Josephine van Dongen

Chapter 6 146 initial three months post-rotavirus vaccination, and up to eleven months of age, covering the period up to the booster vaccinations of the NIP.We also analysed hospitalizations for infectious diseases separately, using Cox modelling as described above. For all models, we considered the following covariates: sex, type of medical risk condition, gestational age, breastfeeding, daycare attendance, type of hospital care (i.e. academic versus general), family educational level, parental origin, NIP vaccination status, presence of siblings in the household and season with high rates of respiratory infections (from October until April). All analyses were performed on according to protocol cohor ts, where all high-risk infants whose parents indicated willingness to vaccinate their child against rotavirus in the pre- implementation cohor t are compared to infants that received at least one HRV dose in the post-implementation cohor t. This information was based on a parental questionnaire at star t of the study, see chapter 4 . We performed complete case analyses, missing information is documented. As statistical software SPSS version 25 and RStudio version 5.0 were used, with packages MASS, fmsb, survival and survminer . Results A total of 1482 high-risk infants with one or more medical risk conditions were enrolled in the study.The population for analysis included 719 rotavirus vaccinated infants and 496 infants whose parents indicated willing to vaccinate against rotavirus, but who were enrolled before rotavirus vaccination was implemented. In total, 1067 infants (87.8%) were born premature, 373 infants (30.7%) were small for gestational age and 201 infants (16.5%) had at least one congenital condition. Baseline characteristics were comparable between the rotavirus vaccinated and unvaccinated group ( table 1 ), with the exception of follow up, which was complete for 450 vaccinated infants (62.6%) and 380 (76.6%) of unvaccinated infants. The propor tion of observation months during seasons with or without high rates of respiratory infections was not significantly different between the groups, ARIs occurring in-season was also not different.