STEP 7: Sensitivity Analysis (SA)-Vary Cost and Impact Assumptions
Several sensitivity analysis (SA) options can be selected: one variable, multiple variable, and threshold analysis. An SA may vary intervention cost, and impact rates, or conduct an analysis to determine at what point cost and impact reach a “threshold.” Personnel and material costs for an HP-DP program will differ by program infrastructure, type of staff, location, from state to state, and by levels of behavioral impact or health outcome rates. The cost of staff and variations in effectiveness of a program in Boston, Massachusestts, will be very different from those of a program in Columbia, South Carolina.
An organization may decide in an SA to vary staff or costs by 10% or 20% and effectiveness by 10% or 20%. By varying impact levels and cost assumptions, using low-medium-high estimates, an SA documents what changes would occur in the conclusions with each variation. In performing an SA, it is important to emphasize that variations of assumptions should be rational. The following five case studies provide examples of the application of the steps and methods used to conduct a CEA and CBA, including performing sensitivity analyses.