ARCHIVED - Vaccine Coverage of 2-Year-Old Children in Montreal - 2003


Category: Public Health Practice, Immunization Programs

Hudson P, Allard R, Joseph L, Valiquette L


A population-based survey of 2-year-old children in Montreal was undertaken in 2003, the first since 1996. We report the main methodologic observations made on this occasion.


We obtained a random sample of 600 children receiving Quebec Health Insurance who met inclusion criteria for age and address. Parents completed a telephone interview or mailed questionnaire to identify socio-demographic characteristics, sources of health care and vaccine doses recorded in parent-held records. Health professionals from the Public Health Department conducted the interviews. Missing vaccine doses were verified with providers after obtaining written consent.


Telephone interviews were completed for 462 children. Mailed questionnaires were received for 44 others. Thirteen children were ineligible. Twelve respondents refused to participate and 28 could not be located. Overall participation was 86%, or 91% of those who were located. Two parents reported that their child had received no vaccine. We validated records with providers for 84 children. We could not validate records for 20 children, including 20 for whom we could not access any vaccination information. The proportion of children vaccinated by age 2 with four doses of diphtheria, pertussis, tetanus and Haemophilus influenzae type b, three doses of polio and one dose each of measles, rubella and mumps vaccines was 83.1% (79.5% to 86.3%). To better quantify uncertainty due to response bias, we computed bounds for the true vaccine coverage by including nonrespondents in our sample and assuming that vaccine doses were all not given (lower bound = 69.8%) or all given (upper bound = 88.8%). Comparable results for our 1996 study showed lower and upper bounds of 57.0% and 80.1%, respectively.


The smaller sample in the 2003 study allowed more creativity in locating subjects. Questionnaire delivery was confirmed by registered mailings to nonrespondents. Highly trained interviews elicited vaccine information at interview, reducing the need for validation. We halved the proportion of unvalidated records at study end compared to the 1996 study. This approach provided a less costly alternative to estimating vaccine coverage, with less overall uncertainty (including random variation and possible response bias) than in the 1996 study.

Report a problem or mistake on this page
Please select all that apply:

Thank you for your help!

You will not receive a reply. For enquiries, contact us.

Date modified: