Versions of this story were co-published with NPR and Stars and Stripes. For more coverage, listen to NPR's All Things Considered.

During the past few decades, scientists have become increasingly persuaded that people who suffer brain injuries benefit from what is called cognitive rehabilitation therapy -- a lengthy, painstaking process in which patients relearn basic life tasks such as counting, cooking or remembering directions to get home.

Many neurologists, several major insurance companies and even some medical facilities run by the Pentagon agree that the therapy can help people whose functioning has been diminished by blows to the head.

But despite pressure from Congress and the recommendations of military and civilian experts, the Pentagon's health plan for troops and many veterans refuses to cover the treatment -- a decision that could affect the tens of thousands of service members who have suffered brain damage while fighting in Iraq and Afghanistan.

Tricare, an insurance-style program covering nearly 4 million active-duty military and retirees, says the scientific evidence does not justify providing comprehensive cognitive rehabilitation. Tricare officials say an assessment of the available research that they commissioned last year shows that the therapy is not well proven.

But an investigation by NPR and ProPublica found that internal and external reviewers of the Tricare-funded assessment criticized it as fundamentally misguided. Confidential documents obtained by NPR and ProPublica show that reviewers called the Tricare study "deeply flawed," "unacceptable" and "dismaying." One top scientist called the assessment a "misuse" of science designed to deny treatment for service members.

Tricare's stance is also at odds with some medical groups, years of research and even other branches of the Pentagon. Last year, a panel of 50 civilian and military brain specialists convened by the Pentagon unanimously concluded that cognitive therapy was an effective treatment that would help many brain-damaged troops. More than a decade ago, a similar panel convened by the National Institutes of Health reached a similar consensus. Several peer-reviewed studies in the past few years have also endorsed cognitive therapy as a treatment for brain injury.

Tricare officials said their decisions are based on regulations requiring scientific proof of the efficacy and quality of treatment. But our investigation found that Tricare officials have worried in private meetings about the high cost of cognitive rehabilitation, which can cost $15,000 to $50,000 per soldier.

With so many troops and veterans suffering long-term symptoms from head injuries, treatment costs could quickly soar into the hundreds of millions, or even billions of dollars -- a crippling burden to the military's already overtaxed medical system.

The battle over science and money has made it difficult for wounded troops to get a treatment recommended by many doctors for one of the wars' signature injuries, according to the NPR and ProPublica investigation. The six-month investigation was based on scores of interviews with military and civilian doctors and researchers, troops and their families, visits to treatment centers across the country, confidential scientific reviews and documents obtained under the Freedom of Information Act.

"I'm horrified," said James Malec, research director at the Rehabilitation Hospital of Indiana and one of the reviewers of the Tricare study. "I think it's appalling that we're not knocking ourselves out to do the very best" for troops and veterans.

Defense Secretary Robert Gates, who has complained over the past year about the growing cost of the Pentagon's health care budget, declined a request for an interview. George Peach Taylor, the newly appointed acting assistant secretary of defense for health affairs, the top ranking Pentagon health official, also declined repeated interview requests. Tricare officials defended the agency's decision not to cover cognitive rehabilitative therapy and said it was not linked to budget concerns.

Capt. Robert DeMartino, a U.S. Public Health Service official who directs Tricare's behavioral health department, said Tricare is mandated to ensure the quality, consistency and safety of medical care delivered to service members.

He said those standards can be difficult to meet with cognitive rehabilitation. Therapists design highly individualized treatment plans, often relying on a variety of different techniques. The holistic approach and lack of standardization makes it hard to measure the effects of the therapy, he added.

DeMartino noted that the agency covers some types of treatment considered part of cognitive rehabilitative therapy. For instance, Tricare will pay for speech and occupational therapy, which can play a role in cognitive rehabilitation.

DeMartino said cost played no role in the agency's decision, calling such a suggestion "completely wrong." He defended the agency's studies of cognitive rehabilitation, calling them objective scientific reviews designed to ensure troops and retirees receive the best treatment possible.

Cognitive rehabilitation therapy "is a new field for us," DeMartino said. "We don't know what it is. That's really an important thing. You don't want to send people out when you don't know what treatment they're going to get and what the services are going to be."

Officials at the Pentagon are themselves divided on the value of the treatment. A handful of military and veteran facilities provide cognitive rehabilitation therapy, though most do not have the capacity or offer programs of limited scope.

Tricare was designed to fill in such gaps in the military health system by allowing troops and veterans access to civilian medical providers. But since Tricare has a policy against covering cognitive rehabilitation, service members and retirees who seek treatment at one of the nation's hundred of civilian rehabilitation centers could have their claims denied, or only partly paid.

The contradictory policies have resulted in unequal care. Some troops and their families have relied upon high level contacts or fought lengthy bureaucratic battles to gain access to civilian cognitive rehabilitation programs which provide up to 30 hours of therapy a week. Soldiers without strong advocates have been turned away from such programs, or never sought care, due to Tricare's policy of refusing to cover cognitive rehabilitation therapy.

As a result, many soldiers, Marines and sailors with brain injuries wind up in understaffed and underfunded military programs providing only a few hours of therapy a week focused on restoring cognitive deficits.

Sarah Wade's husband, Ted, was a sergeant with the 82nd Airborne Division when a roadside bomb tore through his Humvee in February 2004. The blast severed his right arm above the elbow, shattered his body and left him with severe brain damage.

After the military medically retired her husband later that year, Wade struggled to find appropriate care for him. The closest VA hospital set up to handle such complex injuries was in Richmond, Va., a 320-mile drive from their home in North Carolina.

Tricare, however, would not pay for cognitive rehabilitation at a nearby civilian program. Wade, who once worked as an intern on Capitol Hill, turned herself into a one-woman lobbyist on her husband's behalf. She called her representatives and met with senior VA and DOD officials. She testified before Congress, met President George W. Bush and Gates, and was recently invited to the White House by President Barack Obama for a bill signing ceremony.

Wade managed to set up a special contract between the VA and a local rehabilitation doctor to help her husband. But now she wants to move back to Washington, D.C., to be closer to family.

She must begin her fight all over again -- more phone calls to Tricare, more visits to government offices, more battles to get Ted Wade the care he needs.

"We go to Capitol Hill like some people go to the grocery store," Wade joked one afternoon during a recent visit to Washington. "If we can't figure it out, then probably nobody can."

Brain Campaign

The campaign to persuade Tricare to cover cognitive rehabilitation therapy began in earnest after the scandal at Walter Reed Army Medical Center in Washington in 2007. News reports featured brain-damaged soldiers living in squalid conditions and receiving substandard care.

The Brain Injury Association of America, a grassroots advocacy group for head trauma victims, started lobbying Congress and the Defense Department to order Tricare to cover rehabilitation for service members.

The campaign was a natural extension of the association's mission. Each year, more than 1.4 million American civilians suffer brain injuries in car accidents, strokes and other medical emergencies. They and their families often have to battle private insurance companies for cognitive rehabilitation.

The insurance industry is divided: Five of 12 major carriers will pay for cognitive rehabilitation therapy for head trauma, according to Tricare's study. Aetna, United Healthcare and Humana cite national evidence-based studies and industry-recognized clinical recommendations that point to the therapy's benefits.

The federal Centers for Medicare and Medicaid Services does not have a single national policy on cognitive rehabilitation. Instead, it leaves decisions to local contractors, often insurance carriers who process claims for the agency. The contractors are able to provide the therapy case by case, so long as they determine the treatment is "reasonable and necessary," a Medicare spokesman said.

"The totality of the evidence appears to support the value of cognitive rehabilitation for people with traumatic brain injury in improving their function," said Robert McDonough, the head of clinical policy at Aetna. "We feel on balance the evidence leads us to conclude that cognitive rehabilitation is effective."

Carriers and doctors providing the service can point to a long list of medical associations and scientific studies backing the effectiveness of cognitive therapy: The National Institutes of Health; the National Academy of Neuropsychology and the British Society of Rehabilitation Medicine, among others, have weighed in supporting the treatment.

Armed with such evidence, brain injury association lobbyists did not have much trouble finding support in Congress. By 2008, more than 70 House and Senate members had signed letters to Gates asking him to support funding for cognitive rehabilitation therapy. Then-Sen. Obama led a group of 10 senators urging Tricare to pay for therapy.

They noted that the Pentagon and the VA have improved their efforts to treat brain injury, including increases in the number of doctors and therapists available at facilities.

But the military needed to do more, they said. They wrote that Tricare should cover cognitive rehabilitation so all troops "can benefit from the best brain injury care this country has to offer."

"Given the prevalence of TBI among returning service personnel, it is difficult to comprehend why the military's managed healthcare plan does not cover the very therapies that give our soldiers the best opportunities to recover and live full and productive lives," the letter said.

A response letter from the Pentagon told the representatives that Tricare officials had not been convinced by available evidence. "The rigor of the research ... has not yet met the required standard," wrote Gordon England, then the deputy defense secretary.

Everyone Agrees

On an unusually hot spring day in April 2009, 50 of America's leading brain specialists gathered for two days in a sterile hotel ballroom in suburban Washington, D.C.

The Defense Centers of Excellence for Psychological Health and Traumatic Brain Injury, the Pentagon's lead program for the treatment of brain injury, convened the conference to help settle the debate about cognitive rehabilitation therapy.

The participants were top researchers and doctors from the military and civilian world: neurologists, neuropsychologists, psychiatrists, therapists, family doctors and rehabilitation experts.

After two days of discussion, the group hammered out a consensus report, representing the combined wisdom of the field. Their unanimous conclusion: Cognitive therapy improved the thinking skills and quality of life for people suffering from severe and moderate head injuries. Troops with lingering problems from a mild traumatic brain injury, or concussion, also could benefit from the therapy, the experts said.

The consensus was not binding. But those in attendance believed that their opinion -- based on the decades of combined clinical experience and academic study present in the room -- would lead to troops' receiving better treatment.

"When you get the right people in the right room at the right time, you'd expect it would influence the decision makers," said Maria Mouratidis, chairwoman of psychology and sociology at the College of Notre Dame in Baltimore and a conference participant.

Shortly after the conference ended, however, a handful of top officials from the military's medical system met to discuss the findings at Tricare's headquarters, an anonymous sprawl of office buildings in Falls Church, Va., known as Skyline 5.

One person familiar with the discussion, who did not want to be identified for fear of reprisal, said money was part of the debate.

Official Pentagon figures show that 188,000 service members have suffered brain injuries since 2000. Of those, 44,000 suffered moderate or severe head injuries. Another 144,000 had mild traumatic brain injuries. However, previous ProPublica and NPR reports showed that number likely understates the true toll by tens of thousands of troops. Some estimates put the number of brain injuries at 400,000 service members.

Mild traumatic brain injuries are the most common head trauma in Iraq and Afghanistan. Commonly caused by blast waves from roadside bombs, such injuries are defined as a blow to the head resulting in an alteration or loss of consciousness of less than 30 minutes. Studies suggest that while most troops with concussions heal quickly, some 5 percent to 15 percent go on to suffer lasting difficulties in memory, concentration and multitasking.

For the military's health system, the costs of treating brain damaged soldiers with cognitive rehabilitative therapy added up quickly. If tens of thousands of service members and veterans were authorized to receive such treatment, the bill might be in the billions, using high-end estimates for the cost of treatment from the Brain Injury Association.

The costs could swell the Pentagon's annual $50 billion health budget -- at a time when Gates has said the military is being "eaten alive" by skyrocketing medical bills.

Tricare "is basically an insurance company. They'll take no action to provide more service," said the person familiar with the conversation, who would only discuss it in general terms. "If they do it, it's an enormous cost."

At the meeting following the consensus conference, the person said, Tricare staked out its own position: "They had already decided not to do it," the person said.

NPR and ProPublica contacted two others who attended the meeting. Jack Smith, Tricare's acting chief medical officer, said through a spokesman that he could not recall the meeting, but "can't say for sure there wasn't one." Rear Adm. David J. Smith, the joint staff surgeon, declined comment through a spokesman.

The Contract

Soon after the meeting, Tricare sprang into action. In May 2009, records show, it issued a $21,000 contract to the ECRI Institute, a respected nonprofit research center best known for evaluating the safety of medical devices.

The contract called for ECRI to review the available scientific literature to weigh the evidence for whether cognitive rehabilitation therapy helped improve patients with traumatic brain injuries.

Tricare routinely hires contractors to carry out assessments to help determine which medical treatments to fund. But in selecting ECRI, Tricare had a pretty good idea of the response it would receive. ECRI had conducted a similar review for Tricare in 2007 that cast doubts on the evidence supporting cognitive rehabilitation therapy.

To carry out the new review, ECRI followed its standard protocol. It chose to include only randomized, controlled studies. Such studies randomly divide patients into groups that receive different treatments in order to compare their effects.

ECRI gave more credence to blind studies, meaning that patients did not know whether they were receiving genuine therapy or a placebo -- a fake treatment. Blinding reduces bias and is considered one of the most rigorous standards that can be used in scientific testing.

ECRI also excluded studies deemed irrelevant; those studies with fewer than 10 patients; and studies where 15 percent or more of the patients were injured from a nontraumatic blow, such as stroke.

The criteria resulted in the elimination of much of the published scientific literature on cognitive rehabilitative therapy. Before applying the protocol, ECRI identified 318 articles as potential sources of information about cognitive rehabilitative therapy. The firm's final report examined 18.

Based on this limited pool, ECRI graded the evidence for the benefits of cognitive therapy as being "inconclusive" or offering only "low" or "moderate" support of improvement in patients' cognitive functions.

The final report, delivered to Tricare in October 2009, noted some areas of benefit. For instance, "tentative" evidence showed cognitive therapy significantly improved quality of life for brain-damaged patients.

ECRI's review wasn't limited only to science. The review noted one study that found that comprehensive cognitive rehabilitative therapy could cost as much as $51,480 per patient. By contrast, sending patients home from the hospital to get a weekly phone call from a therapist amounted to only $504 per patient.

Overall, the report concluded, the evidence for most benefits from cognitive rehabilitation therapy remained inconclusive, especially when compared to cheaper programs.

"The evidence is insufficient to determine if comprehensive, holistic (cognitive rehabilitation therapy) is more effective than less intensive care" in helping patients, the 2009 report concluded.

Tricare Criticized

By the summer 2009, ECRI researchers had finished a draft of the study. ECRI, later joined by Tricare, asked outside scientific experts to review it.

The reviews, according to interviews and copies obtained by NPR and ProPublica, were uniformly critical.

(NPR and ProPublica obtained a copy of the ECRI reports through the Freedom of Information Act. However, Tricare denied access to reviews of the reports. ProPublica and NPR have appealed the request, but obtained copies of the reports and information on the reports from sources.)

The reviewers acknowledged that more research was needed on cognitive rehabilitation therapy. However, they noted that the Tricare report ran counter to several other so-called meta-analyses, which combine multiple, individual scientific studies to achieve greater statistical reliability.

For instance, a 2005 article in the Archives of Physical Medicine and Rehabilitation, a peer-reviewed journal that is one of the mostly widely respected in the field, examined 258 studies. It concluded that "substantial evidence" supported cognitive rehabilitation. The review included 46 randomized control studies -- more than double the number in the Tricare study.

Reviewer Keith Cicerone, a leading civilian researcher who runs the JFK Johnson Rehabilitation Institute's Center for Head Injuries in New Jersey, disputed Tricare's contention that the treatment was new and untested.

"We have a significant body of evidence describing cognitive rehabilitation and showing what works in cognitive rehabilitation," Cicerone said. "The idea that cognitive rehabilitation is new and untested is simply not true. It's got a better evidence base than most things that we do in rehabilitation."

Asked to explain in plain terms, Cicerone grew animated: "The arguments that are being made against" cognitive rehabilitation "in terms of the level of research that has been conducted are hooey," he said. "It is baloney."

The outside experts also attacked Tricare and ECRI for relying upon a methodology that ruled out important research. ECRI's protocols, they acknowledged, are well-suited for drug studies, where it is easy to prevent patients from knowing which pill they are receiving.

But ECRI's protocols do a poor job in assessing rehabilitation therapy where patients and doctors constantly interact in face-to-face treatment sessions. Other well-accepted methodologies, they said, have been designed to examine the benefits of therapeutic interventions.

They also questioned the reasons for excluding studies with a small number of patients, or with differing causes for brain injury, since a stroke can produce the same types of symptoms as a blow to the head.

Malec, the research director at the Rehabilitation Hospital of Indiana, said Tricare's study sounded like it came from a private insurance company seeking to cut costs. His review said that Tricare's study "fails to represent the evidence relevant to evaluating the effectiveness of cognitive rehabilitation after traumatic brain injury."

In an interview, he said Tricare's demand for conclusive evidence was understandable, but ill-advised. While research continues, existing evidence indicates that the therapy helps, with no studies showing that it harms troops.

"They missed the forest for the trees. They missed the big picture," he said.

Some of the researchers accused Tricare of using ECRI's strict assessment protocols as a cover to justify denying troops' coverage.

Wayne Gordon, director of rehabilitation psychology and neuropsychology services at Mt. Sinai School of Medicine in New York, called the review "dismaying" and "unacceptable." He compared it to tobacco companies that dismissed studies that showed a link between smoking and cancer.

"The ECRI Institute seems to be stating that, while sufficient evidence exists for there to be consensus among diverse groups that cognitive rehabilitation is a useful service, this evidence is 'not good enough' for Tricare," wrote Gordon, who declined to explain his comments further in an interview. He wrote that the ECRI study was "designed to reach a negative conclusion."

ECRI also asked two additional researchers to examine the report, John Corrigan, director of the Ohio Valley Center for Brain Injury Prevention and Rehabilitation in Columbus, and John Whyte, the director of Moss Rehabilitation Research Institute in Pennsylvania, both leading researchers in the field.

Both men declined to comment, citing their contractual obligations with ECRI, and Tricare declined to release their reviews. People familiar with their contents said Corrigan and Whyte closely mirrored the views of their fellow critics. They recommended that ECRI use a different method to judge studies of cognitive therapy, but the institute refused.

ECRI "said thank you very much, but we're not changing anything," said one person familiar with the review process.

More Studies, More Waiting

In an interview, ECRI Institute officials defended their firm's methodology. The system is designed to provide a rigorous review free from researchers' bias, they said.

Karen Schoelles, ECRI's medical director for the health technology assessment group, acknowledged that some of the institute's criteria -- such as accepting only studies with 10 or more patients -- were "arbitrary." But she said they were widely accepted in the assessment industry.

She also noted that Tricare officials were aware of the criteria and made no attempt to change or adjust them. Tricare used ECRI Institute for almost 10 years to carry out health reviews, though the agency recently terminated the contract and selected a new firm to carry out assessments.

Cognitive rehabilitation "may be on to something," Schoelles said. "But it needs more research."

Schoelles acknowledged that ECRI's own reviewers had criticized the report. ECRI offered to provide copies of the reviews, but later said that Tricare ordered them not to release them.

Stacey Uhl, the lead researcher on the review, said the criticism did not change her view that randomized controlled trials were the best way to assess the quality of evidence.

She noted the review found evidence that cognitive therapy did help in some way and said she would not rule out seeking such care for a loved one.

"I as a parent would want my child to receive all available therapies," she said.

DeMartino, the Tricare official who commissioned the report, acknowledged the outside reviewers had "very, very strong opinions" that were "of concern."

He said Tricare was conducting a review to determine whether ECRI's techniques were best suited to measure cognitive therapy's benefits. He denied submitting cognitive therapy to overly-strict review standards.

"You get what you ask for," DeMartino said. "They tell us what they're going to give us, and it's our job to sort of say, 'Okay, we understand that within the limitations of their methodology, this is the information that we get.'"

He added: "The better the information you have, the better that you can move forward and do the best thing." The Tricare reports, coupled with high cost projections, ended the legislative push to get cognitive rehabilitation for service members and veterans.

Last year, Congress ordered the Pentagon to conduct further studies to review the effectiveness of the therapy, but those studies have not yet begun and results are not expected for several years.

Tricare said it would conduct regular reviews to monitor developments in the field. DeMartino first said Tricare would carry out a new review beginning in September. A spokesman later clarified that the National Academy of Sciences Institutes of Medicine would perform the review. It is scheduled to be completed by the end of 2011.

Susan Connors, president of the brain injury association, said she was stunned by the need for legislation at all. As the Pentagon conducts yet more studies, thousands of troops and veterans may be going without the best known treatment available. Thousands more would have to rely on military hospitals or veterans clinics far from their homes, or with substandard programs. The Tricare refusal shut down access to the hundreds of civilian rehabilitation clinics nationwide.

"I'm very disappointed by the resistance," she said. "The military should want to do this."

Struggling for Care

Tricare's stance has not made it impossible to get cognitive rehabilitative. But it has discouraged civilian clinics from treating soldiers.

In interviews, several clinic owners and medical directors described their frustrations.

On some occasions, they were paid after developing relationships with individual Tricare claims processors or case managers, only to have the arrangements fall apart if the person left.

"We have tried to get Tricare and just beat our head against the wall," said Brent Masel, the president of the Transitional Learning Center in Galveston, Texas. "It took forever to get paid. It was always a fight."

Mark Ashley, the president of the Centre for Neuro Skills, a chain of rehabilitation clinics, said Tricare and other insurance providers were unwilling to pay because those with brain injuries can often perform basic functions that let them get through their daily lives.

They are "able to walk around, able to maneuver, but can't function cognitively in a manner that's safe, appropriate or competent," said Ashley, a past president of the brain injury association. "We can fix much of that, but it takes an exhaustive amount of time and effort. That's where the payers are out of touch."

One of the nation's top brain injury centers set up a charity program to help cover gaps left by Tricare. Susan Johnson, who runs Project Share at the Shepherd Center in Atlanta, said Tricare pays only about 40 cents of each dollar of care provided for the type of comprehensive program that the clinic has found successful. The rest comes from Bernie Marcus, a billionaire philanthropist, and income from inpatient services.

"These guys go and they put their lives on the line and we put them in this situation that's difficult for some and less difficult for others to get care," Johnson said. "I find it frustrating."

Other clinic owners said they were able to game the system by providing cognitive therapy, but billing for other Tricare-covered services -- putting them at risk of being accused of false billing.

One clinic manager acknowledged being "creative" when submitting bills to Tricare. He said that he submitted bills to Tricare for occupational therapy when the treatment focused more on improving memory.

"They won't pay for this, but they will pay for that," said the manager, who did not want to be identified for fear of damaging his ability to receive payments. "You just have to figure out how to work the system."

Soldiers and families agreed that Tricare's stance has made getting care a battle.

Sarah Wade said she patched together adequate care for Ted, arranging for him to go to a VA hospital for some services and to travel to Walter Reed Army Medical Hospital for others.

Tricare would have paid for some things, such as a physical therapist to help him learn to walk again. But she has had no luck trying to persuade Tricare to pay to treat his brain injury.

In frustration, Wade personally visited a high-ranking official at the Veterans Affairs Department. He, in turn, ordered a VA hospital to fund a special contract with a local civilian rehabilitation doctor near the Wades' North Carolina home.

"Yes, we have been able to get [cognitive rehabilitation] paid for, but it's been with a lot fighting, red tape, and bureaucracy," Sarah Wade said. "It's his greatest injury and the one that impacts his life the most, that impacts his ability to be a human." She added, "It shouldn't be this hard."

The Wades credit the rehabilitation that Ted has received with markedly improving his cognitive problems. After his 2004 injury, Ted spent months regaining consciousness. Doctors were unsure about his mental state, not certain he would ever talk or even think rationally.

Today, Ted speaks in slow, sure sentences, even cracking jokes. He can make decisions -- choices that seem simple enough to someone with normal cognitive skills, but which often stymie those with brain injury.

He knows, for example, to buy cherry tomatoes at the store rather than big tomatoes, which are hard for him to chop and slice with only one arm. He can read through a menu, and pick food that's nutritious. He can wash and fold his own laundry.

One recent day after dining at a Mexican restaurant in Washington, Ted smiled when Sarah reminded him that he was once unable to figure out whether he liked hot sauce on his tacos.

"It's been a long, slow process," he said.

Inform our investigations: Do you have information or expertise relevant to this story? Help us and journalists around the country by sharing your stories and experiences.

Get ProPublica's investigations and major stories delivered to your inbox.