Poison center exposure calls predict mortality due to prescription opioid poisoning

Results from national substance use reporting systems are of limited value because they are not geographically specific and their publication is delayed by years. Previous studies suggested a small proportion of poisoning deaths are reported to poison centers (PC). However, human exposure calls to PCs may be an alternate indicator of mortality. To investigate this, we used call counts for prescription opioids handled by 15 PCs in 17 states that were reported to the Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS^®) System in 2003 and 2004. Death certificate data from the National Center for Health Statistics were used to identify deaths due to prescription opioids based on ICD-10 toxicology codes for short-acting opioids (T40.2) and methadone (T40.3). Linear correlation and negative binomial regression were used to compare counts from PC calls and death certificates, by state and quarter. R-square correlations were 0.72 and 0.42, for short-acting opioids and methadone respectively. Zero-inflated models were not favored; adjustment for PC penetrance improved model fit. Negative binomial regression models had better predictive capability for short-acting opioids than methadone, when controlling for penetrance and calendar quarter. Our results suggest that calls to PCs for short-acting prescription opioids are strongly correlated with mortality due to poisoning from short-acting opioids as identified on death certificates. Including penetrance in regression models yielded improved fit and should be considered when analyzing PC data. Calls received by poison centers may be used for timely and geographically-specific surveillance of mortality due to prescription opioids.