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:Cambridge Statistics Discussion Group (CSDG)
SUMMARY:Working with Epidemiologists - Nick Galwey\, Glaxo
SmithKline
DTSTART;TZID=Europe/London:20140327T191500
DTEND;TZID=Europe/London:20140327T213000
UID:TALK47526AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/47526
DESCRIPTION:This presentation first explores the motivation fo
r conducting epidemiological research: such resear
ch ranks well below the ‘gold standard’ of randomi
sed clinical trials in the conventional hierarchy
of evidence\, yet observational studies (sometimes
on >100 million patients) provide valuable inform
ation that is rarely obtained from such trials. H
ence the problems of confounding and bias (two sid
es of the same coin) that beset observational stud
ies are well worth addressing\, and in this presen
tation the statistical tools that epidemiologists
use for the purpose – adjustment for covariates\,
case-control matching etc. – are reviewed. The s
tatistics that epidemiologists typically present a
re incidence and prevalence rates\, and their rati
os between exposed and unexposed patients. These
statistics are interpreted in terms of the Poisso
n and binomial distributions\, rather than the Nor
mal distribution used for continuous variables\, s
o the statistician must be familiar with the metho
ds for fitting models using these distributions –
Poisson regression with overdispersion\, logistic
regression\, etc. He or she also needs to be awa
re of the difficulty of specifying confidence inte
rvals for parameter estimates based on these disco
ntinuous\, asymmetric distributions: there is no s
ingle ‘right answer’. In the author’s experience\,
epidemiologists have breathtakingly well develope
d intuition for the ways in which data can mislead
\, and what to do about it. But the majority do no
t habitually express their ideas in algebra and ge
ometry\, and when the tasks required of them take
them beyond the edge of their statistical comfort
zone they will welcome the support of a statistici
an who is a fast learner and a good communicator.
LOCATION:Institute of Public Health\, Forvie Site\, Addenbr
ookes Hospital
CONTACT:Peter Watson
END:VEVENT
END:VCALENDAR