In many traffic deaths in the United States, death certificates frequently fail to list alcohol as a cause of death.

It is no secret that drinking and driving causes many accidents in the United States. However, a new study found that the role of alcohol in the U.S. traffic deaths may be substantially underreported on death certificates.

More than 450,000 Americans were killed in traffic crashes between 1999 and 2009. Unfortunately, in cases where alcohol was involved, death certificates frequently failed to list alcohol as a cause of death, according to a press statement.

Researchers of the study noted that it's important that these things are clearly mentioned on death certificates so that citizens have a clear idea of alcohol's role in traffic deaths.

"We need to have a handle on what's contributing to the leading cause of death among young people," study author Ralph Hingson said in a press statement. "Researchers need reliable data to study the effects of policies aimed at reducing alcohol-related deaths. You want to know how big the problem is, and if we can track it. Is it going up, or going down? And what policy measures are working?"

Researchers focused on traffic deaths because, of all types of accidental fatalities, that's where researchers have the best data. This is partly because many U.S. states require that fatally injured drivers be tested for blood alcohol levels and nationwide about 70 percent of those drivers are tested.

They found death certificates greatly underreported the role of alcohol in traffic deaths between 1999 and 2009. Just over 3 percent listed alcohol as a contributing cause. But based on the  Fatality Analysis Reporting System (FARS) figures, 21 percent of those deaths were legally drunk.

"Whatever the reasons," Hingson said, "the role of alcohol in injury deaths may be seriously underestimated on death certificates. And the situation is likely worse with other types of accidental deaths, such as falls, drug poisoning/overdoses, and drowning, for which there is no mandatory blood alcohol testing or other reporting systems."

The study was published in the Journal of Studies on Alcohol and Drugs.