HEALTH DATA MONITORED FOR BIOTERROR WARNING
27 Jan 2003
Source: New York Times, January 27, 2003
Health Data Monitored for Bioterror Warning
By WILLIAM J. BROAD and JUDITH MILLER
To secure early warning of a bioterror attack, the government is building a computerized network that will collect and analyze health data of people in eight major cities, administration officials say.
The Centers for Disease Control and Prevention is to lead the multimillion-dollar surveillance effort, which officials expect to become the cornerstone of a national network to spot disease outbreaks by tracking data like doctor reports, emergency room visits and sales of flu medicine. "Our goal is to have a model that any city could pick up and apply," a senior administration official said of the plan.
Officials would not disclose the program's cost or which cities will be involved. But experts say Washington is likely to be one of the eight.
Such surveillance is now possible because of an explosion in commercial medical databases that health authorities, with permission and under strict legal agreements, are starting to mine. In ambition and potential usefulness, the health network goes far beyond an environmental surveillance system, disclosed by the administration last week, that will sniff the air for dangerous germs.
The emerging health monitoring network, officials and experts say, will provide information that could save lives if terrorists strike with deadly germs like smallpox or anthrax. In detecting attacks, a head start of even a day or two can greatly lower death rates by letting doctors treat rapidly and prevent an isolated outbreak from becoming an epidemic. A senior official said President Bush was expected to refer to these new bioterrorism defenses in his State of the Union address.
The disease centers' initiative represents a sharp swing to civilian leadership in a field the military pioneered and once dominated. But even in civilian hands, the emerging network has raised concerns that such surveillance may violate individual medical privacy rights.
Officials said concerns were initially heightened because of the Pentagon's central role in the genesis of many systems, and especially because Vice Adm. John M. Poindexter, architect of the much-criticized Pentagon computer surveillance effort known as Total Information Awareness, is in charge of the Defense Department agency that finances some of the government's disease monitoring research.
In November, as the Bush administration came under fire for Admiral Poindexter's project, White House officials ordered the military to drop plans to link four cities into a $420 million health monitoring network and shifted responsibility for such work to the new domestic security agency. The transfer was not motivated by privacy concerns, administration officials say, but by a judgment that the military was ill suited to exploit monitoring for public health.
"We all agreed that doing this surveillance in the civilian sector was not the military's job," Dr. Anna Johnson-Winegar, a Pentagon biodefense official, said in an interview.
Experts say the prospect of war with Iraq, and the chance that Baghdad might retaliate with germ weapons, are accelerating the effort to expand and integrate scores of rudimentary disease surveillance systems being developed by cities, states and the federal government. But public health experts argue that even if the United States never suffers another bioterror attack like the anthrax strikes of late 2001, the emerging network can still help doctors better track, treat and prevent natural disease outbreaks.
"We want as much protection as we can afford," said Dr. Daniel M. Sosin, director of public health surveillance at the Centers for Disease Control and Prevention in Atlanta. Dr. Sosin is helping to expand the nation's health surveillance to incorporate the new systems.
Supporters of the emerging surveillance network insist it raises few privacy issues, saying that the data are laundered of names and identifiers. People are not tracked as individuals, they say, but their symptoms are, and often their age, sex and ZIP code as well. But computer surveillance itself has drawn criticism from the American Civil Liberties Union, members of Congress and others.
The system is needed, proponents say, because few cheap, reliable sensors exist for detecting deadly germs in such likely target areas as subways and shopping malls. Sensors are also prone to false positives, or incorrect germ identifications.
Dr. Thomas R. Frieden, the health commissioner of New York City, which has one of the nation's most highly developed rapid surveillance systems, said the emerging network could help authorities gauge the dimensions of germ attacks and reassure the public.
He pointed to a case in November in which a New Mexico man visiting New York was found to have bubonic plague, a deadly contagious disease. "We were concerned this was bioterrorism," Dr. Frieden said. "But we didn't see any signals. We didn't see any alarms. That added to our confidence to rule out bioterrorism."
Experts say most of the new systems, military and civilian, are still experimental. A critical challenge is finding needles in the haystacks of data about common ailments like respiratory infections, which can rise and fall with great suddenness in winter.
Dr. Marcelle Layton, New York City's assistant health commissioner for communicable diseases, said another challenge was ensuring that there are enough public health officials to respond to alarms that the new environmental and medical surveillance systems might sound.
"The best system will be useless if it's only a fire alarm with no firefighters to put out the flames," Dr. Layton said.
Nonetheless, expectations run high.
"We think this will be important," said Dr. Alan P. Zelicoff, a physician at the Sandia National Laboratories who helped develop a widely used surveillance method, the Rapid Syndrome Validation Project, which is now used in California, New Mexico, Texas, Singapore and Australia. "We need to get disease reporting from the 19th to the 21st century."
For decades, disease surveillance has valued accuracy over speed. Nurses, doctors and public health officers gather raw data, often using paper forms sent by mail. In the background, federal, state and private laboratories use advanced technologies to determine the causes of disease and confirm diagnoses. But the process tends to take days or even weeks.
Moreover, the system is narrow, revealing little about the nation's overall health. While the federal disease control agency has more than 100 surveillance systems, most are designed to track a single organism or condition, like heart disease or flu virus. In addition, most are independent of one another.
The system has serious gaps. While laboratories usually comply with federal rules to report certain illnesses to health authorities, physicians often do not.
The military and the national weapons laboratories, increasingly worried about germ attacks, tried a new approach in the late 1990's. To learn of impending trouble quickly, they decided to scrutinize populations for clues of diseases before they were officially diagnosed. Experts zeroed in on how clusters of such symptoms as fever, cough, headache, vomiting, rash and diarrhea could suggest — but not prove — the presence of particular diseases, some of them lethal. The method was called syndromic surveillance.
An early military system was the Electronic Surveillance System for Early Notification of Community-Based Epidemics, or Essence. It drew medical data from some 400,000 members of the military and their dependents who lived in the Washington area — a major potential terrorist target, but hard for civilians to scan medically because of "the numerous city, county and state jurisdictions," according to a Defense Department statement.
After the 2001 terrorist attacks, the Pentagon's Defense Advanced Research Projects Agency put $12 million into an experimental program, Essence 2, which tracked millions of civilians in the Washington area for signs of bioterrorism. The program now reports to Admiral Poindexter, whose Total Information Awareness program was dealt a setback by the Senate late last week, its future now in doubt. Joe Lombardo, a civilian who runs Essence 2, which is based at the Johns Hopkins Applied Physics Laboratory in Maryland, said that although Admiral Poindexter's office finances the system, Essence 2 shares no data with his computer surveillance project. Essence 2, he said, gathers electronic records from drugstore chains, hospitals and physician groups. Mr. Lombardo said about a dozen people were developing the technology and collecting and analyzing the data.
"We're not Big Brother," he said. "Our objective is to support public health. The information we receive has been sanitized by the provider to ensure that it is impossible to identify individuals."
Privacy, though a goal, is apparently not yet guaranteed. A Pentagon planning document on the surveillance effort for fiscal 2002 and 2003 said the Defense Department was working to develop "enhanced automated privacy protection methods" that will "assure the anonymity of records accessed by the data monitoring software."
Experts say that privacy can, in theory, be violated when connections are made between disparate databases — for instance, between those of physician payment and disease diagnosis, or health and law enforcement. They also say the potential for personal identification increases as the surveillance becomes a two-way street in which not only are problems detected but physicians are notified about potential problems involving individual patients.
This fall, the military sought to incorporate the Essence 2 program into an expanded program, the Biodefense Initiative. Costing a projected $420 million, it was to deploy environmental sensors and wire four major cities, including Washington, into a disease-surveillance network.
But after Admiral Poindexter's Total Information Awareness program came under criticism by privacy advocates, White House officials moved the Biodefense Initiative out of military hands. That step, said Dr. William Winkenwerder Jr., the assistant secretary of defense for health affairs, "just seemed to make sense." Dr. Winkenwerder added that the military has often pioneered technologies, like the Internet, that move into the civilian sector.
Privately, some military officials grumble that transferring the Pentagon's effort to civilians will be wasteful. "It could be reinventing the wheel," a senior officer said.
Administration officials say the new eight-city disease control network will deal with the privacy issue directly. "We have to satisfy the legal constraints, and also people's concerns," a senior official said.
Other civilian surveillance systems are emerging quickly. In Boston, the Harvard Medical School faculty and the Massachusetts Department of Public Health are working closely with Harvard Pilgrim Health Care, a health maintenance organization. For more than a year, the team has studied data from 175,000 people in eastern Massachusetts, and it will soon cover as many as 20 million people coast to coast.
In October, the disease control centers awarded the Harvard team $1.2 million to expand its pilot network nationally. The expansion will not monitor cities, but will concentrate on patients calling an after-hours medical advisory service.
Health officials say civilian emphasis in the developing surveillance field will help ensure privacy and enhance routine disease monitoring.
"It's the practical stuff that's most promising," Dr. Sosin of the disease control centers said. "Whether this is going to detect terrorism is unclear. But as a safety net and for tracking an event once it's going on, it's very promising."