Fig 1.
Schematic flow chart between compartments.
Table 1.
Priors of timescales and ratios for the MCMC modeling.
Table 2.
Age-dependent parameters.
Fig 2.
Left: The time series of fatalities for a number of countries. Right: the time derivative of the curve of fatalities is converted into according to Eq (19).
Fig 3.
a) Evolution of the compartments of exposed (E), asymptomatic (A), symptomatic (I) and hospitalized (H). b) The same curve of H, and the fraction of hospitalizations needing ICU (U). The green dashed curve shows the total number of ICU beds in the country. At the current rate, the epidemics should peak in mid-May and collapse of the health care system should happen around May 1st. c) Cumulative number of reported cases and fatalities as orange and black dots, respectively. The number of hospitalizations closely matches the number of confirmed cases.
Fig 4.
Posterior probabilities for the epidemic parameters determined by the MCMC modeling.
Fig 5.
The result of 2187 simulations, where the parameters used in Fig 3 are varied by -2, 0, and 2 standard deviations, as given by the MCMC analysis of Fig 4.
The model of Fig 3 (zero standard deviation on all parameters) is shown as the thick line. The 95% confidence interval brackets about an order of magnitude above or below, or about three weeks left or right of this fiducial model.
Fig 6.
a) Number of hospitalizations. b) ICU cases. c) Fatalities. The three curves are shown as function of the degree of horizontal confinement. To not overwhelm the health care system capacity (≈ 3 × 104) ICU beds, the level of social distancing should be over 70%. Brazil is managing 56%.
Fig 7.
Upper: Number of hospitalizations in horizontal confinement with different proportions of the population, broken down by age. 2nd row: 60% of the population under 40 is confined, the population older than 40 is confined to a higher degree, at 90% (solid blue line) and 99% (dashed blue line). The 3rd, 4th, and 5th rows of plots show the same analysis but confining 60% of the population up to 50, 60, and 70 years old, respectively.
Fig 8.
Middle: Same as Fig 7, but for fraction of hospitalized that need ICU.
Fig 9.
Same as Fig 7, but for the number of fatalities.
Fig 10.
Vertical confinement as exit strategy after a horizontal lockdown of 60% was held, from March 22nd to May 1st.
Upper plots: S and C compartments, normalized by the number of individuals in the respective age bin. 2nd row: number of hospitalizations, 3rd row: number of ICU cases, lower plots: cumulative number of fatalities. The figure shows full release for the population under 50 (dashed line) and under 60 (solid line). The population past this age is kept at 90% confinement.
Table 3.
Input parameters and arrays for MCMC.