “Preliminary
the study. We would also like to thank the participants of the report from the World Health Organisation chest radiography
study and all data collectors who were involved in collecting in epidemiological studies project,” Pediatric Radiology, vol. 47,
data. We are grateful to the radiologists who took their time no. 11, pp. 1399–1404, 2017.
to read all the CXR films. [11] WHO, Standardization of interpretation of chest radio-
graphs for the diagnosis of pneumonia in children, 2001, http://
References appswhoint/iris/bitstream/handle/10665/66956/WHO V and
B 0135pdf;jsessionid=DADB8F5DE73BC588D17BFD88780B8021?
[1] J. Walter, S. Haller, C. Quinten et al., “Healthcare- sequence=1.
associated pneumonia in acute care hospitals in European [12] M. Soudack, S. Plotkin, A. Ben-Shlush et al., “The added
Union/European Economic Area countries: an analysis of data value of the lateral chest radiograph for diagnosing community
8 BioMed Research International

acquired pneumonia in the pediatric emergency department,” [28] G. Enwere, Y. B. Cheung, S. M. A. Zaman et al., “Epidemiology
Israel Medical Association Journal, vol. 20, no. 1, pp. 5–8, 2018. and clinical features of pneumonia according to radiographic
[13] J. K. Leyenaar, C. B. Andrews, E. R. Tyksinski, E. Biondi, findings in Gambian children,” Tropical Medicine & Interna-
K. Parikh, and S. Ralston, “Facilitators of interdepartmental tional Health, vol. 12, no. 11, pp. 1377–1385, 2007.
quality improvement: a mixed-methods analysis of a collab- [29] F. Pervaiz, M. A. Chavez, L. E. Ellington et al., “Building a
orative to improve pediatric community-acquired pneumonia prediction model for radiographically confirmed pneumonia in
management,” BMJ Quality & Safety, 2018. peruvian children,” CHEST, vol. 154, no. 6, pp. 1385–1394, 2018.
[14] S. Andronikou, P. Goussard, and E. Sorantin, “Computed [30] J. C. Buñuel Álvarez, C. Vila Pablos, J. Heredia Quiciós et
tomography in children with community-acquired pneumo- al., “Usefulness of physical examination at a primary health
nia,” Pediatric Radiology, vol. 47, no. 11, pp. 1431–1440, 2017. centre to diagnose infant pneumonia caught in the community,”
[15] G. Iorio, M. Capasso, S. Prisco et al., “Lung Ultrasound Atencion Primaria, vol. 32, no. 6, pp. 349–354, 2003.
Findings Undetectable by Chest Radiography in Children with [31] G. Xavier-Souza, A. L. Vilas-Boas, M.-S. H. Fontoura et al., “The
Community-Acquired Pneumonia,” Ultrasound in Medicine & inter-observer variation of chest radiograph reading in acute
Biology, vol. 44, no. 8, pp. 1687–1693, 2018. lower respiratory tract infection among children,” Pediatric
[16] E. Urbankowska, K. Krenke, Ł. Drobczyński et al., “Lung Pulmonology, vol. 48, no. 5, pp. 464–469, 2013.
ultrasound in the diagnosis and monitoring of community [32] M. H. Rozenbaum, M.-J. J. Mangen, S. M. Huijts, T. S. van
acquired pneumonia in children,” Respiratory Medicine, vol. 109, der Werf, and M. J. Postma, “Incidence, direct costs and
no. 9, pp. 1207–1212, 2015. duration of hospitalization of patients hospitalized with com-
[17] K. K. Tan, D. A. Dang, K. H. Kim et al., “Burden of hospitalized munity acquired pneumonia: A nationwide retrospective claims
childhood community-acquired pneumonia: A retrospective database analysis,” Vaccine, vol. 33, no. 28, pp. 3193–3199, 2015.
cross-sectional study in Vietnam, Malaysia, Indonesia and the [33] D. J. Williams, Y. Zhu, C. G. Grijalva et al., “Predicting severe
Republic of Korea,” Human Vaccines & Immunotherapeutics, pneumonia outcomes in children,” Pediatrics, vol. 138, no. 4,
vol. 14, no. 1, pp. 95–105, 2018. 2016.
[18] J. Ning, X. Shao, Y. Ma, and D. Lv, “Valuable hematological [34] S. Jain, D. J. Williams, S. R. Arnold et al., “Community-acquired
indicators for the diagnosis and severity assessment of Chinese pneumonia requiring hospitalization among U.S. children,” The
children with community-acquired pneumonia,” Medicine, vol. New England Journal of Medicine, vol. 373, no. 24, pp. 835–845,
95, no. 47, p. e5452, 2016. 2015.
[19] M. I. Neuman, M. Hall, S. C. Lipsett et al., “Utility of blood [35] A. M. Bramley, C. Reed, L. Finelli et al., “Relationship between
culture among children hospitalized with community-acquired body mass index and outcomes among hospitalized patients
pneumonia,” Pediatrics, vol. 140, no. 3, p. e20171013, 2017. with community-acquired pneumonia,” The Journal of Infec-
[20] L. Ambroggio, C. Mangeot, E. Murtagh Kurowski et al., tious Diseases, vol. 215, no. 12, pp. 1873–1882, 2017.
“Guideline adoption for community-acquired pneumonia in [36] M. C. Tracy and R. Mathew, “Complicated pneumonia,” Current
the outpatient setting,” Pediatrics, vol. 142, no. 4, p. e20180331, Opinion in Pediatrics, vol. 30, no. 3, pp. 384–392, 2018.
2018.
[21] N. Principi and S. Esposito, “Biomarkers in pediatric
community-acquired pneumonia,” International Journal of
Molecular Sciences, vol. 18, no. 2, article no. 447, 2017.
[22] K. K. Yadav and S. Awasthi, “The current status of community-
acquired pneumonia management and prevention in children
under 5 years of age in India: a review,” Therapeutic Advance in
Infectious Disease, vol. 3, no. 3-4, pp. 83–97, 2016.
[23] J. M. Ooi, K. P. Eg, K. Chinna, A. M. Nathan, J. A. de Bruyne,
and S. Thavagnanam, “Predictive risk factors for complicated
pneumonia in Malaysian children,” Journal of Paediatrics and
Child Health, 2018.
[24] G. Voiriot, M. Chalumeau, J. Messika et al., “Risques associés
à la prise d’anti-inflammatoires non stéroı̈diens au cours de la
pneumonie,” Revue des Maladies Respiratoires, vol. 35, no. 4, pp.
430–440, 2018.
[25] WHO, Multicentre Growth Reference Study Group. WHO
Child Growth Standards: Length/height-for-age, weight-for-age,
weight-for-length, weight-for-height and body mass index-for-
age: Methods and development, World Health Organization,
Geneva, Switzerland, 2006.
[26] R. B. Wolf, K. Edwards, C. G. Grijalva et al., “Time to clinical
stability among children hospitalized with pneumonia,” Journal
of Hospital Medicine, vol. 10, no. 6, pp. 380–383, 2015.
[27] G. Rathman, J. Sillah, P. C. Hill et al., “Clinical and radiological
presentation of 340 adults with smear-positive tuberculosis in
The Gambia,” The International Journal of Tuberculosis and Lung
Disease, vol. 7, no. 10, pp. 942–947, 2003.
 MEDIATORS of

INFLAMMATION

The Scientific Gastroenterology Journal of

World Journal
Hindawi Publishing Corporation
Research and Practice
Hindawi
Hindawi
Diabetes Research
Hindawi
Disease Markers
Hindawi
www.hindawi.com Volume 2018
http://www.hindawi.com
www.hindawi.com Volume 2018
2013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal of International Journal of

Immunology Research
Hindawi
Endocrinology
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Submit your manuscripts at

www.hindawi.com

BioMed
PPAR Research
Hindawi
Research International
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Journal of
Obesity

Evidence-Based
Journal of Stem Cells Complementary and Journal of
Ophthalmology
Hindawi
International
Hindawi
Alternative Medicine
Hindawi Hindawi
Oncology
Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2013

Parkinson’s
Disease

Computational and
Mathematical Methods
in Medicine
Behavioural
Neurology
AIDS
Research and Treatment
Oxidative Medicine and
Cellular Longevity
Hindawi Hindawi Hindawi Hindawi Hindawi
www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018