Globally, disease in children is caused predominantly by group A [11]. The virus is transmitted by the faeco-oral route; from person to person directly or via contaminated fomites, food or water [12]. Peak incidence of clinical disease is 6–24 months of age [13]. Following an incubation period of 1–3 days, it classically KPT-330 nmr presents with sudden onset of vomiting and fever with profuse watery diarrhoea. Symptoms usually last 2–7 days (average 5 days) [12]. Patterns of immunity

are relatively complex: maternal antibodies confer some protection for newborns, neonatal infection is believed to offer protection against disease, and previous infections progressively reduce a child’s risk of rotavirus infection and disease [14]. Based on the findings of Velazquez et al., children with 1, 2 or 3 previous rotavirus infections have 0.62, 0.40 and 0.34 the risk of rotavirus disease relative to children who have no previous infections [15]. We developed a deterministic age-structured dynamic model of rotavirus transmission which included degrees of susceptibility to re-infection in keeping with known patterns of immunity to rotavirus infections. The full model is illustrated by the flow diagram in Fig. 1 with parameters as defined in Table 1. Full model equations are described

in Appendix A. We incorporated the key features of rotavirus epidemiology in the following ways. Newborn infants of immune mothers were protected by maternal antibodies [16]. Therefore, 3-Methyladenine datasheet we assumed that all children were immune at birth and entered a maternally protected class. This immunity waned at a constant rate with a mean duration of 3 months (1/μ), after which individuals moved into the first susceptible class. Individuals in all susceptible classes could be infected at a rate λ, and they recovered from infection at a rate γ. From the literature we have also concluded that at least three re-infections (four susceptible classes) should be distinguished

[15]. The risk of an exposed individual developing an infection (α1–3) and the proportion of individuals assumed to become immune after each infection (1 − α1–3) varied depending on the number of previous infections. We assumed that the risk of infection was 62% after one infection, 65% (=0.40/0.62) after two and 85% (=0.34/0.40) after three, based on the findings of Velazquez et al. [15] and supported by others [17] and [18]. After four infections, all individuals became immune and entered the recovered class. Also based on Velazquez et al. [15], we assumed that 47% of first, 25% of second, 32% of third and 20% of fourth infections were symptomatic. Once individuals entered the recovered class, they were assumed to be temporarily but completely immune to re-infection. This immunity waned at a rate (ω) and individuals then entered the fourth susceptible class from which they could be re-infected at a rate λ.