Dr. Susan Harkema became the face of one of the University of Louisville’s splashiest research successes the moment one of her paralyzed patients wiggled his toe.
Her name was in Time Magazine. She was interviewed on “Good Morning America” and CNN. The notoriety brought more funding and patients to U of L with hopes that revolutionary studies would help the paralyzed walk again.
But in March, a federal agency took the unusual and drastic move of withdrawing its funding from one of her studies, citing concerns about the validity of the data and unresolved problems with oversight. Meanwhile, the federal Office for Human Research Protections is also conducting its own review, a spokeswoman confirmed.
It’s the latest in a series of issues that have brought outside scrutiny on the school’s medical program, and it could have lasting ramifications on future research.
“This is pretty egregious,” said Milton Corn, deputy director for research and education at the National Library of Medicine.
Corn said he can’t comment on U of L’s research, and he’s not connected to the study’s funding agency. But discontinuing any study is “draconian,” he said.
“It’s much more common to try to work with the investigators and help them repair something that’s flawed and probably can be fixed,” he said. “I think to say we want to discontinue this implies they really were very upset.”
Harkema is a superstar in the world of paralysis research, and her clout has certainly
brought more private money to U of L’s research funding. But ethicists and research watchdogs, when contacted by KyCIR, said the suspension of federal funding raises big questions.
A U of L review board found numerous problems with the study, which measured the effect of a muscle relaxer on paralyzed patients on treadmills. Still, the board concluded that Harkema’s study should continue unless the federal agency shut it down.
Among U of L’s own findings:
- Investigators found evidence of 151 possible “adverse events,” but the researchers didn’t report a single one, including a study participant who reported severe right knee pain that was later deemed a fracture.
- “Inconsistencies” led some patients to pay for therapy when insurance stopped covering part of the treatment.
- Some participants signed the wrong consent forms.
- Research records had no follow-up data or information about patients’ medication compliance.
In discontinuing the $914,000 study it funded, the National Institute on Disability, Independent Living and Rehabilitation Research cited the school’s audit and “numerous instances of non-compliance and serious non-compliance” of protocol.
Elizabeth Woeckner, president of Citizens for Responsible Care and Research, a nonprofit watchdog group, said she’s also rarely seen a federal agency pull funding.
“In my experience, it’s a cold day in hell,” said Woeckner.
Susan Craddock, a University of Minnesota professor who specializes in bioethics, reviewed some documents tied to the study at KyCIR’s request. She said flaws in the study began with its design: U of L doctors recruited their own patients, who are already especially vulnerable because they have the strongest motivation to agree with their doctors.
They want to walk again.
In a June interview with KyCIR, Harkema said patients were not put at risk. She downplayed the impact of the federal action, saying she voluntarily put the study on hold and has since addressed record-keeping issues.
But Harkema said concerns about her study were overblown. The scrutiny only arose because a few disgruntled ex-colleagues blitzed regulators with exaggerated accusations, she said. Two doctors involved with the study sent complaints to federal and U of L regulators.
“I think that these are very disgruntled, unhappy individuals, who are making some incredibly bad choices out of their own selfishness,” Harkema said.
Harkema disagreed that her data was compromised. She said follow-up records were disorganized, not nonexistent.
“Every allegation related to patient safety or risk was unfounded,” Harkema said.
William Pierce, the executive vice president for research and innovation at U of L, declined to comment for this story. Laura Clark, chair of the U of L Institutional Review Board that audited Harkema’s study, also declined to comment.
The U of L School of Medicine has endured major scrutiny in recent years. Problems with the med school curriculum and too-small classrooms jeopardized its accreditation in 2014, although the issues have since been corrected. Last year, a whistleblower filed a federal lawsuit alleging problems with safety procedures in U of L’s germ lab.
David Dunn, U of L’s executive vice president for health affairs, has been on paid leave since December while he is under investigation by the FBI for possible misuse of federal funds.
University’s Fortunes Tied To Research
As U of L aims to become a premier research university, research endeavors like Harkema’s are among its best-known — and most well-funded. And the university’s research success is closely tied to the success of major medical studies.
Harkema, a U of L professor of neurological surgery and director of research at the Frazier Rehab Institute, has earned the most accolades for her research into epidural stimulation implants. When stimulated, patients who were otherwise paralyzed could move and stand unassisted, and have seen improved bowel and sexual functions. Harkema also played a big role in bringing treadmill therapy known as locomotor training into the mainstream.
The late Christopher Reeve participated in some of her early spinal cord injury studies at UCLA, and The Christopher and Dana Reeve Foundation has financially supported some of her research. Harkema directs its NeuroRecovery Network, a consortium of clinical and wellness facilities with programs for spinal cord injured patients. Frazier Rehab is among 11 locations in the consortium.
Last year, Helmsley Charitable Trust awarded the university foundation a $15 million grant for a spinal cord research initiative.
Harkema, who is paid about $285,000 a year at U of L, said all the private foundations that have partnered with her over the years remain her partners.
Concerns Over Research, Ethics
Harkema’s tainted study began recruiting patients in 2012. Its goal was to measure outcomes of patients who perform locomotor training, which uses harnesses to help paralyzed patients stand and step on a treadmill, against those who also took a muscle relaxer called Baclofen.
The study sought to determine if patients using the muscle relaxer saw significantly different outcomes than patients who didn’t use the drug during 80 sessions of treadmill training over four months.
Steve Williams, a physician who worked with Harkema on the study, said in an interview that he was concerned about Harkema’s ethics and research. Last fall, he sent letters to U of L’s Institutional Review Board and to the National Institute on Disability, Independent Living and Rehabilitation Research about four of Harkema’s studies.
Williams, who has since left U of L for a new job at the University of Washington, said he believes U of L officials weren’t willing to admit Harkema’s studies were sloppy, and that “she was taking advantage of a vulnerable population.”
“The research enterprise did not do their job protecting human subjects, and they let their concerns about other research funding … bias what they knew they should do, to do the right thing to protect people,” Williams said.
Only the Baclofen study used federal funds; Harkema said she voluntarily stopped enrolling patients in all four of her studies after Williams and another researcher complained. Enrollment in two of them has since resumed after Harkema said the U of L auditors and private funders found no reason to stop them.
In its letter to U of L, National Institute on Disability, Independent Living and Rehabilitation Research director John Tschida cited problems outlined in the U of L review. They raised “critical questions about the investigators’ ability to reliably measure one of the study’s two primary independent variables [use of Baclofen].”
Pay To Stay
Among his allegations, Williams said some patients were hoodwinked into paying to stay in the research study after their insurance stopped covering training sessions.
Harkema she wasn’t aware of any patients being asked to pay to be in the study, although she claims they knew it was possible they’d have to pay for therapy.
Harkema contended that this allegation, like other problems identified in the audit, stem from a misunderstanding of her research. She said the study was really about the medication, and the treadmill training is more like a prerequisite than part of the study.
“It never came to my attention that anyone was confused,” Harkema said.
Documents from the study, however, show that the research focused on both the training and medication use.
The consent form, obtained by KyCIR, makes no mention of costs for training. A section that does address costs only mentions expenses related to study injuries. “Otherwise, there will be no additional cost to you,” the form reads. (Read the consent form)
Harkema said her staffers explained the potential costs verbally and in a separate handout.
Woeckner, the research watchdog, reviewed the consent form and said it would make no sense to expect a patient to read it and be prepared to pay.
“I would have no idea I was being told I might be charged,” Woeckner said.
Patients reasonably expect a research study to be free or even compensated, because it nearly always is, said Corn of the National Library of Medicine.
“When there are charges, it is really stated and expressed very boldly up front, but most commonly, the study is without expense to the patient,” Corn said.
For now, Harkema is awaiting clearance from the U.S. Office For Human Research Protections. If she gets it, Harkema said she’ll resume the study, albeit without any federal money.
Craddock, the interim director of the Center for Bioethics at University of Minnesota, is wary. She said it’s a waste of a participants’ time to resume a study so deeply flawed.
“Resolving issues in the midst of a study — the kind of issues that should not have happened from the beginning — is odd, in my opinion,” Craddock said.
Kate Howard can be reached at firstname.lastname@example.org and (502) 814.6546.