Demographic and seasonal characteristics of respiratory pathogens in neonates and infants aged 0 to 12 months in the Central‐East region of Tunisia

Background This study aimed to characterize the epidemiology of pathogenic respiratory agents in patients aged 0 to 12 months and hospitalized for acute respiratory infections in Tunisia between 2013 and 2014. Methods A total of 20 pathogens, including viruses, Mycoplasma pneumoniae, and Streptococcus pneumoniae, were detected using molecular sensitive assays, and their associations with the patient’s demographic data and season were analyzed. Results Viral infectious agents were found in 449 (87.2%) of 515 specimens. Dual and multiple infectious agents were detected in 31.4% and 18.6% of the samples, respectively. Viral infection was predominant in the pediatric environment (90.8%, P < 0.001), male patients (88.0%), and spring (93.8%). Rhinovirus was the most detected virus (51.8%) followed by respiratory syncytial virus A/B (34.4%), coronavirus group (18.5%), adenovirus (17.9%), and parainfluenza viruses 1‐4 (10.9%). Respiratory Syncytial virus A/B was significantly associated with gender (38.0% male cases vs 28.3% female cases, P = 0.02). Infections by Adenovirus, Bocavirus, and Metapneumovirus A/B increased with increasing age of patients (predominated cases aged 6‐12 months, P < 0.001). S. pneumoniae was detected in 30.9% of th tested samples. In 18.2% of the negative viral infections, only S. pneumoniae was identified. Conclusion A predominance of the rhinovirus infection was observed in this study. Coronavirus subtypes were described for the first time in Tunisia. The observed different pathogenic profiles across age groups could be helpful to avoid the misclassification of patients presenting with ARIs at the triage level when no standardized protocol is available. This study will provide clues for physicians informing decisions regarding preventive strategies and medication in Tunisia.


| INTRODUCTION
The World Health Organization estimates that acute respiratory infections (ARIs) are the second leading cause of childhood death, and about 70% of these infections occur in Africa and Southeast Asia. 1 ARIs are responsible for 20% to 40% of children hospitalizations, and their course of infection is more severe in children aged less than 5 years. 2 ARIs are caused by a variety of pathogenic infectious agents including viruses, bacteria, and fungi, of which viruses are the most predominant. The viruses associated with ARIs are classified as "old respiratory viruses" and "novel respiratory viruses." The "old  6 Fungal ARIs in children are rare and are found mainly in immunocompromised patients. 7 The acquisition and spreading of ARIs vary among study populations in different countries. This may be due to children's malnutrition, the low socioeconomic status of the country, cultural habits in the community, patient's immune deficiency, local/regional geographical and climate change, and the variability of the health care systems in public and private hospitals. Consequently, the morbidity related to ARIs is mostly observed in the developing world. 8 On the other hand, the expanding number of pathogens identified as a causative agent of ARIs might be due to the introduction of molecular techniques such as the multiplex real-time polymerase chain reactions (qPCR, qRT-PCR) in routine. These assays have increased the sensitivity and specificity of detection and allow the identification of a wide range of viruses and bacteria in one reaction. This will be helpful for the better understanding of the pathogenesis of ARIs. 9 However, in most developing countries, the application of multiplex assays has not been fully implicated in the routine laboratory testing due to their high costs. As far as the Arab Maghreb is concerned, laboratories perform mostly enzyme immunoassays, rapid diagnostic tests, or latex agglutination techniques for some frequent pathogens, but no routine tests using qPCR assays are generally available. Owing to the absence of detailed diagnostic testing, infections are managed empirically by antibiotic regimens, resulting in an overuse of broadspectrum antibiotics, favoring the development of bacteria (bacterial resistance). 10 In Tunisia, studies on the etiology of a large range of respiratory microbes across the seasons, among various ages and genders-and with the application of molecular and sensitive assays-are limited. A Tunisian study published by el Moussi et al in 2013 reports the distribution of the most common respiratory viruses using a multiplex assay. However, no association between viral infection and sociodemographic situation of patients or the seasonal distribution of detected pathogens was observed by this study. In addition, bacterial identification was excluded. 11 In light of this, in the present study, a total of 18 viruses associated with ARIs were analyzed using a sensitive multiplex molecular assay. In addition, detection of the bacterial genomes M. pneumoniae and S. pneumoniae was performed. Moreover, the demographic characteristics of the detected pathogens, as well as their seasonality, were analyzed. This study aimed to gain a better understanding of the epidemiology of the most common respiratory microbes in the area of Sousse, Tunisia. Such findings will help to reduce the use of unnecessary antibiotics, thus, avoiding bacterial resistance and contributing to the implementation of prevention and control systems in Tunisia.

| Ethics approval
The present study was approved with a formal authorization by the Scientific and Ethical Committee of Farhat Hached University-Hospital (FH-UH) of Sousse, Tunisia. The approval number is IRB 00008931, provided by OHRP.

| Subjects
A total of 515 subjects aged 0 to 12 months and hospitalized for upper and/or lower ARIs in the pediatric (PP) and the neonatology (NN) wards of FH-UH of Sousse, Tunisia, were enrolled. Demographic data, including age, ward of admission, date of hospitalization, season, and gender, were obtained using a full-filled information form. The study population was divided into three age groups. The first group G1 included neonates aged 0 to 28 days and the second and third groups contained infants. The groups of infants were described as follows: group G2 was characterized by patients aged 28 days to 6 months, and the third group G3 comprised subjects aged 6 to 12 months (Table 1).

| Statistical analysis
The statistical calculations were evaluated using the Statistical Package

| Age and gender characteristics of the tested pathogens
The distribution of the pathogenic infectious agents, including viruses and S. pneumoniae, by age and gender is described in Tables 3 and 4 and Supporting Information

| Seasonal distribution of respiratory agents
The  Table 3).

| Single and mixed pathogenic infections
Within subjects tested for viral infection, 37.0% were single positive, 31.4% were double infections, and 18.6% were multiple infections (3-6 detected viruses per samples).

| Single and multiple infections
Among 12/66 of the negative viral infections (18.2%), S. pneumoniae was the single detected pathogen ( Table 4). The possible association between single/double/multiple viral infections and demographic characteristics, seasonal distribution, as well as the individual viral infection is described in Table 5. Male patients were more frequently detected within multiple infections (69.5% of multiple infections).
Infected female subjects were more found in single infections (40.8%  The OR for the category "Jul-Sep 2014" of the variable "season" was left blank as a value of zero was obtained. * Describes the reference groups used for the statistical calculations. ** P value was calculated using the X 2 test or the Fisher's exact test on SPSS. A value of P < = 0.05 was considered as significant and is represented in bold.

| 575
The multiplex qRT-PCR detected a high rate of viral infection, estimated at 87.2%. Comparable studies have shown variable detection rates ranging from less than 50.0% to more than 85.0%. [14][15][16] The wide differences observed in the detection rates of infectious pathogenic agents could be interpreted by the heterogeneity of the study population, the genetic variability between populations, the nature of biological material used for the detection of respiratory pathogens by multiplex assays, the technique itself, and the climate/regional factors that influence the appearance of respiratory agents. 16,17 In this study, RV and RSV A/B were the most frequently detected viruses and were found in 51.8% and 34.4% of tested specimens, respectively. Both are known as the crucial causes of ARIs in children. 18,19 In addition, a male predominance was observed within    MPV A/B was followed by BoV, indicating a 7.1% infection rate.
Infection by BoV has recently attracted attention worldwide because the incidence and respiratory infection of this pathogen vary widely and often involve co/multiple infections with other potential pathogens, 32,33 which is in agreement with the present study.
Members of the Picornaviridae family, PeV, and EV, were screened. Although, PeV is frequently isolated from patients with respiratory infection and is thought to be associated with disease.
The qRT-PCR was shown as the best way to detect an EV or PeV, but still little is known about the prognosis of these viruses. 34

| CONCLUSION
In summary, the high prevalence of etiology associated to respiratory infections of relevant pathogens highlights the importance of sustaining national surveillance of ARIs to clearly estimate the role of associated pathogens and establish the burden of disease. The findings of this study address several important gaps in the literature.

The CoVs and AdV ARIs were frequent in neonates and infants in
Tunisia. The CoVs subtypes were described for the first time in Tunisia and revealed that CoV-229E was the most frequently detected. In spite of the common knowledge that RSV A/B is the most common respiratory agent responsible for ARIs, RV was found to be more prevalent in this population, which highlights the interest of cost benefits studies concerning its implementation in the routine laboratory testing. Given the multiplicity of these findings, this study provides a useful starting point for a better understanding about the circulation of these pathogens in the area of Sousse and the seasonal/ demographic attributable to viral/bacterial infection. It will contribute data relevant to the development of prevention strategies, which will help in clinical decisions and allow considering antiviral therapy in clinical routine. The availability of vaccines in Tunisia only against InfVs infections will afford ongoing vaccine research and development on many other leading viral pathogens in future.