Good follow-up results included a 100% follow-up rate, adequate corrections, little correction loss, lower Oswestry scores, and a high satisfaction rate in both groups. The anteroposterior treatment did not influence the curve morphometry more than posterior fusion only. In reducing postoperative sagittal malalignment, the authors believe that surgical management should aim at a correction within the high normal kyphosis range of 40 degrees to 50 degrees, consequently providing good results and, particularly in flexible adolescents and young adults, minimizing the necessity for an anterior release.

High complication rates in this group of patients are partly due to the difficult surgery but also to the underlying disease. The surgery should be concentrated in specialized centers.

Tight hamstrings can be considered as an important factor in the surgical management of thoracic Scheuermann kyphosis. Tight hamstring patients can be classified as "lumbar compensators" and as such are prone to overcorrection and imbalance. Preoperative assessment of the lumbar-pelvic range of motion and tight hamstrings should therefore be advised. Extensive fusion of the lumbar segments might compromise the lumbar compensation mechanism and induces further risk of imbalance.

To determine the effectiveness of posterior H-frame instrumentation for the surgical treatment of idiopathic scoliosis, 36 patients were studied. The patients underwent surgery between 1989 and 1993 and were evaluated for curve correction, hump correction, vertebral rotation, fusion level and complications. Average age at surgery was 19 years. Duration of follow-up averaged 2.5 years. Mean primary curve correction in patients with a King type I curve was 44.8% (n = 4) and in patients with a type II curve 67.3% (n = 9). Patients with King type III (n = 17) and IV (n = 6) curves achieved respective mean curve corrections of 67.8% and 63.9%. During follow-up there was a mean correction loss of 0.8% in type I, 5.4% in type II, 10.1% in type III and 2.4% in type IV curves. No significant derotation of the primary curves was noted. Rib hump correction and rotational changes of the unfused compensatory curves were significant. Fusion levels extended beyond L2 in six cases. Major neurological problems did not occur. Pseudoarthrosis developed in one patient and imbalance in two patients. The H-frame system satisfactorily achieves curve and rib hump correction with little correction loss.

In this study a series of 32 patients with idiopathic scoliosis, managed with selective thoracic fusion, was reviewed. Classified according to King and instrumented with the H-frame, the patients were evaluated for curve correction, rib hump correction and postoperative shift in lumbar rotation. Age and follow-up average 19.4 and 2.4 years, respectively. The 32 patients had an average primary and lumbar curve correction of, respectively, 66% (6.0% correction loss) and 53% (3.4% correction loss). The respective values for postoperative rib hump correction and shift in apical lumbar rotation averaged 8 degrees and 9.4 degrees in type II King curves 4.4 degrees and 3.5 degrees in type III and 11 degrees and -5 degrees in Type IV. Significant differences were noted between the curve types in rib hump correction and shift in lumbar rotation. The study showed that en bloc postoperative rotation of the compensatory lumbar segment, directed towards the rib hump, positively influences rib hump correction. This en bloc rotation of the unfused lumbar segments is induced by the correcting forces applied by the instrumentation. The unfused lumbar spine of a patient with a King type II curve shows a larger lumbar rotation shift and subsequent rib hump correction than that of a patient with a King type III curve. Together with factors such as lateral angulation, rib-vertebra angles and structural limitations, the rotational dynamics of the unfused lumbar spine seem to form an important component in the under-standing and surgical management of scoliosis.