HDRN Canada seeks multi-region studies for Algorithm Inventory
Lisa Lix, HDRN Canada Lead
HDRN Canada is calling on health researchers to submit their published multi-region studies to be considered for inclusion in its Algorithm Inventory. The Algorithm Inventory is an online collection of published algorithms measuring population health, health service use, and the determinants of health that have been validated or tested for feasibility of implementation in two or more Canadian provinces and territories, or nationally.
The only tool of its kind in Canada, the Algorithm Inventory is an innovation in health research, compiling data from cross-jurisdiction research collaborations into a simplified, summarized and searchable resource. Notably, the inventory facilitates repeatable research and ensures that different researchers or research teams across the country are using the same criteria to identify patients, their outcomes and their characteristics.
“Many research data centres develop and validate algorithms for their own administrative health data, and capture these algorithms in centralized repositories of information, such as the Concept Dictionary at the Manitoba Center to Health Policy,” explained Lisa Lix, HDRN Canada Lead and director of the Data Science Platform at University of Manitoba’s George and Fay Yee Centre for Healthcare Innovation. “But prior to the development of the Algorithm Inventory, there was little attention given to algorithms developed for multi-region studies that use population-based administrative data from more than one jurisdiction or to the central storage of detailed information about these algorithms.”
The Algorithm Inventory was developed by HDRN Canada’s Algorithm and Harmonized Data Team in 2020, using a rigorous process that involved a systematic review of articles that developed or tested multi-region algorithms in the Canadian context. “Once we reviewed articles that measured population health, health service use, and the determinants of health, our team established relevant fields for data extraction,” said HDRN data analyst Saeed Al-Azazi. “We then extracted standardized information about these algorithms and organized them into a searchable format.”
Today, the Algorithm Inventory contains more than 185 peer-reviewed algorithms, which researchers can filter by province or territory and other categories, such as disease and symptoms. In order to continue building the inventory, HDRN Canada is asking researchers to submit their published work, or share other published studies that have developed an algorithm that can be used for multi-region research. “We want the Algorithm Inventory to be as robust and as useful as possible, which is why we’re reaching out to the research community,” said Al-Azazi “Once we receive a study, we’ll review it, verify it and then extract the study and algorithm information before posting it on hdrn.ca.”
If you are a health researcher with a multi-jurisdiction validation and/or a feasibility study to submit to HDRN Canada’s Algorithm Inventory, please contact datainventory@hdrn.ca.