Poison center surveillance data: the good, the bad and ... the flu

Poison center data are increasingly used by state health departments and the Centers for Disease Control and Prevention for public health surveillance. Forrester and colleagues evaluated the ability of 6 Texas poison centers covering a population of 24 million to accurately code and report the number of H1N1 calls received over a 5-month period.

The Texas poison centers generated new coding and began work within 24 h of notification of the surveillance need. No additional staff were added for call management, coding, or quality assurance, and no H1N1 training was provided ahead of time. A triple-redundancy coding method was used to prevent underreporting of calls. This allowed the Texas poison centers to accurately flag over 90% of H1N1 cases. Results were available in real time, allowing day-to-day monitoring by poison centers and the state public health department for surges, location, ages of callers and/or patients, and type of question.

The accuracy of poison center near real-time toxicosurveillance data coding was sufficient to monitor emerging trends. The data generated by poison centers are flexible, immediate, unique from other data sources, and useful for trend monitoring. As health departments and other collaborative partners rely more on the data from poison centers, consideration must be given to appropriate funding to support coding training, monitoring, and quality assurance to further enhance this valuable system.