Molecular detection of respiratory pathogens among children aged younger than 5 years hospitalized with febrile acute respiratory infections: A prospective hospital‐based observational study in Niamey, Niger

Abstract Background and Aims In Niger, acute respiratory infections (ARIs) are the second most common cause of death in children aged younger than 5 years. However, the etiology of ARI is poorly understood in the country. This study aims to describe viral and bacterial infections among children aged younger than 5 years hospitalized with febrile ARI at two hospitals in Niamey, Niger's capital city, and the reported clinical procedures. Methods We conducted a prospective study among children aged younger than 5 years hospitalized with febrile ARI at two national hospitals in Niamey between January and December 2015. Clinical presentation and procedures during admission were documented using a standardized case investigation form. Nasopharyngeal specimens collected from each patient were tested for a panel of respiratory viruses and bacteria using the Fast Track Diagnostic 21 Plus kit. Results We enrolled and tested 638 children aged younger than 5 years, of whom 411 (64.4%) were aged younger than 1 year, and 15 (2.4%) died during the study period. Overall, 496/638 (77.7%) specimens tested positive for at least one respiratory virus or bacterium; of these, 195 (39.3%) tested positive for respiratory viruses, 126 (25.4%) tested positive for respiratory bacteria, and 175 (35.3%) tested positive for both respiratory viruses and bacteria. The predominant viruses detected were respiratory syncytial virus (RSV) (149/638; 23.3%), human parainfluenza virus (HPIV) types 1 to 4 (78/638; 12.2%), human rhinovirus (HRV) (62/638; 9.4%), human adenovirus (HAV) (60/638; 9.4%), and influenza virus (INF) (52/638; 8.1%). Streptococcus pneumoniae (249/638; 39.0%) was the most frequently detected bacterium, followed by Staphylococcus aureus (112/638; 12.2%) and Haemophilus influenzae type B (16/638; 2.5%). Chest X‐rays were performed at the discretion of the attending physician on 301 (47.2%) case patients. Of these patients, 231 (76.7%) had abnormal radiological findings. A total of 135/638 (21.2%) and 572/638 (89.7%) children received antibiotic treatment prior to admission and during admission, respectively. Conclusion A high proportion of respiratory viruses was detected among children aged younger than 5 years with febrile ARI, raising concerns about excessive use of antibiotics in Niger.


| INTRODUCTION
Acute respiratory infections (ARIs) are responsible for elevated childhood morbidity and mortality globally, 1,2 accounting for approximately four million deaths among children aged younger than 5 years in 2010 and resulting in substantial burden of healthcare systems. 3,4 In developing countries, ARIs account for 19% of all deaths among children aged younger than 5 years and 8.2% of all disabilities and premature deaths. 5,6 Therefore, data on the epidemiology and seasonality of ARI are important to develop control and prevention strategies. 7 In Niger, the estimated mortality rate among children aged younger than 5 years was 114 per 1000 in 2012 (UNICEF child mortality estimate), and more than 80% of these deaths were associated with malaria, respiratory infections, and diarrhea.
According to the 2012 Niger health statistics yearbook, the estimated case fatality proportion associated with respiratory infections was 8.7%.
In the absence of laboratory diagnosis, it is difficult to clinically differentiate between ARI-related causative pathogens, due to similarities of symptoms. 8,9 The main causative agents of ARI are viruses and bacteria. 10 [11][12][13] In Niger, there is scarcity of data on ARIs, although a routine influenza surveillance system exists. INF has been detected in 11% of samples from children aged younger than 5 years hospitalized with severe acute respiratory illness (SARI). 14  This study aims to describe the viral and bacterial infections among children aged younger than 5 years hospitalized with febrile ARI at two national hospitals of Niamey, the capital city of Niger, and the reported clinical procedures.

| Study design and setting
We conducted a prospective study among children aged younger than 5 years hospitalized with febrile ARI between January and December 2015. Niger has four distinct seasons as categorized by the national directorate of meteorology: the cold season (mid-December to mid-February), the dry season (mid-February to May), the rainy season (June to September), and the hot season (October to mid-December). 15 This study was part of the "TOTAL Niger Infection Respiratoire Aiguë" (TONIRA) Project, which was funded by the TOTAL Corporate Foundation in order to strengthen medical care of febrile ARI among children aged younger than 5 years. The study was conducted at the pediatric wards of two national tertiary hospitals situated in Niamey, namely, the Hôpital National de Niamey (HNN) and the Hôpital National Lamordé (HNL). These two hospitals provide general health care to the population of Niamey estimated at 1.1 million in 2015, as well as to patients referred from across the country. Laboratory detection of respiratory viruses and bacteria were conducted at the Centre de Recherche Médicale et Sanitaire (CERMES), Niamey, Niger (the National Reference Laboratory for influenza).
Demographic and clinical data, as well as the use of antibiotics before and during admission, were collected using standard data collection forms. Malaria testing was implemented on site using thick blood film microscopy. Chest X-rays were performed at the discretion of the attending physician, and typical abnormal radiological findings assessed included pulmonary opacities, acute bronchiolitis, and enlargement of intercostal spaces. Malnutrition was determined using the ratio of weight for height, and its level was classified as moderate or severe by the attending physician using national standards based on the 2010 Technical Guidelines for Integrated Disease Surveillance and Response in the African Region (Retrieved from https://www. afro.who.int/sites/default/files/2017-06/IDSR-Technical-Guidelines_ Final_2010_0.pdf). Febrile ARI in conjunction with malaria, malnutrition, and diarrhea were assessed during admission using a standardized case investigation form. In-hospital outcome (ie., discharge, referral, or death) was recorded for all enrolled patients.

| Case definition
A febrile ARI case was defined as a hospitalized child aged younger than 5 years with onset of fever 38 C or higher and cough within 10 days prior to admission and at least one of the following signs: inability to drink or breastfeed, lethargy, vomiting, convulsions, nasal flaring, chest indrawing, stridor in a calm child, or tachypnea.

| Sampling and laboratory testing
Nasopharyngeal swabs were collected from all enrolled patients, placed in universal transport medium, stored at 4 to 8 C, and trans-  amplification. The fluorescence was assessed at the amplification step. The positive and negative virus plasmid controls provided in the kit were included in all runs to monitor assay performance. 16

| Statistical analysis
Viral and/or bacterial detection was reported as percentage positive, overall and within selected categories (eg, age groups and seasons).
Children aged younger than 5 years were stratified into two age groups: infants (aged younger than 1 year) and young children (aged 1-4 years), for comparability with other studies from Africa. [17][18][19][20] In addition, the majority of enrolled children were aged younger than 1 year hindering our ability to categorize age in smaller age groups.
Stata version 14.2 (StataCorp, College Station, Texas, USA) was used for the analysis.

| Ethical approval
The protocol was approved by the National Ethical Consultative Com-

| Detection of viral and bacterial agents
Overall, 496/638 (77.7%) specimens tested positive for at least one respiratory virus or bacterium (Table 1)
Of the specimens that tested positive for at least one respiratory bacteria, two or more respiratory bacteria were detected in 81/220 S pneumoniae and S aureus were mostly detected during the rainy season (June to September) ( Table 3 and Figure 3B).

| DISCUSSION
We report the detection of respiratory viruses and bacteria among children aged younger than 5 years hospitalized with febrile ARI in T A B L E 2 (Continued)   21,22 and Asia (range 45%-65%). [23][24][25] The HPIV were also the most predominant viruses detected in children aged younger than 5 years in a previous study conducted in Niger. 8 RSV is considered to be a major cause of ARI in children aged younger than 5 years, 26,[28][29][30] and it was the most commonly detected virus also in this study (149/638; 23.3%).

Characteristics
Among the cases in which HPIV and HCoV were detected, HPIV type 3 and HCoV type OC43 were the predominant types, and this is consistent with previous studies. 27,31 In this study, INFs were detected in 8.5% of the cases, which is in agreement with findings from the national influenza surveillance 14   Our study presents some limitations. First, the study spanned over a period of only 12 months, limiting our ability to fully assess the temporal circulation pattern of the investigated agents. Due to the small sample size, we may also have been underpowered to detect small variations in the temporal distribution of the investigated agents. Second, attribution of causality remains challenging due to the lack of controls in our study. The association of pathogen detection with illness could not be well assessed, although most of the viral and bacterial pathogens identified in this study have been described in previous case control studies as causative agents of ARI. 10,39 In addition, we did not collect blood samples to assess whether the detection of common bacterial colonizers of the nasopharynx, such as S pneumoniae, were associated with invasive disease. Third, we did not record the number of patients with febrile ARI that refused enrollment, hindering our ability to estimate underenrollment. Last, given that only patients with febrile ARI were enrolled, this could have resulted in an underestimation of the RSV burden since a large proportion of RSV-related illness is not associated with fever. 40

| CONCLUSIONS
In this 1-year prospective study, both viral and bacterial pathogens were detected in high proportion among hospitalized children aged younger than 5 years with febrile ARI in Niamey, Niger. However, further investigations should be carried out to assess risk factors and association of pathogens with illness. Although national guidelines for ARI treatment recommend systematic empiric antibiotic therapy, our results suggest that antibiotic use might be unnecessary in most cases, given the predominance of viral infections as potential cause of febrile ARI. Furthermore, other predisposing diseases such as malaria, malnutrition, and diarrhea may be contributory factors to febrile ARI exacerbation among hospitalized children.

DISCLAIMER
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention, USA.

DATA AVAILABILITY STATEMENT
The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.