Syndromic Surveillance of Drug and Alcohol Abuse-Related Emergency Department Visits—Pennsylvania, 2003–2012

BACKGROUND:   Substance dependence or abuse has reached epidemic proportions nationally. In Pennsylvania, the 2010–2011 estimated rate of substance abuse or dependence was 8.9% and the 2010 drug overdose mortality rate (15.3 per 100,000) exceeded the motor vehicle accident death rate.   Drug abuse surveillance is challenging: drug abuse is not a reportable condition; healthcare providers might not call Poison Control Centers for patients intoxicated with common substances; and hospital discharge data are not timely. Pennsylvania initiated collection of de-identified emergency department (ED) visit chief complaint data in 2003 for the primary purpose of bioterrorism event detection; however, syndromic surveillance, now supported by meaningful use requirements, might be useful in describing other conditions.

METHODS:   The Pennsylvania Department of Health (PADOH) analyzed existing (2003–2012) syndromic surveillance data from participating emergency departments (n=152) to describe the epidemiology of drug and alcohol abuse-related visits (DAARVs) in Pennsylvania. PADOH identified suspected DAARVs, in 7 substance categories (alcohol, heroin, PCP, methamphetamines, cocaine, prescription depressants, and prescription opioid/morphine derivatives) using SAS to perform a series of text string searches.

RESULTS:   The analysis identified 291,203 (0.84%) visits with a drug abuse-related chief complaint among 34,608,944 ED visits; a total of 65,361 (22.25%) could be categorized into one of the 7 specified substance categories.   Among the drug-specific DAARVs, alcohol was implicated most frequently (81.25%) followed by heroin (7.19%). In univariate analyses, DAARVs referencing PCP (OR=1.36; 95% CI: 1.32–1.41), methamphetamines (OR=1.21; 95% CI: 1.15–1.26), alcohol (OR=1.07; 95% CI: 1.06–1.07), and prescription depressants (OR=1.05; 95% CI: 1.02–1.07) exhibited a statistically significant increase over the study period; visits referencing cocaine (OR=2.12; 95% CI: 1.96–2.29), alcohol (OR=1.10; 95% CI: 1.08–1.12), prescription opioids (OR=1.17; 95% CI: 1.04–1.32), and prescription depressants (OR=1.15; 95% CI: 1.03–1.29) were more likely among patients >29 years of age; visits referencing methamphetamines (OR=2.30; 95% CI: 1.84–2.86) and prescription opioids (OR=1.54; 95% CI: 1.35–1.76) were more likely among patients living in rural zip codes; visits for all substances except for prescription depressants were more likely among males.

CONCLUSIONS:   These syndromic surveillance data, collected with the support of bioterrorism and meaningful use funding, provide novel insight into the descriptive epidemiology of ED-attended drug and alcohol abuse in Pennsylvania. Further analysis and validation is needed to better understand whether these retrospective data, and the prospective analysis of rapidly available syndromic surveillance data, can inform overdose prevention activities in Pennsylvania.