Reliability of using ICD-10 codes to identify COVID-19 cases in multi-regional data
Headshot of Dr. Sasha Bernatsky
In 2020, during the onset of the COVID-19 pandemic, the World Health Organization (WHO) added two new ICD-10 codes to identify and track COVID-19 cases. ICD codes are used to assign diagnoses in research, surveillance, and health administration settings.
Health Canada and Public Health Agency of Canada (PHAC) policy-makers needed to know whether these new ICD-10 codes could reliably be used to identify cases of COVID-19 within administrative health data (e.g. information generated from hospital admissions, emergency room visits, and potentially other sources). If these ICD-10 codes could be validated, they could be used to monitor infection rates efficiently and even potentially assess the safety and effectiveness of treatments.
Health Canada called upon the Canadian Network for Advanced Interdisciplinary Methods (CAN-AIM) to fulfill this critical COVID-19 mandate, and HDRN Canada’s Data Access Support Hub (DASH) was able to play a role in this important endeavor.
About the study
Led by Drs. Sasha Bernatsky and Cristiano Moura, CAN-AIM undertook a study using data from British Columbia, Alberta, Manitoba, Ontario, and Newfoundland and Labrador to assess how reliable the ICD-10 codes were in identifying COVID-19 cases, not only in one province but across jurisdictions.
The results will help policy-makers understand whether ICD-10 codes can be applied regardless of variations in the population demographics of each province or variations in how people access healthcare in the context of COVID-19.
This work was important not just for COVID-19 case surveillance and drug safety and effectiveness work in Canada but also globally, since Health Canada is leading international efforts to understand how these codes work.
How DASH supported researchers
Given the need for timely results, Dr. Bernatsky and her team greatly appreciated having DASH as a resource. “CAN-AIM generally promises to provide initial answers for policy-makers within months, aiming for a final answer within a year or so,” said Dr. Bernatsky, “We started investigating partnerships with DASH for our projects early on, because we believe that DASH has so much potential to help provide data in a timely manner.”
DASH facilitated communications between provincial data holders and streamlined processes, such as its centralized intake form. The CAN-AIM study is the first project to be completed through DASH and has provided an opportunity to further evolve data access processes that will benefit other researchers using DASH services.
“Regular check-ins with DASH ensured early resolution of queries and helped build confidence about the study’s progress,” said Autumn Neville, who coordinated the massive undertaking. “There is an excellent value in the role that DASH has developed in acting as a facilitator to accessing data, particularly for people inexperienced in accessing data from multiple regions.”
Since then, Dr. Bernatsky and her team have reported to Health Canada, PHAC, and others who will be able to use the results in future COVID-19 surveillance and research in Canada and internationally. Some results were presented at the International Society for Pharmacoepidemiology conference this August and will be presented at upcoming scientific meetings this fall. CAN-AIM is hoping for an open-access publication in the coming months.