(344) Effectiveness Of Pertussis Vaccines In Children Between 2013 And 2019 In France: A Nested Matched Case Control Study Using The National Health Data System

Background: France changed its pertussis vaccination schedule for children in 2013, switching from a “3+1” schedule including a primary series of 3 monthly vaccine administrations at 2, 3 and 4 months of age (MoA) followed by a booster dose at 16-18 MoA, to a “2+1” schedule including primary series at 2 and 4 MoA, a booster dose at 11 MoA, and an additional booster dose at 6 years old.

Objectives: To estimate the vaccine effectiveness (VE) against hospitalized pertussis in children born between 2013 and 2019 in France.

Methods: The study design was an observational, retrospective, matched case-control nested in a source cohort of children born between 01 May 2013 and 31 December 2019 using claims data of the French National Health Data System (SNDS). The source cohort included children who had ≥6 consultations to a general practitioner (GP) or pediatrician at 6 months of age, and who received either 0 dose, or 1 dose, or 2 doses before 6 months of age, or 3 doses before 13 months of age of vaccine to control for vaccinations in other non-registered settings. Children were follow-up until death, end of study period on 31 December 2020 or exit from SNDS, whichever occurred first. Cases of pertussis hospitalizations were identified using any hospital diagnosis ICD-10 codes A37.0 and A37.9. Eligible controls were all children in the source cohort who were not hospitalized for pertussis during follow-up. Cases were matched with up to 100 controls on month-year of birth and department of residence. Exposure to vaccine against pertussis included any pertussis-containing vaccine regardless of manufacturer. Vaccination was assumed administered on the closest consultation to GP/pediatrician following dispensation and effective 2 weeks after administration. VE was computed as 1 – adjusted odds ratio of being vaccinated against pertussis (0-reference; 1; 2; 3 effective doses) between cases and controls, with the odds ratio obtained from a conditional logistic regression model. VE and its 95% confidence interval (CI) were estimated overall and stratified by age groups (1.5-4.5 months, 4.5-11.5 months, and 11.5-60 months).

Results: 1,481,884 children were included in the source cohort, of whom 528 cases were matched with 50,373 controls. Overall VE was 81.5% [95%CI: 51.7%-92.9%]; age-stratified VE were 24.1% [0%-43.1%] for ≥1 dose at 1.5-4.5 months, 78.5% [46.3%-91.4%] for ≥2 doses at 4.5-11.5 months, and 93.4% [74.7%-98.3%] for 3 doses at 11.5-60 months.

Conclusions: High VE was observed after 2+1 schedule, but caution is needed due to potential selection bias and exposure misclassification. The implementation of maternal vaccination to protect newborns in their first months of life could reduce the gap protection induce by vaccine delays uptake.