Skip to main content

Table 5 Sensitivity analyses

From: Financing and cost-effectiveness analysis of public-private partnerships: provision of tuberculosis treatment in South Africa

  PWP model PNP model Purely public model
  Site 1 Site 2 Site 3 Site 4 Site 5 Site6
Alternative estimates of provider costs (staff time)       
• % divergence from base-case provider cost estimate -15% -13% -8% -8% -24% -28%
• % divergence from base-case provider cost-effectiveness estimate -15% -13% -8% -8% -24% -28%
Lower limit patient cost       
• % divergence from base-case patient cost estimate N.A.* N.A. -5% -5% -5% -5%
• % divergence from base-case patient cost-effectiveness estimate N.A. N.A. -4% -3% -1% -1%
Upper limit patient cost       
• % divergence from base-case patient cost estimate N.A. N.A. +5% +5% +5% +5%
• % divergence from base-case patient cost-effectiveness estimate N.A. N.A +3% +0.2% +0.6% +2%
0% discount rate       
• % divergence from base-case provider cost estimate -3% -3% -3% -3% -3% -3%
• % divergence from provider cost-effectiveness estimate -2% -0.6% -0.5% -2% -1% -0.5%
6% discount rate       
• % divergence from base-case provider cost estimate +3% +3% +3% +3% +3% +3%
• % divergence from provider cost-effectiveness estimate +3% +2% +2% +3% +2% +2%
Effectiveness adjusted by the death rate**       
• % divergence from base-case treatment success rate -12% -11% -4% -8% -1% -2%
• % divergence from base-case societal cost-effectiveness estimate (treatment success rate) -12% -11% -4% -8% -1.6% -2.5%
• % divergence from base-case cure rate -12% -11% -4% -8% -1% -2%
• % divergence from base-case societal cost-effectiveness estimate (cure rate) -12% -11% -4% -8% -1% -2%
Multivariate analysis (estimates of provider cost and effectiveness adjusted by the death rate are varied at the same time)       
• % divergence from base-case provider cost estimate -15% -13% -8% -8% -24% -28%
• % divergence from base-case cure rate -12% -11% -4% -8% -1% -2%
• % divergence from base-case societal cost-effectiveness estimate -25% -23% -11% -15% -25% -30%
  1. * N.A. = not applicable.
  2. **Effectiveness was re-calculated by adjusting the total number of patients in the cohort by the total number of deaths in the same cohort. In this case, the treatment success rate = [number of successfully treated patients/(total number of patients – number of patients who died during treatment)] * 100. For example, in site 1, the treatment success rate adjusted by the death rate = [83/(95 - 11)] * 100 = 99%.