A Natural Experiment in Rapid TIA Care with Knowledge Transfer and Exchange


Principal Investigators:

Penn, Andrew M. W. (Vancouver Island Health Authority) and Maclure, Malcolm (University of Victoria)


Harris, Devin R. (St. Paul's Hospital).

A. Statement of the health problem

Stroke occurs when part of the brain dies because of disturbed blood flow. You may die from a stroke, but more likely you will survive, disabled in some way.  This disability may affect how you think, or see, or speak or move. Stroke is the largest cause of disability amongst adults in Canada. In addition, a third of patients who have dementia and were previously categorized as having Alzheimer's disease, are now known to be suffering from the effects of strokes. Recently it has become apparent that strokes, like volcanoes, often give a warning sign. This probably occurs over half of the time. The warning sign is a stroke that goes away within minutes or hours. We call these, transient ischemic attacks (TIAs) or mild strokes. It is now emerging that if we intervene very quickly after this warning sign, we can stop the impending large stroke from happening, 80% of the time. Unfortunately, these warning rumblings have been largely ignored. There is usually no pain, as there is with the crippling chest pain felt before heart attack. For that reason we do not even know how many patients experience these warning events. Recent studies in Oxford however suggest that there are more patients experiencing the rumblings of stroke than the rumblings of heart attack. This is much more frequent than we thought. The Vancouver Island Health Authority has a unique system to see and treat these patients called the Stroke Rapid Assessment Unit (SRAU). We need to prove that the SRAU is effective.

B. Objective 

We need to first measure the true rate of TIA's and mild strokes in Canada. Canada has its own racial blend, environmental factors, dietary traditions and health-care provisions, which make it unlikely that our rates will be the same as Britain's. As health-care regions across the country gear up to tackle the new revelations, they need some estimate of how many patients they may need to deal with. Secondly, we need to confirm that intervening quickly can stop strokes from happening. There are two studies which have suggested this. The same British group that measured TIA, also showed a large reduction in stroke when they created a special unit to intervene rapidly after TIA. They did a "before and after study" and showed an 80% drop after they initiated the high-speed service. This type of study, called longitudinal, is not as reliable as you would think. Another study, done in Paris, compared how well the patients did after visiting their unit with how well they would have been "expected" to do. This is even less reliable. The most reliable study is called a randomized study.  Our objective is to prove that TIA rapid intervention units work, using a type of randomized design. Lastly, we will bring our conclusions back to our BC Stroke Strategy colleagues in the other health regions. Ours is the bellwether region for the province, so our data will be used to plan services for the rest of BC.

C. Approach

Vancouver Island is a natural laboratory to examine healthcare. There is just one health authority (VIHA), which manages all 750,000 people and there is relatively little interchange with other health-care regions. In 2004, we set up the SRAU in Victoria, similar to the Oxford and Paris units, to intervene quickly after TIA to prevent devastating stroke. We see about 900 patients a year, but this is probably only half the true volume on the island, based on the Oxford experience. We now have funding from the BC Ministry of Health (MOH) to double capacity in a pilot project. This includes a northern satellite unit. Our project would first identify all TIA patients on the island, by using surveillance staff to flush out all possible cases for review at one of our two units (~5000 patients over the next 3 years). We would then measure what happens to all these patients in the 90 days after TIA. Finally we would hire data analysts to extract MOH data on death and hospitalization from stroke. This will allow us to determine what happened to the 4000 patients we have already seen, after their TIA. This will give a total of 9000 patients, looking forward and back, enough to perform a sophisticated analysis comparing the destiny of weekday TIAs (when the SRAU is open) vs weekend TIAs (when it is shut). We learn how common TIA is in Canada and whether the SRAU approach works.

D. Innovative aspect

We established the SRAU in 2004 as we determined that patients with TIA and mild stroke needed to be seen quickly with specialized imaging but didn't otherwise need to come into hospital. The solution was a day unit with La-Z-Boy, close to imaging, where all tests could be organized. We also utilized an electronic clinical record system which collects rich data on the patient's history and physical, called the Stroke Guidance System. This means that we have now over 4000 patients accumulated in our database, a unique asset even internationally, but we do not have their outcomes. The system also has the capacity to provide guidance to the doctors on all decision points. The project therefore can capitalize on the existing SRAU infrastructure and can increase capacity relatively easily to provide a surveillance system for stroke in Canada. We are developing a Web based referral system that emergency room doctors and GPs can use directly to refer their patients to us. The quasi randomized controlled trial design is an innovative way of leveraging a naturally occurring randomization of patients based on the day of the week that an event happens. This grant adds the process of measurement so that we may establish whether our innovation is working.

E. Relevance to the objectives of the initiative 

Our project is part of the BC Stroke Strategy (BCSS), itself part of the Canadian Stroke Strategy. The TIA program in our region (VIHA), running out of the SRAU, is the bellwether project for the BCSS Secondary Prevention Theme. We are demonstrating and testing how best to coordinate and organize secondary prevention services. Our region therefore does the measuring for the whole province, and the other regions depend on us to put together the best case for further MOH funding. We have taken breakthrough evidence on best practice for stroke prevention, implemented this in an innovative Canadian solution (the SRAU), with strategies for isolated communities (the northern satellite unit), with an innovative way to gather data (the Stroke Guidance System). This grant will help us add the evidence of outcome which can then be transformed into conclusions on policy and the economic cost and impact of our novel approach. Our unit is soon to be toured by delegations from Manitoba, and the other 4 BC health regions, and has recently been copied by Sunnybrooke hospital in Toronto. With 9000 patients including clinical, imaging, and outcome data, this could be one of the richest databases on TIA in the world.