BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:“Statistical opportunities and challenges in observational healt
 h data: Addressing missingness using two-phase sampling for nonresponse: m
 ethods and benefits&quot\; - Dr Matthew Sperrin\, University of Manchester
DTSTART:20170112T143000Z
DTEND:20170112T153000Z
UID:TALK69772@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Health data are now routinely collected across different parts
  of the health system\, and linked at the patient level across data source
 s.  This suggests an extremely rich resource that we can use to make accur
 ate outcome predictions for patients - hence stratification and treatment 
 targeting. However\, these data are 'messy'\, potentially unstructured\, a
 nd subject to 'informative observation' (since patients only generate data
  when interacting with healthcare services). Therefore\, to fully exploit 
 the vast data being generated in health\, careful statistical planning and
  methodological development is required.\n\nIn this talk I will describe r
 ecent work using observational health data - including some work in which 
 we have begun to unpick informative observation\, and other work around us
 ing\, combining and updating predictive models. I will describe the emergi
 ng opportunity of a learning health system\, in which linked data across t
 he health system will be available for research in real time\, demanding d
 ynamic prediction\, and an opportunity to reduce the 'data action latency'
  the time period between data being available and appropriate action being
  taken.
LOCATION:Large  Seminar Room\, 1st Floor\, Institute of Public Health\, Un
 iversity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR