Influenza in children

Excess mortality associated with annual influenza epidemics is highest among persons over 65 y of age, and therefore influenza is often regarded as an illness of the elderly population. Ample evidence indicates, however, that the burden of influenza is also substantial in children, and that children have a central role in the spread of influenza in the community during epidemics. Vaccination of children against influenza could bring about substantial health benefits not only to children themselves but also to persons in other age groups.


Epidemiology
Extensive population-based studies carried out in different geographic areas have shown consistently that the attack rates of influenza during annual epidemics are highest among preschool and schoolaged children in whom the rates of influenza infection may exceed 30% ( Figure 1) [1,2]. Day-care attendance is considered to increase the risk of contracting influenza, and attack rates of up to 50% have been observed in day-care children during regular influenza seasons [3]. In a study of infants followed from birth to 1 y of age, influenza was documented in one third of the infants [4]. During the influenza epidemic of 2000 Á2001, which was generally considered mild or moderate, 19% of prospectively followed children younger than 13 y of age had a symptomatic culture-confirmed influenza, and influenza accounted for more than 7% of all respiratory infections in children during the winter season [5].
In parallel with the age-related incidence of influenza, the frequency of outpatient visits for acute respiratory disease during influenza outbreaks is also highest in children [6,7]. Studies conducted in different countries have demonstrated the considerable socio-economic impact of influenza on children and their families in terms of children's absenteeism from day care or school, and parental work loss because of influenza in children [8,9]. In a prospective study in Finland, the frequency and duration of parental work loss because of child influenza were highest among children younger than 3 y of age, among whom the mean duration of parental absenteeism in families in which a parent had to stay at home was 3.2 d [8].
Although the vast majority of influenza-infected children are treated as outpatients, infants and young children are frequently admitted to hospital for influenza-associated illnesses. Distinguishing influenza from invasive bacterial infections is extremely difficult in infants with high fever as the leading symptom, and children younger than 1 y of age are hospitalized for influenza-attributable illnesses at rates similar to those for adults at high risk of influenza [7,10]. The age-specific rates of hospitalization with acute respiratory disease during influenza epidemics follow a U-shaped curve, and most children ]. In the United States, the admission rate of children younger than 2 y of age without any high-risk conditions was reported to approach that of children 5Á17 y of age with highrisk conditions [14]. The most frequent reasons for hospitalization of children with influenza are lower respiratory tract infections, suspected sepsis, or exacerbation of asthma [15].
In addition to the high attack rates of influenza in children, the duration of viral shedding is more prolonged in children than in adults, persisting for up to 10 Á14 d after the onset of symptoms [16]. The titres of recovered viruses are also generally higher in children [16]. All these features contribute to the current concept that children are the main disseminators of influenza both in the household and the entire community during local outbreaks [2,17]. The central role of children in the transmission of influenza in the community was well demonstrated in a surveillance study in the United States ( Figure 3) [2].
During the early stage of the epidemic, most cases of influenza occurred in children, especially in those aged 5 Á19 y, while the relative proportion of children decreased and that of the adult population increased towards the end of the outbreak. Further, school absenteeism and the numbers of paediatric admissions for pneumonia peaked approximately 2 wk earlier than did industrial absenteeism or adult admissions for pneumonia [2].
Despite the high incidence of influenza and high rates of influenza-associated hospitalizations in children, mortality due to influenza is strikingly low in this age group. Recent estimates of the mortality rates in children during interpandemic outbreaks have ranged between 0.2 and 0.8 per 100 000 person-years [7,18]. However, it is important to appreciate that deaths due to influenza do occur in children, as evidenced by reports of 153 laboratory-confirmed influenza-related deaths in children in the United States during the 2003 Á2004 influenza epidemic [19]. In the age group 2 Á17 y, 70% of influenza-associated  deaths occurred in children who did not have an underlying medical condition that would have placed them at increased risk for influenza-related complications.

Clinical manifestations
The spectrum of influenza illness ranges from an asymptomatic infection to severe life-threatening forms of the disease. Most data on the clinical manifestations of influenza in children are derived from hospitalized patients, which may emphasize the most severe forms of the illness and thus distort our understanding of the clinical presentation in an average child with influenza. It is obvious that hospitalized children represent only a small proportion of all infected children, and the vast majority of children with influenza are treated as outpatients in whom specific aetiological diagnosis is rarely made ( Figure 4). The clinical manifestations of influenza in children are to a great extent similar to those seen in adults, including sudden onset of fever, cough and sore throat [13]. It is presumable that complaints that are typically associated with influenza in adults, such as headache, myalgia, malaise or dizziness, also occur in children, but the presence of these symptoms is difficult to determine in young children whose capability to express themselves is limited. Compared to adults, however, rhinorrhoea and gastrointestinal complaints are more frequent in children [13,16,20]. The duration of illness in children is usually 5 Á8 d [20].
The clinical impact of influenza is not limited to viral infection in the respiratory tract. Influenza frequently gives rise to bacterial complications and also, occasionally, other organ systems may be affected by influenza virus infection. The most common complication of influenza in children is acute otitis media, which occurs in 20% to even 60% of cases, depending on the age of the child and other risk factors involved [8,20 Á23]. Pneumonia and sinusitis are more infrequent complications in children [8,24], but acute exacerbations of asthma are common in asthmatic children with influenza [25]. Influenza infections have also been associated with a higher incidence of febrile seizures than adenovirus or parainfluenza virus infections [26]. In children hospitalized for croup, the course of illness is more severe in children infected with influenza than in those with parainfluenza virus infections [27]. Although viraemia during influenza has been rarely documented, influenza may be associated with complications such as myocarditis, myositis, encephalitis or encephalopathy [16]. Increased numbers of influenzaassociated encephalopathy have recently been reported in young children, especially in Japan [28]. The incidence of Reye's syndrome has decreased substantially since the use of salicylates was banned in children [29].

Diagnosis
The advent of new specific antiviral treatments for influenza has increased the importance of accurate diagnosis of the illness. In addition to potential    initiation of antiviral therapy, positive testing for influenza has been shown to result in decreased use of antibiotics and ancillary testing in children in emergency departments [30]. In adults, the combination of fever and cough during a local influenza outbreak has been reported to predict influenza with over 70% accuracy [31]. In children, however, correct identification of influenza on clinical grounds alone is seriously hampered by the cocirculation of other respiratory viruses during influenza epidemics and the frequent occurrence of fever in children with any viral infection [21,32]. A large study in general practice demonstrated that, in children younger than 5 y of age with influenza-like illness, respiratory syncytial virus was the cause of the illness in 35% of the cases, whereas influenza was detected in only 29% of the children [33]. In a recent study among outpatient children, the overall sensitivity of the clinical diagnosis of influenza made by physicians was 38%, and the positive predictive value 32% [34]. Although culture of influenza viruses remains the gold standard for the diagnosis of influenza, and several other assays based on antigen detection and molecular methods are available for laboratory confirmation of the illness, the optimal use of influenzaspecific antiviral drugs requires a definite diagnosis to be made rapidly at the point of care during the visit to a physician [35]. Several rapid tests for influenza have been developed that can provide results within 15 Á30 min, but the sensitivity and specificity of the currently available tests vary substantially [36,37], and it is obvious that the test results are highly dependent on the quality of the specimen. The optimum site and method of sampling for viral detection have not been determined, and they may also differ between various viruses. However, specimens obtained from the nasopharynx may be superior to throat swabs for the detection of influenza viruses [38]. Nasopharyngeal aspirates or nasal wash specimens are usually considered the specimens of choice for respiratory viruses [39], but the feasibility of these sampling methods may be limited in everyday clinical practice. Collection of a simple nasal swab is likely to be the easiest and most convenient way to obtain a specimen for viral diagnosis, at least in the outpatient setting. Recent studies in children suggest that the sensitivity of nasal swabs for the detection of influenza by viral culture or antigen detection methods is approximately 90% in comparison with nasopharyngeal aspirates [40,41].

Treatment
The management of children with uncomplicated influenza has traditionally consisted of rest and symptomatic treatment only. Although the antiviral drug amantadine has been available since the 1960s for the treatment of influenza, even in young children (and rimantadine in some countries for adolescents), paediatric use of these drugs has been rare. General undervaluation of the burden of influenza in children has probably been the primary reason for the low usage of amantadine in this age group, but other concerns related to this agent include side effects, inefficacy against influenza B viruses, and *most importantly*the rapid development of resistant strains of influenza viruses during treatment [42].
Neuraminidase inhibitors, zanamivir and oseltamivir, are a new class of influenza antivirals (Table I). These agents are effective against both influenza A and B viruses, and the development of resistance against them has been generally low, although a recent study in Japanese children indicated that oseltamivirresistant viruses may arise during treatment more frequently than previously anticipated [42 Á44]. Zanamivir is administered by inhalation using a specific device, which would limit its use in young children even if it were approved for use in them. Zanamivir is generally well tolerated, but some cases of bronchospasm occurring after the inhalation of the drug have been reported [43]. Oseltamivir is an oral drug that is available as both capsules and liquid suspension, and it can be used for treating influenza in children older than 1 y. The most common complaints during oseltamivir treatment are nausea and vomiting, but these side effects are usually mild and subside during the continuation of the treatment, and they can be decreased by taking the drug with food [42,43]. Given the high frequency of acute otitis media as a complication of influenza in young children, it is noteworthy that oseltamivir treatment initiated within 48 h of the onset of influenza symptoms has been reported to decrease the development of acute otitis media by approximately 40% [20].

Prevention
There is a wide consensus that vaccination is the primary approach in the prevention of influenza in all age groups. Among children, annual influenza vaccination is usually recommended for those older than 6 mo of age who are at increased risk for complications from influenza, for instance children with chronic pulmonary, cardiac or metabolic diseases [45]. Influenza vaccine is, however, safe and well tolerated in children, and it can be given to any child 6 mo of age or older to reduce the probability of contracting influenza [46,47]. Accumulating evidence for the substantial burden of influenza on children and the central role of children in the transmission of influenza in the community have initiated an increasing discussion about more widespread vaccination of children against influenza [7,14,48]. Considering the 70 Á 80% protective efficacy of the inactivated influenza vaccine in children [21,46], general vaccination of children could be expected to bring about substantial reductions in the rates of hospitalizations and outpatient visits as well as in the incidence of influenza-associated complications such as acute otitis media [21,49 Á51]. Besides providing protection for children themselves, widespread vaccination of children might also reduce influenza-related morbidity and mortality in other age groups, especially among the elderly population [52Á54]. In the United States and Canada, influenza vaccination is already recommended for all children 6 Á23 mo of age [19,55]. However, because of great differences in the healthcare systems between many countries, vaccination policies adopted in one country are not always directly applicable to other countries, and potential development of new vaccination strategies in any country requires careful evaluation of the risks and benefits, inconveniences, logistics, and cost-effectiveness of the vaccination.
The influenza vaccines currently licensed for use in children are trivalent inactivated subunit vaccines (and inactivated virosomal vaccines in some European countries) [19]. These vaccines are administered intramuscularly, and young children receiving the vaccine for the first time should receive two doses 1 mo apart. Considering the substantial number of injections already included in the routine childhood immunization schedule, the addition of any extra injections might not sound attractive to parents *or to children. Recent advances in influenza vaccination include the development of a trivalent live attenuated, cold-adapted vaccine that is administered intranasally [56]. In addition to eliciting an antibody response in the serum, live intranasal vaccines also induce local IgA production in the nasal mucosa [57]. Clinical trials with the live attenuated influenza vaccine have demonstrated a protective efficacy exceeding 90% against culture-confirmed influenza in children [56], and there is also some evidence for heterotypic protection against strains of influenza not contained in the vaccine [58,59]. It could be anticipated that the intranasal route of vaccine administration might lower the threshold of influenza immunization, especially in children, but because of the concern about potential increased asthma encounters in young children receiving the live intranasal vaccine, this vaccine is currently licensed in the United States only for use in persons 5 Á49 y of age. However, a recent large safety study indicated that the live vaccine is also safe in children aged 18 mo to 4 y [60], and it is therefore possible that the lower age limit for the use of live intranasal influenza vaccine will be reconsidered in the near future.