Skip to main content
. 2021 Apr 20;22(5):711–721. doi: 10.1007/s10198-021-01285-1

Box 1.

Algorithm used for bootstrapping

Step 1: Draw a random sample (with replacement) of the 53 patients from the SAVANT data for the SoC + Sativex arm of the randomisation period (phase 3 of the trial)
Draw a random sample (with replacement) of the 53 patients from the SAVANT data for the for the SoC + Placebo arm of the randomisation period (phase 3 of the trial)
Draw a random sample (with replacement) of the remaining patients from the SAVANT data that only completed phase 1 and 2 of the trial
Step 2: Combine the 3 bootstrapped samples from step 1 and calculate the transition probabilities and utilities
Step 3: Draw a random sample (with replacement) of the 8 Delphi panel members for resource use estimates
Step 4: Calculate costs based on the resource use estimates obtained from step 3
Step 5: Calculate and record the total costs and QALYs for both arms for the time horizon of the model
Step 6: Repeat 1,000 times to obtain the probabilistic results

SoC standard of care; SAVANT Sativex as Add-on therapy vs further optimized first-line antispastics clinical trial; QALYs quality adjusted life years