Predictors of incomplete immunisation in Victorian children

Objective: Using probabilistic record linkage, this study aimed to determine the predictors of incomplete immunisation in Victorian children. Methods: Records of all births in Victoria in 1998 were linked with records from the Australian Childhood Immunisation Register (ACIR). The outcomes of interest were the predictors of immunisation status at 12 and 24 months of age. Results: A total of 91.1% of birth records were linked with records from the ACIR (n=55,129). Immunisation coverage was 92.8% at 12 months and 89.2% at 24 months. After multivariate analysis, the following maternal factors were significant predictors of incomplete immunisation in children aged 12 months: higher parity, 12-23 months interval between pregnancies, not being married, Aboriginal or Torres Strait Islander or born overseas, younger age, no private health insurance, home birth, metropolitan place of birth, and being in the highest or lowest socioeconomic quintiles. Low birthweight and singleton birth were also significant predictors. All of these factors were also predictive of incomplete immunisation in children aged 24 months, with the exception of low birth weight, which was only significant in the very low birth weight category. Congenital malformations or prematurity were not significant predictors. Conclusions and Implications: Record linkage is an effective means of determining the relationship between immunisation status and socio-demographic and health-related factors. It enabled small groups within a large population to be studied, confirming findings from previous survey studies and indicating that several population groups were at increased risk of incomplete immunisation. The main limitation of the study was the under-reporting of immunisation status to the ACIR.