Spinal fusion has been shown to be effective in patients with lytic spondylolisthesis (16,17) and in trauma, scoliosis, infection, or tumor (24). Instability is difficult to demonstrate in a condition in which similar changes are seen in asymptomatic individuals. The cause or causes of LBP remain unknown, involving interrelating physical, psychological, social, and occupational factors (11). A Cochrane review has reported little evidence of efficacy for spinal fusion, with a complete absence of randomized controlled trials (10), but four randomized trials of spinal fusion versus nonoperative care have been reported subsequently (1,5,8,16,17). Möller and Hedlund (16,17) reported a trial in isthmic spondylolisthesis, with 77 patients randomized to different forms of surgery and 34 patients randomized to an exercise program. The patients allocated to surgery appeared to do better than those allocated to exercise, but instrumentation and bone grafting was not found to produce an advantage over bone grafting alone. The Swedish trial compared three different surgical regimes with “physiotherapy,” which was essentially “usual medical care” (8). They randomized 222 patients to surgery and 72 to “physiotherapy” (8), with the surgery group randomized between three operative groups of equal size. Disability and other measures were significantly better in the surgical
compared with the nonsurgical group, but no difference in these outcomes was seen between the different surgical techniques. These authors have later claimed that at least some of the patients were in an intensive physiotherapy program, although this was not in the protocol.

Meantime in Norway another trial compared posterolateral fusion with instru-mentation with intensive rehabilitation. Sixty-four patients were compared with instrumented posterior fusion and a rehabilitation program followed to 12 months. They found a similar treatment effect in both groups. One surgery technique was used (posterolateral fusion with instrumentation), and all the rehabilitation patients attended the same program. The numbers were small, but the circumstances were well controlled (1). No advantage to surgery was demonstrated.

Another trial in the United Kingdom, with the type of operation open to surgeons’ choice of optimal method, again showed no advantage to surgery over an intensive rehabilitation program. The UK study was a multicenter randomized trial of 349 patients aged 18 to 55 years with chronic LBP of at least 1-year duration considered candidates for spinal fusion. They were randomized to a surgical group for spinal fusion of the lumbar spine (of the surgeons’ choice, 98% instrumented) or a 3-week intensive rehabilitation program based on cognitive behavioral principles. The surgical group was statistically better (by only 4.5 Oswestry points) than the rehabilitation group at 2 years (28% of the rehab group later had surgery, when the authors expected at least 50% to do this). The economic analysis showed the surgical arm to be twice as expensive as the rehabilitation arm (5,21).

There are many ramifications to this issue including choice of patient and the skills and experience of the surgeon, but the bottom line is that intensive rehabilitation is a powerful method for treating back pain, with few complications, and surgery has little better results with more complications.

Other investigators have shown that instrumentation increased the fusion rate (7,8,24); it is more expensive because of implant costs, duration of operation, and, in some studies, complications (24). Lumbar fusions have higher complication rates than other forms of spine surgery (13). Observational data may suggest that pseudoarthrosis (failure of fusion) may have a bad effect on outcome, but this was not confirmed in a randomized controlled trial (RCT) of two fusion techniques (22).

All these trials have been criticized for poor patient selection, but there is little basis for this criticism. The treatment effect in the surgery arm was similar in all three trials (about 12–15 Oswestry points).

Nonfusion surgery covers various types of flexible fixation and disc replacement. These are covered in detail elsewhere in this book. The published indications for surgery are not totally consistent. The following are quotations from four sources concerning disc replacement: