A national strategy for waiting-times research?

The First Ministers' Conference on Health Care in the fall of 2004 resulted in a federal government pledge of $4.5 billion over 6 years to reduce waiting times for cancer care, coronary angioplasty, bypass surgery and angiography, diagnostic testing (MRI and CT scans), hip and knee replacement, and cataract surgery. As part of this strategy, each province must create electronic patient registries to track waiting times. Although the immediate priority is to manage waiting lists and monitor progress in the short term, the potential to use these registries as a research tool to improve health care in the future is plain. One issue that must first be resolved is how personal patient information will be handled and protected. Many jurisdictions already have informal waiting lists developed by individual physicians, group practices, hospital clinics and diagnostic centres. These lists vary greatly in the comprehensiveness of the clinical information contained, how the list is used, and legal provisions for access to the data for research purposes. Currently, most waiting-list registries focus on the provision of hospital services and do not collect long-term data on patient outcomes. Several Canadian provinces have waiting-list Web sites that provide information on the average interval between specialist consultation and procedure performance. Unfortunately, the data are often reported by fiscal or calendar quarters, whether rolling (always ending 3 months ago) or fixed, rather than being presented in "real time." Some researchers have challenged the feasibility of ethical and legal protection of personal health information. For example, in anticipation of new Canadian legislation, researchers involved in the Registry of the Canadian Stroke Network made determined efforts from the beginning to obtain informed consent from all stroke victims (or their surrogate decision-makers) to participate in the registry. Unfortunately, this approach resulted in a major selection bias between those who did and did not agree to participate.1 Despite a significant investment of public funds (roughly $500 000), the patients who consented to be in the registry were unrepresentative of typical stroke patients, which hindered the generalizability of the data except in certain subsets. In epidemiologic and health services research, incomplete or unrepresentative data can do greater harm than any potential compromise of the privacy of individual patients, particularly in today's con