Abstract
OBJECTIVES: There is a lack of information about household factors associated with delayed measles, mumps and rubella (MMR) vaccination. We examined whether timeliness of first MMR (MMR1) receipt is associated with sharing a household with an older child with non-receipt of MMR1 independent of household composition and size. DESIGN: Longitudinal observational study using linked electronic health records. SETTING: North East London, UK. PARTICIPANTS: The index cohort comprised 71 509 children (51.0% males) eligible to receive MMR1 between 1 January 2014 and 28 February 2020. METHODS: The primary outcome was MMR1 receipt between 12 months and 24 months of age. The explanatory variable was non-receipt of MMR1 between age 12 months and 24 months in the oldest child sharing the same household. We examined the likelihood of MMR1 receipt in index children sharing a household with an older child with non-receipt of MMR1 between 12 months and 24 months using logistic regression to estimate ORs and 95% CIs before and after adjustment for individual-level, household-level and area-level covariates. We carried out sensitivity analyses excluding households with an age interval between oldest and youngest child greater than 5 years. RESULTS: 59 851 (83.6%) index children received MMR1 between 12 months and 24 months of age. After adjustment for household composition and size, MMR1 receipt was less likely in index children sharing a household with an older child with non-receipt of MMR1 between 12 months and 24 months of age: OR: 0.19 (95% CI: 0.18, 0.20). This association strengthened after excluding households with an age interval greater than 5 years: OR: 0.14 (0.13, 0.15). CONCLUSIONS: There is strong concordance within households of delay in MMR1 receipt independent of household size and composition. Lack of timely protection within households increases the risk of measles outbreaks. There is a need for household-based interventions to improve MMR1 timeliness.