BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:MRC Biostatistics Unit Seminars
SUMMARY:Virtual BSU Seminar: 'Stochastic treatment interve
ntions in causal survival analysis' - Dr Lan Wen\,
Harvard University
DTSTART;TZID=Europe/London:20201013T140000
DTEND;TZID=Europe/London:20201013T150000
UID:TALK150814AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/150814
DESCRIPTION:Several methods are available for estimating the c
ausal effect of time-varying treatment strategies
on survival outcomes in observational studies. The
se include singly robust methods such as inverse p
robability weighting (IPW) that requires a sequenc
e of correctly specified models of the observed tr
eatment distribution (the propensity score)\, and
iterative conditional expectation (ICE) that requi
re a sequence of correctly specified models of the
nested conditional outcome means. Alternatively\,
doubly robust estimators that combine IPW and ICE
require that only one of the sequences of models
be correctly specified\, and thus offer more than
one opportunity for valid estimation. In recent ye
ars\, these methods have been generalized to accom
modate effects of stochastic strategies such that
treatment assignment at each time is non-determini
stic within levels of the measured past. Many aut
hors have considered stochastic strategies that de
pend on the propensity score which would suggest t
hat doubly robust estimators are not possible to c
onstruct. However\, this is not the case. In this
talk\, I will give an intuition into why some stra
tegies that depend on the propensity score can sti
ll be estimated by doubly robust estimators\, and
describe a class of stochastic treatment intervent
ions that will always have doubly robust estimator
s in point treatment processes and multiply robust
estimators in longitudinal observational studies.
I also propose a new stochastic treatment interve
ntion dependent on the propensity score motivated
by an application to Pre-Exposure Prophylaxis (PrE
P) initiation studies that allows doubly and multi
ply robust estimators.
LOCATION:Virtual Seminar
CONTACT:Alison Quenault
END:VEVENT
END:VCALENDAR