There is a high incidence of spinal deformity in children with the Prader-Willi syndrome. We have encountered major complications following spinal surgery in this condition. We report our experience and conclude that spinal surgery is a formidable undertaking and the risks should be appreciated by the surgeon and the parents.

We used a modification of the wake-up test to monitor spinal cord function in 102 consecutive scoliotic patients undergoing Harrington instrumentation. Four patients were found to have a neural deficit when they were awakened during the operation. Three recovered immediately after partial release of the distracting force; one required immediate removal of the rod and was left with a slight neural deficit. Using our method, we have encountered no problem in performing the wake-up test, although attention is drawn to the difficulty in repeating the test if the patient is re-anaesthetised with diazepam. There were no false negative results in this series.

A prospective survey was carried out on all cases of irritable hip presenting at the Royal Liverpool Children's Hospital over a period of one year. All children had a radioisotope scan of the hips and were then followed for one year by serial radiography. Five of the 50 children seen during the one year had areas of ischaemia in the capital femoral epiphysis demonstrated on the scan. all five developed radiological signs of Perthes' disease within sic months. The remaining 45 had radiographically normal hips at one year.

This prospective study was performed to determine the true incidence of deep vein thrombosis of the lower limb in children who had undergone halo-femoral traction to correct scoliosis before operation. Bilateral ascending phlebography of the lower limbs was performed on 54 children two days before spinal fusion and Harrington rod instrumentation. Two patients developed clinical evidence of thrombosis whilst on traction. This diagnosis was confirmed by phlebography. The other 52 children had no clinical evidence of thrombosis and their venograms were normal. The incidence of thrombosis was 3.7 per cent and clinically silent thrombosis did not occur.