The data

View the most recent version.

Data were collected using Ontario's Rapid Risk Factor Surveillance System (RRFSS) infrastructure.  RRFSS is an ongoing, random digit-dialing telephone survey of the adult population in private households in 18 of the province's 36 health regions. RRFSS gathers surveillance data, monitors public opinion on key public health issues, and collects information on emerging issues of importance to public health in Ontario. More information about RRFSS can be found at www.rrfss.ca.

The 18 health units that regularly participate in RRFSS each obtain information from approximately 400 households for every four-month data collection cycle; surveys are conducted by the Institute for Social Research at York University. For this study of pH1N1 vaccination uptake, Public Health Ontario funded an augmentation of the sample to include households in the 18 health units not usually involved in RRFSS.

A two-stage stratified cluster sampling design was used, with households as the first stage and household members as the second. In the first stage, telephone numbers of private households were chosen from telephone books and commercially available compiled lists.⁷ The sample included numbers on either side of those that were selected, thereby ensuring that numbers that were not "listed" would be included. This made it possible to access cellphones as well as land lines. Residents of long-term care facilities, penitentiaries and other institutions were not included. In the second stage, an adult in the selected household was chosen. In households with more than one adult, the person whose birthday came next was selected. If children younger than 18 lived in the household, the adult respondent answered questions about the child with the next birthday. Interviews were conducted from January 14 through May 4, 2010.

Respondents were asked, "Since October 2009, have you received the H1N1 flu shot?" Those who reported having had a flu shot were asked, "In what month was that?" For children and adolescents younger than 18, the same questions were asked of the adults responding on their behalf. Respondents were asked if a doctor or other health care professional had ever told them they had any of the following disorders: high blood pressure, asthma, diabetes, or any other chronic disease including but not limited to heart disease, cancer or thyroid disorder. Those who reported having been diagnosed with asthma, diabetes, cancer, heart, lung or kidney disease or an immune or blood disorder were classified as having a high-risk chronic disease.⁴

A total of 9,010 adults participated in the survey (participation rate= 57.7%). Those whose vaccination month was incompatible with the month in which they were interviewed were dropped (n=36; 0.4%), as were those with a refused/don't know response (n=45; 0.5%) to the vaccination question. In total, 8,929 adults were retained for analysis.

As well, 2,867 adults who were interviewed lived in a household with at least one child. Of these, 11 (0.4%) reported a vaccination date for the child incompatible with their interview month, and 65 (2.3%) were not aware of or refused to provide information about the child's vaccination status. Consequently, 2,791 children and adolescents younger than age 18 were retained for analysis.

Results for people aged 18 or older were weighted to account for the number of adults in the household and the population in each health unit area, and were then post-stratified by the age and sex distribution of the 2009 Ontario population. Adult weights were normalized to maintain the effective sample size and incorporated into all analyses. Weights could not be calculated for respondents missing age information (n=208); they were assigned a weight of 1 and included in the final analysis. A sensitivity analysis removing individuals with missing age was also conducted. Results for children and adolescents younger than 18 were weighted to account for the number of children in the household. Those missing the number of children in the household (n=7) were assigned a weight of 1 and included in the final analysis. Sensitivity analyses were also conducted. Analyses were carried out using SAS Statistical Software (Version 9.2, Cary, NC) and PASW Statistics 18, Release 18.0.0 (SPSS Inc., Chicago, IL). All analyses were weighted, and two-sided 95% confidence intervals were calculated using normal approximation.

The relatively high participation rate and the small percentage excluded because of incomplete responses on the outcome variable tend to increase confidence in the results. Nonetheless, vaccine uptake may have been underestimated. Respondents might have been in the process of obtaining a pH1N1 vaccination while the interviews were being conducted. However, the majority of Ontario's mass vaccination clinics were closed in December.⁸ While the possibility of recall bias exists, the novel nature of the immunization campaign and the media attention it received suggest that the risk of recall bias is minimal. Age was reported in years, so for children younger than 1 year of age, it is not clear if the child was more than 6 months old and unvaccinated, or under 6 months and not eligible. The small number (n=126) to whom this possibility applied minimizes this limitation. Because the survey was self-reported, the possibility of misclassification of vaccine status cannot be ruled out. Finally, by definition, a telephone survey excludes people without telephones, such as those who are homeless or live in remote areas. Statistics Canada estimates that 0.9% of Canadian households do not have telephone service.⁹