Operative vs. Nonoperative Treatment of Thoracolumbar Burst Fractures Without Neurological Deficit: A Randomized, Prospective Study

R. Jhanjee,
K. Wood,
G. Buttermann,
T. Garvey,
R. Kane,
V. Sechreist,
A. Mehbod
University of Minnesota, Minneapolis, Minnesota, USA

PURPOSE:
To our knowledge there has never been a true prospective randomized study comparing these two treatment options.

METHODS:
From 1994 – 1998, 55 consecutive patients with a stable thoracolumbar burst fracture and no neurologic deficit were randomized to two treatment groups: Operative (posterior or anterior fusion plus instrumentation) or non–operative (body cast or orthosis). Radiographs and CT scans were analyzed for sagittal alignment, and canal compromise. All answered a pre–fracture disability questionnaire and a visual pain scale. Post–treatment, patients completed pain scales, functional capacity questionnaires, SF–36, Oswestry, and a satisfaction assessment. All patients also answered special self–normalization pain questionnaires and were compared with uninjured controls. Statistics were ANOVA and regression analysis.

RESULTS:
Ave. follow–up was 47 months. There were 39 men and 16 women. There was no statistical difference between the two groups in terms of pre–injury demographics. For the operative group (Group A; 30 pts), the average fracture kyphosis was 12º and 12º at final follow–up. Average canal compromise on admission was 38%. In group B (non–operative; 25 pts), the admission kyphosis was 9º, and 11º at follow–up. Ave. admission canal compromise was 45%. There was no statistical difference between the two groups in terms of return to work or type of work. Average pain and Roland and Morris scores on admission and at final follow–up were similar for both groups. Final SF–36 and Oswestry scores were also similar. Patient satisfaction was high in both groups. Complications were in more frequent in Group A: 2 wound infections, 2 instrumentation failures; 4 patients had elective hardware removal. One group B patient (Parkinson's dz) converted to surgery for failing to control the kyphosis with a cast. CONC. We found no significant long–term advantage to operative treatment of stable thoracolumbar burst fractures.