AGGREGATE DATA FROM EUROPEAN COUNTRIES: THE EURO-PERISTAT PROJECT

Get Complete Project Material File(s) Now! »

Aggregate data from the US, Canada and Japan: The PREBIC project

The collaboration between Euro-Peristat and PREBIC started in 2012 and focused on a shared methodology for preterm birth analyses using international data. To this aim, the Euro-Peristat data collection protocol was expanded to countries participating in PREBIC: the US, Canada and Japan. The PREBIC Epidemiology Working Group is a multi-disciplinary network of scientific experts focused on the prevention of preterm birth through basic, epidemiologic, and applied clinical research. The project was initiated with support from the World Health Organization and the March of Dimes foundation. For this thesis, national experts provided data on the gestational age distribution of births in their country and country-specific input regarding the interpretation of the analyses.

Aggregate data from Australia: New South Wales

Aggregate data from New South Wales, which represents a third of births in Australia were obtained from Dr. Natasha Nassar who had access to the NSW Births database at the Menzies Center of Health and Policy. These were also collected using the Euro-Peristat protocol.

Individual-level data from France: French National Perinatal Survey 2010

In addition to data provided to the Euro-Peristat project in aggregate form, individual level data from the French Perinatal Survey in 2010 (Enquête Nationale Périnatale 2010 ) were used. The French Perinatal Surveys are population-based representative surveys which are conducted over the course of one week in public and private maternity units in Metropolitan France. Data collection is coordinated by the INSERM UMR1153-EPOPé research team.

Data and definitions

Aggregate data on the gestational age distribution, defined as the number of births at each completed week of gestation, were collected using a common protocol on births in 2010 (26 EU Member States and Norway, Switzerland and Iceland), 2004 (25 EU Member States and Norway) and 2000 (15 EU Member States). Aggregate data were stratified by multiplicity (singleton/multiples) and vital status (stillbirth/live birth). In a separate study on preterm birth, Euro-Peristat collected information on the number of singleton live births by completed week of gestation by method of delivery and mode of onset in 17 countries/regions for the years 1996, 2000, 2004 and 2008 (124).
Aggregate data items included in these 2 databases were collected from other participating countries using the Euro-Peristat data collection protocol. In all studies, gestational age was requested using the best obstetric estimate. When gestational age was missing, we asked countries to include births if birth weight was 500 grams or more. If countries could not provide data using these criteria, they were asked to provide available data using their own inclusion criteria, and to specify their inclusion thresholds for births and deaths. We also asked for separate data on late terminations of pregnancy (TOP), when these were included in the data sources and could be differentiated from spontaneous stillbirths. Data were available on mode of delivery and mode of onset of labour. Indicated deliveries were defined as those that were provider-initiated including: inductions, prelabor and emergency cesarean sections, based on national classifications.

Study population

Population-based data on the gestational age distribution were available from 34 countries in North America, Europe, Japan and Australia in at least one time point for which data collection was undertaken (1996, 2000, 2004, 2008 and 2010). However, not all countries could provide all data points and these differences in data availability are accounted for in our analyses as detailed in the specific chapters. Also, some countries only had data available for selected regions. Data for Belgium (BE) came from Brussels, Wallonia and Flanders; data from the United Kingdom were provided separately by the UK’s constituent countries: England and Wales combined, Northern Ireland, and Scotland. In 1996-2000, Germany provided data from three Länder: Hesse, Lower Saxony and Bavaria. In 2008, data from Canada included all provinces and territories except the province of Ontario; in 2010, data from all provinces were included except for Québec. Some countries had different years of data available. Data from Cyprus were from 2007. In Malta and Sweden data were provided for 2009. In France, data were from 1995, 1998, 2003 and 2010; in Canada and UK: England and Wales, data were from 2005 instead of 2004.

READ  GI-LR framework

Analysis strategy

Data from the Euro-Peristat project were aggregated at the country-level therefore we conducted ecological analyses. Ecological analyses are a powerful tool to test hypotheses about the broader determinants of disease when there are many observations, and these are clustered into groups. These analyses rely on assessing associations between exposures and outcomes measured at the population level. We used both Spearman’s non-parametric and Pearson’s parametric tests.
In Chapter 4, our objective was to assess the robustness of preterm birth country rankings using data from 32 countries in 2008, and we used Spearman’s rank test to relate VPT birth rates based on all births starting at 22 weeks to rates based on other reporting criteria (i.e. all births starting at 24 weeks GA). Spearman’s rho ρ is the correlation coefficient between the ranked variables and therefore is less sensitive to outlying values; it is generally considered a more conservative estimate than Pearson’s r and recommended for use in ecological studies.
In Chapter 5, our objective was to relate changes in preterm birth rates to broader changes in other GA subgroups and we used Pearson’s test to examine possible associations. We conducted a time series analysis of preterm birth rates (using data on multiple time points within each country between 1996 and 2010), therefore we selected the Pearson’s test because it was possible, using Lorenz’s formulae to compute a marginal association measure that remains valid under the clustered data framework (i.e. PTB rates are clustered by country over time). Whereas both Pearson’s and Spearman’s test assume the observations are independent from one another; the adjusted Pearson’s estimate takes into account the variance within and across clusters, and their size (64). For paired data (X, Y) collected on a set of M clusters, let (Xij, Yij) be the jth observation (ie. PTB rate in each study year) for cluster I (i.e. Country), where 1 ≤ i ≤ M refers to the number of countries and 1 ≤ j ≤ ni refers to the number of observations within countries.

Table of contents :

CHAPTER 1: INTRODUCTION
1.BACKGROUND
2.WORK PLAN
CHAPTER 2: STATE OF THE ART
CHAPTER 3: METHODS
1.DATA SOURCES
AGGREGATE DATA FROM EUROPEAN COUNTRIES: THE EURO-PERISTAT PROJECT
AGGREGATE DATA FROM THE US, CANADA AND JAPAN: THE PREBIC PROJECT
AGGREGATE DATA FROM AUSTRALIA: NEW SOUTH WALES
INDIVIDUAL-LEVEL DATA FROM FRANCE: FRENCH NATIONAL PERINATAL SURVEY 2010
2.DATA AND DEFINITIONS
3.STUDY POPULATION
4.ANALYSIS STRATEGY
CHAPTER 4: ARE VALID INTERNATIONAL COMPARISONS POSSIBLE USING ROUTINE GA
DATA?
CHAPTER 5: CHARACTERIZING VARIATIONS IN THE PRETERM BIRTH RATE
CHAPTER 6: TARGETS FOR EARLY DELIVERY PREVENTION
CHAPTER 7: SUMMARY AND FINAL DISCUSSION

GET THE COMPLETE PROJECT

Related Posts