Does pneumococcal conjugate vaccination affect onset and risk of first acute otitis media and recurrences? A primary care-based cohort study

Background: It has been hypothesized that widespread implementation of pneumococcal conjugate vaccination (PCV) in infancy reduces early AOM and thereby prevents further AOM episodes and associated health care resource use. Methods: We tested this hypothesis by applying an extension of the original Cox proportional hazards model (Prentice, Williams and Petersons’ total time) to individual AOM episodes recorded in pseudonymised primary care electronic health records of 18,237 Dutch children born between 2004 and 2015. Children were assigned to three groups: no-PCV (January 2004-March 2006), PCV7 (April 2006-February 2011) and PCV10 (March 2011-February 2015). Results: Of the 18,237 newborns, 6967 (38%) experienced at least one GP-diagnosed AOM episode up to the age of four years (median age at first AOM: 12 months, interquartile range: 12; total number of AOM episodes: 14,689). Time-to-first AOM was longest in the PCV10 group compared with the PCV7 and no-PCV groups (log rank test: P < 0.001); in these groups 30% had experienced a first AOM at 20, 17 and 15 months, respectively. Children in the PCV10 group had a 21% lower risk of experiencing a first AOM episode than those in the no-PCV group (hazard ratio (HR): 0.79, 95% confidence interval (CI): 0.72–0.86), while the effect was less pronounced for the PCV7 group (HR: 0.94, 95% CI: 0.87–1.02). Neither PCV7 nor PCV10 reduced the risk of AOM recurrences. Compared to no-PCV, HRs for overall AOM were 1.00 (95% CI: 0.95–1.06) and 0.89 (95% CI: 0.84–0.95) for PCV7 and PCV10, respectively. Conclusion: Our cohort study suggests that PCV postpones the onset and reduces the risk of first AOM without affecting recurrences. The impact of PCV on overall AOM in children up to the age of four years seems therefore largely attributable to the prevention of a first AOM episode.