As a first-year medical student at the University of California, San Diego, Jessica Tang already has an impressive CV. Her name has appeared on ten papers in the medical literature, including three in the Journal of Neurosurgery: Spine. On one of these she was the sole author.
Except that one doesn’t exist anymore. But the reason for the retraction does not appear to involve shoddy work by the researcher. Rather, Tang failed to appreciate the politics of the lab in which she worked — and it cost her.
Tang was working at the Taylor Collaboration, an orthopedics program at the University of California, San Francisco led by Jeremi Leasure:
The focus of The Taylor Collaboration is principally in Orthopaedics, but we also support biomechanics-focused projects in pediatric surgery, podiatry, general surgery, and neurological surgery. The laboratory specializes in mechanical testing of medical devices, with a particular emphasis on orthopaedic implants.
Tang says she conceived of a project to study “the changes in forces and moments for procedures that induce more drastic curvatures, such as pedicle subtraction osteotomies” — a form of spine surgery in which wedges of vertebra are removed to cause a bend in the column.
According to the abstract of her article, “Comparison of a novel pedicle subtraction osteotomy model using the traditional American Society of Testing and Materials standard for spinal biomechanics fatigue testing. Laboratory investigation”:
There is currently an internationally recognized standard (F1717) provided by the American Society of Testing and Materials (ASTM) for fatigue testing of spinal fixation constructs, assuming relatively straight posterior rods between 2 vertebral segments. However, there is currently no standard that effectively describes the changes in forces and moments for procedures that induce more drastic curvatures, such as pedicle subtraction osteotomies (PSOs). In this study, the author proposes a modified version of the ASTM F1717 standard to compensate for changes in loading conditions for PSO constructs.
Twelve specimens were divided into 2 groups: 6 modeled after the original ASTM standard and 6 after the modified version. Three specimens each from the 2 groups had rods contoured to corresponding PSO angles of 20° and 60°, respectively. Specimens were cycled at 4 Hz at a 400 N/40 N or 700 N/70 N load ratio until failure was observed or run-out (testing cycle end point) was reached at 2,000,000 cycles. Cox proportional hazards regression was used to analyze the effect of rod curvature on the fatigue strength of the 2 different models.
Results indicated that contouring rods from a PSO angle of 20° to 60° significantly increased the fatigue life of the screw-rod construct (hazard ratio [HR] 1.57, p = 0.0144) for the original model, but had the opposite effect of decreasing the fatigue life for the modified model (HR 0.64, p = 0.0144)
Because there is extensive data showing that contouring rods to more extreme angles significantly lowers their fatigue life, the modified ASTM model may be more accurate for simulating constructs that assume insignificant rod bending.
Tang tells us that she
did everything on my own
for the study and didn’t think she needed to include any co-authors:
I was trying to think what the qualifications for authorship were. I didn’t think anyone else was qualified [to appear on the paper].
Meanwhile, Tang was getting ready to leave the lab, which she did before the article appeared. When Leasure saw the paper, Tang says, he was none too pleased:
I think that may have contributed to the not so warm, pleasant feelings
that now exist between her and her former boss.
Leasure, she says, insisted that she retract the paper because it did not give him credit — although she is listed as being affiliated with the Taylor Collaboration in the author information on the article. Tang did not want to comply,but eventually bent to the pressure, with the following result:
To The Editor: I am writing to request the retraction of my article, “Comparison of a novel pedicle subtraction osteotomy model using the traditional American Society of Testing and Materials standard for spinal biomechanics fatigue testing. Laboratory investigation,” which was published online in the Journal of Neurosurgery: Spine on October 26, 2012; DOI: 10.3171/2012.9.SPINE12687.
There was a misunderstanding about the ownership of the data collected for this study. To preserve the academic integrity of published results, I request that the article be retracted.
I apologize for the inconvenience of the reviewers, the editor, and readers of the Journal of Neurosurgery: Spine. The article was retracted on December 7, 2012.
Jessica A. Tang, B.S.
University of California, San Diego School of Medicine
San Diego, California
We left word for Leasure for a comment but have yet to hear from him. We’re hoping he has an explanation of events. On the one hand, we appreciate that his lab was housing Tang when he conducted her study. But, as many commenters have argued on this site, that doesn’t necessarily entitle lab heads to use the Royal We for work in which they were not involved. Assuming Tang is telling the truth, how would Leasure have signed an author attestation form (again, assuming such a form would have come his way) for the nature of his contribution to the study?
Meanwhile, we spoke with Joanne Eliason, a spokeswoman for the journal, who could not confirm Tang’s story but says that had Leasure contacted the publication directly demanding a retraction:
that would have been an entirely different situation.
Tang, for her part, says the experience has taught her a valuable lesson about lab politics:
I think I will be a lot more careful [in the future].