An Evaluation of Carbon Monoxide Exposure Reporting to New York City's Poison Control Center, 2012-2013

BACKGROUND:  Carbon monoxide (CO) exposure can lead to serious health outcomes including death. In 2004, New York City (NYC) updated the Health Code to make all CO poisoning cases treated at a hospital immediately reportable to the NYC Poison Control Center (PCC). We evaluated compliance to this reporting requirement and examined demographic, spatial, and temporal patterns in CO cases across data systems.

METHODS:  To assess compliance with the Health Code, we compared CO exposure cases in the NYC PCC call data to cases from the New York Statewide Planning and Research Cooperative System (SPARCS) 2012 and preliminary2013 ED and hospitalization billing data. Cases identified from the PCC included suspected CO exposures of all intents that were being reported from a NYC hospital. CO poisonings within ED and hospitalization SPARCS were identified using ICD-9-CM diagnosis and external cause of injury codes. We calculated descriptive statistics including counts and proportions by age and gender for CO exposures identified in each data system for comparison.

RESULTS:  Overall, there were 2.4 times as many CO cases identified within SPARCS (n=1,320) as within PCC (n=556) from 2012-2013. This discrepancy was similar by patient gender, with 2.5 and 2.3 times as many male and female CO cases identified within SPARCS as within PCC respectively. By age group, this difference ranged from 1 (among those 90 years or greater) to 3.2 (among those 30-39yrs) times as many CO cases identified within SPARCS as within PCC. This discrepancy varied substantially by hospital county, ranging from 1.4 (among hospitals in the Bronx) to 5.9 (among hospitals in Queens) times as many CO cases identified within SPARCS as within PCC. Of the 556 PCC cases, 445 (80%) reported the carboxyhemoglobin level. When case counts were compared by specific facility, reports to PCC ranged from 0-2.2 times as many as in SPARCS. On average, hospitals had 70% fewer reports in PCC than in SPARCS. Temporal trends were observed in these data, with peak number of cases in PCC and SPARCS occurring from November-February, while the discrepancy in case counts between PCC and SPARCS was lowest from September-November. 

CONCLUSIONS:   Reporting of CO poisonings to the PCC allows for immediate action to prevent further poisonings and is required by the Health Code. Overall and health care facility reporting of CO poisoning to the PCC was low, demonstrating the need to encourage compliance of NYC Health Code expectations at local hospitals.