Attrition and Data Handeling Flashcards
Identify where in the methods data handling is described
-look at the SUPPLEMENT
-statistical analysis
define attritation
-the action or process of gradually REDUCING the strength or effectiveness of someone/something
-loss of participants from a randomized control trial
effects of attritation
-introduction of bias
-impact on trial results
-benefits of randomization are lost
-no specific level where it becomes a problem
Key questions for attrition
-Are the reasons for incomplete/missing data adequatly explained?
-Do the authors’ attempt to address attrition in analysis
reporting attrition
-standard: include table with baseline characteristics of trial participants
-Ideal: present baseline and those lost seperatly
true or false: ITT increases internal validity and PP does not
False. both decrease internal validity
Intention to treat (ITT)
-analyzed as randomized
-conservative estimates in the difference between treatments (less favorable to intervention)
-preserves sample size which has more statistical power
per protocol analysis
-as treated
-analyze only study participants who complete the full study
-benefits of randomization are lost
-more generous estimate between treatments
True or false: for noninferiority trials, both ITT and PP analyses should be presented
true