Table 2

Results from the deterministic sensitivity analysis

ScenarioExpected NPV
(in million Euros)
Benefit-to-cost ratioPr(cost-beneficial),
mean (min–max)
Base-case result (for reference)1192.62 (1061.77, 1338.12)87.44 (77.69, 98.31)0.991 (0.989, 0.992)
Increase cost to highest reported cost per procured ice cleat (€27.9 per pair)1141.55 (1012.03, 1286.41)23.55 (20.93, 26.49)0.988 (0.981, 0.99)
Reduce WTP per QALY gained to match healthcare sector (€50 000 per QALY)245.25 (217.05, 276.71)18.98 (16.85, 21.35)0.987 (0.977, 0.989)
Reduce baseline risk by a factor of 0.59 to match warmest year between 2001 and 2019688.77 (611.34, 774.01)51.5 (45.72, 57.89)0.990 (0.987, 0.991)
Reduce initial compliance rate from 25% to 5% on average281.86 (245.93, 325.25)21.66 (18.96, 24.91)0.986 (0.978, 0.989)
Reduce RR of ice cleat use by half (RR=0.73)405.86 (361.81, 453.98)30.75 (27.44, 34.38)0.989 (0.985, 0.991)
Reduce RR of ice cleat use by three quarters (RR=0.875)166.5 (148.28, 186.02)13.21 (11.82, 14.69)0.985 (0.974, 0.988)
Limit intervention effect to first year443.49 (391.82, 502.75)33.51 (29.64, 37.93)0.989 (0.985, 0.991)
Increase discount rate to 5%1160.59 (1032.03, 1303.57)86.09 (76.49, 96.80)0.991 (0.989, 0.992)
Cost-minimisation analysis (ignore WTP per averted injury)−0.65 (−1.76, 0.62)0.95 (0.85, 1.07)0.405 (0.004, 0.720)
  • The table shows the results from additional scenarios to test the sensitivity of the base-case results to deterministic variations in key assumptions and input parameters. Details and rationale for each scenario is presented in the online supplemental file to this article. The estimates in the table reflect totals or means for all Swedish municipalities averaged across 100 000 simulations, with the 2.5th and 97.5th percentiles of the simulated estimates in parentheses unless otherwise noted. The simulations are run over a 4-year period. The net present value (NPV) is given by Equation (1), which, if positive, implies that the interventions are cost-beneficial. The benefit-to-cost ratio expresses how much the estimated benefits outweigh the costs in relative terms. Pr(cost-beneficial) is the proportion of the 100 000 simulations in which the NPV is positive, which gives an overall estimate of how likely it is that an ice cleat programme would be cost-beneficial according to the model (for this parameter, the numbers in parentheses reflect the least to most certain municipal-specific estimate).

  • Pr, probability; QALY, quality-adjusted life years; RR, relative risk; WTP, willingness to pay.