When Navy SEAL Robert Guzzo returned from Iraq, he feared seeking treatment for PTSD would endanger his career. The Fold devotes today’s show to telling his story. (The Fold/The Washington Post)

Every day about 22 veterans in the United States kill themselves, a rate that is about 20 percent higher than the Department of Veterans Affairs’ 2007 estimate, according to a two-year study by a VA researcher.

The VA study indicates that more than two-thirds of the veterans who commit suicide are 50 or older, suggesting that the increase in veterans’ suicides is not primarily driven by those returning from the wars in Iraq and Afghanistan.

“There is a perception that we have a veterans’ suicide epidemic on our hands. I don’t think that is true,” said Robert Bossarte, an epidemiologist with the VA who did the study. “The rate is going up in the country, and veterans are a part of it.” The number of suicides overall in the United States increased by nearly 11 percent between 2007 and 2010, the study says.

As a result, the percentage of veterans who die by suicide has decreased slightly since 1999, even though the total number of veterans who kill themselves has gone up, the study says.

VA Secretary Eric K. Shinseki said his agency would continue to strengthen suicide prevention efforts. “The mental health and well-being of our courageous men and women who have served the nation is the highest priority for VA, and even one suicide is one too many,” he said in a statement.

U.S. veteran suicides, by day

The study follows long-standing criticism that the agency has moved far too slowly even to figure out how many veterans kill themselves. “If the VA wants to get its arms around this problem, why does it have such a small number of people working on it?” asked retired Col. Elspeth Cameron Ritchie, a former Army psychiatrist. “This is a start, but it is a faint start. It is not enough.”

Bossarte said much work remains to be done to understand the data, especially concerning the suicide risk among Iraq and Afghanistan war veterans. They constitute a minority of an overall veteran population that skews older, but recent studies have suggested that those who served in recent conflicts are 30 percent to 200 percent more likely to commit suicide than their ­non-veteran peers.

An earlier VA estimate of 18 veterans’ suicides a day, which was disclosed during a 2008 lawsuit, has long been cited by lawmakers and the department’s critics as evidence of the agency’s failings. A federal appeals court pointed to it as evidence of the VA’s “unchecked incompetence.” The VA countered that the number, based on old and incomplete data, was not reliable.

To calculate the veterans’ suicide rate, Bossarte and his sole assistant spent more than two years, starting in October 2010, cajoling state governments to turn over death certificates for the more than 400,000 Americans who have killed themselves since 1999. Forty-two states have provided data or agreed to do so; the study is based on information from 21 that has been assembled into a database.

Bossarte said that men in their 50s — a group that includes a large percentage of the veteran population— have been especially hard-hit by the national increase in suicide. The veterans’ suicide rate is about three times the overall national rate, but about the same percentage of male veterans in their 50s kill themselves as do non-veteran men of that age, according to the VA data.

The VA’s hope is that Bossarte’s work will offer the seeds of a solution to the veterans’ suicide problem. He plans to include in the database death certificate data from all 50 states, Pentagon service records, VA hospital data and information from the VA’s crisis line.

Once the database is complete, Bossarte will know how many veterans served in combat, the year they left the military and the jobs they held in uniform. He will know how many of the deceased veterans called the crisis line for help and what kinds of treatment they were receiving from VA doctors before taking their lives.

“The big problem we face in suicide prevention is that we have so little empirical evidence to tell us which programs work,” Bossarte said.

In an effort to improve veterans’ access to mental health care, the VA committed last year to hiring more than 1,900 mental-health-care workers, all of whom will be on the job this summer. Bossarte hopes that his database will be able to help the VA determine where it needs more mental-health-care workers and how best to train them.

His database could help counselors know how long they should wait before following up with veterans and how many times they should call. Is one call sufficient, or are counselors better off making several calls over a period of several months?

Recently Bossarte noticed an interesting trend that followed the VA’s move to change the name of its call center from “suicide hotline” to “crisis line.” Overall call volume spiked in the months after the name change, and emergency rescues, which accounted for about 5 percent of all calls in 2010, dropped to less than 2 percent.

The figures suggest that more veterans are calling the line before their despair reaches suicidal levels. “We are getting them earlier, and that is a good thing,” Bossarte said.

Studies of Vietnam War-era veterans found that their suicide risk was highest in their first five years after leaving active duty. “I want to know everything I can find out about that five-year window for today’s soldiers,” Bossarte said. “I want to know whether it is five years or three years or six years.” The answer could help the VA tailor intervention strategies for troops who saw combat over the past decade.

Another question that has long vexed doctors is whether hospitalizing high-risk patients reduces the likelihood that they will attempt suicide over the long term. Bossarte’s database should help VA doctors know how long they should hold patients and how often they should follow up with them after their release.

In recent years, national suicide rates have been highest in the rural Mountain West states. Epidemiologists have speculated that the availability of guns, the lack of mental health services and a rugged independent mind-set that stigmatizes seeking help could all be driving up the rate. Bossarte’s data could help pinpoint the culprit and the cure.

“The VA is a great laboratory,” Bossarte said. “We have a population we can track. We have a national medical system in which we can do enhancements. We may be able to make recommendations that are transportable far beyond the VA’s walls.”