![[ Visit Client Website ]](images/banner.png)
BACKGROUND: From 1999-2010, unintentional poisoning mortality rates increased nearly fivefold in Oklahoma. In 2010, Oklahoma had the fifth highest poisoning mortality rate in the United States (22.1 deaths per 100,000 population). From 2007-2011, 3,198 Oklahomans died as a result of an unintentional poisoning (17.4 per 100,000 population). The Injury Prevention Service (IPS) of the Oklahoma State Department of Health maintains a medical examiner (ME)-based unintentional poisoning surveillance system. Through crosschecking and linkage with Vital Statistics death data, the number of appropriate cases to review increased, reporting accuracy improved for certain variables (e.g., race, ethnicity), and additional variables of interest were made available (e.g., marital status, ICD-10 cause of death codes).
METHODS: The Office of the Chief Medical Examiner (OCME) serves as the centralized ME system for Oklahoma. The IPS receives reports from the OCME for all non-natural deaths occurring in Oklahoma. Deaths with a manner of ‘Accident’ and mention of a poisoning in the cause of death were included for analysis. Cases were crosschecked with Vital Statistics death data (ICD-10 underlying cause of death codes X40-X49) and the OCME annual database to ensure surveillance data were as complete as possible. After finalization of both databases and crosschecking, ME cases were matched to their respective death certificates.
RESULTS: Crosschecking 2007-2011 ME data with Vital Statistics unintentional poisoning deaths resulted in a possible 224 additional cases. Of these, 53% (n=118) were added to the IPS surveillance system. Of the 106 excluded cases, a poisoning was not listed as the cause of death for 62 (58%) cases. The most common cause of death listed in excluded cases was smoke inhalation. Nearly one third (32%, n=34) of cases were excluded because the manner of death on the ME report was ‘Suicide’ (n=12), ‘Natural’ (n=19), or ‘Unknown’ (n=3). Ten additional cases were excluded due to residence or ME jurisdiction. A death certificate could not be located for <1% (n=27) of cases. Racial and ethnic misclassification on ME reports was most common for American Indian and Hispanic populations, for which the number of cases increased by 64% and 97%, respectively.
CONCLUSIONS: Use of ME data for surveillance purposes increased timeliness and availability of data compared to death certificate data. However, use of ME data alone resulted in unidentified cases and misclassified demographic variables. Death certificate data provided an important supplement data for an ME-based surveillance system, but resulted in an overestimate of the number of unintentional poisoning-related deaths.