PubMed1985
A long-term follow-up study has been made of 276 patients with thoracic or double major idiopathic curves treated by Harrington rodding between 1963 and 1971. 82 patients were able to be followed up to the present date, including 24 double major idiopathic curves and 58 thoracic or thoracolumbar curves. The average follow-up was 15 years with a maximum of 20 years. There were 71 female and 11 male patients. It was concluded that the lumbar curve associated with a thoracic or thoraco-lumbar curve lost about 9.9 degrees when the lower point of insertion of the rod was higher than L3. When it was at L3 or L4 loss of correction was only 3.6 degrees. During the same period the loss of correction of the main grafted curve was 13.6 degrees. In double major curves, the best results were obtained when the lumbar curve was completely or almost completely fused, with an average correction of 17.2 degrees. The results were good when the fixation extended to the apex of the lumbar curve, with an average correction of 11.6 degrees and was moderate when it extended to a higher level with an average correction of 10.9 degrees. It is concluded that in almost all cases, instrumentation needs to be extended to a low level at L3 or L4 to provide greater stability. A study on all 276 cases showed that the morphological adaptation beneath the grafts was satisfactory and long lasting. When the instrumentation stopped at the level of T12, L1 or L2 there was a danger of the development of secondary lumbar kyphosis. Low back pain was not made worse in any cases of low lumbar fixation.