![[ Visit Client Website ]](images/banner.jpg)
Each year in Maine approximately 150-200 people receive care for unintentional non-fire (UNF) related exposure to Carbon Monoxide (CO.) One to five people die each year in Maine from exposure to CO.
Health effects associated with exposure to CO vary, depending on the amount and duration of exposure, and may be mild to severe in nature.
In 2008, CO poisoning was made a notifiable condition in Maine, and a CO case-based reporting surveillance system (COPSS) was established. Detailed data on the characteristics of UNF CO poisoning cases and their exposure source are collected and analyzed to support targeted public health interventions. As few states operate a COPSS, we present findings from the evaluation of reporting source completeness and accuracy.
METHODS:
COPSS receives reports from multiple sources, such as, healthcare facilities (hospitals, provider offices/clinics), laboratories, Poison Control Center, Vital Statistics, and Hyperbaric Chamber units. Completeness was measured by estimating the number of reports not received by individual healthcare facilities (HF) through a medical record review and case matching process. Through this same process, accuracy was assessed by measuring the sensitivity and positive predictive value (PPV) of reporting sources and then, separately, for administrative billing data using the Nationally Consistent Data and Measure (NCDM) definition developed by the National Environmental Public Health Tracking Network (NEPHT) for UNF related CO poisoning. Medical record and other reporting information is used to define cases within Maine’s case based surveillance system and is considered the ‘gold standard’ for assessing accuracy of NCDM defined billing data.
RESULTS:
We identified 180 cases of CO exposures or poisonings that would have been missed due to non-report, representing 60.4% of overall missed reports by HF. HF reporting source sensitivity was 75.6% overall and 36.1% as a unique case reporting source. Laboratory and Poison Control Center (PCC) reporting produced similar overall sensitivity (41.4% and 31.2%, respectively); as a unique reporting source sensitivity changed dramatically, 0.4% and 19.9%, respectively. HF and PCC had similar PPV of 63.8% and 58.9%, respectively.
Using the NCDM definition for UNF CO poisoning, administrative billing data sensitivity was 73.8% and 72.1% for PPV. Almost 53.0% of administrative billing data identified as ‘unknown’ for intent and fire-relatedness were actually UNF CO poisonings.
CONCLUSIONS:
From this evaluation Maine was able to identify significant under-reporting of CO poisoning. Additionally, although widely used, administrative billing data may overestimate the burden of CO poisoning when applying the NCDM definition for UNF CO poisoning.