Table 4

Sensitivity analysis results

Modified parameterBase caseβ multiplied by 0.68β doubleds0=34%s0=100%Infectious period doubledLatent period halvedRDT sensitivity halved
r valuer=5r=3.4r=10r=3.4r=10r=10r=5r=5
Use case 1: surveillance
 Time available for boosting (respectively ICU capacity building) with surveillance17 (47)28 (57)8 (37)25 (55)9 (39)14 (46)11 (41)14 (44)
 Time available for boosting (respectively ICU capacity building) without surveillance3 (33)9 (39)0 (28)7 (36)0 (30)1 (32)0 (30)3 (33)
 Time gained for boosting or ICU capacity building through the use of surveillance14199199141111
Use case 2a: TTI* (without surveillance)
 Time to boost (respectively build ICU capacity)1 (2)5 (5)1 (1)3 (3)1 (1)1 (1)2 (2)1 (1)
 Percentage reduction in peak ICU demand5%17%2%16%2%3%7%2%
Use case 2b: TTI* (with surveillance)
 Time to boost (respectively build ICU capacity)4 (4)12 (13)3 (3)11 (10)3 (3)4 (5)5 (6)2 (2)
 Percentage reduction in peak ICU demand6%20%2%20%2%4%8%3%
Use case 3: health facilities
 Percentage reduction in ICU admissionsHigher r lowers the percentage reduction in nosocomial COVID-19 achievable through screening but increases nosocomial COVID-19 as a share of hospitalised COVID-19 cases.Lower benefits
Use case 4: test and treat
 Percentage reduction in hospital admissionsBenefits unchanged if the percentage of cases reached and linked to care does not change.Benefits halved
 Percentage reduction in total deaths
  • Source: authors’ simulations using the model in online supplemental appendix A.

  • *The values used for these simulations correspond to an optimal testing, tracing and isolation scenario (country type F or 80% tested, 75% isolating and 75% traced).

  • ICU, intensive care unit; RDT, rapid diagnostic test; TTI, testing, tracing and isolation .