Hartshill fixation of thoracolumbar burst fractures is justified on grounds of providing early stabilization and correction of deformity. It is important to answer the question of whether it is able to sustain correction so that the costs and risks of surgery of this disabling problem can be justified. Reliable answers are also needed in view of the costs and risks of surgery prognostication. This before-after study is on 49/61 consecutive cases of burst fractures of thoracolumbar spine, admitted during October 1998 to November 2000. Patients were treated with Hartshill segmental spinal fixation two segments above and below the fracture and bone grafting, and followed-up for a mean of 28 months (18–43 months) to determine the fate of radiological parameters indicating correction of kyphosis and clinical indicators predicting neurological recovery. The mean pre-operative kyphosis angle of 19.59° (17.93–21.24) initially improved by 9.73° (8.07–11.39), p < 0.05 but reverted to 19.71° (17.73–21.69) during follow-up. Vertebral height measurements indicated similar findings. Mild pain persisted in 48/49 patients while 1/49 needed analgesics. Improvement of neural function was seen in 37/44 patients having a pre-operative deficit. 6/16 variables had a significant univariate relation with neurological recovery at p < 0.3 levels; namely initial Frankel's grade 1.85 (0.81–4.22), p = 0.14; fracture level 3.83 (0.58–25.10), p = 0.16; initial kyphosis 1.11 (0.93–1.32), p = 0.24; age 1.03 (0.96–1.11), p = 0.28; sex 0.39 (0.06–2.23), p = 0.29 and post-operative kyphosis 1.07 (0.93–1.22), p = 0.30. When combined, these variables were able to correctly classify recovery in 77.78% of cases at the probability cut-off of 0.75, sensitivity of 0.77 and specificity of 0.75, while the pre-test probability of recovery of 0.84 improved to 0.93 post-test, Negative Predictive Value (NPV) was 0.42., +Likelihood Ratio (LR) was 3.11 and -LR was 0.29. 77% of the area was under the Receiver Operator Characteristic (ROC) Curve. Although Hartshill fixation with segmental sub-laminar wiring failed to maintain kyphosis correction in thoracolumbar burst fractures, it resulted in reducing the recumbence and hospital stay, and facilitating early rehabilitation. However, the costs and the risks of such procedures have to be carefully weighed before they are recommended for wider use. We also have a rule for predicting neurological recovery in such patients.
Modified pedicle screw placement at the fracture level for treatment of thoracolumbar burst fractures: a study protocol of a randomised controlled trial