RCT class #10 Flashcards
3 reasons to stop a trial due to efficacy reasons
- benefit
- harm
- impossibilty of finishing due to low recruitment.
Who addresses the questions of ongoing data analysis
Data and Saftey Monitoring board
T/F Stopping rules are usually applied to the primary efficacy outcome only.
T
DSMB makes a recommendation or definitive statement
recommendation.
Interim monitoring in trials is ___ ____
group sequential meaning interm data looks are taken after groups of patients have completed their follow up.
group sequential monitoring
take interm looks at data after groups of patients have completed their follow up. For this we use repeated sinificance testing.
Repeated significance testing
when do number of looks need to adjust power calcilation. # of final and # of interum for a total
how to word group sequential monitoring
of final + # of interim for a total # of looks
Z score shap
Side ways funnel. Need collosal effect size in the beginning (giving up info so it had better be good).
Who started repeated significance testing and what did they do
Pocock and O’brien-Fleming. prespecified equeally spaced looks.
Pocock/ Obrien/ Flemmings repeated significance testing had problems a.)name some
b.) how was this changed and by whom and when
a. ) 1. want to vary amount of information time with DSMB meeting times. stistically adjust stopping boundary
2. DSMB meets and wants to look early then schedules and additional look…
b) error spending (alpha spending strategy by Lan and Demets.
Lan and demets alpha spending strategy
this is the most widely currently used. DSMBs have meetings and want to use calendar time rather than information fraction. uses flexible sequential methods to
In PASS what is the alpha spending called
obrien flemming (meaning alpha spending appicaltion of pbrien fleming
