Skip to main content

Table 1 Summary of simulation variables

From: Assessing methods for dealing with treatment switching in randomised controlled trials: a simulation study

Variable

Scenarios

Details

Sample size

1

500 patients, 250 in each treatment arm

Weibull shape parameter γ

1

0.5, to represent mortality rate decreasing over time

Weibull scale parameter λ

1

1.33, chosen such that 90% of patients have died after 3 years of follow-up

Probability of patient having 'good' prognosis

2

30% or 75%

Difference in survival between 'good' and 'poor' Prognosis groups

2

Survival times of 'good' prognosis group multiplied by a factor of either 1.2 or 3

Probability of switching treatment dependent on prognosis group

2

10% ('good' prognosis) and 25% ('poor' prognosis) or 50% ('good' prognosis) and 75% ('poor' prognosis)

Switching time

1

Generated from a Uniform distribution

Initial treatment effect

2

Hazard ratio of 0.9 or 0.7