Cost Effectiveness of Intermediate-Risk Squamous Cell Carcinoma Treated with Mohs Micrographic Surgery Compared with Wide Local Excision

Published: 29 July 2021| Version 2 | DOI: 10.17632/j6r3rd5jhv.2
Jeremy Udkoff,
Brandon Beal,
David Brodland,
Thomas Knackstedt


Mendeley Supplemental Figure 1: Tornado Diagram. This diagram shows the sensitivity of the model to a +/- 50% change in each variable. Increasing the value of a probability, cost, or effectiveness variable (red colored bar) results in a certain change in the incremental cost-effectiveness ratio (ICER, displayed on the X-axis) output of the model while decreasing the value (blue colored bar) results in an alternative change. The baseline model demonstrated that MMS was superior to WLE in both cost and effectiveness resulting in a negative ICER (WLE was dominated). The model’s ICER output was most sensitive to the probability of NR after WLE (first row in the diagram). However, as discussed in the manuscript, a 50% (or even 100%) reduction in the variable from 3.97% to 1.99% (or 3.97% to 0.00%) would not result in WLE being cost-effective; a value of 1.99% results in an ICER of around $0, so costs could be equivalent but there is increased effectiveness with MMS. If one were to test extreme alterations in the variables, the X-axis could be expanded to the right to include $50,000 (the willingness to pay threshold for our model). This point would be where one would consider whether the proposed intervention (MMS) was worth paying for. However, we did not encounter this point of equivalence during sensitivity analysis.