Searching for the Gulf War Syndrome - Dr. Haley's Factor Analysis

[Guest guest]

http://www.computoredge.com/sandiego/Editorial/inside1.htm

Searching for the Gulf War Syndrome

.. . . Or Dr. Haley Puts Away His Stethoscope

By Kirk Kirksey

Why is the United States government angry at Dr. Haley about the Gulf

War Syndrome, and what does all this have to do with computers anyway? Sit

back; make yourself comfortable; get ready for a story in the best tradition

of Sam Spade and Bond.

It all started back in 1994. Soon after the Persian Gulf War ended, vets

across the country began complaining of a vague mishmash of seemingly

unconnected symptoms. Some said they suffered from dizziness and confusion.

Chronic fatigue was a frequent complaint. One said his teeth fell out and he

had to have a jaw removed. According to a report by CNN, after the war, Army

Colonel Herb has had days " when he was in so much pain, he could only

lie in bed while tears ran down his face. " Doctors using traditional

diagnostic methods found nothing. No nasty viruses. No signs of injury

caused by chemical agents. No parasites. The feds chalked the whole thing up

to combat stress.

Gulf War Syndrome—Fact or Fiction?

Dr. Haley is director of the Division of Epidemiology in the Department of

Internal Medicine at the University of Texas Southwestern Medical Center at

Dallas. As an epidemiologist, Dr. Haley studies populations rather than

individual patients by analyzing data and looking for trends of infection

and disease. For years after the war, scientists and veterans alike

theorized that Gulf War Syndrome was caused by exposure to agents in the

Iraqi chemical warfare arsenal or the smoke from the oil fires. Trouble was,

no one had any proof.

Dr. Haley and his team wanted to look for a smoking gun. They decided to

study members of the 24th Reserve Naval Mobile Construction Battalion

(RNMCB-24). Members of this unit were near Khafji, a village near the

Saudi-Kuwait border, the day Czech experts detected low levels of Sarin and

Mustard Gas. Word of the project went out, and 249 members of the 24th,

spread across five southeastern states, volunteered for Dr. Haley’s study.

Dr. Haley’s team put together an extensive survey that examined the full

array of symptoms described by the veterans. Physical or psychological,

Haley’s team wanted it all. Vets would be asked about their war time

activity and their health since coming home. Haley wanted to know about

numbness, coordination, sleeping problems, fatigue, swollen glands and a

host of other characteristics. If trends were found, the team was prepared

to do more extensive research with vets, and perform experiments in the

laboratory. Surveys were administered in December 1994 and January 1995. The

hunt for the Gulf War Syndrome was on.

Number Crunching

To you and me, Dr. Haley’s survey would probably look like a hodgepodge. Who

could make sense of it all? Dr. Haley and his computers, that’s who. He

decided to use a statistical technique called factor analysis. Here’s how it

works:

In many projects, researchers gather lots of information they believe will

give them an answer. Dr. Haley studied factors he believed were indicators

of health problems associated with the Gulf War. An appliance manufacturer,

on the other hand, might want to know why customers select a particular

brand of refrigerator. Factors like price, color, availability, reliability

and more may all be important to buyers. But which product characteristics

are most important? Are certain combinations of product characteristics more

important than others? Factor analysis is a technique used by researchers to

find hidden or latent variables by examining the relationships between

observed or measured factors.

Using computer analysis, the team split ambiguous descriptions of symptoms

into precise components. " This was the first breakthrough, " Haley said. " All

this time researchers have been trying to identify syndromes from ambiguous

symptoms. Only when we disentangled the different meanings of each symptom

did the real syndromes jump out at us. "

Haley believes his team has discovered not one, but three Gulf War

syndromes. Each is associated with the interaction of chemicals including

low-level nerve agents, anti-nerve gas pills issued by the military, and

chemicals found in the flea collars worn by some personnel as protection

against insects. When Haley’s findings were published in the Journal of the

American Medical Association, the Feds went ballistic. Combat stress, they

said, was still the culprit. Although the government position has softened,

the controversy still rages. For more information about Dr. Haley and

The Gulf War Syndrome Studies, visit http://www.swmed.edu/home_pages/l.

And the Winner Is . . .

Factor analysis isn’t new. Computers have been around for 50 years. But in

the past, many medical studies were based on information that had to be

manually encoded from paper records. Cost and time restricted the population

sizes and the amount of information that could be studied.

Today, we are beginning to build large—very large—warehouses of medical

information. Data from historically disparate systems can be combined into

repositories containing records for millions of patients. Although many of

these repositories or warehouses are being used to replace the paper medical

record, some institutions are applying advanced statistical techniques (like

factor analysis) to study things like disease patterns and the human genome.

The Internet and Internet2 add even more possibilities.

If the U.S. government’s reaction to Haley’s findings is any indication, the

dream has a dark side. We are in an era of constricted health care. In the

world of managed care, there are no more patients—only market share. Despite

large billboards and sappy radio spots, cost control is king. Every hospital

administrator in America is worried about getting patients out of the

hospital as quickly as possible.

As research technologies evolve, will managed care companies and insurance

conglomerates recognize and pay for the treatment of illnesses found by

scientists like Dr. Haley? Will the computer and statistical analysis take

their place as diagnostic tools alongside the x-ray, test tube, and petri

dish? The fight will be long and hard, because in the end, we’re not talking

about caring for the sick. We’re talking about money.

Copyright © 1997 The Byte Buyer, Inc.

ComputorEdge Magazine, 3655 Ruffin Rd Suite 100, San Diego, CA 92123. (619)

573-0315