-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Possible inconsistency in getSimulationSurvival() with specified thetaH0 and thetaH1 #25
Comments
Hi Tim, |
Issue was fixed in branch dev/3.5.2 |
…ecalculation rules to the setting of binary endpoints according to [Bokelmann et al. (2024)](https://doi.org/10.1186/s12874-024-02150-4) * The `getSimulationMultiArmMeans()`, `getSimulationMultiArmRates()`, and `getSimulationMultiArmSurvival()` functions now support an enhanced `selectArmsFunction` argument. Previously, only `effectVector` and `stage` were allowed as arguments. Now, users can optionally utilize additional arguments for more powerful custom function implementations, including `conditionalPower`, `conditionalCriticalValue`, `plannedSubjects/plannedEvents`, `allocationRatioPlanned`, `selectedArms`, `thetaH1` (for means and survival), `stDevH1` (for means), `overallEffects`, and for rates additionally: `piTreatmentsH1`, `piControlH1`, `overallRates`, and `overallRatesControl`. * Same as below for`getSimulationEnrichmentMeans()`, `getSimulationEnrichmentRates()`, and `getSimulationEnrichmentSurvival()`. Specifically, support for population selection with `selectPopulationsFunction` argument based on predictive/posterior probabilities added (see [#32](#32)) * Issues [#25](#25), [#35](#35), and [#36](#36) fixed * Minor improvements
Hello,
I am trying to understand thetaH1 in getSimulationSurvival. I am interested in an adaptive design involving a thetaH0 bound of 0.9. Note that I am interested in testing H0: HR >= 0.9 vs. H1: HR < 0.9, that means the 0.9 bound is not a non-inferiority bound in the classical sense. In the first scenario I ran a simulation without specifying thetaH1, which should be using the test-statistics of stage 1 and 2 and the observed hazard ratio of stage 2 to reassess the required events for stage 3. I picked out an iteration which had a hazard ratio of 0.4503625 at stage 2. Then I re-ran the simulation, using the same seed, with thetaH1 = 0.4503625 and looked at the same iteration again. Given I used the same seed, it was expected that the results including stage 2 test-statistic and observed hazard ratio are the same. However, this time, the reassessement, which in my understanding only requires the test-statistics of the previous stages (which here are the same between the two scenarios) and an assumed hazard ratio, which I also made sure to be the same given my specification of thetaH1, reported an amount of events for stage 3 which differs from the first scenario.
My code is below. Unless my understanding of the subject is incorrect, I do suspect some kind of inconsistency in regards the thetah0 bound between the two scenarios.
Kind regards
Tim
The text was updated successfully, but these errors were encountered: