CSTE logo
This website uses cookies to store information on your computer. Some of these cookies are used for visitor analysis, others are essential to making our site function properly and improve the user experience. By using this site, you consent to the placement of these cookies. Click Accept to consent and dismiss this message or Deny to leave this website. Read our Privacy Statement for more.
CSTE Features
Blog Home All Blogs
Search all posts for:   


View all (177) posts »

“Cloudy with a chance of…” Classification of Emergency Department Visits related to Extreme Weather

Posted By Teresa J. Hamby, Hui Gu, and Stella Tsai, Friday, April 29, 2016
Updated: Thursday, April 28, 2016

In New Jersey, real-time emergency department (ED) data are received from 78 of 80 EDs by Health Monitoring Systems Inc.’s (HMS) EpiCenter system, which collects, manages, and analyzes ED registration data for syndromic surveillance.

Hurricane ‘Superstorm’ Sandy struck October 29, 2012, causing harm to New Jersey residents and extensive damage to businesses, transportation, and infrastructure. Monitoring health outcomes for increased illness and injury due to a severe weather event is important in measuring the severity of conditions and the efficacy of state response, as well as in emergency response preparations for future severe weather events. After Hurricane Sandy, the need to be prepared for future severe weather events prompted the New Jersey Department of Health (NJDOH) to develop a suite of 19 syndromic surveillance classifications for extreme weather-related conditions in EpiCenter. Examples include carbon monoxide poisonings resulting from generator misuse and disrupted medical care where patients needed emergency visits for medicine refills after losing their medicines in the flood or running out with no pharmacy available, and the need for oxygen or dialysis due to power outages at homes and procedure locations.

The development of these classifications followed a two-stage validation of keyword lists using diagnostic codes. First, staff identified possible inclusion keywords using records with ICD codes that met case definition. Then, exclusion text was determined by evaluation of cases with keywords of interest but without ICDs meeting case definition. Sensitivity and positive predictive values were computed for both the initial keyword list and the final keyword list to ensure the keywords were a good fit for the process.

NJDOH has since used these classifiers in more recent events to monitor for weather-related visits to storm-affected area ED’s. In June, 2015, a squall line of damaging thunderstorms, known as a “bow echo,” caused downed wires and power outages in two southern New Jersey counties. In the aftermath, there was a spike in the rate of visits for disrupted medical care, in particular for oxygen needs. In January, 2016, Winter Storm Jonas dropped more than a foot of snow over New Jersey. During and after that storm, carbon monoxide poisoning visits spiked, likely due to the misuse of generators, as did visits for medication refills.

While not every classification would be relevant in every extreme weather event, having the elements available provides tools for state and local users to monitor storm impacts locally and at the state level.

Teresa Hamby, MSPH is a data analyst, Hui Gu, MS is a health data specialist, and Stella Tsai, PhD, CIH is a research scientist at the New Jersey Department of Health. For more information about disaster epidemiology, join the Disaster Epidemiology Subcommittee.

This post has not been tagged.

Share |
Permalink | Comments (0)