Interim estimates of the effectiveness of influenza vaccination against influenza‐associated hospitalization in children in Hong Kong, 2015–16

From 1 September 2015 through 31 January 2016, we enrolled 2068 children 6 months to 17 years of age admitted to hospital with a febrile acute respiratory infection in our test‐negative study. Information on receipt of 2015–16 northern hemisphere inactivated influenza vaccination was elicited from parents or legal guardians. Using conditional logistic regression adjusting for age and matching on calendar time, we estimated influenza vaccine effectiveness against hospitalization with influenza A or B to be 79.2% (95% confidence interval: 42.0%–92.4%). Annual influenza vaccination should be more widely used in children in Hong Kong.


| INTRODUCTION
Influenza viruses circulate each year, causing a considerable burden to public health. Children typically face the greatest risk of influenza virus infection among all age groups in a population, 1 and young children have high rates of hospitalizations associated with influenza. 2,3 Influenza vaccination is the most effective measure to prevent infection and recommended by the World Health Organization for many high-risk groups including young children from 6 to 59 months of age. 4 In Hong Kong, a subtropical city on the south coast of China, all children between 6 months and 6 years of age and children 6 years of age or above with an underlying condition associated with an increased risk for influenza complications are recommended to receive influenza vaccination each year. 5 The Hong Kong government provides a subsidy for influenza vaccination for children between 6 months and 6 years of age and older children from lowincome families, while children between 6 months and 17 years of age with an underlying medical condition can receive free influenza vaccination. 6 It is informative to evaluate influenza vaccination effectiveness (VE) each season to confirm that vaccination is providing adequate protection, to inform risk communication and to provide evidence for vaccination strategies and other public health measures. 7 Timely interim, or mid-season, estimates of influenza VE can be particularly useful to guide resource allocation or implement additional preventive measures if VE is low. 8

| Study design
We used the test-negative study design, which is a type of casecontrol design, to estimate influenza VE. 6,7,[9][10][11] We enrolled chil- to the general wards of these hospitals with a febrile acute respiratory infection, defined as fever measured ≥38°C with any respiratory symptom such as cough, runny nose or sore throat, were eligible for inclusion in this study. Children with risk factors for potentially severe respiratory infections such as prematurity or chronic lung disease were not excluded. Nasopharyngeal aspirates were obtained from all patients and tested for influenza A and B virus by direct immunofluorescence assay (for rapid diagnosis) followed by reverse transcriptase polymerase chain reaction for seasonal influenza A and B viruses using laboratory methods as previously described. 6,11 Influenza vaccination history within 6 months of hospitalization was elicited from the parents or legal guardians of patients using a standardized questionnaire administered by research personnel.
Vaccinated children were those who had received influenza vaccination within the 6 months prior to admission in a regimen and dosage appropriate for age and influenza vaccination history according to the recommendations of the Advisory Committee on Immunization Practices, with the last dose more than 2 weeks before hospitalization. 12 Children who should receive 2 doses of influenza vaccination, but only received 1 dose or were vaccinated within 2 weeks of hospitalization were categorized as unvaccinated. The northern hemisphere formulation of trivalent and quadrivalent inactivated influenza vaccines were used during our study period. Analysis was performed by R 3.1.1.

| ETHICAL APPROVAL
The study protocol was approved by the Institutional Review Board of the University of Hong Kong and that of the Kowloon West Cluster Research Ethics Committee which waived the need for written consent as viral investigation was a routine diagnostic test carried out as part of routine care, any patient information was delinked from individual patient identification to maintain patient confidentiality and participation by responding to the questionnaire was voluntary and indicative of consent.

| Statistical analysis
Following the analytic approach used in previous years, 6

| RESULTS
We We did not have sufficient data to estimate VE against influenza A subtypes. There are some limitations of our study. We estimated VE against hospitalization, and this will differ from VE against infection or medically attended infection if vaccination modifies disease severity. 10 In a recent review, estimates of VE against hospitalization were very similar to estimates of VE against medically attended illness, for studies conducted in the same populations, years and age groups. 17 Influenza vaccination history of 6 months was self-reported, and misclassification of vaccination status may have biased estimates of VE. 18 Information was not available on potential confounders such as underlying medical conditions. The sample size was large enough to confirm statistically significant overall VE, but not sufficient to provide precise estimates of VE in some age groups. Finally, our interim estimates of VE may differ from the future end-of-season estimate, although usually interim estimates are quite similar to final end-of-season estimates in test-negative studies. 8,19,20 Final estimates for 2015-16 in Hong Kong will be available around the end of 2016.

| DISCUSSION
In conclusion, we documented that influenza vaccination was associated with good protection against influenza-associated  F I G U R E 2 Interim estimates of influenza vaccine effectiveness against influenza viruses overall and by influenza type and age group in Hong Kong, 2015-16. Vaccination effectiveness estimates of 100% with wide confidence intervals are from subgroups in which there were no vaccinated children in the test-positive group and are shown in grey