Two weeks ago, Kurcharski et al., from the at the London School of Hygiene and Tropical Medicine, published on medrXiv a statistical analysis via a stochastic SEIR model of the evolution of the 2019-2020 nCoV epidemics, with prediction of a peak outbreak by late February in Wuhan and a past outbreak abroad. Here are some further details on the modelling:
Transmission was modelled as a geometric random walk process, and we used sequential Monte Carlo to infer the transmission rate over time, as well as the resulting number of cases and the time-varying reproduction number, R, defined as the average number of secondary cases generated by a typical infectious individual on each day.
To calculate the likelihood, we used a Poisson observation model fitted jointly to expected values based on three model outputs. To calculate the daily expectation for each Poisson observation process, we converted these outputs into new case onset and new reported cases inside Wuhan and travelling internationally. We assumed a different relative reporting probability for Wuhan cases compared to international cases, as assumed only a proportion of confirmed Wuhan cases had known onset dates (fixed at 0.15 based on available line list data). As destination country was known for confirmed exported cases, we used 20 time series for cases exported (or not) to most at-risk countries each day and calculated the probability of obtaining each of these datasets given the model outputs. International onset data was not disaggregated by country and so we used the total daily exported cases in our Poisson probability calculation.
I did not look much further into the medrXiv document but the model may be too simplistic as it does not seem to account for the potential under-reporting within China and the impact of the severe quarantine imposed by Chinese authorities which may mean a new outbreak as soon as the confinement is lifted.