SMS scnews item created by Michael Stewart at Mon 19 May 2014 1848
Type: Seminar
Distribution: World
Expiry: 21 May 2014
Calendar1: 20 May 2014 1800-1930
CalLoc1: Gregg Lecture Theatre, Level 2, Edward Ford Building (A27), School of Public Health
Auth: michaels@pmichaels.pc (assumed)

NSW Stat Society Monthly Talk: Simpson -- Analysis of Linked Data in Perinatal Population Health Research

This month’s NSW Statistical Society talk is on Tuesday May 20 and is being given by 
Professor Judy Simpson from the School of Public Health at the University of Sydney.  

There will be light refreshments from 6pm, and the talk is scheduled to run from 6.30 to
7.30.  

The venue is the Gregg Lecture Theatre, Level 2, Edward Ford Building (A27), School of
Public Health, University of Sydney.  

The title, abstract and a short bio appear below.  

Cheers, 

Michael 

-- 

 Analysis of Linked Data in Perinatal Population Health Research 

Perinatal researchers study what happens to mothers and babies before, during and after
birth.  This research has been greatly facilitated by the availability of data linkage,
enabling hospital records of mothers to be linked to those of their babies (including
twins) and of their own previous and subsequent pregnancies and other hospital
admissions and registries.  

The NSW Perinatal Data Collection (PDC) is a legislated population-based surveillance
system which monitors patterns of pregnancy care, services and pregnancy outcomes.  It
covers all births in NSW, i.e., live born babies regardless of gestational age and
stillbirths of at least 20 weeks gestation or 400 grams birth weight.  For multiple
births, a separate record is completed for each baby.  Each record contains details of
the mother, her previous pregnancies, this pregnancy, labour and delivery, the baby, and
postnatal care.  

Examples will be described of the use of linked PDC data to address different types of
research question.  These include: examining the impact of the ‘baby bonus’ on birth
rate by age, parity, socioeconomic status and geography; explaining increases over time
in adverse maternal and infant outcomes using logistic regression; estimating the risk
of complications in a second pregnancy following caesarean section in the first
pregnancy; using longitudinally linked data to validate cross-sectional reporting of
previous caesarean section; using linked data to obtain outcomes for a randomised trial;
examining risk factors for late preterm singleton birth, taking into account multiple
observations for the same mother using generalised estimating equation (GEE) methods;
and investigating variations in outcomes among hospitals.  

Biography of Prof. Judy Simpson: Judy Simpson is Professor of Biostatistics in the
School of Public Health at the University of Sydney with 30 years’ postdoctoral
experience in applying biostatistical techniques to the design and analysis of research
in public health and clinical medicine.  During her career as a biostatistician, she has
published extensively both nationally and internationally, including over 200 articles
in peer-reviewed journals.  Her main research interests are: design and analysis of
randomised controlled trials, especially cluster-randomised trials; statistical
modelling, including hierarchical (multilevel) modeling; survival analysis, especially
analysis of multiple failure time data; and analysis of linked data sets.  She has been
collaborating for several years in the analysis of linked data with Associate Professors
Christine Roberts and Jane Ford, perinatal epidemiologists.