Twenty-year perspective of randomized controlled trials for surgery of chronic nonspecific low back pain: citation bias and tangential knowledge

Background context: After decades of clinical research, the role of surgery for chronic nonspecific low back pain (CNLBP) remains equivocal. Despite significant intellectual, human, and economic investments into randomized controlled trials (RCTs) in the past two decades, the role of surgery in the treatment for CNLBP has not been clarified.

Purpose: To delineate the historical research agenda of surgical RCTs for CNLBP performed between 1993 and 2012 investigating whether conclusions from earlier published trials influenced the choice of research questions of subsequent RCTs on elucidating the role of surgery in the management of CNLBP.

Study design: Literature review.

Methods: We searched the literature for all RCTs involving surgery for CNLBP. We reviewed relevant studies to identify the study question, comparator arms, and sample size. Randomized controlled trials were classified as "indication" trials if they evaluated the effectiveness of surgical therapy versus nonoperative care or as "technical" if they compared different surgical techniques, adjuncts, or procedures. We used citation analysis to determine the impact of trials on subsequent research in the field.

Results: Altogether 33 technical RCTs (3,790 patients) and 6 indication RCTs (981 patients) have been performed. Since 2007, despite the unclear benefits of surgery reported by the first four indication trials published in 2001 to 2006, technical trials have continued to predominate (16 vs. 2). Of the technical trials, types of instrumentation (13 trials, 1,332 patients), bone graft materials and substitutes (11 trials, 833 patients), and disc arthroplasty versus fusion (5 trials, 1,337 patients) were the most common comparisons made. Surgeon authors have predominantly cited one of the indication trials that reported more favorable results for surgery, despite a lack of superior methodology or sample size. Trials evaluating bone morphogenic protein, instrumentation, and disc arthroplasty were all cited more frequently than the largest trial of surgical versus nonsurgical therapy.

Conclusions: The research agenda of RCTs for surgery of CNLBP has not changed substantially in the last 20 years. Technical trials evaluating nuances of surgical techniques significantly predominate. Despite the publication of four RCTs reporting equivocal benefits of surgery for CNLBP between 2001 and 2006, there was no change in the research agenda of subsequent RCTs, and technical trials continued to outnumber indication trials. Rather than clarifying what, if any, indications for surgery exist, investigators in the field continue to analyze variations in surgical technique, which will probably have relatively little impact on patient outcomes. As a result, clinicians unfortunately have little evidence to advise patients regarding surgical intervention for CNLBP.