Introduction to Unit 3: Public Health Data Systems
This is the third unit in the module ‘GIS for Analysis of Health’. This unit explores spatial data sources for health analysis within GIS. Types of data are considered within the framework for disease causation introduced in the previous units. In each case, examples are given of the data sources available in different contexts. Data acquisition still comprises the largest challenge to many GIS implementations and this is particularly the case where relevant data sources span many different organizations.
There are 8 other subsections in this unit:
- Individual vs aggregate data
- Confidentiality of health data
- Health event data
- Predisposing factors data
- Behavioural factors data
- Environmental factors data
- Exposure data
- Vulnerability data
‘Individual vs aggregate data’ introduces the very important differences which apply when we use health data at individual and aggregate levels. The analytical methods that can be applied and inferences drawn from aggregate analyses are inherently limited by the inability to observe individual-level relationships, yet there are many situations in which we must work with aggregate data.
Some of the most important constraints on the use of individual level health data are those relating to confidentiality. Even within health authorities and national statistical organizations, codes of practice designed to protect the confidentiality of individuals may restrict the analyses which can be undertaken or may require approval of research designs by ethical committees or data custodians before individual-level data can be used. This subsection introduces these issues and some of their implications for GIS use of health data.
The next six subsections address each of the principal data elements of the disease causation diagram in turn. Each involves a review of the key characteristics of a specific type of spatial data. Data for health GIS are drawn from different sources, with information about health events usually being sourced from within the health care system, but predisposing and behavioural data often obtained from censuses and surveys. A full understanding of data source considerations is critical to the establishment of appropriate linkages between health event data and appropriate denominators. Data relating to physical and social environments are different again may come from a wide variety of sources.
The final activity (vulnerability data) includes a practical exercise looking at heart disease in the Cardiff Bay area of the UK.
Expect to spend about 2 weeks working through these materials.