About 151,000 Iraqis died from violence in the three years after the United States invaded, concludes the best effort yet to count deaths one that still may not settle the fierce debate over the war's true toll on civilians and others.
The estimate comes from projections by the World Health Organization and the Iraqi government, based on door-to-door surveys of nearly 10,000 households. Experts called it the largest and most scientific study of the Iraqi death toll since the war began.
Its bottom line is far lower than the 600,000 deaths reported in an earlier study but higher than numbers from other groups tracking the count.
The new estimate covers a period from the start of the war in March 2003 through June 2006. It closely mirrors the tally Iraq's health minister gave in late 2006, based on 100 bodies a day arriving at morgues and hospitals. His number shocked people in and outside Iraq, because it was so much higher than previously accepted estimates.
No official count has ever been available. While the U.S. military says it does not track Iraqi deaths, it has challenged some news reports of tolls from shootings and bombings as exaggerated indicating it does in fact monitor fatalities.
In November, a U.S. military official said the Pentagon was working with Iraqi authorities to better track civilian casualties. One goal is to avoid duplicate reports, said Col. Bill Rapp, a senior aide to the top U.S. commander in Iraq, Gen. David Petraeus.
The true toll may never be known because many deaths go unreported in the chaos that has gripped the country, or the numbers may be tainted by sectarian bias. The Iraqi security forces and government are led by Shiites. Muslim burial traditions add to difficulties many families are believed to simply bury loved ones before sundown on the day of death without ever reporting the fatality.
Still, Iraq's minister of health, Dr. Salih Mahdi Motlab Al-Hasnawi, defended the new estimate in a telephone interview with reporters Wednesday.
"This is a very sound survey" with a large sample and good methods, he said.
Richard Brennan of the New York-Based International Rescue Committee, which has done similar research in Kosovo, Uganda and Congo, agreed.
"The goal is not to give an absolute, precise number of deaths. The goal is to give a sense of the magnitude of the problem," he said.
White House deputy press secretary Tony Fratto said White House officials had not seen the study, but called the deaths of Iraqi citizens or any troops "tragic."
"We mourn the deaths of all people in Iraq as the country fights to defeat extremists ...," he said, contending that last year's surge of troops is reducing civilian deaths.
The United Nations paid more than $1.6 million for the new study. Results were published online Wednesday by the New England Journal of Medicine.
By any count, the toll is "massive," wrote Catherine and John Brownstein, statistics experts at Yale University and Harvard Medical School, respectively, in an accompanying essay. It likely still is low, because many Iraqis have fled and aren't there to report deaths and because Iraq is too dangerous to survey some areas.
A poignant example: One statistician was killed during the project and another, shortly afterward.
The survey was done by Iraq Ministry of Health employees during late 2006 and early 2007 in all 18 provinces, divided to get a valid sample of each area. But Iraqis hold a deep distrust of central authority, given the tribal nature of their society and the years they lived under Saddam Hussein, whose grip on power was built partially on a web of informers.
"We are dealing with surveys in a country where there is unrest and high insecurity situations," said Dr. Ties Boerma, a WHO official. "Surveys are imperfect, no matter how well we do it."
Researchers asked families whether any deaths had occurred in their households, recorded details like age and time and place of death, and assigned deaths as violence-related or not.
However, road accidents were not counted unless they were caused by a bomb one of many ways that surveyors could have underestimated the true toll, some experts said.
Limiting the study to the time since the invasion in March 2003, and extrapolating results to the whole country, researchers arrived at the 151,000 estimate. The study authors say they are 95 percent certain that the true number is between 104,000 and 223,000. Iraq's population is roughly 26 million.
That seems low, especially because the new survey saw no increase in deaths in recent years, as previous surveys did, said Columbia University's Dr. Ronald Waldman, who has long done humanitarian research for WHO and others.
More than 100 neighborhoods, mostly in Baghdad and Anbar, could not be visited for safety reasons. So researchers estimated deaths in those areas by using a formula based on information from another group that tallies fatalities, the British-based Iraq Body Count.
The Body Count project bases its figures mostly on media reports a method known to underestimate deaths because many go unreported. That group listed 47,668 civilian deaths from violence during the period studied in the WHO survey, and between 80,331 and 87,742 to date since the war began.
The group's numbers do not include deaths of fighters, but the WHO survey and an earlier one published in the journal Lancet in 2006 do.
The Lancet study, by Johns Hopkins University in Baltimore and the Al-Mustansiriya University in Baghdad, drew wide criticism, partly because it came out just before the 2006 congressional elections. It surveyed 1,849 households and concluded that 600,000 Iraqis had died from violence, mostly gunfire, and roughly 50,000 more from other causes like heart disease and cancer.
The WHO survey tallied only violence-related deaths, but researchers plan future reports on other health measures.
Les Roberts, a Columbia University epidemiologist involved in an even earlier survey in 2004 when he was at Johns Hopkins, believes the new toll is too low.
"This is consistent with family members not wanting to tell the government about violent deaths," he said.
The Associated Press began tracking civilian deaths after the new Iraqi government took office on April 28, 2005. Since then, at least 37,547 Iraqis have lost their lives due to war-related violence, according to the AP toll, which is considered a minimum since many killings go unreported or uncounted. It's compiled from police, hospital officials, morgue workers and verifiable witness accounts, and reporters and photographers at the scenes. Insurgent deaths are not included.