Carla McCabe spent a decade building nuclear bombs at the sprawling Rocky Flats complex near Denver. When she developed a brain tumor and asked for help, federal officials told her that none of the toxic substances used at the top-secret bomb factory could have caused her cancer.

Now, on the eighth anniversary of the federal program created to help sick nuclear weapons workers, the man who until recently was the program's top doctor says that McCabe, now 55, and many others like her are being improperly rejected.

The doctor, Eugene Schwartz, recently resigned (PDF) and in his first interview since quitting, he said many of the complaints that workers, advocates and lawmakers have leveled at the controversial program are valid. For instance, Schwartz said he repeatedly warned the U.S. Department of Labor that it is ignoring established medical knowledge about the dangers of bomb work.

"I was muzzled," said Schwartz, a Harvard-trained doctor with a master's degree in nuclear engineering, whose job was overseeing medical decisions at the federal compensation program.

The Labor Department took charge of the program, the Energy Employees Occupational Illness Compensation Program, at its creation on July 31, 2001. Today, it boasts that it has paid $5 billion in compensation and medical costs to more than 52,600 former workers or their survivors. That averages to $95,000 each.

But sick workers, who have banded together in multiple advocacy groups across the nation, point out that the Labor Department has denied nearly three out of four claims — 127,000 filed on behalf of sick nuclear weapons workers or their survivors in the past eight years.

The sick workers and their advocates say they feel vindicated that Schwartz confirms many of the complaints they've raised previously about waste, bias and bad science within the program.

"He is saying what we've been saying is true," said Harry Williams, a sick worker from Oak Ridge, Tenn., who helped found the national Alliance for Nuclear Workers Advocacy Groups and has spent more than a decade trying to help others. "With his credentials, the people in power will definitely have to pay attention."

Schwartz said Labor continues to incorrectly tell weapons workers with multiple diseases — including cancers of the brain, breast or bones — that radiation and other toxic substances that permeated the bomb factories could not have made them sick.

"That's madness," said Dr. Daniel Teitelbaum, a nationally recognized toxicologist who helped review cases when the federal government began compensating its sick nuclear weapons workers in 2001. In addition to working for the government, Teitlebaum also has testified on behalf of sick workers in other industries.

Teitelbaum and Schwartz said that multiple studies have found links between brain, breast and bone cancers and exposure to toxic substances such as plutonium, PCBs, and mixtures of chemicals and radiation — all key bomb ingredients.

Those studies were subjected to scrutiny by other experts, what is known as "peer review," before they were published in scientific journals.

But it has not been enough for the Labor Department.

When an official punched McCabe's diagnosis into the Labor Department's $11 million database of nuclear-related illnesses, nothing came back, her records show.

A Labor Department document regarding her case reads: "The research did not identify any toxic substances at the Rocky Flats Plant which are linked to ... any form of brain tumor."

The database did not include several peer-reviewed studies that have linked glial brain tumors like McCabe's to radiation and chemicals known to have been used at Rocky Flats. At least one of those studies, Teitelbaum said, was done on workers at the now-demolished Rocky Flats site itself. McCabe said she knows of at least eight coworkers who suffered similar brain tumors.

Program officials said they would send her case to another agency for analysis of her radiation exposures. But she's not holding her breath. That process takes on average two years, and none of her coworkers with brain tumors — most of whom are now dead — have been compensated through that process either.

The Labor Department declined to answer questions about its database or make public the grounds on which it includes some diseases and excludes others. It could not be learned how much consideration the peer-reviewed studies of brain cancers received.

"Somebody needs to do something about this," Teitelbaum said.

Schwartz said he tried but failed.

The McCabe case, he said, illustrates the flaws he tried to address. Schwartz said he began documenting his concerns shortly after he began work as medical director at the compensation program in March 2008.

A former epidemiologist for the World Health Organization, the physician and nuclear engineer from South Hadley, Mass., spent more than 30 years working in his field, which assesses the effects of industrial processes on workers. His research has focused on the causes and prevention of cancer and occupational diseases.

But Schwartz said his bosses at the Labor Department largely ignored the issues he raised, and then tried to silence him.

"The program needs scientific oversight," Schwartz said. "I was told they're not going to do that — repeatedly."

U.S. Sen. Mark Udall, a freshman Democrat from Colorado, helped push for the program as a congressman in 2000. This year, he introduced legislation to reform it. He said Schwartz's statements confirm what he and others have long believed.

"It is what I've suspected all along — an attempt by the agencies to delay and deny benefits to workers of nuclear facilities," Udall said.

The most serious allegations Schwartz raised involve what are arguably the two most important steps on the path to compensation and medical care.

One involves the very first step: Determining whether a worker was exposed to toxic substances that could have caused the disease in question. The other involves what is often the last step: Seeing whether a government-hired doctor agrees that the worker's exposures did, in fact, cause his or her disease.

In the first step, the Labor Department is supposed to use its database of diseases and exposures to screen which cases deserve further investigation. That initial screening is performed by a Labor Department employee, not a doctor.

Sick workers have previously complained that some cases are wrongly rejected for compensation because the database, created in 2006, is missing information about both exposures and their links to disease.

Schwartz said they have a point.

"Claims are being denied because of these problems," Schwartz said.

Labor Department rules say the database should be used for guidance and that the claims examiners should dig deeper if they suspect an illness arose from work at a bomb factory.

Schwartz, however, said some claims examiners told him they had used the database to decide cases without further review.

"The issue is how it is used — or misused," Schwartz said.

According to Schwartz, the disease-exposure links in the database were decided by one doctor: Jay Brown, an occupational medicine physician from Tacoma, Wash., who has a background in family medicine.

While the database has "this aura of scientific validity," Schwartz said, it has never been peer reviewed and does not take into account the combined effects of low-level radiation and toxic chemicals.

When asked about Schwartz's concerns, Brown replied via e-mail: "I don't work on SEM," the Labor Department's acronym for the database, called the Site Exposure Matrix. He referred all further questions to the Labor Department.

Labor Department officials would comment only through a spokeswoman, who said in an e-mail that Schwartz's concerns "were reviewed, discussed, and addressed appropriately."

The agency spokeswoman added that Labor pays Brown to conduct "independent research to identify established chemical-disease links specifically focused on materials used in the Department of Energy complex." Brown puts the links into a database he developed, called Haz-Map. He began compiling Haz-Map in 1991 as a way to identify and prevent occupational disease, according to his Web site. Haz-Map is published there and on the National Library of Medicine Web site.

The Labor Department says Haz-Map automatically feeds its information into the Labor database. Beyond that, Labor has released very little information about it. Officials there declined to release Brown's contract or give details about it.

Schwartz said that, before he resigned, he learned that the Labor Department was poised to remove from the database more than 100 toxic exposures it previously considered linked to disease.

"I asked about a half-dozen times to see the more than 100 links being removed," Schwartz said. "I was rebuffed." He added, sardonically: "Talk about scientific validity and transparency. Based on what science?"

More than half a million people have worked to build the nation's nuclear arsenal since World War II. Less than 15 percent of them have filed claims for aid, but that's still more than 180,000 former nuclear weapons workers — or their survivors — and they are from every state in the nation.

The compensation program was created after the federal government admitted that nuclear weapons workers had been exposed to dangerous levels of toxic substances at more than 300 weapons sites across the country.

But only 29 percent of those who applied have been approved for aid, and many of those received aid only after years of appealing.

The Labor Department, citing national security, has declined to provide the entire database to sick workers who've asked for it.

As for the government-contracted doctors who help decide who gets compensation, Schwartz said he examined the cases handled by doctors who review claims and found a pattern in the denials from some high-volume doctors. But he says Labor never investigated potential bias among these so-called "district medical consultants."

Labor has never checked the credentials of these 80 or so medical consultants, Schwartz said. He also says Labor was overstating the expertise of some doctors.

"This is no small issue," Schwartz said. "If a doctor were being hired by a hospital, his name would be checked through the National Practitioner Data Bank, state licensure and board certification would be checked, and he'd be asked to provide information on malpractice claims and Medicare sanctions.

"This program hasn't done that."

The compensation program divides sick workers into two groups. Scientists assess whether workers were exposed to high enough doses of radiation so that it was "at least as likely as not" the cause of their cancer.

The cases of workers who believe their cancer or other diseases were caused by chemical exposure — or a combination of chemicals and radiation — are weighed separately.

Schwartz said this second category is where he found most of the problems. According to Schwartz, the Labor Department has decided on its own that radiation exposures can be ignored if they weren't high enough to be the sole cause of a cancer.

Labor's rule book says "DOL has not found scientific evidence to date establishing a synergistic or additive effect" between exposure to radiation and other toxic chemicals.

Schwartz says this flies in the face of known science about how toxic chemicals and radiation can work together.

"It is generally accepted that the effect of combined exposure will likely be greater than either exposure separately," he said.

Schwartz showed ProPublica a document he says Labor sent to a medical contractor who was reviewing the case of a worker with skin cancer that could not be traced to radiation exposure alone.

The Feb. 19, 2009, letter (PDF) said that in such an instance: "Radiation should not be considered a toxic substance for any cancer and should not be a part of the medical opinion that is being requested of you."

Advocates for the sick workers say that instruction is illegal because the law requires the contribution of all toxic substances be considered — and that includes radiation.

"It is against the law," said Terrie Barrie, who leads the national Alliance for Nuclear Workers Advocacy Groups, from her home in Craig, Colo. "Congress has told them that. But it's like talking to a brick wall."

The Obama administration's new Labor secretary, Hilda Solis, received a detailed letter recently from Barrie's group complaining about what they see as "woefully inaccurate" information being used to wrongly deny claims.

One such case was that of Melissa Webb, whose job at the Mound nuclear facility near Dayton, Ohio, involved taking samples from 55-gallon drums of nuclear weapons waste at a top-secret facility five stories underground.

Webb was 33 when she learned she had Parkinson's disease. When she filed for compensation in 2007, she listed exposure to the solvent carbon disulfide as a possible cause. The Labor Department sent her a letter saying its records showed she was indeed exposed to carbon disulfide at her Ohio nuclear weapons site, but that there were no known links between the poisonous substance and Parkinson's.

However, the Labor Department's own Occupational Safety and Health Administration recognizes carbon disulfide's link to Parkinson's. And a bulletin issued by the same program that denied Webb's claim lists carbon disulfide at the top of its list of toxic links to Parkinson's.

"You'd think mine would be an open and shut case," said Webb, now 48, who first filed for compensation in 2004.

On Jan. 9 of this year, the same day the Labor Department sent Webb a letter saying it could find no toxic link to Parkinson's, Schwartz penned a memo (PDF) to his bosses at the Labor Department. One of his top concerns was the "scientific integrity and validity" of the database because it failed to find well-known links between toxic exposure and multiple diseases.

In the memo, Schwartz also called for the Labor Department to submit the program to outside peer review, in which independent experts would validate the conclusions in the database.

Schwartz said none of his bosses, all longtime program employees, ever responded.

But after he sent the memo, Labor Department officials put new demands on his work schedule and, records show, planned to have a manager accompany Schwartz on training trips to make sure he didn't raise issues about the scientific validity of the program.

Then, on April 16, Schwartz gave testimony about his allegations to the Government Accountability Office, the investigative arm of Congress, which has a continuing probe into problems with the compensation program.

The day after his GAO testimony, Schwartz got a memo from one of his bosses, policy chief Mike Chance, who had listened in on the GAO call. Chance told Schwartz that Labor was changing his previous work agreement, which had guaranteed he would not be sent on the road more than three times a month. The change was "non-negotiable," Schwartz said he was told.

"My current situation is the response" to his memo, Schwartz said. "I believe I was forced out."

When asked for a response, Rachel Leiton, director of the compensation program, sent this statement through a spokeswoman:

"While it is not our policy to discuss personnel matters, in this case we can emphatically state that Dr. Schwartz was not forced to resign. Rather, he submitted his letter of resignation on a completely voluntary basis."

When asked for elaboration, the same spokeswoman did not reply. But she did send out a press release on the amount of money collected so far by claimants in Colorado, where a bipartisan group of Congress members has introduced legislation to reform the national program.

The payments — some $400 million — are a lot of money, said Carla McCabe, the former bomb builder from Rocky Flats suffering from a brain tumor she believes is linked to her work.

"But they're denying people like crazy," she said. "I don't know how they keep denying us. I hope finally a whistleblower can help us get some answers."