Fusion for chronic low back pain. Treatment effects, complications and cost-effectiveness. The Swedish lumbar spine study

Abstract: Chronic low back pain is rarely fatal or completely cured but constitutes a considerable suffering from those afflicted by it. When conservative treatment fails, lumbar fusion may occasionally be performed in order to reduce pain and decrease disability. Surgical intervention in this patient category is however controversial, and the efficiency is questioned. With the primary objective to compare lumbar fusion with non-surgical treatment in patients with CLBP using a valid scientific approach, the Swedish Lumbar Spine Study Group conducted a multicenter prospective randomized controlled trial with two years follows up by an independent observer.The studies included 294 patients with a mean age of 43 years (range 25-65), men and women in equal proportions. Non-surgical treatment should have been unsuccessful and radiographic changes should be present at L4-L5 and/or L5-S1. Patients were randomized to one of three surgical groups, 1. Posterolateral fusion without instrumentation (PLF n=73), 2. PLF + instrumentation (n=74), 3. Instrumentation + anterior lumbar interbody fusion or posterior lumbar interbody fusion, a so-called "360-degree" fusion (n=75). There were 72 patients randomized to commonly used non-surgical treatments. The primary outcome measurements were the patient's global assessment, pain, disability, and work. The technical measurements were different hospitalization variables, complications and fusion. Costs were evaluated and related to treatment effects, and the patient characteristics were compared with an age and sex matched random sample from the general population.This multicenter randomized controlled study showed that more patients treated with fusion were improved after two years, 63% vs. 29% in the control group (p<0.0001). Pain and disability was reduced by 25-33% vs. 4-8% (p<0.002). Return to work was 36% vs. 13% (p=0.002). The outcome did not differ among the surgical groups, but the power to detect differences was low (<10%). Complications increased significantly with more demanding surgical procedures: PLF- instrumented PLF -"360": 11-22-40% (p=0.0003), but most complications left no sequelae for the patient. The re-intervention rate increased with instrumentation compared with non-instrumented fusion (p=0.020). The cost differences in a societal perspective was of random significance between surgery and conservative treatment after two years: SEK 711 000 ± 48 500 vs. SEK 626 000 ± 81 000 (p=0.104). In a health care perspective, surgery was more costly: SEK 131 000 ± 9 500 vs. 55 000 ± 6 500, (p<0.0001). The direct incremental cost for the health care sector to get one patient back to work using fusion instead of non-surgical treatment was SEK 329 000 ±122 000. The costs for production losses, or indirect costs, for one patient on sick leave was SEK 464 000 ±182 000. The surgical candidates in this study had more severe back pain, but otherwise resembled the average Swedish citizen with back pain. They did not resemble patients in rehabilitation centers and we conclude that our results are probably not generally applicable to every patient with chronic low back pain.