This survey and user-needs assessment on privacy and public health shows a definite requirement by public health professionals – in various fields and positions in both Canada and the UK – for personally identifiable data, including spatial data. The requirement for this spatial data is at its most granular level – latitude and longitude, or exact street address – which necessarily compromises patient privacy. It is not surprising, therefore, that public health professionals perceive privacy to be a significant obstacle to public health practice.
There are those who would argue that this perception is the product of a lack of understanding of the legislation and regulations by the public health community. The results of this research, however, indicate the contrary. Not only did public health professionals in both countries generally rate themselves high on knowledge of privacy legislation and related issues, but those with the highest self-rated scores also tended to rate privacy as more of an obstacle. That these self-ratings of knowledge are not representative of actual knowledge remains possible.
Participants perceived the most critical obstacles to sharing or acquisition of health data with PID to be bureaucracy, followed by legislation.
Bureaucracy surrounding health research in both Canada and the UK generally revolves around data ownership, academic competitiveness, ethics review boards or committees, and in particular, requirements for informed consent, even if they compromise public health, or are not in the best interests of the patients involved [40–42]. Since seeking subject consent with every new hypothesis to be tested or model to be developed is an impossible task, some have suggested that thought be given to "blanket" consent. At the Canadian Institutes for Health Research (CIHI) 2003 workshop on the legal and ethical issues facing the Canadian Lifelong Health Initiative [43], participants spent some time discussing such issues, only to emphasise the importance of the establishment of ethical governance and structure; essentially, more necessary bureaucracy. Interestingly, while the debate continues, a relatively recent survey found that most of the British public did not consider the use of their National Cancer Registry PID for public health research and surveillance to be an invasion of their privacy [30]. While the ethics of blanket consent are not discussed in this study, it is nonetheless offered as a potential solution in light of the requirements of the public health community. This does not, however, address other issues of data ownership and control that contribute to the bureaucratic debate.
While many individuals recognised the importance of privacy legislation, participants generally indicated a concern and, in some cases, first-hand frustration that legislation unduly restricts public health activities, compromising surveillance and research. Many phrases were used by respondents to describe the implications of privacy legislation on public health, including, among others: "increasingly restrictive;" "serious;" "incomplete;" "fuzzy;" "does more harm than good;" "two-edged sword;" "causes challenges;" "delays and restricts access [to data];" " [is a] hindrance to the improvement and efficiency of public health;" "disappointing;" "frustrating;" "difficult to interpret;" "very worrisome;" "disadvantages the public interest;" "not properly understood;" "over-protective;" "limiting;" "hinders knowledge;" and "used as an excuse not to share data." A large proportion of the public health community represented in this sample clearly expressed major concerns with the impact of privacy legislation on their work – both in Canada, and in the UK – in spite of having a good understanding and acceptance of its purpose and necessity. It is also important for legislation to be written in an unambiguous manner that is clearly understood by both public health professionals and the general public [4].
Public health professionals are largely of the opinion that the general public's level of awareness of the impact of restricted access to PID on public health practice is extremely low. Surveys by the Office of the Privacy Commissioner in Canada [44] repeatedly show that the majority of Canadians surveyed (up to 80%) place an extremely high level of importance on strong laws to protect personal information, particularly health information, and that they feel that the level of protection of their personal information has declined over the past ten years. Yet interestingly, only 20% are clearly aware of existing laws, and even fewer (12%) are aware of their rights around the collection, use and disclosure of this information. The "need to raise Canadians' awareness about the current laws in place and what their rights are" [44] must therefore be coupled with the corresponding need to address this from within the context of public health requirements.
Educating the public, therefore, as well as practitioners, data users, policy makers and politicians, was not surprisingly identified by participants as a potential solution. Participants put emphasis on the utilisation of the media to educate and increase awareness, as well as demonstrating the impact of a lack of data, and the benefits of its use when available. Demonstration of the benefits to the individual (e.g. streamlining of the system, not being asked for personal information with every visit to a new clinician, improved dissemination of public health information and intelligence directly to the public) was also offered as a solution, and summed up by one participant in the phrase "seeing is believing". It is worth noting, however, that a number of participants displayed a certain level of pessimism that until a crisis or extreme event occurs, no amount of education or awareness-increasing activities would make a difference.
Public health professionals generally prefer disaggregate, case-level data, but access to this data is an issue. The limitations imposed by privacy on public health have resulted in the development of a variety of techniques for data anonymisation [15, 23, 45]. However, all unavoidably have their issues, risks and limitations, and there is currently no framework to guide public health professionals in their appropriate use and interpretation.
Generalisability
Although the findings of this paper may be generalisable to public health professionals in Canada and the UK, issues of privacy and public health are not unique to these countries. Privacy is defined as a fundamental human right in the legislation of many countries, and the concept is enshrined in Article 12 of the United Nations' Universal Declaration of Human Rights [46] and Article 8 of the European Convention on Human Rights [47]. Similarly, public health is an international discipline; both diseases and information are ubiquitous, and neither is constrained by political boundaries and oceans. The increasing requirement for spatial data and its inherent clash with privacy legislation therefore extend beyond the UK and Canadian contexts, and the results, requirements and conclusions drawn from this research can be generalised to wherever such a clash exists. The implementation of solutions by national governments may be further exacerbated by issues of social political trust. General public distrust in government initiatives and motives, such as in most countries of the European Union, Canada, and the United States [48, 49], complicates changes that may be perceived by the public to be intrusions of privacy. Such issues may currently be less of a concern in countries such as Finland, Sweden, Denmark, and the Netherlands, where social political trust, though declining, has traditionally tended to be much higher [50–53]. However, even in such nations where privacy and health have traditionally not clashed, increased international data sharing requirements and spatial data implications may pose unanticipated and challenging obstacles.
Limitations
No comprehensive lists of public health and health GIS professionals were found in either country, so it was not possible to invite a random sample. In addition, the response rate in the UK was relatively low, and it is therefore uncertain that the sample is representative of all public health professionals in the two countries. However, responses between the two countries were consistent, with no significant differences.
Since knowledge of privacy legislation and policies was based on self-rated scores, a thorough review and assessment of privacy legislation as it pertains to public health practice is required in both Canada and the UK to validate the findings of this survey.
A number of limitations and issues pertaining to the web-survey were identified. Most notable of these was the presence of a scroll bar in sections II and III which most participants missed, thereby eliminating the ability to capture items in reference to "place", such as usefulness. However, these items were also captured more broadly in other sections of the survey. Other issues involved the inability of the architecture to support various designs and types of questions that would have facilitated the completion of the survey, and shortened the length of time required. Participants also noted frustration with the navigation and structure of the survey pages. A document outlining these issues and others was submitted to the ALPHA team after the initial pilot for future enhancements to the architecture.