Four Projects Selected to Develop New Multi-Jurisdictional Health Algorithms
In 2020, HDRN Canada put out a national Call for Proposals for research teams to validate or test new algorithms to measure population health, health services use, and determinants of health across multiple provinces and/or territories.
An algorithm is a set of rules for measurement. For example, an algorithm can be used to identify individuals with a chronic health condition. An algorithm often includes diagnosis codes recorded with the International Classification of Diseases (ICD), observation period (i.e., number of years) to which the rules are applied, and other inclusion and exclusion criteria.
Four projects, led by teams of researchers from across Canada, have been selected as part of the Projects to Advance the Algorithm Inventory (PAAI):
- External Validation of the Passive Surveillance Stroke Severity (PaSSV) Score
- Predicting Hospital Admissions for Ambulatory Care Sensitive Conditions Using Primary Care Electronic Medical Record Data: A Multi-jurisdictional Feasibility and Validation Study
- Validation of Algorithm to Identify Juvenile Idiopathic Arthritis (JIA) across the Provinces of Alberta, Manitoba, Ontario, Nova Scotia, PEI and New Brunswick
- Development of a Validated, Patient-level Definition of Chronic Pain for Use in Administrative Health Data
HDRN Canada’s Algorithm Inventory was initially developed to provide a repository of published algorithms that have been validated and/or tested in two or more provinces or territories. In recognizing the need to expand the number of algorithms, particularly around new health conditions and high-priority measures, research teams were invited to submit proposals to lead a multi-jurisdiction validation or feasibility study.
“The Working Group members have done an excellent job to guide this initiative,” said Dr. Lisa Lix, Chair of HDRN Canada’s Algorithms and Harmonized Data Working Group. “These four projects will add new knowledge to the Algorithm Inventory that will benefit researchers who conduct multi-jurisdiction studies.”
As part of the PAAI, HDRN Canada will support research teams in navigating the data access request processes across multiple jurisdictions, extracting data, and conducting analyses. HDRN Canada member organizations have also received funding to provide research teams with analytic support. The research teams are providing scientific guidance and oversight for the projects.
“We are excited to collaborate with HDRN Canada on this project,” said Dr. Deborah Marshall, Principal Investigator of the study on juvenile idiopathic arthritis (JIA). “We are currently unable to obtain comparable estimates of the prevalence and incidence of JIA in the population or assess the burden of JIA supported by related health services, health outcomes and health economic research. With the increased use of costly biologics to treat JIA and improve outcomes, a validated algorithm would allow us to assess treatment patterns and economic analyses across provinces.”
Additional details about the selected projects are outlined below.
External Validation of the Passive Surveillance Stroke Severity (PaSSV) Score
Principal Investigator: Amy Y. X. Yu, University of Toronto (Sunnybrook)
Project Summary: Stroke is an important cause of mortality and morbidity and monitoring stroke outcomes is important for ensuring excellence in care. Stroke severity is one of the most important predictors of outcomes, but it is often missing from routinely collected administrative data, thus limiting population-based stroke outcomes research. Our team derived the Passive Surveillance Stroke SeVerity (PaSSV) score using administrative data in Ontario. We aim to externally validate PaSSV in British Columbia, Alberta, and Nova Scotia. The availability of a measure of stroke severity derived using administrative data is crucial for monitoring and improving quality of care, clinical research, as well as facilitating inter-provincial comparisons across Canada.
Predicting Hospital Admissions for Ambulatory Care Sensitive Conditions Using Primary Care Electronic Medical Record Data: A Multi-jurisdictional Feasibility and Validation Study
Principal Investigator: Andrew Pinto, University of Toronto
Project Summary: Hospitalizations related to ambulatory care sensitive conditions (ACSC) are common, costly, to some degree preventable with high-quality primary care, and are thus used as a health system performance indicator in Canada. We aim to develop and evaluate predictive algorithms for one-year non-elective ACSC hospitalizations among community dwelling adults using primary care EMR data in British Columbia, Manitoba and Ontario. Our algorithm could be used by primary care providers to identify high-risk patients who may benefit from more proactive models of care aimed at preventing health decline and the undue distress of an avoidable hospitalization.
Validation of Algorithm to Identify Juvenile Idiopathic Arthritis (JIA) Across the Provinces of Alberta, Manitoba, Ontario, Nova Scotia, PEI and New Brunswick
Principal Investigator: Deborah Marshall, University of Calgary
Project Summary: The project is a multi-jurisdictional validation study of administrative case ascertainment algorithms for Juvenile Idiopathic Arthritis (JIA). Case ascertainment algorithms for JIA have been developed and validated in Manitoba, Canada, but the diagnostic accuracy of these algorithms in other provinces is unclear. JIA clinical cohorts from sites in Manitoba, Alberta, Ontario, and Nova Scotia (which serves patients from PEI and New Brunswick as well) will be used to validate case algorithms using administrative health data. Measures of diagnostic accuracy will be calculated, and refinements made to the algorithms to ensure the best possible balance of diagnostic accuracy for each province.
Development of A Validated, Patient-level Definition of Chronic Pain for Use in Administrative Health Data
Principal Investigator: Morgan Slater, Queen’s University
Project Summary: Approximately 20% of Canadians live with chronic pain and the cost of caring for patients with chronic pain outweighs the cost of cardiovascular disease, cancer or diabetes. Despite this, we know little about the extent of chronic pain in Canada and the healthcare needs of those living with pain. We will determine the most appropriate combination of diagnostic codes, medications and treatments that are representative of chronic pain and use patient-reported chronic pain as the reference standard to test the accuracy of the algorithm. This study will be the first to develop valid case-finding algorithms to identify patients with chronic pain in pan-Canadian administrative health data.