The Health Outcomes and Health Service Use of Immigrants in New Brunswick: An Investigation Using Linked Administrative Data
Research Team: Ted McDonald, University of New Brunswick; Brent Cruikshank, University of New Brunswick; Zikuan Liu, New Brunswick Institute for Research, Data and Training (NB-IRDT)
Institutional Partners: Charles Ayles, Government of New Brunswick – Population Growth Division; Nancy Roberts, Government of New Brunswick – Department of Health
There has been great interest in the health of immigrants to Canada and other immigrant-receiving countries because of the large flows of immigrants to those countries and the fact that the wide range of source countries provides great variation in immigrant characteristics and experiences. This has led to a voluminous body of work that has considered health status, health service use and health-related behaviors of immigrants, usually in comparison to non-immigrants. A consistent finding is that immigrants are found to be in significantly better health (variously measured) than comparable non-immigrant residents but that this advantage is lost with additional time in the host countries. However with few exceptions this work has relied on self-reported survey data, and this is potentially problematic given the well documented reporting biases for both chronic conditions and health service use. Furthermore, reporting biases may vary by immigrant characteristics such as country of birth (reflecting language and cultural differences and other factors), giving rise to misleading results about actual differences in health. Survey datasets are typically also cross-sectional, which limits the extent to which health transitions can be observed over time. Longitudinal datasets in Canada either have a sample of immigrants that is too small (NPHS) or have poor information on health (LSIC). In contrast, administrative health data are generally available but offer only very limited demographic information such as age, sex and location of residence. As a result there is still much that is not known about the health of immigrants, how it compares to non-immigrants, and how it evolves with time in the host country.
This project will take advantage of the availability of provincial administrative data through the New Brunswick Institute for Research, Data and Training, a provincial data custodian at the University of New Brunswick. From the Population Growth Division of the Government of New Brunswick (GNB), detailed information on all provincial nominees arriving in New Brunswick over the previous 10 years will be linked with individual level data on physician billing and hospital discharge records for each year in New Brunswick. These health datasets offer very detailed information on all health services accessed through the provincial system and a range of chronic conditions can be identified using diagnostic codes. Health outcomes and health service use for these immigrants will be tracked over time and contrasted with age-sex-location-year matched non-immigrant New Brunswick residents. Statistical analysis will depend on the particular variable of interest but will typically be Logistic models of incidence or count models of health service use frequency. These methods will also reflect the longitudinal nature of the available data.