BACKGROUND: At 10:52AM on 10/21/2016, the Kansas Department of Health and Environment’s syndromic surveillance system (SyS) was notified of a chemical release. Information shared showed the incident occurring at 8:02AM when sulfuric acid and sodium hypochloride were accidentally mixed at a NE Kansas grain products plant. The reaction produced a thick, white, gaseous chlorine cloud covering a large portion of the town. By 11:00AM, authorities gave the all clear to residents.
METHODS: Initial syndrome definitions based on key terms related to chlorine release and exposure did not yield accurate results. Hospital A’s small size allowed KDHE SyS to quickly view the data line-by-line and determine more appropriate syndrome definition. Data were gathered during the incident by KDHE’s SyS Program and medical records were abstracted post-incident for assessment and final reporting. KDHE SyS received the majority of relevant medical records from (Hospital A), proximal to the event through test data quality feeds. SyS data were processed by SAS and frequency analysis was performed in SAS. Messages were sent via the Kansas Health Alert Network to hospitals statewide requesting patients seen due to the incident. Seven area hospitals reported event-related patients and KDHE requested these medical records. A modified Assessment of Chemical Exposures Medical Chart Abstraction form was used and data were entered into Microsoft Excel. Finally, SyS data gathered during the incident were compared to medical records abstracted post-event.
RESULTS: Hospital A’s daily patient count doubling within the first 6 hours was the earliest signal of an emerging event detected through SyS. Detailed SyS information was not available in the first 6 hours due to paper triage notes and extreme patient load. Initial syndrome definitions yielded very few results. Further inspection of the data showed many unexpected search terms, which were included in future searches. Medical records reviewed after the event identified additional terms to educate on future emerging events viewed through SyS.
CONCLUSIONS: Conversations with Hospital A after the event revealed that, due to the overwhelming patient load, triage was performed with paper records and entered into EHR systems at a later point. This event demonstrates syndromic surveillance’s usefulness for rapid situational awareness in emerging health events. Hospital A’s work with KDHE showed the importance of coverage from even small facilities in a SyS program.