After years of planning, the Spine Patient Outcomes Research Trial (SPORT) was funded in 1998 by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) to compare surgical and nonsurgical outcomes for intervertebral disk herniation (IDH), spinal stenosis (SpS), and degenerative spondylolisthesis (DS) using both randomized and observational study designs. The results were so significant that the paper written by James N. Weinstein, DO, MS, summarizing them was awarded the 2014 Orthopaedic Research and Education Foundation (OREF) Clinical Research Award.
In the 1990s, he wrote, the spine community became aware of increasing lumbar surgery rates, with IDH, SpS, and DS being the most common diagnoses leading to an operation. From 1979 to 1987, laminectomy rates increased 25 percent, diskectomy rates increased 75 percent, and fusion rates increased 200 percent. The societal cost of low back pain was estimated at $30 billion to $70 billion annually.
Data on outcomes for IDH, SpS, and DS surgery were lacking, until the publication of a seminal randomized controlled trial (RCT) on lumbar IDH in 1983. That study found that patients treated with discectomy improved more than patients treated nonsurgically at 1 year, but the differences were no longer statistically significant at 4 and 10 years.
One attempted meta-analysis on surgery for SpS found that the low quality of the literature, primarily comprising case series with surgeon-determined outcomes, precluded any meaningful conclusions. DS patients had generally been included in SpS studies despite the fact that DS patients typically were treated with decompression and fusion, while SpS patients without listhesis were usually treated with decompression alone.
With this background of increasing surgery rates, increasing costs associated with the treatment of lumbar spine disorders, and “a profound lack of quality evidence demonstrating the effectiveness of lumbar spine surgery,” the National Spine Network (NSN) was created as a multicenter collaborative to study outcomes for these diagnoses. Members of the NSN gathered observational data on lumbar spine patients as they underwent treatment outside of a formal study.
Although the NSN database provided further observational data on spine patients, the significant limitations inherent in such methodology precluded definitive conclusions about surgical effectiveness. As such, Dr. Weinstein wrote, NSN members “were motivated to perform RCTs and prospective observational studies comparing surgical and nonsurgical outcomes for IDH, SpS, and DS.”
Additionally, concerns about the rapidly escalating cost of spine treatment served as the impetus to perform cost-effectiveness analyses (CEAs) alongside the prospective trials.
To test the hypothesis of no outcome differences between surgical and nonsurgical treatment for IDH, SpS, or DS on standardized measures of symptoms, physical function, health-related quality of life, cost, and satisfaction, SPORT had the following aims:
- To use the NSN to simultaneously conduct three multicenter RCTs of surgical versus nonsurgical treatment for patients with IDH, SpS, or DS with longitudinal outcomes measurement out to at least 4 years.
- To perform a simultaneous observational cohort study that would include patients who declined to participate in the RCTs but agreed to be followed as part of an observational cohort, evaluating their outcomes in the same fashion as patients enrolled in the RCTs.
- To synthesize the results from the RCT and the observational cohort to formally estimate the cost-effectiveness of surgical versus nonsurgical interventions for IDH, SpS, and DS.
SPORT was conducted at 13 multidisciplinary spine practices in 11 states. After patients were identified and informed about the study, they could elect to enroll in the RCT and be randomized to surgical or nonsurgical treatment or to enroll in the observational cohort and make the treatment decision with their physician.
For IDH, patients 18 years and older diagnosed with IDH by participating physicians were considered for inclusion if their symptoms persisted for at least 6 weeks, despite nonsurgical treatment. The content of pre-enrollment nonsurgical care was not prespecified in the protocol. Specific inclusion criteria at enrollment were radicular pain, evidence of nerve root irritation with a positive straight leg raise or positive femoral tension, or a corresponding neurological deficit (asymmetric depressed reflex, decreased sensation in a dermatomal distribution, or weakness in a myotomal distribution).
Additionally, all participants had cross-sectional imaging showing a herniated disk at a level and side corresponding to the clinical symptoms. Patients with multiple herniations were included if only one of the herniations was considered symptomatic (ie, if only one was planned for surgery).
For SpS and DS, patients were at least 40 years old and were diagnosed with SpS with or without DS. All patients had neurogenic claudication and/or radicular leg symptoms for at least 12 weeks despite nonsurgical treatment. The content of pre enrollment nonsurgical care was not prespecified, and there was no requirement for a neurologic deficit. All patients had confirmatory cross-sectional imaging showing lumbar spinal stenosis at one or more levels. Patients found to have a degenerative anterolisthesis on the standing lateral radiograph were classified as DS and those without anterolisthesis as SpS. Patients with a spondylosis or isthmic spondylolisthesis were excluded.
The impact of SPORT
SPORT was funded by NIH in 1999 and, with nearly $30 million of direct funding, remains the largest project ever funded by the NIH for outcomes assessment in orthopaedic surgery. SPORT has been successful not only in meeting its original aims but also in establishing new paradigms for conducting outcomes research and quality-of-care evaluation. It has served as a model for other important NIH-supported trials in orthopaedics. It was the first study to include both an RCT and a concurrent observational cohort study, which allowed for the inclusion of many patients in the observational study who declined participation in the RCT.
Although 37 percent of eligible patients enrolled in the RCT, an impressive 87 percent of eligible patients were included in either the RCT or observational cohort study. In addition to the benefit of patient enrollment, this design also allowed for comparison of both patient characteristics and outcomes between the RCT and the observational cohort, which allowed for the evaluation of the generalizability of the RCT.
SPORT simultaneously studied three related but clinically distinct patient cohorts using a common infrastructure and research methodology. The primary research question focused on the most fundamental question in spine surgery: Does surgery improve outcomes compared to nonsurgical treatment for the three most prevalent conditions leading to lumbar spine surgery?
SPORT was also the first trial to use shared decision making as a standard part of the enrollment and consent process. It established new standards for analytical methods to deal with complex longitudinal data, repeated measures, treatment crossovers, and missing data. It has also provided benchmarks for comparisons for surgical complications, quality of care, and reliability of claims data in evaluating outcomes.
National media attention to the results from SPORT continues today. More than 40 original articles have been published based on the SPORT data, and they have subsequently been cited by more than 780 studies. SPORT has provided quantitative information on outcomes and cost-effectiveness, demonstrated the successful integration of shared decision making into clinical practice, and improved techniques of real-time outcomes measurement in the clinical setting, all of which have changed clinical practice and health policy.
Several large programs, such as the High Value Healthcare Collaborative (HVHC) and an NIH-sponsored Multidisciplinary Clinical Research Center (MCRC), both located at Dartmouth, grew out of the SPORT experience. The HVHC currently includes more than 70,000 physicians and more than 70 million patients and has the goal of improving healthcare by first identifying and then rapidly disseminating best practices that lead to better outcomes and higher value. The MCRC is a P60 research center organized around a methodology core, which uses data collection and analytical techniques developed in SPORT.
“SPORT was one of the largest and most important trials ever conducted in orthopaedics,” Dr. Weinstein wrote, “ and not only demonstrated the clear benefit of surgery for patients who met strict indications but will also serve as a model for study design, data collection, and analysis in the future.”
Disclosure information: James N. Weinstein, MD, MS; Adam M. Pearson, MD, MS; Tor D. Tosteson, ScD; Anna N.A. Tosteson, ScD; William A. Abdu, MD, MS; Sohail K. Mirza, MD, MPH—no conflicts; Jon D. Lurie, MD, MS—National Institutes of Health; AHRQ; Baxano; FzioMed; NewVert; Blue Cross Blue Shield.