Sirens and blood. When there's a pileup on I-95, the National Highway Traffic Safety Administration (NHTSA) gathers up all the details -- the make of the vehicles, the time of day, the place on the highway, whether the victims were wearing seatbelts or the drivers were drunk.

Blood and sirens. When a person blows his head off in the kitchen or the woods, or when someone is murdered in an alleyway, well -- forget it. There's no equivalent system that collects the details of the death and analyzes the patterns of these violent injuries.

This is the Fatal Data Gap. One kind of sudden violent death is routinely studied on a nationwide basis; another kind is not. Yet about 50,000 Americans suffer a nontraffic violent death every year, almost all from suicide and homicide.

To reduce the toll, researchers need to know how and why these fatalities occur. What percent of young people who take their own lives are intoxicated at the time of death? How often does a child die from abuse -- every two hours or two days?

Nobody really knows. The National Center for Health Statistics collects information from death certificates. The FBI collects information about offenders and victims of crime. But the two systems aren't linked. The circumstances that led to the violence are often not recorded. Information is scattered among different local and federal agencies. There's no national reporting system on these deaths like NHTSA's Fatality Analysis Reporting System (FARS).

Now that's about to change. At President Bush's request, the Centers for Disease Control and Prevention (CDC) has started a small $1.5 million project to set up a National Violent Death Reporting System.

It builds on the innovative statewide reporting system in Wisconsin and a pilot program sponsored by Harvard's School of Public Health in seven areas of the country. The goal is to analyze these fatalities so that public health officials will have a scientific basis for designing preventive strategies.

"Our knowledge of these events is fragmented. That's why it's so exciting to see the dots being connected," says Stephen W. Hargarten, who heads the emergency medicine department at the Medical College of Wisconsin in Milwaukee. As he makes clear: "I'm a physician. This is a health problem."

You notice that so far I have avoided the mention of one key word: guns.

Nobody wants to rouse the gun lobby and give it a reason to block this initiative. Language is important. The project is not called the firearm fatality surveillance program. It is the National Violent Death Reporting System that will collect data regardless of the weapon used -- a gun, a fist or a knife.

"Everyone can be for it," says David Hemenway, director of Harvard's injury control research center. "The more data, the better. This is not only more politically correct. It makes a lot more sense in terms of the science."

For decades public health officials tried to address gun violence the way they tackled tobacco use and other leading causes of death. Former surgeon general C. Everett Koop labeled the bullet a pathogen. But targeting gun injuries can be hazardous to research. A pilot project at CDC in the 1990s to monitor firearm fatalities drew ire from gun advocates and was stopped after three years. Now all money appropriated to CDC to study injuries comes with a stipulation from Congress that the funds cannot be used to advocate for gun control.

Still, guns are the significant factor. Easier access to guns may explain why homicide rates are so much higher in the United States than in other high-income countries. The overall crime rates for robbery, car theft and sexual assault are about the same. But the homicide rate is about five times higher in this country. Among children aged 5 to 14, the gun suicide rate is 10 times higher.

Even within the United States, many more gun suicides occur in "high gun" states -- places such as Louisiana, Wyoming and West Virginia, where gun ownership is more common -- than in "low gun" states like Massachusetts, Rhode Island and New Jersey.

By not focusing on guns, the new initiative makes an effort to create some common ground between pro-gun and anti-gun groups. As James A. Mercy of CDC's injury prevention program says: "People with all points of view on guns can support collecting objective data on this problem."

The reporting system is modeled after the national traffic safety program, which analyzes all the factors that lead to fatal injuries. Out of this information has come seat belt and motorcycle helmet laws, better car designs, campaigns against drunk driving. Automobile fatalities have fallen about 80 percent per mile driven over 50 years.

The same bonanza in saved lives could come from a national reporting system on violent death. There are already tantalizing clues from Harvard's pilot program. Location is a key determinant of violent death, for example. More than 65 percent of women who are murdered are killed at home, while more than 80 percent of men are murdered away from home. The implication: Putting more police on the streets is not likely to protect women, but home visits by social workers might.

A national reporting system could also evaluate interventions. In Wisconsin, researchers found that guns recovered in local buyback campaigns were not the kind of guns used in homicides -- which suggests that buybacks may not reduce the toll of gun fatalities.

All in all, good data will lead to better public health.

© 2002 The Washington Post Company