Editor's note: The Atlanta Journal-Constitution ran a shorter version of this story in their print edition

A FEMA trailer in Coden, Ala. (Bill Starling/ProPublica)

Update 10/05/08 5:15 PM: A newer version of the congressional report examining CDC's and ATSDR's response (PDF) replaces the version previously linked to in this story.

The Centers for Disease Control study (PDF) sounded reassuring when it was made public in 2007. Hurricane Katrina survivors didn't have to worry about reports that there were harmful levels of formaldehyde in their trailers. The air was safe to breathe and the contamination would not reach a "level of concern" as long as they kept the windows open.

Today, senior CDC officials acknowledge that the study was based on a fundamental error.

An agency standard says that people exposed to as a little as 30 parts of formaldehyde per billion parts of air (ppb) for more than two weeks can suffer constricted airways, headaches and rashes. The trailers all measured above that level.

But the scientists who conducted the study used a much higher agency standard to evaluate the formaldehyde in the trailers: instead of 30 parts per billion, they said health dangers wouldn't occur until the substance reached 300 ppb, 10 times greater than the long-term standard. According to the CDC, people exposed to that amount for just a few hours can suffer respiratory problems and other ailments.

The story of the Katrina survivors and the trailers has been told many times in Congressional hearings and in the media. But it has been unclear until now why government officials continued reassuring residents the trailers were safe, at least a year after they should have been warning them to get out.

A reconstruction of how CDC and other government agencies handled the formaldehyde problem, drawn from documents, interviews and a new congressional report (PDF), suggests that top government officials were worried from the beginning about lawsuits by the people living in the trailers. Communications among government agencies broke down, so much so that the CDC wasn't aware that other government agencies were continuing to rely on a flawed study.

CDC’s reaction to the formaldehyde problem was “marred by scientific flaws, ineffective leadership, a sluggish response to inform trailer residents of the potential risks they faced, and a lack of urgency to actually remove them from harm’s way,” concludes the 40-page report, scheduled to be released this week by Democrats on the Science and Technology Committee’s subcommittee on investigations and oversight for the U.S. House of Representatives.

The report also chronicles the futile efforts of Christopher De Rosa, a senior CDC toxicologist, to warn top officials of another problem with the 2007 study: It failed to mention that formaldehyde can cause cancer.

Other clues were found by ProPublica, an investigative journalism organization based in New York City, which examined hundreds of pages of e-mails and other documents and interviewed former and current CDC scientists and officials.

The story that emerged is of a government bureaucracy that remained silent as the formaldehyde crisis mounted, straying from its mission to serve the public by "providing trusted health information to prevent harmful exposures and disease related to toxic substances."

Joe Little, one of the ATSDR scientists who conducted the study, said they chose the higher 300 parts per billion standard because it is the lowest level that is likely to cause a health "effect." The 30 ppb level, he said, is a "risk" level, meaning that illness is less certain. "Risk and having an effect are two different things," he said.

A year after the CDC issued its first study, it conducted new tests in occupied trailers. The results were clear. Formaldehyde levels surpassed 100 ppb -- more than three times the 30 ppb standard for year-long exposure -- in 41 percent of the trailers tested.

"It would be wise for people to be relocated" from the trailers before summer, CDC director Julie Gerberding said in a prepared statement to the media on Feb. 14, 2008.

A CDC official later told Congress that he should have noticed that the first study used the wrong standard.

“I believe everybody who reviewed that document had the opportunity to see that, and we missed that,” said Tom Sinks, deputy director of the CDC’s Agency for Toxic Substances for Disease Registry, or ATSDR.

Early warnings, delayed testing

Katrina survivors started complaining about the air in their trailers almost as soon as they began moving into them, in the fall of 2005. Formaldehyde was quickly pinpointed as a possible cause, because it's often in the glue used to make plywood and particleboard, which are found in most trailers.

Some people described an almost overpowering "new car smell" in the trailers, which were supplied to them by the Federal Emergency Management Agency. Paul Nelson, who moved into a FEMA trailer that September, described it as "kind of a musky smell, but it burns your eyes and sinuses."

"It gives you excruciating headaches," he said.

By April 2006, complaints from trailer occupants had grown so loud that the Sierra Club stepped in and tested dozens of units. It found levels as high as 340 ppb in some of the trailers, more than 10 times the amount considered safe for long-term exposure.

FEMA employees working in the hurricane area asked for government testing. But the agency's lawyers initially resisted. Documents obtained by the House Committee on Oversight and Government Reform show that litigation was a major concern.

"While I agree that we should conduct testing we should not do so until we are fully prepared to respond to the results," FEMA attorney Rick Preston said in an e-mail (PDF) to another FEMA employee. "Once you get results, and should they indicate some problem, the clock is running on our duty to respond to them."

In a series of conference calls (PDF) that began in June 2006, FEMA's lawyers discussed the problem with representatives of the Environmental Protection Agency and the Agency for Toxic Substances for Disease Registry, a little-known division of the CDC that studies how environmental hazards can harm people. One of the ATSDR scientists on the calls later told congressional investigators, "To say initially it was a P.R. problem is probably accurate."

In September and October of 2006, the EPA gathered air samples in 96 unoccupied trailers. According to an email written by a participant on the conference calls, an EPA official warned that formaldehyde levels in the units might turn out to be "far above" the acceptable level "even after we ventilate them."

Preston told ATSDR to analyze the samples.

But the instructions given to the two ATSDR scientists assigned to the task -- Joe Wright and Scott Little -- seemed odd. FEMA told them to establish "a frame of reference," not to assess the actual health consequences, Little told ProPublica.

When Preston gave the samples to Wright and Little, he attached a letter saying, "No information should be released to any third party without my express permission."

FEMA did not respond to ProPublica's questions on the trailers.

It referred ProPublica to previous news releases in which the agency denied influencing the CDC study.

Mistaken standards

One of the first things Wright and Little had to do was decide how much formaldehyde was too much.

Federal agencies use more than half a dozen different formaldehyde standards for different segments of the population, creating confusion for lawmakers, manufacturers, the public and the agencies themselves.

ATSDR itself recommends several standards, including the 30 ppb standard suggested for people exposed to formaldehyde regularly for up to a year, and the 300 ppb standard Wright and Little chose. The higher standard is intended as a guide for physicians treating patients exposed to formaldehyde during a chemical spill or other emergency.

The scientists bolstered their choice of 300 ppb by saying that is the standard recommended by the American Conference of Governmental Industrial Hygenists. But the group's chairman, Dr. Terry Gordon, said 300 ppb is intended to be a guide for the maximum amount of formaldehyde that a worker should ever be exposed to for any period of time.

"They should not have applied a worker ceiling limit to 24-hour, around the clock living," said Gordon, whose organization is comprised of industrial hygienists throughout the nation.

The samples Little and Wright analyzed came from two sets of unoccupied trailers: one with all the windows and vents open, and the other with the air conditioning running and the bathroom vent open. The report (PDF) found that if all the windows and vents were left open for four days, the average formaldehyde levels fall beneath 300 ppb, although they were still above 30 ppb. But in the set with the air conditioning running and the bathroom vents open, formaldehyde levels dropped beneath that threshold on only two days of the 14 day study.

Glen Nowak, a CDC spokesman, emphasized that FEMA hadn't asked ATSDR to analyze the impact of the trailers' air on public health. Despite the publicity swirling around the trailers at this point, he said he didn't think Little and Wright knew people would be moving into the trailers they were testing.

Vincent Garry, director of environmental medicine at the University of Minnesota, read Little's and Wright's report at the request of the subcommittee and told ProPublica he "found it to be naïve."

In a letter (PDF) requested by the subcommittee, Garry, who peer-reviewed ATSDR's guidelines on formaldehyde exposure, said the report did not "take into account that unlike occupational exposures this is a 24 hr per day 7 days per week exposure for children and the old who are sensitive to the chemical for different reasons."

"I think the report gave us what we were looking for"

In early 2007 a draft of the ATSDR report went to the office of Howard Frumkin, ATSDR's director, where Wright and Little said it was revised at least four times. Wright's handwritten notes obtained by the subcommittee show that Sinks, ATSDR's deputy director, made numerous comments. But nobody questioned the 300 ppb standard or the absence of any mention of formaldehyde’s link to cancer.

Preston, the FEMA lawyer, received the report on Feb. 1, 2007 with a cover letter (PDF) from Little and Wright summing up their findings: opening windows and vents would bring formaldehyde levels "below levels of concern."

The two scientists assured Preston they hadn't shown the letter to anyone else and "as requested" they did not evaluate "health concerns related to potential exposures." They also warned him that the report was "not intended to establish FEMA's future policy concerning temporary units."

Almost immediately, however, FEMA employees began using the report to tell occupants that the formaldehyde did not pose a health threat.

"Thanks, Rick, I think the report gave us what we were looking for," said an emergency management program specialist said in an e-mail (PDF) to Preston 11 days after the report was delivered. "Changing air via external venting is effective in reducing the formaldehyde levels."

A month later, a FEMA trailer maintenance coordinator, told staff in an email (PDF) that FEMA would use 300 ppb "as a guide in our housing program."

A whistleblower ignored

The report by Wright and Little showed up on Christopher De Rosa’s desk on February 27, 2007, nearly four weeks after it went to FEMA. Throughout his 17-year career with the CDC, De Rosa had received positive performance evaluations. A few months earlier, however, his bosses had criticized his work on two controversial projects, one on industrial waste in the Great Lakes and the other on a cancer-causing chemical found in some cosmetics.

Although Little and Wright normally reported to De Rosa, they told the subcommittee they sent all their Katrina work directly to Frumkin's office because they were following a new chain of command Frumkin had developed to get those reports out faster. But Frumkin told ProPublica that the scientists still should have sent the report to De Rosa. And Frumkin told Congress that De Rosa had missed several opportunities to be more involved in the study.

When De Rosa skimmed the report, he immediately called Sinks, ATSDR's deputy director.

The report downplayed the health risks, he told Sinks, because it omitted the long-term and potentially cancer-causing effects of formaldehyde exposure. He repeated his warning in an email to Sinks and Frumkin that day and attached a letter that he suggested sending to Rick Preston, the FEMA lawyer who had asked ATSDR for the report. It said that "failure to communicate this issue is possibly misleading, and a threat to public health."

De Rosa told Congress he was so alarmed by those omissions that he missed the more serious flaw in the 12-page document: that it used the wrong safety standard.

When De Rosa hadn't heard from Frumkin or Sinks a week later, he sent them another email (PDF). If he didn't hear otherwise by the end of the next day, he told them he would notify FEMA of the report's flaws himself.

A month later, De Rosa's letter made its way to Preston, the FEMA lawyer. But Preston told the subcommittee he stuck the letter in a file and never shared it with anyone. He also told Congress he had never told Wright and Little not to study the health effects of formaldehyde in the trailers.

"Everything in that letter was already known to FEMA," he told subcommittee investigators. Preston has since left FEMA and could not be reached for comment.

Not long after Preston put the warning in his file, FEMA began drafting a news release announcing the results of ATSDR's study. As FEMA staffers scrutinized the release, one of them spotted a problem.

"I guess I'm a little concerned about this paragraph. [300 ppb] is high according to information available to the public," the employee said in an email (PDF) to his colleagues, pointing to the 300 ppb standard used by ATSDR. This is “high according to information available to the public.” In a follow-up email he listed the much lower standards made public on Web sites from other agencies.

But FEMA decided to release the report anyway.

For the next several months, FEMA officials – including the agency's chief, R. David Paulison, and its press spokesman, Aaron Walker – assured the public that the trailers were safe as long as the windows were kept open.

"We have no need, and we see no need, to question the reliability and safety of the trailers," Aaron Walker told the media on May 10, 2007.

"We've been told that the formaldehyde does not present a health hazard," Paulison told Congress at a May 15 hearing (PDF).

Less than a week after the hearing, an ATSDR employee sent Frumkin an email that included a link to a FEMA press release announcing that ATSDR had said formaldehyde levels were beneath the “level of concern” if the windows were open. On May 30, Frumkin received another email from someone outside the agency that included a copy of a story quoting Paulison making a similar assurance.

But ATSDR didn’t notify FEMA that it was still using the flawed study, even though Frumkin received several emails mentioning FEMA's public statements.

“I was not aware of how FEMA was using the information in our report until the middle of ’07,” Frumkin told ProPublica. “We really don’t and can’t routinely monitor what other agencies and organizations say or do with our information.”

Barry Johnson, who ran ATSDR in the 1990s, told ProPublica his staff regularly updated him on media coverage of health studies like the one produced by Wright and Little. "We were able to track the impact of those consultations closely," he said.

Frumkin also said FEMA’s public statements didn’t come up in the weekly meetings he held with his top staff. Yet documents show references to FEMA trailers, formaldehyde, or news reports about the issue on more than a dozen meeting agendas between January and July 2007. Frumkin said those discussions focused only on an inquiry from a Mississippi congressman about formaldehyde complaints from trailer occupants in his district.

A former high-level CDC employee, who asked not to be named because the individual still does business with the agency, told ProPublica that the CDC normally keeps close tabs on the media, including Web searches for relevant news.

The former employee said Sinks and Frumkin are genuinely dedicated scientists, but that getting the CDC to admit it made a mistake would be extremely difficult, because the agency is hobbled by top-heavy, overly cautious management.

“The agency as a whole should have stepped in,” the former official said.

Unsatisfactory performance

It wasn’t until July 19, 2007 that Congress and the public learned that FEMA had been using flawed information to assure trailer occupants the trailers were safe.

Mary DeVany, an industrial hygienist, was called in as an expert witness by the House Oversight and Government Reform committee, which was holding hearings on FEMA’s response to the formaldehyde issue.

She testified that ATSDR had “arbitrarily” chosen the 300 ppb limit “and applied this high level to the results as if it were a safe and applicable limit,” said DeVany, who has spent much of the last three years testing the trailers for a lawsuit against the trailer manufacturers. She told ProPublica that her company’s employees’ constant exposure to the chemical has left some of them suffering the same ailments as the trailer residents.

After that hearing, Frumkin, Sinks and other top CDC officials began talking about doing more tests.

But De Rosa thought immediate action should be taken. He sent another e-mail (PDF), this time addressing it to CDC officials as well as to his bosses at ATSDR.

Yes, more testing was needed, he said. But what was being done in the meantime to make the trailer occupants safe?

"I am concerned that the reported clinical signs are a harbinger of an impending public health disaster," he said in the July 24, 2007 message.

Mike McGeehin, director of the CDC's division of environmental health hazards, answered De Rosa in an e-mail (PDF).

"The type of data on which to base a decision whether to uproot 66,000 families is lacking," McGeehin wrote. He asked De Rosa to discuss future concerns over the phone or in person. "Emails can be interpreted so many ways by different readers inside and outside the agency," McGeehin said.

On Aug. 10, 2007, Gerberding, the CDC director, echoed De Rosa's concerns in an e-mail (PDF) to Frumkin. "I realize that good science takes time, and good regulations can take an eternity," she said. "But in the meantime, this issue is festering, and these people are suffering."

By then De Rosa's career with the CDC was crumbling.

In October, he and Frumkin attended a conference for the Collegium Ramazzini, a body of 180 recognized experts in occupational and environmental health. It was held in Carpi, Italy. De Rosa took along his father, who is Italian and had never been to Italy.

At an awards ceremony during the conference, Frumkin walked over to De Rosa and his father and handed the toxicologist an unsatisfactory performance evaluation, the first in his 27-year government career. A memo (PDF) attached to the performance evaluation told him he was being reassigned.

When questioned about De Rosa's status, Frumkin told Congress that "the reassignment of Dr. De Rosa was not in any way a retaliation for his actions in this case. His reassignment was a result of personnel actions that are best not discussed in a public forum like this."

Several former and current CDC scientists interpreted De Rosa's reassignment as a message that CDC employees should be wary of criticizing CDC projects.

Susan Kess, a former senior medical officer in ATSDR's division of toxicology, who left the agency in 2005, said that by removing De Rosa from his position, "They are squelching scientific integrity within the agency, and that is an injustice to the public."

De Rosa is now contesting his reassignment in mediation hearings.

Two years later, risk officially revealed

In December 2007 and January 2008, two years after FEMA trailer occupants had begun complaining about burning eyes and labored breathing, the CDC conducted the federal government's first formaldehyde tests on trailers where people were actually living.

In February the CDC announced that preliminary results showed the formaldehyde levels in many of the trailers were high enough to increase the risk of cancer and respiratory illnesses. FEMA began urging people to move out.

The CDC has asked its independent Board of Scientific Counselors to investigate how the flawed formaldehyde report slipped through the agency's bureaucracy. The agency also has agreed to study how the children who lived in FEMA trailers may have been affected by their exposure to formaldehyde.

More than 100,000 of the trailers now sit, unused, at sites throughout the country.