Journal to scrutinize hip fracture study
Inquiry follows allegation of ethical breach
A leading medical journal is launching an investigation into the work of a research team led by a Harvard doctor, after federal health regulators accused the scientists of failing to inform elderly nursing home residents of serious health risks discovered during a hip fracture study.
Editors at the Journal of the American Medical Association, which published a 2007 article by the scientists that included research from that hip fracture study, will be reviewing the scientists’ work and supporting documents, journal spokeswoman Jann Ingmire said yesterday.
“They want to talk to the study authors,’’ Ingmire said.
At the same time, the Harvard-affiliated Institute for Aging Research at Hebrew SeniorLife in Boston, where Harvard Medical School gerontologist Dr. Douglas P. Kiel presided over the hip study, said it would also investigate.
“We will conduct an objective, independent review of the [federal regulators’] findings because this is the best way to assure our overriding commitment to patient safety and scientific integrity,’’ Len Fishman, Hebrew SeniorLife’s chief executive, said in a prepared statement.
“We have the highest respect for the scientists at the Institute for Aging Research, and deeply support the institute’s ongoing mission to conduct studies that improve the standard of care for seniors,’’ Fishman said.
Kiel’s team tested whether padded underwear could protect the hips of nursing home patients from shattering during falls. Typically, the underwear is padded on both hips, but in their study, researchers assessed garments padded on just one side.
After they started enrolling patients in 2002, data from the study indicated that seniors were more often having serious falls on the padded side than the one that was unpadded.
But the scientists failed to disclose that risk to patients; to the National Institutes of Health, which provided $8.6 million funding for the study; and to other officials overseeing the research, according to findings issued June 23 by the Department of Health and Human Services. An HHS spokeswoman declined to comment, citing her agency’s ongoing investigation.
A Harvard Medical School spokesman referred questions to Hebrew SeniorLife.
A Canton businessman who alleged the scientists were concealing information from officials said his complaint to federal regulators sparked their inquiry.
Ed Goodwin, president of HipSavers Inc., a maker of padded hip protectors, sued Kiel in 2008. Goodwin’s lawsuit accuses Kiel of unfairly disparaging Goodwin’s product, because Kiel’s team concluded in its 2007 study that they could find no evidence that padded undergarments protected patients from fractures. The team did not study products made by Goodwin.
As Goodwin’s lawyer collected evidence, he uncovered many private e-mails the researchers had exchanged that showed they knew of problems in the study and that exposed their decision not to disclose problems to patients or officials overseeing the research.
One of Kiel’s coauthors, Dr. Stanley Birge, an associate professor of medicine at Washington University School of Medicine in St. Louis, said in an interview yesterday that he and fellow researchers took every precaution in designing and implementing the study.
“When we looked at the data carefully, we just weren’t convinced that it was due to the pad,’’ he said of the disproportionate number of falls on the hip that was protected. “These hip pads weigh less than the average man’s wallet - just several ounces. So we think it’s very unlikely that they’re going to cause some kind of imbalance.’’
But Michael Carome, a deputy director at the Washington-based consumer watchdog group Public Citizen, blasted such explanations as a “post-hoc attempt to explain away the findings’’ and said the researchers’ conduct was a clear breach of ethics. “It’s a problem which they basically tried to cover up.’’
Hip pads rose in popularity more than a decade ago when small studies in nursing homes showed they prevented hip fractures. But the pads dropped out of favor as larger studies produced murkier results.
Kiel’s team, which enrolled more than 2,000 nursing home patients in Boston, St. Louis, and Baltimore between 2002 and 2006, aimed to provide clearer evidence.
In an e-mailed response to questions from the Globe yesterday, Kiel explained researchers’ rationale for conducting their unusual study. Kiel wrote that more traditional studies that included patients wearing two-sided hip protectors, comparing them with patients with no protection, may have been biased. In those studies, he asserted, nursing home staff may have paid more attention to patients wearing the protectors, aware of their risk of falling.
Kiel said his team chose the one-sided garments so that the results would be more pure, with findings attributable to the protector, rather than differences between residents who did or did not use them. And his team’s design, he said, reduced the prospects that staff behavior could influence the outcome.
The 2007 study published in the Journal of the American Medical Association tracked 1,042 patients enrolled by Kiel’s team before October 2004. The results were striking: There were twice as many fractures on the side that was padded compared with the side that was not. The actual number of fractures overall was small, with 13 among protected hips and seven on the unprotected side.
When journal editors were reviewing the researchers’ manuscript, they asked the scientists if they had observed whether the unusual single-padded undergarment may have thrown off patients’ gait, causing falls.
Federal regulators noted that the researchers told the editors “we have no information regarding whether wearing one hip protector influences gait or falling.’’
Yet the regulators noted in their report that the researchers, in fact, had information from a smaller, similar study one of the scientists conducted in 2002 to gain federal financing for the larger study. That earlier study showed that patients were more than twice as likely to fall on the side that had padding.
“It does not appear that [journal editors] were told about the data on this issue generated by the [earlier] study,’’ the regulators concluded.
Carolyn Y. Johnson of the Globe staff contributed to this report.
Kay Lazar can be reached at email@example.com.