Josephine van Dongen

Chapter 3 62 The propor tion of rotavirus hospitalizations attributable to nosocomial or community-acquired infections and also the ratio of healthy vs risk-group children were originally derived from the RoHo study. Propor tions were updated based on results from active AGE surveillance conducted in 12 Dutch hospitals between November 2014 and November 2016 18, 19. Collected data include age, sex, rotavirus presence in stool, type of infection (community- acquired or nosocomial), and the presence of medical risk conditions. As the active surveillance only included children <2 years of age, propor tions for older children were kept consistent to what was found in the RoHo study. Other parameter updates Each model input parameter and assumption was checked for potential updates by screening the literature and checking available data from ongoing surveillance.As we outline in this section, this yielded new and improved data on the impact of rotavirus disease and vaccination, and we updated our parameters accordingly. A recent UK study estimated the quality-adjusted life year (QALY) loss due to severe rotavirus AGE 20 .These estimates were applied as QALY loss for community-acquired rotavirus hospitalizations ( Table 1 ). For rotavirus episodes without medical care and those requiring GP visits, we updated our previous estimates on parental productivity losses due to work absence based on results from a prospective household study on AGE among 289 Dutch families with young children conducted between January and May 2016 21 . (See Table 1 and Additional file 1 for details.) All costs — healthcare costs, patient and family costs, and productivity losses — were updated to 2016 cost prices using Dutch consumer price indexes and recent reference prices ( Table 1 ) 22 . Herd protection as a result of infant rotavirus vaccination, where rotavirus AGE in unvaccinated children is reduced, has been widely observed post-implementation in high-income, high- coverage settings 23-27 . We therefore incorporated herd-protection effects in our base case for universal infant rotavirus vaccination. We stratified herd-protection levels by age and by vaccinated vs unvaccinated cohor ts (Additional file 1: Table S1 ). Unvaccinated age cohor ts were assumed to be ineligible for vaccination based on age at the time of implementation, but may still benefit from herd effects.The available studies on herd-protection levels used historical pre-vaccination cohor ts as a comparator in settings where annual rotavirus epidemics occurred 23-27 .To account for the presence of a biennial epidemic pattern in the current pre-vaccination setting in the Netherlands, we lowered study estimates by 50% for our analysis. This assumes that relevant reductions due to herd effects only occur every other year.We assumed no effect on adult rotavirus infections from any of the infant vaccination strategies 11 and no herd effects for targeted vaccination, as this would result in a maximum vaccine coverage of 8% in the infant