Investigation reveals widespread double dipping in NIH program to pay off school debt

Joshua Bauml, a University of Pennsylvania (UPenn) oncologist, received welcome news last year: He had won a coveted spot in the National Institutes of Health (NIH) Loan Repayment Program (LRP). Since 1988, the program has aimed to keep promising young biomedical scientists in academic research by helping repay school loans that can run up to hundreds of thousands of dollars. Without it, supporters say, many of these researchers might have chosen lucrative slots at pharmaceutical companies or in private practice.

Bauml, who finished medical school in 2008, says he will get $75,000 in tax-free loan repayments during his first two LRP years and hopes a renewal will further pay down his debt. He also became a program "ambassador"—a model of success who counsels applicants.

Bauml has another distinction, too. A Science investigation has shown that he, like more than one-third of the 182 clinical scientist ambassadors whose years of participation could be determined, broke the program's rules against certain forms of industry funding. (NIH's own analysis finds a smaller number of violations.) The extensive rule breaking—inadvertent in Bauml's case, he says—is leading some observers to question the public purpose and fairness of a generous taxpayer subsidy for young scientists. "In this presumably exemplar program … you see widespread flouting of the rules, and no one seems to care," says Vinay Prasad, a hematologist-oncologist at Oregon Health & Science University in Portland who studies industry influence on medical research.

Since 2004, the LRP has paid nearly $1.1 billion for more than 22,000 new or renewal awards to researchers, including about $72 million in fiscal year 2018. Ericka Boone, its director for the past 4 years, said via email and in interviews that the program supports "mission-critical" NIH priorities by preventing school debt from creating "a vacuum of research innovation and discovery." Recipients, she added, have excelled in gaining funding; publishing research in basic science, clinical trials, and public health; and serving in institutional leadership roles up to 15 years after receiving federal benefits.

Other taxpayer-funded student loan forgiveness plans have a similar rationale: They free recipients from debt so that they can serve the public interest. But these programs tend to be highly competitive and forcefully policed, and often require lengthy government or public service commitments. A Department of Education program for public servants requires at least 10 years in government or nonprofit jobs and awards benefits very selectively.

The LRP offers a sharp counterpoint. Applicants compete on scientific merit, but half win acceptance. The LRP generally does not consider financial need. Although many of the academic-scientist recipients earn relatively modest salaries, taxpayers also repay school loans for physician researchers who rank in the top 1% to 2% of all U.S. wage earners.

Over its history, the program has at times allowed recipients to accept industry support—either by direct payments or research funding—in unlimited amounts. As a result, the LRP in effect subsidizes some young scientists as they build strong ties to pharma. In addition, lax policing and confusing rules about acceptable forms of corporate support mean many recipients end up reaping even more industry largesse than the rules allow.

Labyrinth of rules

Bauml accepted more than $28,000 in pharmaceutical industry money, mostly for consulting or speaking at pharma events, during his first 6 months in the program. That was within the rules: Although the program previously barred such payments from drug companies, NIH reversed the policy in 2017. During the same period, industry also underwrote $473,000 of Bauml's research. For Bauml, this was business as usual. Drug firms had previously given at least $3.2 million for his work, in a steady flow leading up to his LRP award. The practice was allowed until 2017, but the latest program rules banned it, placing him in violation.

Bauml says he was shocked to learn from Science that he had broken LRP rules. He says he disclosed everything to NIH, and heard no objection. (NIH declined to verify disclosures by any recipient.)

Prasad calls the LRP's permissive approach "ethically problematic." He and others say the program should be more vigilant about industry payments to the researchers it helps. "Something designed to keep early-career investigators in academia can instead have the function of basically being a supply pipeline to industry," says Leigh Turner, a bioethicist at the University of Minnesota in Minneapolis. The result, he says, can be long-term financial dependence, regardless of whether a researcher stays in academia.

The program is a lost opportunity, critics say. "Industry influence on young scientists is vast and pervasive," Prasad says. "The intent of the LRP should be to assist people who pursue something that is very challenging in academic medicine—building a career that is truly independent of conflicted bodies."

Although the LRP supports a variety of biomedical fields, Science's investigation focused on physicians because drug industry payments they receive must be publicly disclosed. Doctors make up more than half of recipients, and the relevant records—beginning in mid-2013—are posted on the federal Open Payments website. Disclosure statements on articles in journals, which often require authors to declare industry support, provided additional data.

Before 2011, LRP participants were free to accept all industry funding. Stephen Wiviott, a Harvard University cardiologist and an LRP awardee in the mid-2000s, wrote in an email that the LRP paid his entire $100,000 medical school debt, helping him buy a home and start a family. It "allowed me to spend more time on my research … and to build a portfolio of publications that facilitated my first job, and ultimately set me on the path for a lifetime career," he wrote. At the same time, Wiviott earned consulting, advisory, and speaking fees from at least 12 companies, and worked on research underwritten by at least 24 firms, according to published disclosures, which did not specify amounts received.

Later, the program imposed a changing array of restrictions, whose evolution can be seen on archival versions of the program website. From 2011 to 2014 it continued to permit any research funding from for-profit sources. In 2014, the Code of Federal Regulations barred such funding for LRP recipients. But the program continued to allow corporate research support, capping it at less than half of a recipient's total, until it finally adopted the required ban in 2017. Personal income from for-profit sources followed a reverse pattern: From 2011 until 2017, it was strictly off limits, but since then, any amount has been allowed.

Boone declined to explain the policy shifts. But she said allowing awardees to accept industry money is not in conflict with the program's goal of keeping scientists in academic research. Current LRP restrictions on industry funding, Boone added, simply follow federal requirements.

She gave shifting accounts of past rules, however, insisting at first that the LRP has never permitted industry support for research or personal income. In later written comments, Boone acknowledged that assertion was wrong. But she continued to contradict the program's rules as published on its website. She said the LRP did not restrict moonlighting income from 2011 to 2017. Yet published rules for 2011 to 2017 barred income of any kind from for-profit sources or private clinical practice, and stated that income from nonprofits other than the recipient's regular employer could not be "significant."

For this pattern [of violations] to emerge, many people have to be ignorant of or indifferent to the rules.

Boone also seemed uncertain about the purpose of the program. She initially told Science that the LRP looks for scientists "who really had much more financial-based need—where it would really be important for them to receive this assistance or they would probably end up … getting out of research."

Yet, other than a track for individuals from disadvantaged backgrounds—less than 2% of awards since 2004—the LRP uses no means test other than total debt of at least 20% of a recipient's base salary when accepted to the program. Salaries for many recipients at public universities exceed $200,000, and elite private institutions, which employ many LRP recipients, often pay far more. In later, written comments, Boone conceded that payments from the program "are not need-based awards."

The rule changes may explain some of the widespread violations apparent in records from mid-2013 through 2018. Among 182 LRP ambassadors, Science found nine violations of program rules involving research funds—from about $7000 to $473,000—and more than 75 involving consulting fees or other personal income, including a dozen from about $12,000 to $59,000. Academic papers by some LRP recipients also disclosed improper payments in unspecified amounts prior to 2013.

Boone would not comment on specific findings, but in an interview she told Science that violations are rare. The program relies on self-reporting by recipients, and regular validations from their research mentors and employers. "No system is foolproof. But we have a pretty good system," she said. In recent years, violations have been found only once or twice, and were due to circumstances beyond recipients' control, she said, such as a nonprofit hospital employer being taken over by a for-profit chain. Boone said the LRP has found no violators since 2017, despite the numerous cases Science identified.

Later, Boone reversed herself, telling Science that by her reckoning—which relied on a chronology of rules that is at odds with those the LRP published at the time—34 of the roughly 400 LRP ambassadors who could be tracked on Open Payments were in violation since 2014. When Science applied Boone's chronology to the sample of 182 ambassadors, the number of violators and the amounts accepted improperly for research actually jumped sharply. (After Science asked about apparent violations by some publicly posted ambassadors and "success stories," NIH protected their web pages behind passwords.)

Genevieve Kanter, an economist and ethicist at UPenn's Perelman School of Medicine, whom Science asked to consider its findings and LRP's policies, says she isn't surprised that both administrators and awardees are confused when rules are "chaotic" and seem to lack "underlying principles."

"For this pattern [of violations] to emerge, many people have to be ignorant of or indifferent to the rules," Prasad adds. "Not only the beneficiaries. It's also the … people who are signing off on these agreements with NIH, and NIH officers who may even be aware of some of this and turn a blind eye," says Prasad, who benefited from a small, parallel LRP as an NIH employee.

Adriane Fugh-Berman of Georgetown University in Washington, D.C., who studies how the pharmaceutical industry influences medicine, says that by permitting industry funding and failing to enforce limited restrictions consistently, the LRP effectively subsidizes industry research. "When the biomedical industry controls the research agenda, it restricts the questions being asked to those with answers that are potentially profitable … excluding research on older therapies, on generic therapies, on nonpharmacologic therapies, on social influences on illness," Fugh-Berman says.

Bauml and LRP ambassador Scott Tagawa at Weill Cornell Medicine in New York City argue that industry money is crucial to certain lines of medical research, and that the LRP should actually relax its current restrictions on industry support for recipients' work. Tagawa, for example, studies cancer drugs for which industry may be the only research funder. Often, a company provides funds for a clinical study or donates a drug so study subjects don't have to pay out of pocket, Tagawa says. Keeping such support off limits to LRP recipients "doesn't make a lot of sense to me," he says.

Tagawa had robust ties to industry while he was an LRP recipient, from 2009 through fiscal year 2013. In 2011 and 2012, Sanofi U.S., Amgen, and Janssen paid him consulting or speaking fees in undisclosed amounts, and in 2013 he made more than $36,000, mostly for consulting and speeches, according to published disclosures and Open Payments. Companies "made the presentation slides" for his talks, he says, calling it "basically the same thing as being a sales rep." LRP rules at that time barred such income—something Tagawa says he didn't realize, noting that he fully disclosed his earnings. He says he stopped giving the speeches after Cornell proscribed them, but doesn't regret taking payments for those "educational" talks.

LRP alumni often retain close ties to industry later in their careers. Science's investigation identified substantial industry support for dozens of LRP ambassadors, including Tagawa, after their loans were repaid—in amounts per person of up to $29 million for research and $159,000 in personal income—according to Open Payments data from mid-2013 through 2018.

Yet some LRP recipients have built robust research careers without turning to industry. Another LRP ambassador and current beneficiary, Adrienne Boire, is a biochemist and neuro-oncologist at Memorial Sloan Kettering Cancer Center in New York City who studies brain cancer cells within spinal fluid. Throughout 14 years of postgraduate training, during which she gave birth to her daughter, Boire amassed debts of $250,000. "The loan repayments made all the difference," she says.

Boire, whose current clinical trial involves an off-patent drug, says she avoided industry research funding because "I really like … the intellectual freedom you have when you're in a pure academic system."

This story was supported by the Science Fund for Investigative Reporting.