Based on an original page posted by Nick Grassly on the H1N1 pandemic website.
The basic reproduction number of the swine influenza epidemic, \(R_{0}\) can be estimated from its initial rate of spread. If we assume roughly exponential growth then the basic reproductive number is related to the growth rate by the so-called Lotka-Euler estimating equation:
\[R_{0}=\int_{0}^{\infty} \frac{1}{w(\tau)e^{-rt}} d\tau\]where \( r \) is the rate of exponential growth and \( w(\tau) \) the generation time distribution [1]. The generation time distribution can be thought of as the probability density function describing the distribution of times between successive infections in a chain of transmission.
The estimate of the basic reproductive number is therefore dependent not just on an estimate of \( r \) , but also a good estimate of the generation time distribution [2]. In the case of swine influenza the generation time distribution is unclear, but data appear consistent with seasonal influenza that has a mean generation time of approximately 3 days.
Analytical solutions for \( R_{0} \) can be derived for different assumed generation time distributions using the Lotka-Euler estimating equation (which is essentially a moment generating function). If we assume a generation time distribution that follows the gamma distribution, then
\[ R_{0}=\left(1+\frac{r}{b}\right)^{a} \]
where a and b are the parameters of the gamma distribution (\( a = m^{2}/s^{2} \) and \( b = m/s^{2} \) where \( m \) and \( s \) are the mean and standard deviation of the distribution respectively).
Estimates of \( R_{0} \) based on the estimates of \( r \) reported by Andrew Rambaut are given in Table 1. Obviously \( r \) can also be estimated from epidemiological case data and this may give different results.
| \( r \) (per day with 95% HPD) | generation time distribution (and parameters in days) | \( R_{0} \) |
|---|---|---|
| 0.053 (0.0014, 0.12) | gamma (\( m=3 \), \( s=2 \)) | 1.17 (1.00 - 1.40) |
| 0.053 (0.0014, 0.12) | exponential (\( m=3 \)) (i.e. SIR model) | 1.16 (1.00 - 1.36) |
Table 1 | Estimates of \( R_{0} \) from the coalescent growth rate, \( r \) for the early period of Pandemic H1N1.
Citations
- Wallinga J, Lipsitch M. (2007) How generation intervals shape the relationship between growth rates and reproductive numbers. Proc Roy Soc Lond B 274: 599-604
- Grassly NC, Fraser C. (2008) Mathematical models of infectious disease transmission. Nat Rev Microbiol. 6: 477-487