At MCHP, our research focuses on the question: “What makes people healthy?” We use data in the Manitoba Population Research Data Repository to answer this question. The Repository is a collection of more than 90 different databases that contain information on the health and social well-being of everyone in Manitoba. Some of the databases go back as far as the 1970s. Other data are new to the Repository. But all of the data can be linked together to give us a picture of how Manitobans are faring.
What data are in the Repository?
Most of the databases in the Repository consist of administrative data. These data are created when people use public systems and services in Manitoba. For example, records are kept of every doctor’s visit and prescription filled. When someone graduates from high school, a record is kept. These data are called administrative data. Some examples of administrative data in the Repository are:
- From the health care system:
- Doctor’s visits and diagnoses
- Hospital admissions and care Manitobans receive in hospital
- Drugs prescribed to Manitobans
- From social services:
- Receiving Employment and Income Assistance
- Living in social housing
- Receiving the provincial child care subsidy
- From the education system:
- School enrollment, course marks and assessments
- From the justice system:
- Involvement in and charges for justice system-related incidents (e.g., assault, theft, domestic incidents)
Some databases in the Repository are registries. One example is the Manitoba Health Insurance Registry. The Manitoba Health Insurance Registry has demographic data (e.g., date of birth, sex) on nearly everybody in the province. The Repository also holds registries on First Nation and Metis populations in Manitoba. These registries are only used with permission from those groups.
The Repository also contains survey data. Survey data are collected directly from people. One example is Manitobans’ responses to the Canada Census.
Another type of data in the Repository is clinical data. Clinical data describe the results of laboratory or medical tests. One example is data on Manitoban’s bone density.
What about personal health information and data privacy?
Protecting Manitobans’ personal information and privacy is a priority for MCHP. We follow the rules for privacy and protection of personal information outlined in Manitoba’s Personal Health Information Act (PHIA) and Freedom of Information and Protection of Privacy Act (FIPPA). Some of the precautions we take are:
- All of the data in the Repository are de-identified. Personal information, such as names and addresses, are removed from each record before being put in the Repository. None the people who work with the data at MCHP ever know who the individual Manitobans in their studies are.
- All projects using Repository data are reviewed by the Manitoba Government’s Health Information Privacy Committee and the Health Research Ethics Board at the University of Manitoba to ensure peoples’ privacy is protected.
- Access to the data is provided only to individuals who have completed an accreditation course about data processes and protections at MCHP and have signed a pledge of confidentiality. These users only have permission to access the data needed for their projects while those projects are active.
More information about data privacy measures at MCHP are available in the publication “Population Data Centre Profile: The Manitoba Centre for Health Policy”: https://ijpds.org/article/view/1131/2498
How are the data linked together?
Before data go into the Repository, identifying information is removed. A number replaces the identifying information. This number makes it possible to link data from different sources without knowing who a person is. The linking of data makes it possible to look at data from different sources so that we can learn more about Manitobans. One example of this is linking education and head injury data to see if test scores on average are lower for individuals who have had a head injury.
If you are interested in learning more about MCHP’s data linkage processes, please click here to read “Population Data Centre Profile: The Manitoba Centre for Health Policy”: https://ijpds.org/article/view/1131/2498
What kind of statistical analyses can we do with the Repository data?
The analysts who work with the Repository data are trained in advanced statistical methods.
- We can describe the characteristics of a population, including age, sex, income, family characteristics, level of education, and many other characteristics.
- We can make comparisons between two or more groups of people. For example, we can compare any of the characteristics above to see whether they are different between people living in urban areas of Manitoba and people living in rural areas.
- We can use statistical tests to say whether there is a significant difference between the groups. A statistically significant difference between the two groups means that we are confident that the difference is real and not just due to chance.
- We can use advanced statistical methods like modeling to measure the size of the difference between groups, or to balance the differences between groups so we can make fair comparisons between them. Sometimes modeling is used to look at past trends and predict future ones. There are many other types of analysis that we can use, depending on what questions we want to answer.
What kinds of things should we be cautious about when interpreting the results of our analyses?
The Repository is a powerful tool for learning about Manitobans. It can help us develop policies for better health and social services. But every type of research has its limitations. Here are some things that we should be cautious about when using the data in the Repository:
- All of the studies that we do with Repository data are observational studies. We can describe associations between two events. But we can’t say that one event causes the other.
- The data in the Repository tell us a lot about services and programs that Manitobans have used. They also tell us which Manitobans do not access those services and programs, which has been useful in many studies. But we should keep in mind that when people do not use services, we don’t know whether this is because they don’t need these services or because they are unable to access them. Access to services needs to be considered when looking at whether a service or program is meeting the needs of the population it is intended for.
- Some research studies ask questions that need more in-depth data to answer them well. The numbers often tell only one part of the story. To understand the full story and to answer the question “why are we seeing these trends?” we need to consider the historical, cultural and social context. When we ask questions like this, we often involve people in our research teams who can provide this context. These people may be topic experts who know a lot about the data and how the data are collected, community leaders who know the people and programs the study is about, or government staff who can help us focus on priorities. But even with their help, there are still limits on what conclusions we can draw from the study results.