The government is going to take a new look at the claims of about a third of the military veterans who now get disability payments for post-traumatic stress disorder (search).

The Veterans Affairs Department will begin a yearlong review next month of 72,000 cases after an internal study found inconsistencies in the way the claims were decided, including many cases approved though they lacked required medical evidence.

Millions of dollars a year could be involved.

Post-traumatic stress disorder, a mental illness characterized by subjective symptoms like flashbacks and nightmares, can be difficult to diagnose and quantify.

The review will cover veterans whose claims were approved between 1999 and 2004 and who receive full disability benefits -- $2,299 a month -- for PTSD alone or in combination with other conditions, said VA spokesman Phil Budahn.

The review won't consider the other side of the question, whether some veterans were wrongly denied benefits, and that has some critics upset.

"We need to pay as much attention to improper denials as we do to improper grants," said Illinois Rep. Lane Evans, ranking Democrat on the House Veterans' Affairs Committee (search).

Budahn said the VA could look at that later, "but first we're just going to keep our focus on the problem that's been identified."

Randy Reese, national service director for Disabled American Veterans, complained that the review would divert resources from the VA's backlog of several hundred thousand disability claims yet to be processed.

Last year, the VA spent $4.3 billion on PTSD disability payments, not including medical care.

The number of vets receiving compensation benefits for the illness rose nearly 80 percent between 1999 and 2004 -- from 120,265 to 215,871. During the same period, benefits for all types of disabilities grew by just 12 percent, to about 2.5 million.

So far, about 10 percent of the stress syndrome increase comes from veterans of the Iraq and Afghanistan wars, Budahn said.

Some experts say PTSD is diagnosed too readily. "PTSD went from being problematic being accepted as a condition, to being almost too easily accepted," said Wilbur Scott, a University of Oklahoma sociologist who has written about the disorder.

Budahn attributed the increase mostly to rising awareness of the illness, thanks to education efforts by the VA, veterans' organizations and health care workers.

If the current review finds a veteran's claim does not include adequate proof of a "stressor" -- the specific event or events that trigger the mental illness -- the veteran will be asked to provide more information and could ultimately lose benefits.

Proof of a service-related triggering event is one of the VA's few hard and fast requirements for PTSD disability benefits. The inspector general's report last May found that more than 25 percent of the PTSD cases reviewed lacked adequate proof of a service-related stressor.

"Some disabilities are inherently prone to subjective rating decisions, especially conditions such as PTSD where much of the information needed to make a rating decision is not physically apparent," VA inspector general Richard Griffin said. "This subjectivity leads to inconsistency."

Valid stressors can be difficult to prove.

"Sometimes it's a bureaucratic nightmare to get the evidence. People in wartime aren't sitting there with a steno pad keeping good records," said Mary Ellen McCarthy, Democratic staff director for the House Veterans subcommittee on disability assistance.

The VA said it would work with vets to help them prove their cases. It will also be on guard for possible fraud -- one vet in the sample review last spring submitted as evidence a personal account written by someone else and published on the Internet.

Some Iraq war vets with PTSD say identifying a single, specific stressor can be arbitrary.

"I think the whole year over there (in Iraq) was my stressor, but they actually wanted a specific incident," said Jesus Bocanegra, 23, of McAllen, Texas. His disability status won't be part of the review because it was approved this year. "I just gave them two."

Col. Charles Hoge, chief of psychiatry and behavior services at Walter Reed Army Institute of Research, said that while most service members with PTSD can identify incidents that affected them deeply -- for instance, their worst firefight, losing a buddy or seeing injured children -- "the reality is that there are also stressors that are ongoing."

Hoge has found evidence of PTSD symptoms in about 15 percent to 17 percent of service members returning from Iraq and Afghanistan.

"For most people who go into a war environment, it will affect them in some way," Hoge said. "An important minority of people are going to be affected to the degree that they need treatment."