Feds, health insurers focus on sharing bioterror data

The U.S. Department of Homeland Security Monday plans to begin a five-day exercise to test government and private-sector information-sharing in response to mock terrorist attacks involving weapons of mass destruction.

The White House is billing the exercise, called TopOff 2 (for Top Officials 2), as the most comprehensive terrorism-response exercise ever undertaken in the U.S. It will include 19 federal agencies, the American Red Cross, and officials from Washington, Illinois, the District of Columbia and Canada. The scenario will consist of near-simultaneous mock attacks involving a radiological device in Seattle and the covert release of a biological agent in Chicago.

A key aspect of the exercise will be the ability of state and local officials in the U.S. and Canada to identify medical patients complaining of symptoms that indicate exposure to a biological agent and to communicate that information in a timely manner to other federal and state officials.

The exercise comes as the U.S. health insurance industry nears the completion of a pilot project that aims to create a nationwide data mining, surveillance and information-sharing system for the type of regional health crisis envisioned in TopOff 2.

Relying on a mix of private funding and a US$1.2 million grant from the Atlanta-based Centers for Disease Control and Prevention, four member organizations of the American Association of Health Plans (AAHP) are testing a national bioterrorism syndromic surveillance system that uses real-time data collected from more than 20 million people in all 50 states. The primary goal of the program is to “develop and implement standards, protocols, infrastructure and analytic tools for detecting and reporting unusual geographic clusters of symptoms or complaints” of acute illness that might indicate that a covert bioterrorism attack has taken place, said AAHP President and CEO Karen Ignagni, during a May 5 House of Representatives hearing.

Dr. Jim Norton, program manager at project participant HealthPartners Inc., said his Bloomington, Minn., organization’s mainframe-based research database wasn’t timely enough to meet the 24-hour reporting requirement of the surveillance system. As a result, a significant amount of programming was required to pull the data out of the medical operations mainframe, put it into a standard file format, strip out all personally identifying information and assign geographic and demographic codes to each patient.

Transmitting the data to the central server operated by Harvard Pilgrim Health Care Inc. in Wellesley, Mass., remains a 36-hour manual process, he said. Once fully automated, reporting will occur every 24-hours. Pattern-recognition software and trend analysis are also still works in progress, said Norton. Some algorithms have been developed, but officials aren’t satisfied with how they handle geographic analysis of outbreaks.

Debra Ritzwoller, clinical research investigator at Kaiser Permanente Colorado in Denver, said daily encounters with patients are scanned for a set of 700 diagnosis codes. That data is then grouped into syndromes and further classified by age, sex and ZIP code.

“We expect to be able to pick up something two or three days sooner with this system than you would by waiting for people to come into emergency rooms,” said Norton. “That will be critical to preventing the spread of these diseases.”