Tracking a Synthetic Cannabinoid Overdose Outbreak in Real Time, Connecticut, August 2018

On the morning of August 15, 2018, the Connecticut Department of Public Health learned from news reports that several people had been found unresponsive in a city park in New Haven County after using synthetic cannabinoids — manmade psychoactive substances with unpredictable, sometimes severe effects. Acute poisonings from them had been rising across the country.

The department turned to syndromic surveillance: emergency department records arriving in near real time, used to track outbreaks and give health officials a picture of suspected overdoses as they happen.

The system

Since January 2018, the department had been collecting visit data from all 38 emergency departments in Connecticut through the EpiCenter system. Using Health Level Seven messaging — a national standard for exchanging health data — each emergency department sends details such as sex, age, ZIP code, chief complaint and triage notes at registration and discharge, within 5 minutes.

Sizing the outbreak as it happened

• Within 20 minutes of the first news report, the department wrote a search definition to find suspected synthetic cannabinoid overdoses, built from keywords in the chief complaint — first drug terms such as "K2," "spice" or "weed," then place terms such as "green," "bench" or "park," refined against the news and the records.
• By midday August 15 it had found 25 likely outbreak visits; by 5 p.m. on August 16, 55 — all in New Haven County. Leadership and the local health department got regular e-mail updates.
• The response ended on August 17. The Drug Enforcement Administration found that the drugs contained AMB-FUBINACA, an ultrapotent synthetic cannabinoid with strong depressant effects.

Looking back

On August 20, the department searched triage notes as well as chief complaints to find visits the real-time search had missed. Of 2,086 emergency visits in New Haven County on August 15–16, 72 fit the outbreak definition:

• 53 different patients — 12 of them came back up to five times, suggesting they were exposed again
• median age 43 (half between 35 and 51); 41 (77%) were male
• of 63 visits with a recorded outcome, 57 (90%) ended in discharge and 6 (10%) in the patient leaving without being seen; nobody died

What changed

Real-time data gave health departments a timely sense of the outbreak's size, and the follow-up showed its full extent and confirmed it was over. Afterwards the department:

• wrote more drug- and place-specific overdose definitions to catch future events
• built a dashboard for near real-time analysis
• set standard guidelines for sharing information and resources with partners during overdose events
• shared its methods with the National Syndromic Surveillance Program Community of Practice

Syndromic surveillance, the report concludes, could be an important way to keep officials aware of harm from substance use as it happens.

Sources

Based on Jones SA, Soto K, Grogan E, et al., "Notes from the Field: Syndromic Surveillance Used To Monitor Emergency Department Visits During a Synthetic Cannabinoid Overdose Outbreak — Connecticut, August 2018," MMWR Morbidity and Mortality Weekly Report, volume 69, Centers for Disease Control and Prevention; a work of the United States government in the public domain.