THURSDAY, June 19 (HealthDay News) -- Wars around the world have killed three times more people over the past half-century than previously estimated, a new study suggests.

The finding supports the notion of armed conflict as a "public health problem" whose instability leads not only to violent deaths, but to indirect deaths from infectious disease and other causes, experts add.

"War kills more people than we had previously thought," said lead researcher Ziad Obermeyer, a research scientist at Brigham & Women's Hospital, in Boston. "And that has to be taken into account when we're looking historically, and it's important for people and policy makers to know when they're looking at the consequences of the war. It's important that there's an awareness of how many people actually die."

In the study, Obermeyer's group compared data on war deaths from eyewitnesses and the media from 13 countries over the past 50 years with peacetime data in the United Nations World Health Surveys, which was collected after the end of the wars.

This method avoids problems collecting data during active combat, and also reduces counting deaths twice or exaggerating the number, Obermeyer said.

The researchers estimate that 5.4 million people died from 1955 to 2002 as a result of wars in 13 countries. These deaths range from 7,000 in the Democratic Republic of Congo to 3.8 million in Vietnam.

According to Obermeyer, the estimates are three times higher than those of previous reports. Data from this new study also suggests that 378,000 people worldwide died a violent death in war each year between 1985 and 1994, compared with 137,000 estimated at the time.

The biggest differences were seen in Bangladesh, where 269,000 people died during that country's struggle for independence, compared with previous estimates of 58,000, the report shows. In Zimbabwe, the researchers estimate that 130,000 people have died in times of conflict, compared with earlier estimates of 28,000.

The findings are published in the June 20 online edition of the British Medical Journal.

According to the authors, current methods of collecting data on those killed during war are plagued by biases that produce inaccuracies and underestimate the number of people actually killed. This can lead to widely varying casualty estimates. For example, in Iraq, a report published in the medical journal The Lancet in 2006 estimated that 650,000 Iraqi civilians had been killed by that time since the start of the war -- a claim disputed by the White House, whose own estimates put the death toll at 30,000.

In their study, Obermeyer's team drew on several sources to try to more accurately estimate the number of military and civilian deaths from recent wars. Their estimates do not include people who died during the war from starvation, sickness or other conditions indirectly caused by war.

"There is a notion in political thought that the number of deaths due to war has been declining in recent years," Obermeyer noted. "That is attributed to a lot of different things, but among them technological innovations like 'smart' bombs and different strategic priorities. This idea appears to be supported by media reports. But what we are finding is these reports are not a reflection of reality."

Contemporary media reports of deaths are not to be fully trusted, Obermeyer addeds. "The reason we should be skeptical of media reports is that they are subject to political pressures and cannot always be verified," he said. "These numbers can be pushed up or down, depending upon what kind of political pressure is being exerted."

Richard Garfield, a professor of clinical international nursing at Columbia University in New York City and the author of an accompanying editorial in the journal, said that even this method underestimates the number of people killed in wars.

"Even though the data on war deaths is not very good, it is much better . . . in poor developing countries -- where virtually all wars now are -- than it was 10 or 20 years ago," Garfield said.

However, all deaths because of war are not being counted, Garfield said, since even Obermeyer's team left out the more indirect deaths from starvation, infectious disease and other illnesses, and forms of injury not directly linked to armed combat.

"We are counting more of the violent deaths, but we only irregularly address indirect deaths, which may be far greater than combatant deaths," he added.

More information

The Iraq Coalition Casualty Count offers their tally of lives lost in the current Iraq conflict on their Web site.

SOURCES: Ziad Obermeyer, M.Phil., research scientist, Brigham & Women's Hospital, Boston; Richard Garfield, Dr.PH., professor of clinical international nursing, Columbia University, New York City; June 20, 2008, online edition, British Medical Journal