Rate of complications in scoliosis surgery - a systematic review of the Pub Med literature

Abstract

Background: Spinal fusion surgery is currently recommended when curve magnitude exceeds 40-45 degrees. Early attempts at spinal fusion surgery which were aimed to leave the patients with a mild residual deformity, failed to meet such expectations. These aims have since been revised to the more modest goals of preventing progression, restoring 'acceptability' of the clinical deformity and reducing curvature.In view of the fact that there is no evidence that health related signs and symptoms of scoliosis can be altered by spinal fusion in the long-term, a clear medical indication for this treatment cannot be derived. Knowledge concerning the rate of complications of scoliosis surgery may enable us to establish a cost/benefit relation of this intervention and to improve the standard of the information and advice given to patients. It is also hoped that this study will help to answer questions in relation to the limiting choice between the risks of surgery and the "wait and see - observation only until surgery might be recommended", strategy widely used. The purpose of this review is to present the actual data available on the rate of complications in scoliosis surgery.

Materials and methods: Search strategy for identification of studies; Pub Med and the SOSORT scoliosis library, limited to English language and bibliographies of all reviewed articles. The search strategy included the terms; 'scoliosis'; 'rate of complications'; 'spine surgery'; 'scoliosis surgery'; 'spondylodesis'; 'spinal instrumentation' and 'spine fusion'.

Results: The electronic search carried out on the 1st February 2008 with the key words "scoliosis", "surgery", "complications" revealed 2590 titles, which not necessarily attributed to our quest for the term "rate of complications". 287 titles were found when the term "rate of complications" was used as a key word. Rates of complication varied between 0 and 89% depending on the aetiology of the entity investigated. Long-term rates of complications have not yet been reported upon.

Conclusion: Scoliosis surgery has a varying but high rate of complications. A medical indication for this treatment cannot be established in view of the lack of evidence. The rate of complications may even be higher than reported. Long-term risks of scoliosis surgery have not yet been reported upon in research. Mandatory reporting for all spinal implants in a standardized way using a spreadsheet list of all recognised complications to reveal a 2-year, 5-year, 10-year and 20-year rate of complications should be established. Trials with untreated control groups in the field of scoliosis raise ethical issues, as the control group could be exposed to the risks of undergoing such surgery.

Figure 1

Similar Cobb angles clinically may look different depending on curve pattern. All patients on this figure have a Cobb angle of 40 degrees. As can be seen, the more decompensated a curve, the more visible the deformation. Double major curvatures are compensated; the most stable curves present after the end of growth [4] and therefore rarely requiring surgical treatment.

Figure 2

Failure of spinal fusion often requires salvage surgery. Failure of the ventral instrumentation (VDS): An additional dorsal rod was implanted to stabilize the spine.

Figure 3

Ventral decompensation after spinal fusion. After operation this patient was unable to walk upright. The implant fixed the patient in forward bent position.

Figure 4

Lateral decompensation after fusion of the lumbar curve. The lumbar curve has been fused; the thoracic curve progressed leading to a decompensation to the thoracic convex side. Because of the imbalanced appearance the patient was dissatisfied. There was no cosmetic/psychological benefit in this case and therefore this surgery should perhaps not have been performed.

Figure 5

Balanced appearance with marked rib-hump after surgery. Although a marked rib-hump is clearly visible after surgery the patient is satisfied with the operation. The rib-hump reappeared after 5 years, however compensation has been maintained. The best cosmetic result was achieved directly after surgery.

Figure 6

Excellent clinical result without patient satisfaction. Excellent clinical result 20 years after Harrington instrumentation. The patient is without pain, however suffers from lack of spinal function although the lumbar spine remained unfused. Additionally, the patient, operated on at the age of 13 years, complained that her parents made the decision. After operation the patient experienced significant functional problems she feels unable to cope with.

Figure 7

Excellent radiological result without patient satisfaction. An 'excellent radiological result' one year after ventral fusion. But the patient complained about the decompensation (clinical overcorrection) and the visual prominence of the shoulder blade.

Figure 8

Not the best clinical result with patient satisfaction. This patient was satisfied although two operations have been necessary and the rib-hump and decompensation are still visible. This satisfaction may be the result of the dissonance effect [242].