Epidemiology of respiratory viral infections in children enrolled in a study of influenza vaccine effectiveness

Abstract

Influenza-like illness (ILI) confers a high annual morbidity in young children. We report the epidemiology of ILIs in children who participated in an influenza vaccine effectiveness study during the 2010 Southern Hemisphere influenza season in Sydney, Australia.
Children aged 0·5-3 years were prospectively recruited from child care centres (CCCs). We classified them as fully vaccinated, partially vaccinated and unvaccinated according to their receipt of unadjuvanted vaccines containing influenza A (H1N1)pdm09. For 13 weeks commencing 30 July 2010, parents reported when their children developed an ILI (fever ≥37·8°C/feverishness plus ≥1 respiratory symptom) and collected nose and/or throat swabs for multiplex respiratory virus polymerase chain reaction (PCR) testing. Health impacts were assessed by telephone interview at enrolment and two weeks after each ILI.
There were 124 ILIs reported in 105 of 381 enrolled children. Swabs were taken in 117 ILIs: 175 viruses were identified from 103 swabs. Adeno- and rhinoviruses were most frequently identified; 44% of swabs yielded multiple viruses. No virus was associated with more severe symptoms, although rhinovirus-related ILIs lasted longer. Nose swabs had a higher virus detection rate than throat swabs. Influenza-vaccinated children were 1·6 times (P = 0·001) more likely than unvaccinated children to have a non-influenza ILI.
Adeno- and rhinoviruses were the most common viruses causing ILI. Swabs taken by parents are an effective method for sample collection. Influenza-like illness was more common in children vaccinated against influenza in this observational study, but prior health-seeking behaviour may have contributed to this difference.

Full text

Epidemiology of respiratory viral infections in children enrolled in
a study of influenza vaccine effectiveness
Alexa Dierig, a,b Leon G. Heron, a,c,d Stephen B. Lambert, e,f Jiehui Kevin Yin, a,c Julie Leask, a,c,d Maria Yui Kwan Chow, a,c
Theo P. Sloots, e Michael D. Nissen, e Iman Ridda, c Robert Booy a,c,d
a National Centre for Immunisation Research and Surveillan ce, T he Children’s Hospital at West mead, Westmead, NSW, Australia.
b University Children’s Hospit al bot h Basel, Basel, Switzerland.
c Sydney Medical School, The Universit y of Sydney, Sydney, NSW, Australia.
d Marie Bashir Instit ute, Sydney, NSW, Australia.
e Queensland P aediatric Infectious Disease Laborat ory , Queensland Children’s Medical Research Inst itute, Queensland Children’s Health
Service, Brisbane, Qld, Australia.
f Clinical and Stat ewide Services, P athology Queensland Cent ral, Herston, Qld, Australia.
Correspondence: Alexa Dierig, University Children′s Hospital Basel, Spitalstr. 33 , 4031 Basel, Switzerland. E-mail: alexadierig@web.de
Introduction
Acute respiratory infection (ARI) is among the major causes of death in young children worldwide. 1 In Australia, ARI is
the main cause for short-term illness in children aged 0–14 years. 2 The number of newly identified viruses in respiratory
tract specimens, including the recently discovered polyomaviruses WUV and KIV, 3,4 is increas ing. Influenza causes a
substantial health burden with direct and indirect costs, including hospitalisations and loss of productivity. 5–8
Inactivated and live-attenuated influenza vaccines offer both direct and herd benefits to vaccinated children, their
contacts and the broader community. 9–11 Several studies have shown that children attending child care centres (CCCs)
are at greater risk of ARI including influenza. 12–15 In 2008, formal child care was undertaken by 9% of Australian children
aged <1 year, 35% aged 1 year and 47% aged 2–3 years. 16 In order to determine the health, s ocial and economic effects
of influenza vaccination in young children, we planned a randomised controlled trial (RCT) of an unadjuvanted trivalent
influenza vaccine in children aged 6–35 months who attended a CCC in metropolitan Sydney during 2010. However,
because of the 2009 pandemic, the Australian government recommended and funded universal use of inactivated
pandemic influenza A(H1N1)pdm09 vaccination for those aged >6 months in 2010. Hence, a RCT design became
unethical. The study was restructured to a prospective cohort design addres sing the epidemiology of ILIs among young
children.

Methods
Study cohorts
Children were recruited through 90 CCCs and one general practitioner with a s pecial interest in paediatrics. Informed cons ent was
obtained from a parent or legal guardian. To meet inclusion criteria, children needed to be aged ≥6– 35 months on 1 March 2010.
Exclusions were for known allergy to any component of the influenza vaccine, a history of Guillain–Barre s yndrome, a bleeding
disorder, an unstable chronic illness or enrolment in another trial. Parents reported influenza vaccinations their children had
received and, where possible, the influenza vaccination status was validated from vaccination records . The children were divided
into three cohorts: fully vaccinated (usually two doses in 2010), partially vaccinated (usually one dose in 2010) and unvaccinated
according to their receipt of vaccines that contained influenza A (H1N1)pdm09 – full definitions are in the footnotes to Table 1.
ILI case reporting
We planned to commence ILI case reporting as soon as an upswing in influenza cases in Sydney was recognised through
laboratory surveillance. During the ILI follow-up period, parents were asked to report to us whenever a subject child developed an
ILI, defined as fever ≥37Á8°C or feeling feverishness according to the carer’s assessment, plus at least one of the following
symptoms: cough, rhinorrhoea/nasal congestion, sore throat. During the ILI-reporting period, each family received a weekly e-mail,
text message or telephone call to remind them to contact the study team immediately if a child developed an ILI. Before the ILI-
reporting period, e-mail addresses and mobile telephone numbers of all parents/guardians were confirmed with parents/guardians to
ensure that they received messages. In addition, during the 1 week of the ILI-reporting period, all parents/guardians were
contacted/recontacted until they indicated that they had received the message that the ILI- reporting period had commenced.
Parents/guardians were provided with plastic shaft rayon-budded swabs and plastic trans port tubes with a foam pad reservoir
soaked in viral trans port medium (Virocult MW950; Medical Wire & Equipment, Wiltshire, UK). They were given verbal and written
instructions on how to collect a nose swab and a throat swab from the subject whenever an ILI occurred. We as ked for a nos e swab
to be done first and did not ins ist on a throat swab if the parents felt uncomfortable to collect one. Parents mailed swabs to the
Queens land Paediatric Infectious Diseases Laboratory (QPID), where they were stored at 80°C until tested.
Laboratory methods
Swabs were tes ted for 19 respiratory viruses by qualitative real-time PCR, 17–19 including influenza viruses A and B (Flu A, Flu B),
adenovirus (AV), human rhinovirus (HRV), polyomaviruses (JCV, BKV, WUV, KIV), parainfluenza viruses 1, 2, 3 (PIV1, PIV2, PIV3),
coronaviruses (HCoV- OC43, HCoV-NL63, HCoV-229E, HCoV-HKU1), human metapneumovirus (HMPV), bocavirus (BV) and
human respiratory s yncytial viruses A and B (hRSV A, hRSV B). All RNA virus assays used Qiagen One-Step RT-PCR, Qiagen
(Melbourne, Victoria, Australia), and all DNA virus assays used Qiagen Quantitect Probe PCR Mix, Qiagen, Australia. In order to
asses s extraction quality, specimens were spiked with equine herpes virus-1 (EHV-1) and tes ted for EHV-1 using a duplex real-time
PCR as say. Any samples that failed the EHV- 1 quality control were re-extracted. In order to monitor quality of s pecimen collection,
specimens were tested for human endogenous retrovirus 3 (ERV3) using a duplex real-time PCR assay. In addition, the positive
controls included in each PCR run were monitored for any shift in cycle thres hold values to detect problems within individual runs.
ILI outcome assessment
In order to determine health impacts of the ILI event upon the child and the household, study staff interviewed the child’s
parent/guardian by telephone 2 weeks after the onset of each ILI in subject children, and if, at that time, the subject child still had
ILI symptoms other than a dry cough, another telephone interview was arranged and conducted 2 weeks later. Data collected
included the nature and duration of symptoms, severity of illness , intrahousehold spread, visits to healthcare providers and
medication usage. Severe ILIs were defined as having at least one of the following features: fever ≥5 days, any s ymptom other than
dry cough persisting more than 14 days, otitis media, suspected bacterial respiratory infection or admiss ion to hos pital.
ILIs in household members
During ILI outcome as sess ment interviews, parents were as ked to report ILIs in household members in the week before and the
week after the onset of an ILI in the subject children. ILI attack rates in hous ehold members were calculated for the week following
ILI onset in subject children.
Statistical analysis
Statistical tests used were one-way ANOVA for continuous variables, chi-square tests for categorical data (SPSS 19, Chicago, IL,
USA) and t-tes ts for normally distributed data to compare means. Poisson regression (STATA/SE 120, StataCorp LP, TX, USA) was
used to compare incidence rates. We used person-year in the calculation.
2

3

For children who became (fully) vaccinated in the few days after the formal start date of follow-up (30 July 2010), we deducted the
number of days until the children became (fully) vaccinated from the surveillance time. Meta-Analyst 3Á13 (Tufts Medical Centre,
Boston, MA, USA) was used to calculate the probability of a virus being the sole agent identified from nose/throat s wabs during
an ILI episode (binary analysis with model type Random (D/L) and random method Der-Simonian Laird).
Ethical approval
The study was approved by the Human Research Ethics Committee at The Children’s Hospital at Westmead and was registered
with the Australian New Zealand Clinical Trials Registry (ANZCTR, ACTRN12610000319077).
Results
The cohorts
Between March and August 2010, we enrolled 399 children, of which 95Á6% completed follow-up (exclusions: 9 were of incorrect
age and nine others withdrew without contributing to the ILI-reporting period); therefore, 381 children (208 males) from 358
households participated. The ILI-reporting period was 30 July to 31 October 2010. There were just four subjects who were enrolled
slightly late during the first week of the ILI-reporting period – their person-year contribution to the ILI follow-up period was
adjusted to be from the time of joining the cohorts (one was fully vaccinated, one partially vaccinated and two unvaccinated). The
majority (89%) of enrolled children attended CCCs. On commencement of ILI surveillance, the mean age of enrolled children was
2Á3 years (0Á9–3Á4 years). At that time, 83 (22%) children fulfilled the criteria for ‘fully vaccinated’ against influenza
A/California/7/ 2009 (H1N1) (A(H1N1)pdm09); 60 (16%) were ‘partially vaccinated’, and 238 (62%) were unvaccinated (see Table 1).
The great majority given influenza vaccine (94Á4%) com- pleted vaccination by the beginning of the formal ILI follow-up period;
eight subjects received their s econd dose of influenza vaccine after follow-up began (between 1 August and 3 September). These
subjects were primarily assigned to the fully vaccinated cohort, and for incidence rate calculations, their person-time contributions
to the partially and fully vaccinated cohorts were determined using 1 week after the date of receipt of the second dose of vaccine
as the time-point at which they changed s tatus. All vaccines given were licensed unadjuvanted inactivated split virion vaccines.
There were no demographic differences between the vaccinated children, partially vaccinated children and unvaccinated children
(Table 1).
ILI episodes
During the 13 weeks 30 July to 31 October 2010, parents/ guardians reported a total of 124 ILI episodes in 105 children (13 had two
ILIs, three had three). Symptomatic ILIs were reported significantly more commonly in recipients of influenza vaccination (Table 1).
Non-influenza ILIs were more common among fully vaccinated subjects (33 non-influenza ILIs, 1Á59/person-year of observation)
and partially vaccinated subjects (20 non-influenza ILIs, 1Á54/ person-year) than among unvaccinated subjects (59 non- influenza
ILIs, 0Á99/person-year, P = 0Á001, rate ratio 1Á6, vaccinated versus unvaccinated, Table 1). Excluding ILIs from which no virus
was identified made no significant difference to this finding. No particular res piratory virus, with the exception of AV, was found
less frequently in ILI episodes among unvaccinated subjects compared to fully or partially vaccinated subjects (P = 0Á04, data not
shown). The vaccination s tatus of subjects was not correlated with the mean number of doctor (GP, emergency department or
specialist) visits made in response to non-influenza ILIs (P = 0Á45, data not shown). Nor were there significant differences in the
mean duration of ILIs (P = 0Á95) or use of antibiotics (P = 0Á92) for non-influenza ILIs between fully or partially vaccinated
subjects and unvaccinated subjects (data not shown). However, before enrolment in the study, there was evidence for an
increased use of healthcare services in both the partially and fully vaccinated groups with significantly higher rates of prior
hos pitalisation, hearing tests and grommet insertion, whereas the incidence of past otitis media was not significantly different
between the groups (Table 1).
4

Increased use of thes e healthcare services did not, however, prove significant when added to a multivariate model to predict non-
influenza ILI in study subjects (data not shown). The apparent greater risk of non-influenza ILI in influenza-vaccinated s ubjects did not
vary significantly over time, compared with non-vaccinated participants. For example, the rates of non-influenza ILI between the two
groups (vaccinated versus non-vaccinated) were 1Á88/person-year and 1Á09/person-year, respectively (rate ratio = 1Á72, P < 0Á001) in
the first half of follow-up period, while the rates in the second half were 1Á40/person-year and 0Á97/ person-year, respectively (rate ratio
= 1Á45, P = 0Á02).
Follow-up of ILIs
Telephone follow-up 2 weeks after ILI ons et was conducted for all 124 ILIs. Symptoms other than dry cough persisted in 29 ILIs at that
time. At the 4-week telephone interview, symptoms other than dry cough still persisted in eight ILIs. Data on ILI duration were not
available for five ILIs. The most commonly reported s ymptoms (data available for 121 of the 124 episodes ) were rhinorrhoea 92% (111),
cough 63% (76), decreased activity 27% (33), gas trointestinal symptoms 22% (abdominal pain, diarrhoea, vomiting; 27) and s ore throat
20% (24). Fever was documented in 84 ILIs (68%). The mean temperature was 38Á7°C. The frequency of documented fevers was similar
in each of the three cohorts – see Table 1. Managing the ILIs required 134 GP visits (for 70 ILIs, 35 of which required more than one GP
visit), 106 pharmacy visits (64 ILIs), five emergency department visits and three specialist visits – evenly distributed between the cohorts
(data not shown). No hospitalisations were reported. Antibiotics were used for 52 ILIs and 73 were treated with analgesic/antipyretics –
evenly distributed between the cohorts (data not shown). The median duration of ILIs was 8 days , but 16 (13%) lasted more than 28
days. Forty-four ILI episodes (35%) met our definition of ‘severe ILI’: 15 had otitis media, 31 had symptoms other than post-ILI dry
cough persisting >14 days, and 9 had fever persisting ≥5 days (some overlap).
Virus identification
Swab samples were available for 117 (94%) of the 124 ILI episodes, both nose and throat (69 ILIs) or nose only (48 ILIs). The quality of
samples was high in terms of extraction and cell collection; only one sample failed EHV testing and ERV3 was detected in all but one
specimen, and this s pecimen was negative of all other viruses. A total of 175 viruses were identified from 103 ILIs (see Figure 1). Multiple
viruses were detected in 52 (44%) of the swabbed ILIs – 38 ILIs yielded two viruses each, nine yielded three viruses , four yielded four,
and one yielded five. The probability of a virus being the sole agent identified from nose/throat swabs during an ILI episode is s hown in
Figure 2. Influenza A(H1N1)pdm09, which was the sole virus causing 5 ILIs, was the only virus consistently identified as the sole agent
from all ILIs with which it was associated. Although coronavirus NL63 (3 of 5 ILIs in which they were identified) and rhinovirus (15 of 39
ILIs in which they were identified) were frequently identified as sole agents of ILIs, that tendency was not statistically different to the
probabilities of the other non-influenza viruses being solely identified. No particular virus or virus combination or multiplicity of virus
infection was associated with any particular symptom or combination of symptoms or with greater frequency of antibiotic or
analges ic/antipyretic use, GP visits or other healthcare service usage. One or more of the polyomaviruses WUV and KIV were more
commonly identified in children aged <2 years (P = 0Á05), and adenoviruses were more common in females (P = 0Á03) – data not shown.
Rhinovirus alone or in combination with other viruses was as sociated with longer duration of ILI than other viruses (P = 0Á02). None of
these values were corrected for multiple testing. Of the five children who had influenza A(H1N1)pdm09 infection, one was fully
vaccinated, one was partially vaccinated (1 dose of Panvax, CSL, in October 2009), and three were not vaccinated against influenza. The
management of the A(H1N1) pdm09 infections required GP visits for four of the children; three received antipyretic/analgesic
medications, and two received antibiotics.
Yield of viruses by swabbing site
Nose swabs were collected from 117 swabbed ILIs, while throat swabs were collected from 69. The use of neither nose nor throat swabs
was not s ignificantly differently distributed acros s the three cohorts. Furthermore, there was no statistical difference in the number of
throat s wabs collected from the three study groups (P = 0Á20). Swabs were not combined prior to tes ting. One or more viruses were
detected in 88% of swabbed ILIs. Nos e swabs more often yielded viruses – 102/117 (87%) – than did throat swabs – 45/69 (65%), P <
0Á001.
5

Nose swabs yielded more viruses per swab than did
throat swabs (161 virus identities from 117 nose s wabs
= 1Á38 viruses/swab compared to 59 virus identities
from 69 throat swabs = 0Á86 viruses/swab, P < 0Á001).
Limiting the comparison of virus yields from nos e
versus throat swabs to ILIs from which both nose and
throat swabs were taken (n = 69) gave the same rates of
virus identification (1Á37 viruses per swab for nose
swabs, 0Á86 viruses per swab for throat swabs, P =
0Á001) with 61 (88%) of 69 nose swabs yielding a virus
and 45 (65%) of 69 throat swabs yielding a virus, P =
0Á002.
ILI transmission within households
In the week before onset of their ILIs, only nine subjects were exposed to one or more household members with ILIs (three other
children and eight adults). In the week after onset of the subject’s ILI, eight other children (154 exposed, 5% attack rate) and 38
adults (244 exposed, 16% attack rate) in the ill subjects’ household reported ILIs. Adult household members more often developed
an ILI in the week after ILI onset in subject children than did child members of the households, P = 0Á001 (asymptomatic carriage
and trans miss ion were not taken into account as it could not be identified). No virus was more likely than any other to be
transmitted from the ill subject to members of the household.
Discussion
In the 2010 Southern Hemisphere influenza season in Sydney, Australia, we found that young children suffered relatively often
from ILIs, but less than in previous studies. 20 The ILIs were caus ed by many different viruses, most commonly rhinoviruses and
adenoviruses. Adenovirus was more commonly found in females, an association which has not been reported previously 21,22 and
may be due to chance as no correction for multiple testing was performed. In contrast to others, we detected few RSV infections
14,20,23,24 probably because RSV infections peaked in Sydney during June and July 2010 and had declined significantly in frequency
by the time we commenced obs ervations for ILIs in the study participants (from 30 July 2010). We found only 5 ILIs caused by
influenza viruses – all A (H1N1) pdm09. Their illnesses were little different to the ILIs experienced by the children from whom other
viruses were identified (data not shown). However, the small number of influenza infections is consistent with the low degree of
influenza activity during 2010, 25 limiting the power of this study to detect differences in influenza infection rates. We did, however,
unexpectedly find that non-influenza ILI occurred about 1Á6 times more commonly in children vaccinated with one or two doses of
the influenza vaccine than in unvaccinated children.
6

These results support the findings of a recent RCT reported by Cowling et al. 26 Cowling’s study in Hong Kong concluded that non-
influenza ARI may be detected at a higher rate in children for a short period after they received influenza vaccine. The non- influenza
virus incident rate ratio (IRR) was higher in the Hong Kong s tudy (4Á4 versus 1Á6), but there are some key differences to our study,
including age of subjects, follow-up period, proportion of illnesses swabbed and proportion of swabs yielding viruses. As with all
obs ervational studies, bias must be cons idered. Vaccinated and unvaccinated cohorts in Sydney were demographically s imilar (Table 1);
however, lack of blinding by vaccination status makes it difficult to rule out selection or measurement bias. We could find no evidence of
different parental responses to ILIs in vaccinated and unvaccinated children: parents of vaccinated children were no more likely to seek
medical care during an ILI. However, we did find that health-seeking behaviours, recorded on enrolment (before the ILI observation
period), s uch as hospitalis ation (any cause), hearing tests and grommet insertion were significantly more common in the vaccinated
groups, suggesting that families that vaccinate children have a prior preference for greater healthcare service usage. This may be a
partial explanation for the obs erved difference in ILI frequency between the groups; however, prior access to any of these healthcare
services did not predict the frequency of reported ILI. Cowling et al. proposed pos sible explanations ranging from an unknown biological
mechanism by which vaccine-induced immunity to influenza was accompanied by decreased immunity to other respiratory virus to a
temporary non-specific immunity (interferon- and/or cell-mediated related) to other respiratory viruses after wild influenza infection. A
formal biological explanation is lacking. A recent US obs ervational (case–control) study has not found an association between influenza
vaccination and detection of non-influenza respiratory viruses 27 . Re-analysis of observational studies or preferably new RCTs with high
parent-collected specimen availability is required to further examine this phenomenon. It should be a priority to determine whether a
caus al as sociation exists, whether it is consistent across vaccines and populations and whether any observed increase in the rate of
non-influenza respiratory virus identification outweighs the benefit of seasonal TIVs in children. In nearly half (44%) of the nose/throat
swabs, multiple viruses were detected. Other studies in different s ettings, us ing a variety of definitions for ILI, have reported diverse
virus aetiologies for ILIs, but, in general, with lower frequencies of virus co-infection than we report. 14,20,23,24,28 Our finding may be
explained by the large number of viruses for which we tested and also by our high rate of swabbed ILIs being positive (88%).
While we found no increased severity or number of symptoms with multiple virus compared with single virus infection, this study lacked
power to tease apart the relative significance of each virus, virus combinations and multiplicity of infection. A simultaneously sampled
asymptomatic control group would have been us eful to explore further the meaning of multiple virus identification. Interestingly,
influenza A(H1N1)pdm09 was the only virus constantly identified as the sole virus from ILIs. However, there were too few cas es (five
only) to permit a firm conclusion about this. We found that nose swabs were more effective than throat s wabs for detecting respiratory
viruses in young children with ILIs. Also, parents more frequently collected nose swabs than throat swabs from their children,
suggesting that they may be more acceptable. In our study, we detected the polyomaviruses WUV and KIV at higher frequencies
(mainly as co-infections) than other investigators. 29–38 Children aged <2 years were more often infected than older children, but this
barely reached a statistical significance. To date, the pathogenicity and clinical significance of WUV and KIV (discovered in 2007) 3,4
remain unclear and studies, conducted in a variety of settings and with a variety of respiratory disease definitions, have yielded
inconsistent results. Key strengths of our study include the high rate of specimen collection (94% of ILIs were swabbed) and the high
rate of virus detection in swabbed ILIs (88% of all cas es swabbed yielded at least one virus). We believe that parent-collected specimens
combined with mail return of the specimens to the laboratory can be cons idered a reliable means of virus detection for studies s uch as
ours, partly because children are sampled earlier when viral loads may be higher.
Limitations
The participant characteristics are somewhat different to the general Australian population. Participating households had a higher
income (83% of the studies households compared to 30% of Australians had income of $ 2000/week or more), 39 the mothers were slightly
older (33Á1 years compared to 30Á1 years in the general population) 40 and were more likely to live with a partner (married or de facto
97% compared to 88%) 41 and, because we recruited in CCCs, 89% of the study population were in formal child care compared to 35% of
Aus tralian children aged 0–4 years. 42 At the start of the ILI follow-up period, we established that all parents/guardians were receiving
reminder messages, and during the course of the ILI follow-up period, we s poke to parents/guardians of 52% of the enrolled children
(evenly distributed across the cohorts – data not shown). However, we did not contact all parents/guardians after the ILI follow-up
period in order to determine whether they had not reported ILIs. We did
record at enrolment data on prior healthcare service usage (e.g. hospitalisation) and this was higher in vaccinated children; all this might
have biased the research to show a higher frequency of ILI in influenza vaccinees . The open-
label cohort design is also open to unmeasured confounding.
7

Conclusion
Influenza-like illness is common in children, and the burden on their families may be considerable. Many different respiratory
viruses are responsible for ILIs in children. In this study, conducted after the RSV season, adeno- and rhinoviruses were the most
commonly detected viruses . Symptom profiles were similar among the different viruses, and the rate of virus co-infection was high.
Recipients of influenza vaccines had about 1Á6 times more ILI episodes than did unvaccinated children, and although this may be
at least partly explained a healthcare service-seeking bias, further investigations are warranted into whether influenza vaccine
increases the risk of non-influenza ILI, as healthcare-seeking behaviour did not predict ILI in a regression model. Nose swabs
collected by parents had a high yield of respiratory viruses when using multiplex PCR methods and had s ignificantly more viruses
compared to throat swabs . In addition, parents appeared to feel more comfortable in performing nose than throat s wabs. This is of
relevance to future studies requiring parent-collected samples for PCR analysis.
Addendum – List of Authors
Dr. Alexa Dierig contributed substantially to the design of the study, helped with analysis and interpretation of data and wrote and
revised the intellectual content. She was also the study coordinator. Dr. Leon Heron contributed s ubstantially to the concept and
design of the study, helped with analys is and interpretation of data and wrote and revised the intellectual content. A/Prof Stephen
Lambert contributed substantially to the concept and des ign of the study, helped with analysis and interpretation of data and wrote
and revised the intellectual content. Dr. Jiehui Kevin Yin analysed the data, helped with their interpretation and revised the
intellectual content. A/Prof Julie Leask contributed substantially to concept and design of the study and revised the intellectual
content. Maria Yui Kwan Chow helped with analysis of data and revised the intellectual content. Prof Theo Sloots contributed
substantially to the concept and design of the study and helped with the analysis of data. Prof. Michael Nissen contributed
substantially to the concept and design of the study and helped with the analysis of the data. Dr Iman Ridda helped with the
interpretation of data and wrote and revised the clinical content. Prof Robert Booy contributed substantially to concept and design
of the study, helped with analysis and interpretation of data and also with writing and revising of the intellectual content. He was
the supervisor of the whole project. All authors approved the final version.
Funding
This work was supported by a grant from the Australian Research Council and Sanofi Pasteur (industry partner) with kind
assistance from KU Children’s Services.
Acknowledgements
We thank all participating families, the research nurses and the staff at the Queens land Paediatric Infectious Diseases Laboratory.
References
1 World Health Organisation. Initiative for Vaccine Research (VRI). Acute respiratory infections. Available at http://www.who.int/vaccine_research/diseases/ari/en/ (Accessed
4 June 2012). 2 Australian Bureau of Statistics, 2004. 4829.0.55.001–Health of Children, 2004. Available at: http://www.abs.gov.au/AUSSTAT S/
abs@.nsf/Lookup/4829.0. 55.001Main+Features12004. (Accessed 4 June 2012). 3 Allander T, Andreasson K, Gupta S et al. Identification of a third human polyomavirus. J
Virol 2007; 81:4130–4136. 4 Gaynor AM, Nissen MD, Whiley DM et al. Identification of a novel polyomavirus from patients with acute respiratory tract infections. PLoS
Pathog 2007; 3:e64. 5 Lambert SB, Allen KM, Carter RC, Nolan T M. T he cost of community-managed viral respiratory illnesses in a cohort of healthy preschool-aged
children. Respir Res 2008; 9:11. 6 Lambert S, O’Grady KA, Gabriel S, Carter R, Nolan T. The cost of seasonal respiratory illnesses in Australian children: the dominance of
patient and family costs and implications for vaccine use. Commun Dis Intell Q R ep 2004; 28:510–516. 7 Nichol KL. The efficacy, effectiveness and cost-effectiveness of
inactivated influenza virus vaccines. Vaccine 2003; 21:1769–1775. 8 Cohen GM, Nettleman MD. Economic impact of influenza vaccination in preschool children. Pediatrics
2000; 106:973–976. 9 W eycker D, Edelsberg J, Halloran ME et al. Population-wide benefits of routine vaccination of children against influenza. Vaccine 2005; 23:1284–
1293. 10 Jefferson T , Rivetti A, Harnden A, Di PC, Demicheli V. Vaccines for preventing influenza in healthy children. Cochrane Database Syst Rev 2008; 16:C D004879.
11 Vesikari T, Knuf M, Wutzler P et al. Oil-in-water emulsion adjuvant with influenza vaccine in young children. N Engl J Med 2011; 365:1406–1416. 12 Lu N, Samuels
ME, S hi L et al. Child day care risks of common infectious diseases revisited. Child Care Health Dev 2004; 30:361–368. 13 Kamper-Jorgensen M, W ohlfahrt J, Simonsen J,
Gronbaek M, Benn CS. Population-based study of the impact of childcare attendance on hospitalizations for acute respiratory infections. Pediatrics 2006; 118:1439–1446. 14
Lambert SB, Allen KM, Druce JD et al. Community epidemiology of human metapneumovirus, human coronavirus NL63, and other
8

respiratory viruses in healthy preschool-aged children using parent- collected specimens. Pediatrics 2007; 120:17. 15 Dales RE, Cakmak S, Brand K, Judek S. Respiratory illness
in children attending daycare. Pediatr Pulmonol 2004; 38:64–69. 16 Australian Bureau of S tatistics, 2008. C hildhood education and care. Available at
http://www.ausstats.abs.gov.au/Ausstats/subscriber.nsf/ 0/573FA1277E2378E3CA25765700162238/$File/44020_June% 202008%20(reissue).pdf. (Accessed 30 July 2012). 17 Lu
X, Holloway B, Dare RK et al. Real-time reverse transcription-PCR assay for comprehensive detection of human rhinoviruses. J Clin Microbiol 2008; 46:533–539. 18 O’Grady
KA, T orzillo PJ, Rockett RJ et al. Successful application of a simple specimen transport method for the conduct of respiratory virus surveillance in remote Indigenous communities
in Australia. T rop Med Int Health 2011; 16:766–772. 19 Maunula L, Klemola P, Kauppinen A et al. Enteric Viruses in a Large Waterborne Outbreak of Acute Gastroenteritis in
Finland. Food Environ Virol 2009; 1:105–105. 20 Fairchok MP, Martin ET , Chambers S et al. Epidemiology of viral respiratory tract infections in a prospective cohort of infants
and toddlers attending daycare. J Clin Virol 2010; 49:16–20. 21 Zou L, Zhou J, Li H et al. Human adenovirus infection in children with acute respiratory tract disease in
Guangzhou, China. APMIS 2012; 120:683–688. 22 Akarsu D, Altoparlak € U. The investigation of some viral agents in the patients with upper respiratory tract in Erzurum
region. Kafkas € Univ Vet F ak Derg 2009; 15:951–955. 23 Canducci F, Debiaggi M, Sampaolo M et al. Two-year prospective study of single infections and co-infections by
respiratory syncytial virus and viruses identified recently in infants with acute respiratory disease. J Med Virol 2008; 80:716–723. 24 Arden KE, McErlean P, Nissen MD, Sloots
TP, Mackay IM. F requent detection of human rhinoviruses, paramyxoviruses, coronaviruses, and bocavirus during acute respiratory tract infections. J Med Virol 2006; 78:1232–
1240. 25 NNDSS Annual R eport Writing Group. Australia’s notifiable disease status, 2010: annual report of the National Notifiable Diseases Surveillance S ystem. Commun Dis
Intell Q Rep 2012; 35:1–69. 26 Cowling BJ, Fang VJ, Nishiura H et al. Increased risk of noninfluenza respiratory virus infections associated with receipt of inactivated influenza
vaccine. Clin Infect Dis 2012; 54:1778–1783. 27 Sundaram ME, McClure DL, van Wormer JJ et al. Influenza vaccination is not associated with detection of non-influenza
respiratory
viruses in seasonal studies of influenza vaccine effectiveness. Clin Infect Dis 2013; 57:789–793. 28 van der Zalm MM, van Ewijk BE, Wilbrink B et al. Respiratory pathogens in
children with and without respiratory symptoms. J Pediatr 2009; 154:396–400. 29 F oulongne V, Brieu N, Jeziorski E et al. KI and W U polyomaviruses in children, France.
Emerg Infect Dis. 2008;14:523–525. 30 Bialasiewicz S, Whiley DM, Lambert SB et al. Presence of the newly discovered human polyomaviruses KI and WU in Australian patients
with acute respiratory tract infection. J Clin Virol 2008; 41:63–68. 31 Neske F, Blessing K, Ullrich F et al. WU polyomavirus infection in children, Germany: Emerg Infect Dis.
2008;14:680–681. 32 Ren L, Gonzalez R, Xie Z et al. WU and KI polyomavirus present in the respiratory tract of children, but not in immunocompetent adults. J Clin Virol
2008; 43:330–333. 33 Zhuang W L, Lu XD, Lin GY et al. WU polyomavirus infection among children in South China. J Med Virol 2011; 83:1440–1445. 34 Payungporn S,
Chieochansin T, Thongmee C et al. Prevalence and molecular characterization of WU/KI polyomaviruses isolated from pediatric patients with respiratory disease in Thailand. Virus
Res 2008; 135:230–236. 35 Zhao L, Qian Y, Zhu R et al. Identification of WU polyomavirus from pediatric patients with acute respiratory infections in Beijing China. Arch Virol
2010; 155:181–186. 36 Abed Y, Wang D, Boivin G. WU polyomavirus in children Canada. Emerg Infect Dis 2007; 13:1939–1941. 37 Sumino KC, Walter MJ, Mikols CL et al.
Detection of respiratory viruses and the associated chemokine responses in serious acute respiratory illness. Thorax 2010; 65:639–644. 38 Venter M, Visser A, Lassauniere R.
Human polyomaviruses, WU and KI in HIV exposed children with acute lower respiratory tract infections in hospitals in South Africa. J Clin Virol 2009; 44:230–234. 39
Australian B ureau of Statistics. Household income and income distribution 2009–2010. Available at http://www.ausstats.abs.gov.
au/ausstats/subscriber.nsf/0/DBE855896D8CA36DCA2578FB001853 3C/$File/65230_2009-10.pdf (Accessed 14 June 2012). 40 Laws PJ, Sullivan EA. Australia’s mothers and
babies 2008. P erinatal statistics series 2010 no. 24. Cat. no. PeR 50. Canberra: AIHW. 41 Australian Institute of Health and Welfare 2009. A picture of Australia’s children 2009.
Cat. no. PHE 112. Canberra: AIHW. 42 Australian Bureau of Statistics. 4402.0 Childhood Education and Care, Australia, June 2011. Available at http://www.abs.gov.au/
ausstats/abs@.nsf/Products/4402.0. (Accessed 8 January 2014).
9