Table 1

Model parameters, probability distributions and data sources for the decision-analytical model for economic evaluation of municipal ice cleat distribution programmes

Parameter	Average (range, if applicable)	SE	Distribution (probabilistic sensitivity analysis)	Data source
Annual number of snow-related or ice-related fall injuries at baseline as a function of population size and climate ()	21.6 (2.7, 472.6)	Municipality-specific SE from regression prediction	Lognormal	Municipality-specific and age-specific data from National Patient Register21. Population data from Statistics Sweden30. Annual number of snow days from the Swedish Meteorological and Hydrological Institute*
Initial change in ice cleat use as a function of climate ()	0.25 (0.09, 0.35)	Municipality-specific SE from regression prediction	Logit-normal	National survey (random sample, n=4608 aged 65+) conducted in 2007 by the Swedish Civil Contingencies Agency*
Compliance over time, multiplicative scaling factor ()	{	Not available	None	Calibrated* simulation model to results from quasi-experimental evaluation in Gothenburg13
Effect of ice cleat use (RR), log scale	−0.799	0.333	Lognormal	Randomised controlled trial12, RR=0.45 (95% CI: 0.23 to 0.85)†
Total programme cost per purchased ice cleat pair in 2018 Euros (c), log scale	1.998	0.100	Lognormal	Electronic survey sent to all Swedish municipalities (n=34 responses with cost data)*
Benefit per averted injury in 2018 Euros	329 783	Not available	None	Swedish Transport Administration17

*Additional information on data and estimation is provided in the online supplemental file.
†RR for falls (with or without an injurious outcome). Estimate for injurious falls was 0.1 (95% CI: 0.02 to 0.53) but was only based on a total of 11 events (one in the treatment group and 10 in the control group). Our preferred estimate (for falls) is supported by more data and is more conservative.
RR, relative risk.