Introduction: It is the accepted dogma that should paralysis complicate spinal deformity surgery, then the internal fixation should be removed within three hours. This dogma is based on MacEwen’s paper in 1975 which related to the Harrington system and which did not contain statistical analysis (MacEwen G.D. et al, JBJS 557A, 1975,404-8). Since that time spinal cord monitoring systems have been developed and internal fixation systems have become considerably more complex. Does the accepted dogma need to be reviewed? Methods and results: The author has reviewed the literature which contains statistical analysis of risk factors and results in relation to major neurological complications of spinal deformity surgery (Dove J. Résonance Européenes du Rachis 1999, 7[23]961–66). The risk factors are adult scoliosis, congenital and neuromuscular curves, kyphosis, combined anterior and posterior surgery, intra-operative hypertension, distraction and certain types of segmental fixation. Furthermore these risks are additive. MacEwen’s 1975 paper did not include statistical analysis and its conclusions are not borne out by the information within the paper. The only statistical analysis of the management of neurological complications has shown that surgical removal of the internal fixation was not related to neurological recovery (Paonessa K.G., Hutching F. Scoliosis Research Society Meeting. New York. Sept 1998). Conclusion: Based on an analysis of the relevant literature and current clinical practice, the author suggests an algorithm to be followed by the surgeon faced with a major neurological complication of spinal deformity surgery. The author also raises the question as to whether the British Scoliosis Society should make a statement regarding “best practice” in such cases.
Sublaminar wiring with posterior instrumentation is one of the methods used when long fusions involving 10 to 12 thoracolumbar levels are required. Classically wires are used at every consecutive level to make the construct as rigid as possible although complications like dural tears, CSF leak, and neurological deficiet have been reported during their passage. We compared the mechanical stability of five specimens of each of the three construct designs by static and fatigue testing to torsional strain on Electro-servo-hydraulic testing machine. In construct A, a contoured Hartshill rectangle was used from T2 to L2, with sub-laminar wires passed at every level. In construct B, every alternate level was wired. In construct C, every alternate level was wired except at the proximal end two consecutive levels were wired. Industrially fabricated spine models were used to prepare these constructs. The intervertebral motion within the construct was measured using FASTRAK magnetic field sensor device. On static testing, no statistically significant difference was found in the rotational displacement of the three construct designs. On fatigue testing, all samples of construct B consistently failed with breakage of the wire at the most proximal level on the left side. But on adding additional wires to the next level (Construct C), all five samples withstood fatigue testing at 300 Newton load to 3 million cycles. Wiring alternate levels instead of every level, does not compromise the stability of the construct provided the most proximal two levels are consecutively wired. This practise would minimise the risk of dural tears and cord damage during wire passage and reduce surgical time, not to mention the economical benefit.
The aim of this study was to assess the accuracy of pedicle screw placement using NAVITRAK, a system of Computer Assisted Orthopaedic Surgery and conventional fluoroscopic technique. Twelve porcine lumbar spines were scanned pre-operatively by computer tomography for 3-D reconstruction ( 1 mm slice thickness, 1mm increment and 2.5 mm pitch ). Computer randomisation divided the specimens between surgeons of different experience, and the two pedicles of each vertebral level between the two surgical techniques. Stainless steel screws (6.5 spongiosa) were inserted. Post-operatively, fluoroscopic- and CT imaging were blindly assessed for accuracy by two independent observers, and compared to macroscopic dissection of the spinal segments. Of 168 pedicles in 12 porcine specimens, 166 received a pedicle screw. Two pedicle screw placements were abandoned. Sixyty-one screws (73%) were placed satisfactorily with the CAOS system, 56 (67.5%) in the conventional group. In 26 pedicles the screws were placed unsatisfactorily (12 pedicles (46.2%) with the NAVITRAK system and 14 pedicles (53.8%) with the conventional technique. The NAVITRAK system in combination with stainless steel screws showed a difference of 5.5% in misplacement in favour for the computer assisted technique.
The aim of this study was to demonstrate a correlation between FASTRAK readings of spinal movement and established disability scores in-patients undergoing litigation. A retrospective, blind study was conducted on patients who had been evaluated clinically between January 1994-October 1998. Statistical regression analysis between evaluated Oswestry Disability Score (ODS) and MSPQ/Zung questionnaires and the mean ROM was obtained. 49 patients with soft tissue injuries of the cervical (n = 14) and lumbar (n = 34) spine were assessed. All of them were undergoing litigation. A standardised Fastrak trace measuring flexion, extension, right and left bending and rotation of the cervical and lumbar spine was recorded. An ODS and MSPQ/Zung questionnaire was filled in under the supervision of a senior physiotherapist. There was no correlation between the ODS and MSPQ/Zung and mean ROM for the cervical spine. In the lumbar spine, flexion and ODS correlated statistically significantly (p<
0.01) and right rotation with the combined MSPQ/Zung score (p<
0.014). This preliminary study is encouraging in that it demonstrates a direct correlation between FASTRAK measurements and recognised disability scores in the lumbar spine. Further analysis of non- litigation cohorts will contribute to establish these correlations.
The aim of this study was to assess the accuracy of pedicle screw placement comparing Computer Assisted Orthopaedic Surgery equipment with conventional fluoroscopic technique. Twelve porcine cervical spines were scanned pre-operatively by computer tomography for 3D reconstruction (1 mm slice thickness, 1mm increment and 1 mm pitch). Computerised randomisation divided the specimens between surgeons of different experience, and the two pedicles of each vertebral level between the two surgical techniques. Stainless steel screws (6.5 diameter, spongiosa) were inserted. Post-operatively, fluoroscopic imaging was used for accuracy assessment by two independent observers, and findings were compared to macroscopic dissection of the spinal segments. Of 96 pedicles in 12 porcine specimens, 78 received a pedicle screw, 18 screw placements were abandoned, 38 (39.6%) were satisfactorily placed (19 in each, p>
0.05). 40 screws were misplaced, 18 (45%) with the NAVITRAK system vs. 22 (55%) with the conventional technique. These single factor results (all non-significant), were corroborated using a linear logistic regression model. Some heterogeneity in performance was detected between surgeons, independently of the type of technique used. Computer assisted surgery is an aiming device and is not advantageous over conventional methods in spines with high bone density.
