Christina Ölvingson, Niklas Hallberga, Toomas Timpkaaand Kent Lindqvista (Sweden)
Public health is an information-intensive area where the electronic patient records have opened up for detailed surveillance on a population basis. By introducing geographically ref-erenced health data and geographic information systems (GISs) it is possible to study causal-relationships, dispersal patterns for diseases, etc. GISs have been identified as one of the “new core technologies in public health” [1, p xii]. In general, public health professionals are dependent on accurate information for studying trends and relations between different factors. However, the accuracy of the data and technical features are lesser constraints than the pub-lic’s concern for privacy. In spite of this, ethical issues are often neglected. Still, the aspect of confidentiality is of major importance in public health due to the sensitivity of the data in-volved and GISs have the potential to be more threatening to personal integrity than many other information technologies. Violation of privacy occurs when there is an intrusion into the private sphere, and when facts that could be considered sensitive to integrity are distributed. Confidentiality arises when a person is obliged or asked to hold a confidence. It cannot be breached unless it becomes necessary in order to protect the welfare of individuals or com-munity, or if required by law. The objective of this paper is to explore ethical conflicts that occur due to the resolution of geographically referenced data in public health in literature and reality.
This explorative study is based on a literature study and interviews with persons working ei-ther with the injury data or the (geographical) information infrastructure in the WHO Safe Community injury prevention program in Motala Municipality in Östergötland County, Swe-den. The interviews (n=6) were performed in order to explore how ethical issues are handled in public health projects. In the literature study topics of interest were identified. A synthesis was made in order to analyze differences and similarities between issues identified in the lit-erature and the interviews.
Scientific literature study Data with a geographic reference could be an area or a point and, depending on what is stud-ied, the level of resolution varies. In general, as high a resolution as possible is recommend-able. However, as with all sensitive data there is a need to protect confidentiality. Leakage of sensitive data could result in e.g. embarrassment for individuals or loss of jobs. When adding a spatial aspect the problems could even affect large areas and groups of people living in these areas, e.g. redlining.
When using address-matched data it might be possible to backtrack and identify an individual from the information associated to the point. Confidentiality is then betrayed and often fol-lowed by an intrusion of privacy. By presenting data as area measures it is possible to main-tain the confidentiality of the data. However, there are several problems associated with area-measures. The value of subsequent analyzes are reduced if area data are used as original data and it is impossible to identify causal relations.
Ethical problems due to resolution were found to arise if it is possible to identify a single in-dividual either by use of identifiable data or by re-identifying through local knowledge and backtracking. In small areas this problem is more pronounced than in large areas since small areas usually involve fewer cases. The severity of breaches of confidentiality is related to what the data is about. However, what is seen as harmless is individually, which makes it dif-ficult to develop a general policy.
If in-data are at area level, the uncertainty inherited in the data is immense and will only in-crease as analyzes are performed. An ethical dilemma will also arise if not the best data avail-able is used, and to get as valid results as possible in-data with the individual as the informa-tion carrier should be used.
In public health there is a need to address ethical issues in a more complete and rigorous man-ner. The use of GISs in public health will be affected by e.g. how issues like privacy will be resolved. The placing of address-matched records into areas is neither a rational nor efficient method of spatial data encoding, and the development of “new methods for protecting confi-dentiality of health records while conserving their geographic accuracy should have high pri-ority in future GIS-based research in health” [2, p. 65]. Not using the best data available will cause ethical conflict, especially if major errors are inherited in the data. However, the rights of the individual have to be balanced against those of the community.
In terms of ethical issues, there was a significant overlap between theory and what could be observed in practice. The main difference was that in practice, policies were mainly devel-oped in response to articulated problems. However, overlooking ethical issues in the planning phase will systematically delay the progress of public health projects.
4 Discussion and Conclusion
Advances in information technology will increase the demand for more and better data in-cluding geocoded data. The wide array of data used and issues covered in public health result in difficulties in defining general policies on what resolution is suitable. As a result, to maxi-mize the potential benefit a case-to-case-evaluation is recommended. Further, if the results are reported in an appealing way, it will help keep the public’s confidence and hopefully create a positive opinion on the use of individual health data for public health purposes. Continuous research on how to protect privacy while preserving as much information as possible is of ut-most importance. What people desire is perhaps not absolute privacy but rather a reasonable assurance that sensitive information is treated fairly, that high security measures are provided and that the information is used for important health purposes.
 Yasnoff WA and Sondik EJ: Geographic Information Systems (GIS) in Public Health Practice in the New Millennium. In J Public Health Management Practice 1999, 5(4), ix-xii.
 Gatrell A and Löytönen M 1998: GIS and Health. London, Taylor & Francis.