Assessing the epidemiology and seasonality of influenza among children under two hospitalized in Amman, Jordan, 2010‐2013

Abstract Background The disease burden of influenza‐associated hospitalizations among children in Jordan is not well established. We aimed to characterize hospitalizations attributed to influenza in a pediatric population. Methods We conducted a cross‐sectional study from our viral surveillance cohort in children under 2 years hospitalized with acute respiratory symptoms and/or fever from March 2010 to March 2013. We collected demographic and clinical characteristics, and calculated the frequency of children who met the severe acute respiratory illness (SARI) criteria. Nasal specimens were tested using real‐time reverse transcriptase polymerase chain reaction to detect influenza A, B, or C. Further subtyping for influenza A‐positive isolates was conducted. Results Of the 3168 children enrolled in our study, 119 (4%) were influenza‐positive. Influenza types and subtypes varied by season but were predominantly detected between December and February. Codetection of multiple respiratory pathogens was identified in 58% of children with the majority occurring among those <6 months. Bronchopneumonia and rule‐out sepsis were the most common admission diagnoses, with influenza A accounting for over 2/3 of children with a rule‐out sepsis admission status. One‐third of children under 6 months compared to 3/4 of children 6‐23 months met the SARI criteria. Conclusions Influenza was an important cause of acute respiratory illness in children under 2 years. Children <6 months had the highest burden of influenza‐associated hospitalizations and were less likely to meet the SARI global surveillance case definition. Additional surveillance is needed in the Middle East to determine the true influenza burden on a global scale.


| BACKG ROU N D
Acute respiratory infections (ARI) are the leading cause of morbidity and mortality of children under 5 years outside the neonatal period. 1 The majority of ARIs are viral in origin, with influenza substantially contributing to outpatient and emergency department visits, hospitalizations, and even deaths in these children. 1 Influenza is a vaccine-preventable disease and impacts all age-groups, with the highest risk of influenza-related complications in children under 2 years, adults older than 65 years, pregnant women, and individuals with underlying medical conditions. 2 Worldwide, although influenza hospitalizations among the pediatric population are known to inflict a large burden, the exact number of global cases and hospitalizations attributed to influenza is not well described. 1,3 In 2008, a systematic review estimated the global incidence of influenza in children under 5 years to be 90 million cases, with approximately one million severe cases. 1 The uncertainties of global pediatric influenza-associated hospitalization burden estimates may be posited to the lack of standard worldwide reporting and testing. 4 In response to the 2009 H1N1 pandemic and in an attempt to overcome the influenza surveillance gap, the World Health Organization (WHO) launched an initiative in 2011 to develop global standards for influenza surveillance, including a global case definition of severe influenza. 5 The case definition was intended to capture hospitalizations related to influenza and is known as severe acute respiratory infection (SARI), defined as an acute respiratory illness with a measured temperature of ≥38° Celsius and cough, with illness onset within the past 10 days, and hospitalization. 5 Although a standardized approach for influenza surveillance has been developed, the epidemiology in many parts of the world, including the Middle East North Africa (MENA) region, remains unclear, especially among young children. 1,3,4,6,7 Since 2007, Jordan has participated in sentinel-site surveillance with the Eastern Mediterranean Acute Respiratory Infection Surveillance (EMARIS) network. 8 One study conducted in Jordan identified that 9% of all patients who met the SARI case definition were influenza-positive, with 3% of influenza-associated deaths, of which all occurred among pediatric patients. 4 Additional influenza research has been conducted in Jordan, but many of these studies have only included individuals who met the SARI case definition and have not solely focused on a pediatric population. 4 Therefore, our study aimed to evaluate and describe the epidemiology, seasonality, and clinical characteristics of influenza-associated hospitalizations in Jordanian children under 2 years, who presented to a large government hospital with fever and/or respiratory symptoms over three full respiratory seasons.

| Study design
From March 13, 2010, to March 31, 2013, we conducted a prospective year-round viral surveillance study of children <2 years who were hospitalized with acute respiratory symptoms and/or fever within 48-hours of hospitalization at Al-Bashir Hospital in Amman, Jordan. 9 Enrollment occurred Sunday through Thursday and children with chemo-associated neutropenia and/or newborns never discharged from the hospital were excluded from the study (detailed inclusion/exclusion criteria are previously published). 9 Written informed consent was obtained from parents or legal guardians prior to enrollment into our study. 9 The study was approved by the Institutional Review Boards at the University of Jordan, the Jordan Ministry of Health, and Vanderbilt University. 9

| Study site
During the study period, Al-Bashir Hospital had a total of 185 pediatric beds (120 pediatric and 65 neonatal intensive care unit) and 11 230 hospitalizations among children <2 years. 7,[9][10][11] Al-Bashir is one of three major government run hospitals that services the population of Amman (capital and largest city in Jordan [>2 million persons]), with over 60% of the pediatric care occurring at this hospital. 7,[9][10][11] As part of the government policy, all Jordanian children under 6 years are provided no-cost medical care, regardless of insurance status, at Al-Bashir Hospital. 7,[9][10][11]

| Data and specimen collection
After obtaining informed consent, trained research personnel collected nasal and throat swabs from all enrolled children. 7,[9][10][11] Parents/ guardians were interviewed to obtain the child's demographic characteristics and medical and social histories using a standardized questionnaire. 7,[9][10][11] All interviews were conducted in Arabic using a standardized case report form and transcribed into English. 7,[9][10][11] After children were discharged, medical records were abstracted for the following: oxygen use, intensive care unit (ICU) stay, mechanical ventilation, length of stay in the hospital, and discharge status. 10 Complete details on the methods of data collection are explained in a previous publication. 10 We inputted and stored all data in a secure REDCap™ (Research Electronic Data Capture, Vanderbilt University, Nashville, TN, USA) database. 9,12 Data quality checks were performed on a minimum of 10% of the charts, and data from all case report forms were verified after entry. 7,9

| SARI criteria/case definitions
We categorized children into two groups: (a) children who met SARI criteria and (b) children that did not meet SARI criteria. Qualifying characteristics were extracted from the interview-derived questionnaires and medical chart abstractions. The fever component of SARI was met if the child had one of the following: self-reported history of fever during current illness, temperature of ≥38° Celsius recorded at admission, and/or an admission or discharge diagnosis of fever. 10 Children were recorded to meet the cough component if it was selfreported as a symptom and/or was recorded as an admission or discharge diagnosis. Illness duration was captured on the standardized questionnaire, and the duration component of SARI was met if the child had illness duration of 10 days or less at enrollment. 10 Intensive care unit stay included children who were transferred to the ICU during the admission or were admitted directly. Children were categorized as rule-out sepsis (ROS) if they had the admission diagnosis of "rule-out sepsis" or "febrile neonate". 7

| Data analysis
Descriptive statistics are reported as frequency or median and interquartile range where appropriate. We used Pearson chi-square and Fisher's exact tests to compare categorical variables and two-sample t tests allowing for unequal variances for continuous variables.
Seasonality and trends of influenza type and subtype are evaluated using an epidemiologic curve by the date of specimen collection.
All analyses were conducted using statistical software StataIC 16.0 (StatCorp LLC).

| Study population and demographics
From March 2010 to March 2013, we identified 3793 children eligible for enrollment; 618 (16%) children had parent/guardian refuse to study participation, three were deemed ineligible after enrollment due to being older than 2 years, and four children had a diagnosis of meningitis. 7 Our final cohort included 3168 children, of which 119 (4%) were influenza-positive.

| Demographics and clinical characteristics
The most common symptoms reported were cough, fever, wheezing, and shortness of breath (SOB); children primarily had an admission diagnosis of either bronchopneumonia or ROS (Table 1). Compared to influenza-negative children, influenza-positive children were older and more likely to present with fever. Overall, influenza-positive children were less likely to be administered oxygen, but had a higher proportion of death, but these were not statistically significant (Table 1). Of those children who died, two had influenza A H3N2 and one had influenza A H1N1pdm09. Interestingly, two of the children had a serious comorbid condition, osteogenesis imperfecta. Of note, only six children were reported to have influenza vaccination and none were influenza-positive.

| Codetection with other viruses
Among influenza-positive children, 58% had additional co-pathogens detected. Children with influenza codetection with other respiratory viruses were more likely to be male compared to children with influenza detection alone (Table 1). Whereas children with only influenza detection were more likely to present with fever, they were less likely to have SOB. Interestingly, no children with influenza detection alone had an admitting diagnosis of bronchiolitis (Table 1).

| Influenza types and seasonality
Throughout our study, influenza types and subtypes varied by season; however, the majority of influenza-positive cases were detected in the months of December through February (Figure 1). The majority of the cases were influenza A (H1N1pdm09 =35; H3N2 =32; unable to subtype =4), followed by influenza B (n =28), influenza C

| Clinical presentation, codetection, and admission diagnosis stratified by age
We stratified influenza-positive children into three age-groups: under 6 months; 6-11 months; and 12-23 months. Compared to the 6-11 months age-group, children under 6 months with influenza were more likely to have a history of breastfeeding, but less likely to be premature and had higher birthweight ( Table 2). In addition, they were less likely to present with SOB and wheezing and had longer hospital stay. Compared to the 12-23 months of age-group, children under 6 months with influenza were more likely to be male and less likely to have an underlying medical condition and present with vomiting. When compared to both other age-groups, the under 6 months children were less likely to have fever, cough, and meet the   SARI case definition (Table 2). In addition, all three mortalities were in children under 6 months.  In our study, fever was an important clinical criterion with influenza-positive children, particularly in children 6-23 months.

| D ISCUSS I ON
Specifically, fever seems to be a unique component of the presentation to children who only had influenza detected. While cough was another common presenting symptom, it was much less common in children under 6 months was significantly more commonly detected in children with codetection. Moreover, the combination of fever and cough together was rarely seen in these young children. However, both cough and fever plus cough were also less common in children younger than 6 months. Rhinorrhea was much more common in this study (83% compared to 1.7% in our study) but this could be attributed to the wider age range enrolled in this study. In another similar study from China in 480 children with influenza who were <15 years old and hospitalized for ARI, fever was a common presenting symptom in children under 24 months but was less common in younger infants. 16 Therefore, the inclusion of fever or cough for viral surveillance is important but both should not be required for surveillance studies that are trying to estimate the exact influenza burden to avoid underestimation of influenza in young children.

ACK N OWLED G EM ENTS
We would like to thank all the doctors at Al Bashir Hospital and The University of Jordan for their collaboration in this surveillance project. In addition, we would like to thank our research re- Writing-review & editing (supporting).

PEER R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/irv.12813.