Table 1 Summary of model assumptions trialled for norovirus, where each model was fitted to age-specific incidence data in England. Models A0–B20 were taken forward to estimate incidence of norovirus between 2020 and 2022
Model | A0 | A20 | B0 | B20 | D0 | D20 | E0 | F0 |
|---|---|---|---|---|---|---|---|---|
Assumptions | ||||||||
First infection | Always symptomatic | Always symptomatic | Always symptomatic | Always symptomatic | 50% asymptomatic | 50% asymptomatic | 50% asymptomatic | Always symptomatic |
Duration of Asymptomatic infectiousness (days) | 15 | 15 | 20 | 20 | 15 | 15 | 15 | 15 |
Asymptomatic infectiousness relative to symptomatic | 0.05 | 0.05 | 0.05 | 0.05 | 0.05 | 0.05 | 0.5 | 1 |
Under-reporting assumed in fitted data (%) | 0 | 20 | 0 | 20 | 0 | 20 | 0 | 0 |
Results | ||||||||
Probability of transmission given a contact (qs) | 0.1892696 | 0.2070616 | 0.1693301 | 0.1824529 | 0.2183579 | 0.2193598 | 0.374627 | 0.1909186 |
Log-likelihood of age-specific incidence when fitted to data from England [15] | −77.94 | −190.03 | −72.10 | −107.46 | −667.77 | −850.36 | −1458.08 | −1454.57 |
R0 (at endemic equilibrium) | 1.81 | 2.03 | 1.81 | 2.01 | 1.94 | 1.62 | >10 | >10 |
Shedding prevalence (at endemic equilibrium) (%) | 0.18 | 0.25 | 0.303 | 0.41 | 0.718 | 0.70 | 59.44 | 59.77 |
Sero-prevalence (at endemic equilibrium) (%) | 24.64 | 30.55 | 24.54 | 29.92 | 27.73 | 37.80 | 20.49 | 20.17 |
Comments | Good fit; used for 2020–2022 projections | Good fit; used for 2020–2022 projections | Good fit; used for 2020–2022 projections | Good fit; used for 2020–2022 projections | Poor fit to data, incidence in younger ages was under-estimated and incidence in older ages was over-estimated | Poor fit to data, incidence in younger ages was under-estimated and incidence in older ages was over-estimated | Poor fit to the incidence data (high incidence in adults) and unrealistic R0, but better comparison with expected seroprevalence and shedding | Poor fit to the incidence data (high incidence in adults) and unrealistic R0, but better comparison with expected seroprevalence and shedding |