2014.ISDS.Abstracts.Final.pdf


ISDS Annual Conference Proceedings 2014. This is an Open Access article distributed under the terms of the Creative Commons Attribution-
Noncommercial 3.0 Unported License (http://creativecommons.org/licenses/by-nc/3.0/), permitting all non-commercial use, distribution, 
and reproduction in any medium, provided the original work is properly cited.

ISDS 2014 Conference Abstracts

Assessment of National Poison Data System Algorithms 
to identify Public Health Events
Royal K. Law*1, Howard Burkom2, Alvin Bronstein3 and Josh Schier1

1Centers for Disease Control and Prevention, Chamblee, GA, USA; 2Johns Hopkins Applied Physics Laboratory, Laurel, MD, USA; 
3American Association of Poison Control Centers, Alexandria, VA, USA

Objective
To compare the effectiveness of current surveillance algorithms 

used in the National Poison Data System (NPDS) to identify incidents 
of potential public health significance with 1) new algorithms using 
expanded NPDS surveillance capabilities and 2) methods beyond the 
NPDS’ generalized historical limits model.

Introduction
NPDS is a near real-time surveillance system and national database 

operated by the American Association of Poison Control Centers. 
NPDS receives records of all calls made to the 55 regional US poison 
centers (PCs). The Centers for Disease Control and Prevention (CDC) 
use NPDS to 1) provide public health surveillance for chemical, 
radiological and biological exposures and illnesses, 2) identify early 
markers of chemical, radiological, and biological incidents, and 3) 
find potential cases and enhance situational awareness during a known 
incident. Anomalies are reviewed daily by a distributed team of PC 
medical and clinical toxicologists for potential incidents of public 
health significance (IPHS). Information on anomalies elevated to 
IPHS is promptly relayed to state epidemiologists or other designated 
officials for situational awareness and public health response.

Current NPDS surveillance algorithms utilize the Historical Limits 
Method, which identifies a data anomaly when call volumes exceed 
a statistical threshold derived from multiple years of historical data 
[ref]. Alternative analysis tools such as those employed by ESSENCE 
and other computerized data surveillance systems have been sought 
to enhance NPDS signal analysis capability. Technical improvements 
have been implemented in 2013 to expand NPDS surveillance 
capabilities but have not been thoroughly tested. Moreover, other data 
aberration detection algorithms, such as temporal scan statistics, have 
not yet been tested on real-time poison center data.

Methods
The data series for this assessment are 7-year historical time series 

of NPDS hourly call counts. NPDS catalogues confirmed PHS call 
clusters each year for current surveillance algorithms. New algorithm 
strategies will be run on the same data series to compare with current 
algorithm strategies. These new strategies include the use of the 
newly expanded NPDS surveillance capabilities and new methods 
(e.g. scan statistics) beyond NPDS’ historical limits model. The 
same methodology will be used for substance-specific surveillance, 
including carbon monoxide (CO) exposures and food poisonings.

To focus on missed cluster-type concerns, scenario-driven call 
clusters will be injected into the data series. PHS identification 
sensitivity and non-PHS alert rates will be balanced to assess the 
ability to capture historical and injected call clusters of desired 
exposure types for each type of surveillance algorithm. For example, 
performance of an algorithm with higher sensitivity and lower non-
PHS alert rates will be considered superior.

Results
The collected data series includes hourly time series of general 

poison center call volumes and substance-specific subseries, such 

as CO exposure and food poisoning. Selected inject scenarios are 
under discussion, including a latency model to capture the problem 
of delayed uploads. R code implementing the generalized algorithm 
queries is written and will be shared. Summaries of NPDS-specific 
evaluation measures, such as positive predictive value, defined as 
counts of PHS clusters divided by all clusters, will be presented for 
method comparison using both historical IPHS and injected call 
clusters.

Conclusions
Based on the results, the authors will determine the most effective 

algorithms for identifying IPHS without overburdening the NPDS 
surveillance team with high quantities of anomalies. This work will 
improve surveillance activities using NPDS and may be helpful for 
health departments that conduct surveillance using poison center data.

Keywords
toxicosurveillance; poison control center; NPDS

References

Stroup, D.F., et al., Evaluation of a method for detecting aberrations in 
public health surveillance data. Am J Epidemiol, 1993. 137(3): p. 
373-80

*Royal K. Law
E-mail: hua1@cdc.gov    

Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org * (1):e35, 201