June 1996
PHILADELPHIA—A study conducted by an independent research group has determined the "Historical Cohort Study of Pedicle Screw Fixation in Thoracic, Lumbar and Sacral Spinal Fusions" to be scientifically valid. The devices are at least as safe and effective as fixation with non-pedicle screw instrumentation and, therefore, should "have a clear and valid role in the range of techniques available to the surgical community," the group concludes.
The study, compiled by Premier Research Worldwide (PRW), here, was funded by Sofamor Danek, a manufacturer of pedicle screw fixation devices and was included in their comments to the Food and Drug Administration (FDA) on the proposed downclassing of the devices. The historical cohort study was used by the FDA's Orthopedic and Rehabilitation Devices Advisory Panel to support its unanimous recommendation for downclassing the devices. Currently, pedicle screw systems are regulated in class II for indications of severe spondylolisthesis at the L5-S1 level. Downclassification of these devices would extend the clinical indications to include spinal fractures and less severe forms of degenerative spondylolisthesis.
Sofamor Danek sought the independent review of the cohort study following the submission of comments to the FDA by attorneys representing patients who are seeking liability action for alleged damages and injury caused by the implantation of pedicle screw spinal fixation systems. In the comments, the attorneys asserted that, among other things, the cohort study was tainted by a lack of unbiased scientific evidence and undisclosed relationships between certain manufacturers of the devices and some physicians who participated in the study and/or members of the advisory panel.
One aspect of the agreement between Sofamor Danek and PRW was that the research would be included as part of the manufacturer's comment no matter what the conclusion. "We did not want to limit their research," said Ron Pickard, chairman and CEO of Sofamor Danek. "We told them that no matter what they found, we would include it in our comments. That is how confident we are in the strength of the cohort study."
Regarding the appropriateness of the historical cohort design, the PRW investigators wrote, "Given the status of pedicle screw fixation as: an extensively described procedure in the literature, a technique that is broadly taught within the undergraduate and post-graduate medical curriculum and a procedure generally considered to be the accepted standard of care internationally in the orthopedic community, we would foresee immense difficulties in establishing and conducting any form of prospective randomized clinical trial attempting to achieve the same broad participation of investigators or patients. We believe data of high quality can be obtained from a historical cohort design and, in this particular study, has been obtained."
![[bar]](../art/gradient.gif){kind=small}
Researchers Camille Loque, PhD, and Jason S. Zielonka, MD, noted several concerns that have been voiced by other reviewers over the data contained in the cohort study. One of the key questions in evaluating the study, they wrote, is whether the sampling of the two populations (surgeons who performed the spinal fusion procedures and the patients who underwent the procedures) was conducted in an adequate fashion resulting in unbiased population samples. Other questions addressed by the researchers were whether the data contained in the study may be invalid due to financial relationships between investigators and the device manufacturers, whether the patients who were lost to follow-up may have affected the validity of the data collection process and if the low mean rate of patient enrollment per surgeon affected the results.
Of these concerns, the researchers wrote that the majority of the participants (86.8%) were orthopedic surgeons, 74.3% of whom stated that they had no financial interest or other tie to pedicle screw device manufacturers. Of the mean number of patients submitted per surgeon, which was 11.1, the researchers wrote, "No surgeon submitted sufficient subjects to be a disproportionate influence on the data, in fact, no single surgeon submitted more than 5% of the total cases." Furthermore, the researchers noted that, despite the rate of submission per surgeon, overall the study included more than 60% of the potentially eligible U.S. population.
As for the patients lost to follow up, the researchers wrote, "the assumption being made here is that 'the patients with poor results may never had returned to the surgeons who caused them such pain.' In fact, it is more likely that the patients with a satisfactory result did not return since: they had no reason to—they were 'feeling fine'; there would be considerable imposition to them either as lost time from work or with family; there would be a financial burden to them in the expense of seeing the physician and having any diagnostic procedure which might be requested; and in a cost contained environment, they would not obtain the follow-up radiography and any specialized evaluations if they were doing clinically well."
In the statistical analysis, the researchers reported additional survival analyses were performed that examined the rate of fusion among those patients whose physician did or did not report a financial or other relationship to a pedicle screw device manufacturer. The results demonstrated in the degenerative spondylolisthesis patients whose physicians reported a financial or other relationship, the adjusted fusion rate was 80.6% for the pedicle screw fixation patients and 73.2% for the non-instrumented patients (P=0.11). For the degenerative spondylolisthesis patients with surgeons who did not report a relationship with a manufacturer, the adjusted rate of fusion was 81.9% for the pedicle screw patients and 72.1% for the non-instrumented group (P=0.001).
Allowing for the difference in the sizes of populations between the groups, 66% of the patients had physicians claiming no relationships with the manufacturers, which may have resulted in a loss of significance, the researchers claimed. "In any case, these results suggest that financial or other relationship between the investigator and a pedicle screw device manufacturer does not predispose the investigator to reporting a favorable outcome for pedicle screws."
The researchers noted for the study they used the statistical meaning of bias, indicating the presence of a factor other then the independent variables being studied that impacts the relationship between exposure to the test article and the outcome. "This is quite different from the common, everyday use of word, where it implies an attempt by an individual to alter the outcome of a situation to fit a pre-determined belief or desired outcome."
Other reviewers of the study have also criticized the use of fusion as an outcomes measure since the causal relationship between the outcome of fusion and instrumented procedures has not been established. To this, they said for the indications included in the cohort study, degenerative spondylolisthesis and spinal fracture, the assessment of fusion is a valid outcome measurement and has been used as such "in the vast majority of studies of spine surgery involving degenerative spondylolisthesis and other conditions which vertebral instability plays a prominent role in the pathophysiology." Also, they note, "FDA has long considered fusion a key outcome and has required this endpoint in studies conducted for approval of devices intended for treatment of these indications."
Vert Mooney, MD, of the University of California in San Diego and a past president of the North American Spine Society, said he agrees with the researchers conclusions about the cohort study. "It seems very thorough and balanced in the manner that they address various situations and present the contrasting information and opinions. Their review of the cohort study parallels my own."
Mooney said he feels that the issue of bias in regard to financial ties between the manufacturers of the devices and the investigators involved in the study has been "blown out of the water" by the researchers' findings that the surgeons without the ties had slightly better track records for fusion than did those with the alleged ties. He agreed with the investigators' findings that patients lost to follow up would more likely be those with a satisfactory result than an unsatisfactory one. "
Since the review of the cohort study was funded by a device manufacturer, many may question the objectivity of the research. However, as stated by Mooney, "Who else would fund research such as this?" Besides that, he said, these researchers are putting their reputation on the line, and I am sure they would not jeopardize their career for one of their clients.
To obtain a copy of the PRW report contact Sofamor Danek at (901) 396-2695.
The Premier Research Worldwide review of the historical cohort study was submitted to the FDA as part of Sofamor Danek's comment on the proposed downclassing of pedicle screw spinal fixation devices. Sofamor Danek's comments as well as comments from physicians, patients and other manufacturers may be viewed through the Orthopedics Today web site at: http://www.slackinc.com/bone/ortoday/comments/comment.htm
You can express your views on this
article, or other relevant themes, in the
Orthopedics Today Specialty Forums.
