Author Interviews, CMAJ, Opiods / 04.01.2016
Algorithm Can Help Predict Deaths Related to Prescription Opioids
[caption id="attachment_20406" align="alignleft" width="231"]
Tara Gomes[/caption]
Medical Research: What is the background for this study? What are the main findings?
Response: Surveillance of the harms associated with chronic opioid use is imperative for clinicians and policy-makers to rapidly identify emerging issues related to this class of medications. However, data regarding opioid-related deaths is difficult to obtain in Canada as it is collected by local coroners and is not widely available to researchers. We conducted a validation study to evaluate whether regularly collected vital statistics data collected by Statistics Canada can be used to accurately identify opioid-related deaths in Canada.
We compared deaths identified from charts abstracted from the Office of the Chief Coroner of Ontario to those identified using several coding algorithms in the Statistics Canada Vital Statistics database. We found that the optimal algorithm had a sensitivity of 75% and a positive predictive value of 90%. When using this algorithm, the death data obtained from the Vital Statistics database slightly underestimated the number of opioid-related deaths in Ontario, however the trends over time were similar to the data obtained from the coroner’s office.
Tara Gomes[/caption]
Medical Research: What is the background for this study? What are the main findings?
Response: Surveillance of the harms associated with chronic opioid use is imperative for clinicians and policy-makers to rapidly identify emerging issues related to this class of medications. However, data regarding opioid-related deaths is difficult to obtain in Canada as it is collected by local coroners and is not widely available to researchers. We conducted a validation study to evaluate whether regularly collected vital statistics data collected by Statistics Canada can be used to accurately identify opioid-related deaths in Canada.
We compared deaths identified from charts abstracted from the Office of the Chief Coroner of Ontario to those identified using several coding algorithms in the Statistics Canada Vital Statistics database. We found that the optimal algorithm had a sensitivity of 75% and a positive predictive value of 90%. When using this algorithm, the death data obtained from the Vital Statistics database slightly underestimated the number of opioid-related deaths in Ontario, however the trends over time were similar to the data obtained from the coroner’s office.


















