Table 1.
Symbol | Name | Type | Description |
---|---|---|---|
i | Index | Index of individual | |
h | Index | Index of household | |
g | Index | Index of RSV group type, either A or B | |
t | Index | Index of time in days | |
Infectivity | Dataa | Categorical data variable for infectious individuals indicating level of infectivity categorized by viral load and symptom status at time t. The categories are: low viral load and asymptomatic (reference group), high viral load and asymptomatic, low viral load and symptomatic and, high viral load and symptomatic. High viral load is defined as >6 log10 viral copy number. | |
Infection_history | Data | Variable indicating if an individual has experienced and recovered from an infection by a particular RSV group in the current epidemic at time t. | |
Age_groupS | Data | Categorical data variable indicating the susceptibility age group of an individual. The age groups are <1 year (reference group), 1-4 years, 5-14 years and ≥15 years. | |
Data | Binary data variable indicating if an individual is present in the household at time t. Absence from the household means that an individual was not present at the point of sample collection and thus in the model they can only get infection from a community source and not from an infectious housemate (not sampled and not at household risk). Individuals who were present but not sampled are exposed to both household and community source transmission in the models (not sampled but at household risk). | ||
Household_size | Dataa | Binary data variable indicating whether the individual lives in a large or small household. A small household (reference group) has <8 individuals. | |
Age_groupE | Data | Categorical data variable indicating the community exposure age group of an individual. The age groups are <1 year (reference group), 1-4 years and ≥5 years. | |
Sus.age.2 Sus.age.3 Sus.age.4 |
Parameter | Coefficients modifying susceptibility to RSV depending on age, applied to the age group covariate Xi. Sus.age.2 estimates the effect being in age group 1-4 years, Sus.age.3 the effect of group 5-15 and Sus.age.4 of group ≥15 relative to group <1 year. | |
Prev.hom Prev.het |
Parameter | Coefficients modifying susceptibility to infection by a particular RSV group depending on infection history. Prev.hom estimates the effect of a previous homologous group infection, while Prev.het estimates the effect of a previous heterologous group infection. Applied to the categorical covariate Yi(t). | |
HH.size | Parameter | Coefficient modifying the amount of within household exposure by household size. HH.size estimates the effect of being in a large household relative to a small one. Applied to covariate Hi. | |
HH.rsv.a HH.rsv.b |
Parameter | Baseline rate of within household exposure by RSV group | |
High.Asym Low.Sym High.Sym |
Parameter | Coefficients modifying infectiousness by viral load and symptom status. Relative to shedding low viral load and being asymptomatic, High.Asym estimates the effect of shedding high viral load and being asymptomatic, Low.Sym the effect of shedding low viral load and being symptomatic and High.Sym the effect of shedding high viral load and being symptomatic. Applied to the infectivity covariate . | |
Exp.age.2 Exp.age.3 |
Parameter | Coefficients modifying the rate of community exposure by age group. Exp.age.2 estimates the effect being in age group 1-4 years and Exp.age.3 the effect of group ≥5, relative to the <1-year age group. Applied to the age group covariate Ei | |
Comm.rsv.a Comm.rsv.b |
Parameter | Community transmission coefficient by RSV group | |
Estimated | RSV group specific, time-dependent curve modifying the rate of community exposure. | ||
Data | Set of group specific onset days for an individual i in household h used in calculating the likelihood of an individual’s data. |
The choice of cut-off for high viral load and large households was based on initial runs of the inference algorithm that explored different cut-offs for each. The choice of 6 log10 copy number for high viral load and 8 persons for large households led to the best convergence.