How can multi-regional data help researchers better understand stroke care?
Every five minutes, someone in Canada experiences a stroke – one of the leading causes of death and disability across the country. That’s why it’s critical that stroke systems of care are adequately monitored to ensure excellence in treatment, according to Dr. Amy Yu, an adjunct scientist at ICES, a data centre and HDRN Canada member organization. Dr. Yu’s new study uses routinely collected health data from multiple provinces to enhance the research community’s ability to accurately monitor these systems of care.
“As a health services researcher and a stroke neurologist, I want to make sure people experiencing stroke across Canada have access to timely stroke care and are able to recover,” said Dr. Yu. “An improved understanding of stroke care and outcomes can help system planning and optimize stroke systems of care across Canada, creating more opportunities for provinces to learn from each others’ experiences.”
A stroke occurs when blood stops flowing to a part of the brain, causing damage to those brain cells. There are different types of strokes, depending on which part of the brain is impacted and the amount of damage. “Not all strokes are equal,” Dr. Yu said. “Some people may be more severely affected than others, so their starting point is different.” While the severity of a stroke is the most important factor in understanding stroke recovery and patient outcomes, this information is missing from the administrative health data that are routinely collected by hospitals. This significantly limits research on population-based stroke outcomes, explained Dr. Yu.
An improved understanding of stroke care and outcomes can help system planning and optimize stroke systems of care across Canada, creating more opportunities for provinces to learn from each others’ experiences. ~ Dr. Amy Yu
“When we are doing studies involving large populations, there isn’t information on how severe a stroke is to start with.” That’s why Dr. Yu’s team developed the Passive Surveillance Stroke SeVerity (PaSSV) score: “I wanted to use administrative data sets to see if we could come up with a proxy for stroke severity when conducting population-level research.” The PaSSV score was first developed using administrative data in Ontario, but the research team soon became interested in whether the PaSSV score could be used to compare care systems across provinces.
In 2021, Dr. Yu submitted her study, External Validation of the Passive Surveillance Stroke Severity (PaSSV) Score, to HDRN Canada’s Projects to Advance the Algorithms Inventory (PAAI), which invited researchers to submit proposals to lead a multi-regional validation or feasibility study intended to expand HDRN Canada’s Algorithms Inventory. The Algorithms Inventory is an online collection of published algorithms measuring population health, health service use, and the determinants of health. This innovative tool compiles algorithms that have been validated or tested for feasibility of implementation in two or more Canadian provinces and territories or nationally. Making these algorithms available to the research community helps ensure that researchers’ approaches are consistent, reliable and repeatable.
The Algorithms Inventory is a tool of HDRN Canada’s Data Access Support Hub (DASH), a one-stop data access service portal for researchers requiring multi-regional health administrative data in Canada. Selected studies received support from DASH in navigating the data access request processes across multiple provinces, extracting data and conducting analyses. This enabled Dr. Yu to access administrative data from multiple regions in Canada to compare the PaSSV score across provinces. “This was the first time I’ve done research in multiple provinces, so I was very grateful for the support and expertise that HDRN Canada provided,” said Dr. Yu. “The methodologists assigned to my project were very knowledgeable, thorough and helpful in understanding the subtle data differences in each province.”
After receiving administrative data from British Columbia and Nova Scotia, as well as new data from Ontario, the research team was able to determine that the PaSSV score performed similarly in each province. The study is currently under review for publication. Data received through this project has already led to a publication on how the PaSSV score impacts hospital performance ranking when evaluating risk-standardized stroke mortality in Ontario.
Once published, Dr. Yu’s multi-regional study will be available through the Algorithms Inventory for future researchers interested in stroke systems of care to learn from. “Having a measure of stroke severity that can be obtained from administrative data is essential for accurately monitoring and improving the quality of care provided in hospitals,” explained Dr. Yu. “Without a way to measure stroke severity, researchers may not get an accurate representation of how care is being provided.”