Based on detailed Dutch epidemiological data analyzed in an age-structured, discrete-time event, stochastic multi-cohort model, we conclude that targeted RV vaccination of high-risk infants is highly cost-effective and potentially cost-saving in the Netherlands. Cost-effectiveness estimates were most sensitive to RV mortality rates, but targeted RV vaccination remained cost-effective when mortality would be 90% lower than observed.
In our analysis, universal RV vaccination was not considered cost-effective from the healthcare provider perspective and would only become cost-effective when herd-immunity and caretaker QALY losses were included and if vaccine prices would be at most €60/child. These results are in line with results from other European cost-effectiveness analyses that have used comparable methodology and QALY loss estimates for RVGE. Universal RV vaccination was not considered cost-effective from the healthcare provider perspective in Belgium, England and Wales, France, the Netherlands and Ireland, and was cost-effective in Finland only [9, 69]. Our analysis demonstrated that universal RV vaccination could, however, be considered cost-effective from the societal perspective at the €35,000/QALY threshold. These findings differ somewhat from previous economic analyses of universal RV vaccination in the Netherlands [9, 10, 45] which can be explained by the updated and more reliable parameter estimates used. Incidence and costs of RV hospitalizations in the Netherlands determined in our study are comparable to estimates from Germany, Finland and the UK and another recent Dutch observational study [65, 70–72]. Previously, lower incidence and cost estimates were derived for the Netherlands by using indirect methods combining sentinel laboratory data and hospital discharge codes [73, 74]. The accuracy and completeness of methods using discharge codes has been criticized and depends on local coding practices [75–77]. In addition, we could include recent estimates of parental work loss in children hospitalized for RV .
Our analysis did not account for potential costs and QALY losses associated with vaccination induced intussusception. Based on observed intussusception risks attributable to RV vaccination in different populations, 0 to 9 additional cases would occur each year in the Netherlands when universal RV vaccination is implemented [78–86]. Clearly, this could have a negative impact on cost-effectiveness, although overall effects may be small. Furthermore, the recent reports on an increased risk of intussusception after the first dose of RV vaccine may raise concerns about exposing healthy children at low risk of RV-related complications to vaccination risks [87, 88].
Our study confirms that prematurity, LBW and congenital pathology are important risk factors for RV hospitalization and increased healthcare needs. Furthermore, we observed RV mortality exclusively among patients with any of these high-risk conditions. Although absolute numbers were low, similar observations in other European and US studies and the association between diarrhea-related mortality and birth weight confirm the existence of differential mortality risks [17, 19, 30–32]. Of note, in five out of seven patients who succumbed the underlying illness rather than RV was stated as the cause of death in death-records. Yet, in these patients RV caused a profound medical deterioration leading to premature death, as confirmed by expert review of case histories. These findings suggest that among children with severe underlying conditions fatal RV disease is underreported.
Although limited data are available on vaccine safety and efficacy among high risk patients, protection provided by RV vaccine was comparable in premature and non-premature infants without additional safety risk [89–91]. Current recommendations support RV vaccination in preterm infants and also in those with preexisting underlying disease, including gastrointestinal disease, in non-acute phases of illness [3, 5, 92]. Recently, it was shown that RV vaccination among short bowel patients is well tolerated .
Targeted RV vaccination does not offer the potential benefits of herd-protection, which has been described after implementation of universal RV vaccination. Observed effects among unvaccinated individuals ranged from 0 to 72% with substantial differences between consecutive years and effects declining with increasing age [46, 58, 59]. As severe RVGE occurs mainly in those <5 years old, herd-immunity could be a transient effect post-implementation, which disappears when coverage rates among this age-group approach 100%. Therefore, herd-immunity effects on the population level are difficult to predict . Continued surveillance may provide more insights in coming years.
Naturally, our findings and conclusions may not hold for countries with high RV mortality among the general infant population and with higher RVGE incidences. In such countries universal RV vaccination remains the recommended approach.