Figure 2.

Case fatality ratio (CFR) of patients seeking care divided by the CFR if those same Ebola and non-Ebola cases had remained in the community. Each row represents a particular stage in the epidemic: from left to right: early (a,b, c), during the peak (d,e,f), shortly after the peak (g,h,i) and once the epidemic is tailing off(j,k,l). Each column reflects a testing strategy, namely polymerase chain reaction (PCR)-only (a,d,g,j), dual strategy (b,e,h,k) and rapid diagnostic test (RDT)-only (c,f,i,l). White horizontal lines show the threshold bed capacity below which demand cannot be met for PCR-only (same threshold as dual strategy) and RDT-only. Solid grey and black lines (left panels, a,d,g,j) indicate, respectively, where the outcomes of PCR-only and RDT-only are equivalent, and where the outcomes of dual (RDT and PCR) testing and RDT-only are equivalent. Those lines delimit parameter space where (1) dual strategy is best followed by PCR-only and then RDT-only, (2) dual strategy is best followed by RDT-only and then PCR-only and (3) RDT-only is best followed by dual strategy and then PCR-only. On the left of the white solid vertical line (specific for the testing the benefit of care is sufficient to decrease the average CFR among patients seeking care (unaware of their disease status, and assuming hospital infection control has not improved over the course of the epidemic). The black arrows on the right y-axis of the RDT-only plots indicate the likely availability of beds at the corresponding stage of the epidemic (Table 2); however, this is likely to have varied between different health-care units.