Fig 1.
Attrition diagram that presents the process of choosing four outbreaks to study.
Fig 2.
A snapshot of data relating to the outbreak on board Medic.
Plot (A) displays the daily new severe infections among the ship’s population during the voyage and quarantine period in Sydney. Plot (B) displays the removal of mild or severe infectious/ infected and healthy people on board Medic during the quarantine period. Plot (C) compares the total infections and deaths that occurred with respect to the initial population size by group.
Table 1.
A summary of the ships’ details.
Fig 3.
Proposed models for Medic, Boonah, Devon and Manuka.
The compartments of the models are shown for group i (i=1 (crew); i=2 (passengers)). The population is separated into those who are susceptible (Si), exposed (Ei), asymptomatic (A1i, A2i), pre-symptomatically infectious (IiP), mild (Mi) and severely (IiS) infectious, mild and severely infected (CMi, Ci), severely infectious/infected and quarantined (), and recovered (RAi, Ri, AiQ). The force of infection, λi has terms for both within- and between-group transmission.
Fig 4.
Curve-wise intervals (50%, 75%, and 95%) of conditional re-sampled paths (in green) by the hierarchical estimation algorithm and re-sampled paths (in purple) by the estimated parameters.
Black lines are the median paths, and the blue dot-dashed lines represent the data.
Fig 5.
Posterior distributions of the transmission rates under the hierarchical analysis.
Diagonal: The marginal posterior distributions. Lower triangle: Scatter plots between the transmission rates. Right column: Box-plots of the posterior distributions of the transmission rates.
Table 2.
Posterior medians and 95% HPD intervals of the within and between transmission rates of crew and passengers in the ships.
Fig 6.
Estimated R0 for the ships.
Fig 7.
Interventions vs. no interventions for Medic during their quarantine periods in Australia.
Panels (A) and (D): Black solid line: Observed time-series data. Simulated paths in Grey: Trajectories up to the time the ship arrived at the quarantine Station. Simulated paths in Green: Conditional re-sampled paths. Simulated paths in purple: re-sampled paths. Simulated paths in yellow: Counterfactual paths of no interventions corresponding to conditional re-sampled or re-sampled paths. Panels (B) and (E): Paths of the difference between no interventions and interventions. Panels (C) and (F): Difference in total new infections between no interventions and interventions.
Table 3.
Difference between no interventions and interventions (total infections throughout the epidemic) for Medic.
Fig 8.
Interventions vs. no interventions for Boonah during their quarantine periods in Australia.
Panels (A) and (D): Black solid line: Observed time-series data. Simulated paths in Grey: Trajectories up to the time the ship arrived at the quarantine Station. Simulated paths in Green: Conditional re-sampled paths. Simulated paths in purple: re-sampled paths. Simulated paths in yellow: Counterfactual paths of no interventions corresponding to conditional re-sampled or re-sampled paths. Panels (B) and (E): Paths of the difference between no interventions and interventions. Panels (C) and (F): Difference in total new infections between no interventions and interventions.
Table 4.
Difference between no interventions and interventions (total infections throughout the epidemic) for Boonah.