Intial SPORT Trial Outcomes Subject to Interpretation, but Tend to Favor Surgery
BY JOHN MCCORMICK, NOVEMBER 22, 2006
The first arm of the Spine Patients Outcome Research Trial ("SPORT") study was released by the Journal of the American Medical Association yesterday. The $13.5 million NIH funded SPORT study has been designed to determine the efficacy of spine surgery versus non-operative care and has caused quite a bit of controversy and excitement among surgeons and healthcare providers.
Yesterday's release of the SPORT data was initial and highlighted in two separate papers only one type of surgical procedure: lumbar discectomy. The first of the two papers discusses the results of a randomized trial. The second paper discusses the results of an "observational cohort" study where participants declined randomization.
The results? Interpretations vary widely and people have taken sides. While the New York Times indicates that the study "questions the need to operate", Forbes is giving a cautious nod in favor of surgery and spine industry behemoth Medtronic is applauding the study because it confirms that surgery renders a positive outcome on an absolute and relative basis.
So who is right and, darn it, is spinal surgery better or not? Our most immediate interpretation is that only the lumbar discectomy arm of the SPORT study has now been released and the all-important stenosis and spondylolisthesis studies are still pending. So this is not about spinal surgery generally, but about a specific (albeit common) form of spine surgery. The SPORT study as a whole, therefore, has not been completed.
Additionally, the broad set of interpretations outlined above are due to the studies released to date being extremely complex to execute and rife with data and bias issues. The randomized trial assigned 501 patients to either receive surgery or a menu of non-operative treatments. One of the immediate problems that occurred with the study was that a high percentage (45%) of patients who were randomized to non-operative care ended up "crossing over" to receive the operative treatment when the non-operative treatments failed to relieve their condition. In addition, a likewise high percentage (40%) of patients who were randomized to surgery had less severe symptoms, or improved while waiting for surgery, and many ended up not going through with the surgery. Both groups turned out having substantial and significant improvements, but nonadherence to randomly assigned treatment biases the study and conclusions are not warranted when reviewing the data from the original selection or "intent-to-treat" viewpoint. On an "as-treated" basis researchers found that outcomes for the patients treated with surgery showed "strong, statistically significant advantages" when compared to non-operative treatments, however.
What about the "Observational Cohort"? Did the people who got to choose their own treatment have more distinct outcomes? According to the Observational Cohort paper (also released by JAMA yesterday), both operative and nonoperative patients improved and notably "those who chose operative intervention reported greater improvements than patients who elected nonoperative care".
A key issue to us is that "non-operative" is not clearly defined. First, among others, was the widely varied methods used when implementing non-operative treatments. Anything from narcotics, to TENS units, to counseling to chiropractors were deployed in some but not all of the non-operative treatments. Second, non-operative treatments often involve lengthy periods of rest and/or physical rehabilitation and even if this method is efficacious it may be of no practical use to the individual who has to remain engaged in the work force.
Our own interpretation of the initial SPORT data released yesterday is that the data does tend to favor the lumbar discectomy procedure where the patients, not statisticians and computers, are choosing the treatment.
