A voluntary privacy standard for health services and policy research: legal, ethical and social policy issues in the Canadian context.

• Author: Weisbaum, Karen M.

In this review, the authors describe how one group of Canadian researchers has begun to address current privacy protection challenges with the end goal of developing a Canadian national standard for privacy protection specific to health services and policy research (HSPR.) They provide a concise description of some of the key messages and issues that resulted from a recent series of HSPR workshops and describe some of the workshop outcomes in terms of their legal, ethical and social/policy significance. The review ends with a brief description of some future directions for research and development in this area.

Introduction

Privacy of personal information is a current hot topic across Canada and nowhere does that heat radiate more than in debates about privacy of personal health information, particularly in the area of health research.

In Canada, there are many different kinds of "health research." The Canadian Institutes of Health Research (CIHR) generally classifies research according to four "pillars": 1) basic biomedical; 2) applied clinical research; 3) health services research; and 4) population health. (1) Each category can be distinguished from the others in a variety of ways, including the manners in which data are collected, pooled (or aggregated) and analyzed. While each category of research may involve collection and use of health information, the information is not necessarily of the same scope or depth, or used in the same way for the same research purpose.

For example, clinical trials overlap with clinical care. Data collection in clinical trials involves direct contact between research subjects and members of the research team, so individuals are always known to at least one researcher. (2) Researchers need to know that particular data are about a particular individual. Personal identity is likely not revealed widely within the research team, nor is it necessary for the researcher to know subjects' direct identities, as unique study numbers are usually assigned to data from an individual. Nevertheless, the nature of the research is such that the data are either highly identifiable, or there is at least a significant potential for identification of individual research subjects if appropriate safeguards are not observed.

In stark contrast, HSPR uses health information that is external to the circle of care. It consists mainly of personal health information collected into large databases that are used primarily for administrative purposes, such as physician service claims or provincial drug benefit reimbursement. These large databases are also used for the secondary purpose of scholarly research focused on the health of entire populations, (3) rather than unique individuals. Use of these data for particular studies does not include personal identifiers, such as names and addresses. The data often include the elements of age-in-years and sex, as well as demographic information. The data are never used to make a decision that directly affects a specific individual.

It makes sense, therefore, that different kinds of health research reflect significantly different kinds of privacy issues. There is a corresponding need for different and appropriate privacy protection standards that are tailored to the nature of each category of health research and the personal health information that is used for that research.

Such a need is especially acute in the case of HSPR researchers who find themselves bound by guidelines that flow largely from models developed for clinical trials. (4) Many HSPR researchers already argue that this is inappropriate and adds an extra burden in conducting legitimate and valuable research. In addition, there is currently a patchwork of statutory and regulatory standards across Canada for privacy protection of personal health information, which create extra challenges for HSPR data collection activities that extend beyond provincial borders. As a result of all these factors, HSPR researchers have been left without clarity about what constitutes appropriate or adequate standards for privacy protection in HSPR protocols and programs of research. (5)

Despite this lack of clarity, the HSPR community has made efforts to delineate its own appropriate standards. At a recent national HSPR privacy workshop, a group of Canadian...