Free Website Hosting

Friday, July 3, 2009

Life After Fraud

You put your name into Google, and the first entry is about a transgression from 20 years ago, the penalty for which only lasted three years. Now you can't get a job.

Each year, the U.S. Office of Research Integrity (ORI) investigates dozens of charges of scientific misconduct. And each year, the ORI adds a handful of names to a list of researchers found guilty of falsifying figures, fabricating data, or committing other academic infractions. As of April 1, 2009, this Administrative Actions list, presented on the ORI Web site, carried 38 names. These people are barred from receiving federal funds and/or serving on a Public Health Service committee, typically for a period of 3–5 years. Once the debarment term is up, the name disappears from the list. In theory, the punishment—and the shame—of the ordeal is over.

However, any time the ORI makes a formal ruling of misconduct, that information ends up on the Internet. The ORI's newsletter and annual reports, which used to be hard copies sent to federally funded schools, are now all electronic. The NIH Guide, a weekly report that lists findings of misconduct to help grant reviewers flag scientists who apply for federal funds before their exclusion period is over, is online, too. And the Federal Register, the official publication of every federal agency, is available as a daily email digest. So Google anyone's name who has ever been penalized by the ORI, and even if their debarment was lifted more than a decade ago, even if they signed a document stating they accepted the ORI's decision well before the Internet became such a staple of daily life, the description of the finding against them—and the penalty they received—will pop up. In some cases, it's the first article that appears.

Of the hundreds who faced federal reprimands in the last 20 years, three agreed to speak to The Scientist about their experience. All spoke on the condition that the story would not include their real names. Doing so, of course, would only create one more unpleasant entry in what comes up when you enter their names into an online search engine.

IT WAS THE DEADLINE OF THE MOST IMPORTANT DECISION MOLECULAR biologist Daniel Page had ever had to make in his career, and he didn't know what to do. In his hands were papers listing a series of charges of scientific misconduct against him—penalties his institution, Ohio State University, was asking him to accept. The hardest one: admitting guilt.

The charges were relatively minor. It all came down to company data Page had included in a grant application that he says he believed he had permission to use, but the company says he didn't. A couple of oversights on Page's part led to accusations of plagiarism, falsification of qualifications, and breach of confidentiality. He didn't mind having to take a class on misconduct, withdrawing his application for early tenure, and writing letters of apology to some of the people involved. But even after several months of an emotional investigation, during which he hadn't been sleeping well and his relationship with his fiancée had been deteriorating, he just wasn't sure he was willing to admit guilt to something he says wasn't true.

"So I came into work on the deadline day still not knowing what to do," Page recalls. "I turned on my computer, and a plane hit the World Trade Center. And then another plane hit the World Trade Center. And then another plane hits the Pentagon." It was the morning of September 11, 2001. When he realized the enormity of what was happening, his concerns about signing the paperwork melted away. "I realized there were a lot bigger things in the world." He signed a declaration of guilt, and handed it over to OSU officials that day. The school passed on its findings to the ORI. Eventually, the ORI added Page to its Administrative Actions list.

In 2007, the ORI received 222 allegations of misconduct, opened 14 new cases of misconduct, and closed another 28. Ten of those closed cases resulted in findings of misconduct and/or administrative actions.

Once he signed the document, Page told his department what had happened, met one-on-one with 14 faculty members with whom he had close research collaborations, and explained the situation to his graduate students. "I'd like it to go away, but I haven't tried to hide from it," he says. For the most part, people accepted his side of the story. "The people that work close to me, most of them patted me on the back and said 'we trust you.'" Others were less sympathetic. One faculty member, who works across the hall from Page, still believes he is a liar and a cheat, Page says. Page never shared his story with anyone in his family, out of embarrassment. To this day, he doesn't know if they know about it.

The story began while Page was researching a steroid hormone that appeared to improve the immune response to viral and bacterial infections. A startup company, which we'll call Vaxeen (the real company did not respond to requests for comment), approached him to see if he could provide a wet research lab to do pharmaceutical work in animals, to test the hormone's ability to boost vaccine efficacy. Page agreed, and signed a contract that guaranteed him $280,000 over 2 years in funding, with the promise of more. Page decided he wanted to do a broader project, so he began writing an NIH grant with a Vaxeen scientist (whom he declines to name). This scientist provided Page with some preliminary data from the company, which Page added to the application without attribution, since the Vaxeen scientist would be represented in the list of coauthors.

Many allegations to the ORI involve "honest differences in interpretations or judgments of data," that the agency does not consider misconduct. Similarly, the ORI does not investigate authorship or credit disputes between former collaborators, even if the complainants describe them as plagiarism.

Just days before Page planned to submit the grant, however, the Vaxeen scientist told him that the company did not want his name on the application. Page took him off the author list, but forgot to remove the company's results from the preliminary data section. "Absolutely, if I had thought of it, I would have put [that data] in the background of the grant, and attributed it to the company," he says. "Rarely a week goes by where I don't think about this."

Over time, Page pieced together what happened next. A reviewer of his grant who had also worked with Vaxeen (and declined to be named in this story) recognized the data, recused himself from the review, then likely contacted the company (the reviewer can't recall if he contacted Vaxeen or not). However the company found out, it then asked Page for a copy of his written permission to use the data. Page said he thought he had received permission from his collaborator, without needing a written agreement. "Absolutely, I was caught," he says. OSU investigated—going through his files, computer hard drive, and all communications with the company—and concluded that Page "had committed scientific misconduct under federal and university guidelines," according to a university statement issued at the time. (An OSU spokesperson declined to comment further on the case.) Page thought about leaving OSU, but his colleagues—who largely supported him and believed he didn't intend to mislead—convinced him to stay. "If it wasn't for them, I would have quit," he says. "I'd be teaching at a college."

One year after Page signed the paperwork on that fateful day, he received tenure. Since that time, he has published 35 papers and accumulated a lifetime total of more than 1,200 citations. He now receives more requests to review papers and grants than he did before the incident occurred. For him, the hardest part is meeting other researchers and asking himself: do they know? "When they hear my name, do they go, 'Oh, I know you already.' Do they have a preconceived notion about me? When I interact with people at study sections or when I interact with people at scientific meetings, do they already know who I am before I know them?"

Gerry Levick first realized the gravity of a decision
he'd made 4 years earlier while being cross-examined in a 1998 trial. As a researcher in human performance and consciousness at Touro College in New York, Levick occasionally testifies as a forensic consultant in court cases—in this one, he spoke about whether he believed a driver was paying attention when a car accident occurred.

While being cross-examined, the lawyer asked him if he had ever been found guilty of scientific misconduct. He said no. Had he ever been censured by a professional organization? No. Had he ever been found guilty of misconduct by a professional organization? No again. Then came the most direct question: Had he ever been convicted of professional misconduct by the National Institutes of Health? "And then it hit me. I said, 'Oh my God.'" The attorney immediately presented official documents showing that, in 1994, Levick had agreed to be penalized for misrepresenting his qualifications and expertise on a grant application. Levick had signed a document, but had never admitted guilt, and he believed that the penalty period had expired in 1997.

The judge cleared the jury from the room, and told Levick he could not continue his testimony. Levick stepped down from the witness stand, embarrassed and perplexed. How had the lawyer found out about something that ended a year ago? "I couldn't figure out what had happened, where this came from. I didn't realize it was in the public domain."

The charges stemmed from the wording Levick used on an NIH training grant application he submitted in 1988, when he was 39. He no longer has the original application, and sometimes struggles to remember exactly what he wrote. First, the agency alleges he claimed he had an MD degree from the University of Manchester—Levick admits that he wrote that on his application, but his real degree was an MBChB, a Bachelor of Medicine and Bachelor of Surgery that, in the United Kingdom, represents a combined undergraduate and graduate degree that serves as the initial step students take who want to become doctors. (Levick eventually obtained an MD from a university in Sri Lanka.) He says he wanted to simplify the process since this type of degree doesn't exist in the United States. Second, Levick said he was based at Harvard Medical School (HMS), when his real affiliation was, according to Levick, "the Child Study Unit" at Children's Hospital Boston, a teaching hospital of HMS. His funding, he says, came from the Research Foundation of Harvard University. When describing his role at Harvard, "I think I said that I was, uh, associated. I think the word was associated, or a fellow. I actually don't remember." (A Harvard Medical School spokesperson confessed that there are many groups associated with Harvard, but he had never heard of the Research Foundation.) The final charge is that he falsely claimed to have 13 patents—Levick says he wrote 13 patents "and technologies," representing new tools modeled on older inventions.

"I'm angry as hell. And there's nothing whatsoever I can do about it."

"They were looking to see if the t's were crossed, the i's were dotted. And admittedly sometimes they weren't," he says. "Maybe I wasn't careful enough, maybe I was. But the sum and substance of this stuff has no merit."

"They" in his story represents investigators at the New York Chiropractic College (NYCC), where Levick was affiliated between 1986 and 1990, then based in Long Island. As a member of the board of trustees, Levick says he was privy to heated discussions about the college's decision to move its central campus upstate (a move that ultimately took place), and he suspects that board members decided to investigate him as a way to muscle him out. The college found some discrepancies on the grant application (which was never funded) and passed its conclusions on to the ORI. Given how long ago the events took place, an NYCC spokesperson could only confirm the dates of Levick's appointment, and had no details about the investigation.

In 1994, Levick received a letter from the ORI saying he was being accused of misconduct. It rattled him. "I'm a really strong character, but I was probably nonfunctional for a couple of days. All the blood drained out of the upper portion of my body. I felt pretty helpless." He consulted an attorney, who said that fighting the case would cost $150,000. Levick was in the midst of a divorce, already $120,000 in debt to attorneys. He contacted the ORI. "They said 'well, we can make you an offer.'" If he signed a voluntary exclusion agreement, he would forego federal funds for 3 years, ending in 1997. "And I said, 'and that's the end of that?' They said 'yeah.' I did ask them whether this would appear anyplace. And they said 'no.' And I said 'okay.'"

On average, an ORI investigation takes 19 months overall to conclude. Around one-third of researchers investigated for misconduct by the ORI are eventually found guilty. Of the 10 findings of misconduct in 2007—all of which involved falsification or fabrication of data —seven scientists were barred from receiving federal funds. Two scientists were barred for 5 years; one was barred for life.

Fifteen years later, sitting in his tiny, windowless Long Island office at Touro College on one of the first sunny spring days of the year, surrounded by richly colored paintings and drawings that cover every wall, Levick catalogues the impact that decision has had on his life.

He estimates that he has applied to hundreds of institutions, none of which hired him. He says has published more than 300 academic papers, and comes with millions in funding—student tuition (he works with 13 PhD students), a $650,000 yearly contract for 12 years from the F. R. Carrick Institute, as well as grants over the years from state and federal agencies. Just this past winter, he received an offer of full professor from a university in Israel (he declined to name which one), along with money for travel and to build a lab. "I told them this story, because I wanted there to be no chance of there being a problem. They said 'we're going to investigate,' etcetera. And they did." The faculty senate cleared him, but the president of the university reneged on the offer. "I felt like shit warmed over," Levick says. "It was really painful."

Once he realized that his misconduct was on the Web for all to see, he wrote to then-head of the NIH, Elias Zerhouni, asking if he could take down the information, considering that he was "unemployable as a result." One month later, Levick received an email from a representative from the NIH's Office of the Director. It simply said:

You recently contacted Dr. Zerhouni via e-mail concerning the voluntary exclusion you signed ten years ago. Your concern relates to access to this information on the web. I assume that you refer to the citation in the [date omitted] NIH Guide for Grants and Contracts. If so, please understand that this is a publication and therefore is not subject to redacting. In addition, it is clear from the announcement that the exclusion was for a term of three years and is no longer in effect.

ORI's Response

John Dahlberg, director of the division of investigative oversight at the ORI, did not speak about any of these particular cases, but acknowledges that having an official record of misconduct is a "heavy burden." However, the ORI has no control over the NIH Guide or the Federal Register, and can't take down its annual reports when they contain notices that have expired, he says. The agency does remove the names from the Administrative Actions listing once the penalty period is up, Dahlberg notes, because that information typically comes up very quickly in an online search. Still, he adds, there is a public benefit to making misconduct findings public, and easily retrievable. For every case of misconduct the ORI catches, there are many more the agency misses, Dahlberg says, and announcing every guilty finding sends a message. "It's creating a deterrent effect. When we publicize findings of misconduct, it makes people more aware of the consequences of their actions." But to the scientists at the heart of misconduct findings who want to continue their careers, the public benefit offers little consolation, he adds. "There's collateral damage. I regret it."

If you type his full name into Google, the first article that appears is the notice of misconduct. (In part because of that, Levick now publishes mostly under "Gerry," not "Gerald.") The night he and I met in his office, Levick flew to Israel to deliver a lecture and discuss more details about another offer he just received from a "major university" there. His contract at Touro ends June 30, and he has nothing lined up. Again, that offer from the Israeli university fell through at the 11th hour—the result, Levick believes, of the public information about his voluntary exclusion agreement, which ended 12 years ago. "I'm angry as hell," he says. "And there's nothing whatsoever I can do about it."

After his wife died 3½ years ago, John Franklin, 67, started dating again. On a first date, a woman asked him about something she'd found on Google. As of this spring, if you put his (real) name into the search engine, the eighth entry is a 1990 news article about concerns that a blood test to detect cancer that Franklin developed didn't perform as he claimed. Six entries below that is the NIH Guide's entry about Franklin. It stated that he fudged data in a grant application to show that the test was more accurate, sensitive, and specific than it was. He was barred from federal funding for 3 years.

Franklin explained to the woman that he, in fact, was the person in those articles. Had he known that the details would be so permanently fixed on the Internet, however, he says he never would have signed the document accepting the ORI's ruling of misconduct. But it was the mid-1990s, before the Web became such a fixture itself.

Franklin doesn't dwell on the details of his case or replay the decisions he made. Much of it he can no longer remember, though he does recall contacting the ORI once he realized his case was so prominent and permanent. "I called up the NIH and said it's supposed to be taken down [after three years, after the penalty was up]," he says. "If they say [the record] is going to be expunged in three years, it should be," Franklin says.

The problems began for him while he was an associate professor at Harvard Medical School, working on a technology to diagnose cancer from blood plasma. He found that nuclear magnetic resonance (NMR) scans of blood lipids appeared to spot tumors before X-rays, and months or years before people showed clinical signs. Franklin published his findings in the New England Journal of Medicine; however, soon after, the journal published research by independent groups that were unable to confirm his results.

"I probably wouldn't have signed [the exclusion agreement] had I known it would come up on the internet 10 years later."

Franklin says that elevated levels of blood fat or improper handling or preparation of samples can influence NMR scans and lead to false results. Franklin was consistently able to make the diagnostic work, and he denies ever fudging any data along the way to improve its performance. However, a company that licensed the technology based on Franklin's initial data failed to show in its own research that the technique worked.

In the midst of all that, one Sunday in October 1993, on his way to church, Franklin began feeling chest pains. He went to the doctor and learned that his cholesterol had spiked as a result of stress, and caused an almost total occlusion of a coronary artery. "It was while I was in the hospital that the clinical trial results [from the company] came in. And they were terrible."

After the negative results, the company had to fold. It sued Franklin and Beth Israel Hospital, where Franklin was working; a sheriff came to Franklin's house on a weekend to serve him papers; and the school launched an investigation, consisting of two informal 1-hour meetings with a committee of administrators and professors.

Between 1992 and 2001, an average of 1 in 70 institutions reported investigating misconduct each year, according to a 2004 ORI report. Of these institutions, 55% reported only one suspected misconduct, 29% reported between two and five cases of suspected misconduct, and the rest reported between six and 20 cases.

Before HMS could make a ruling, Franklin left voluntarily. He had lost his funding from the company when it went under, and "I figured, with a heart attack and everything, I needed a change of life." HMS's official statement noted that the school investigated the allegations but Franklin resigned "prior to the completion of the institutional proceedings." The NEJM never corrected, retracted, nor issued an expression of concern about Franklin's paper.

Eventually, Franklin received a letter from the ORI saying it was conducting its own hearing about his NMR data, inviting him to attend and defend himself. Franklin says he could not bring the data with him, though, as Beth Israel was holding onto it due to the ongoing lawsuit with the company, so he didn't attend. Then, when he received a letter about the agency's concerns related to a grant application about NMR, "I wasn't surprised." Even though he says he did nothing wrong, he signed the letter, essentially accepting the agency's ruling and penalty. "I did believe then that the public record would be expunged in three years," he says. "I probably wouldn't have signed if I had known it would come up on the Internet 10 years later."

In the mid-1990s, at the age of 54, Franklin set up a new lab in his kitchen. With a hotplate, a stirrer, and funds from his retirement account, he began experimenting with new ways of delivering drugs topically through the skin. His first target was arginine, an amino acid that the body converts to nitric oxide, which boosts blood flow. (His wife always complained of cold hands.) Franklin found that if he added salt to a charged compound like arginine, the salt helped push the compound into the tissue. Tweaking the pH and changing the ratio of water and fat in the cream made things even easier. A pilot study published in Diabetes Care showed that the cream raised foot temperatures in 13 diabetics (who are prone to circulation problems) by several degrees, and improved blood flow.

Misconduct cases appeared to cluster in the upper echelon of institutions. Between 1992 and 2001, the top 75 institutions (ranked by NIH funding out of a pool of more than 2600) represented 29% of the institutions that investigated misconduct. Only one of the top 50 institutions did not report any possible cases of misconduct during that 10-year period.

Sitting at the head of a shiny cherry-colored table in his new office this past spring, Franklin crosses his legs and describes his life since the ORI's ruling. It's the day of the Boston Marathon, so the streets of Cambridge are unusually quiet outside the window of the small office park. Behind him hang four framed patents, all related to the new transdermal technology. His new company, which he does not want named, now makes up to $1 million each year in mail order sales of the "warming cream" and other topical deliveries of arginine. He has a list of 100 drugs he'd like to develop, funding from "rich private investors" (whom he won't disclose), is in talks with four companies to license the technology, and is at work on an investigational new drug application for topical ibuprofen to treat knee pain, which he'll submit to the FDA this summer. "I don't think I'll ever retire for real," he says.

As for the woman he confessed his story to on their first date, she eventually married him.

Have a comment? Email us at mail@the-scientist.com

No comments: