Multiple Testing in Cardiovascular Medicine Dhruv S Kazi MD MSc AHA Cardiovascular Outcomes Research Fellow Stanford University kazistanfordedu Off hand Id say youre suffering from an arrow through your head but just to play it safe lets get an echo ID: 739772
Download Presentation The PPT/PDF document "Bonferroni: Friend or Foe?" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Bonferroni: Friend or Foe?Multiple Testing in Cardiovascular Medicine
Dhruv S. Kazi, MD, MSc
AHA Cardiovascular Outcomes Research Fellow
Stanford University
kazi@stanford.eduSlide2
“Off hand, I’d say you’re suffering from an arrow through your head, but just to play it safe, let’s get an echo.”Slide3
Death from Cardiovascular Causes,
Nonfatal Myocardial Infarction, or Stroke
= 9 billion dollars
Yusuf, S, et al. N Engl J Med 2001;345:494-502
CURESlide4
Liver
Clopidogrel
Cytochrome P450-dependent oxidation
Binds to P2Y12 Receptor on Platelets
Ticagrelor
Binds to P2Y12 Receptor on Platelets
CYP2C19Slide5
Potential StrategiesClopidogrel
Ticagrelor
Which one would you want?Slide6
Cannon, CP, et al. Lancet 2010; 375: 283-93.
Primary Efficacy Endpoint in the PLATO-Invasive RCTSlide7
Cannon, CP, et al. Lancet 2010; 375: 283-93.Slide8Slide9Slide10Slide11Slide12
MethodsCohort: 100,000 patients who present with ACS and undergo PCI, age at entry – 65 years
Analytic Horizon: Lifetime
Perspective: “Ideal Insurer”
Interventions
DAPT 12 months from last ACS or PCI, whichever is later
Aspirin
monotherapy
for life thereafterSlide13
Possible Explanations?Slide14
Possible Explanations?True Difference
Chance Finding
Fraud?Slide15
The Multiple-Look Problem
Number of analyses
Cumulative
prob
of
a positive associationSlide16Slide17
So how do you get around this?Traditionally, “don’t run multiple subgroups” unless:
The analyses are pre-specified
The analyses are biologically plausible
And if you must, conduct rigorous statistical adjustment!Slide18
Bonferroni Adjustment ConservativeAssumes independence1-(1-
α
)
1/n
~
α/n
But does this make sense? BMJ. 1998 April 18; 316(7139): 1236–1238.Slide19Slide20
How Do We Proceed? (Do you still want the drug?)
Multiple testing is problematic (even if pre-specified)
The challenges of a priori hypothesesSlide21
ConclusionsMultiple testing is a complicated question: with real clinical consequencesStatistical adjustment is a necessary but imperfect solution
Trial and Error. Kaul S, et al. J Am Coll Cardiol 2010;55:415–27Slide22
ConclusionThe p value is no substitute for a brain. Slide23
Thank You!
kazi@stanford.edu