Fourteen patients with acetabular dysplasia were studied by using three-dimensional computed tomography (CT) reconstructions before pelvic osteotomies. Computer manipulation of the data allowed a preoperative visual assessment of acetabular shape, assessment of potential congruency between the femoral head and acetabulum by using a mathematical best-fit sphere, and measurement of surface contact distances that depict joint coverage and relate to concentration of weight-bearing forces. Preoperative evaluation of the three-dimensional images for these 14 patients allowed improved understanding of their abnormal anatomy and better surgical planning.

STUDY DESIGN:To determine the effectiveness of posterior TSRH instrumentation for the treatment of idiopathic scoliosis, 103 patients with a 2-year minimum followup were retrospectively studied.METHODS:Patients who underwent operations between October 1988 and April 1991 were evaluated for curve correction, spinal balance, and complications. Age at surgery averaged 14.3 years. Follow-up averaged 2.5 years.RESULTS:Thoracic curve correction averaged 65% in those with King Type III/IV curves and 54% in those with Type II curves. With follow-up, correction loss averaged approximately 13% for each group. Lumbar curve correction after instrumentation in Type I and II curves averaged 48% postop but lost approximately 20% with follow-up. Trunk balance improved 77% toward midline after surgery in those with Type III/IV curves. Improvement in trunk balance was less impressive in patients with Type II curves, particularly after selective thoracic fusions. Thoracic sagittal contour improved 43% for hypokyphotic (< 20 degrees) patients but, in the remainder, no significant radiographic change was evident. No neurologic complications occurred. Delayed deep infections developed in ten patients (10%) between 11 and 45 months postoperative. Cultures eventually grew Propionibacterium acnes, staph epidermidis, or staph coagulase negative in eight patients. Two patients had pseudarthroses.CONCLUSIONS:Frontal and sagittal thoracic curve correction can be satisfactorily obtained using TSRH instrumentation. Continued efforts are being made to improve lumbar hook patterns and technique to achieve and maintain better lumbar curve correction.

Eleven children in whom a tibial fracture occurred after minor trauma had pre-existing dysplastic changes evident radiographically. These changes included cortical tapering, sclerosis, and formation of a cyst in the region of the medullary canal. Ten of the eleven patients had had no more fractures an average of fifteen years after the most recent fracture. Six of the fractures healed following prolonged immobilization in a cast, but four of the six tibiae were abnormally bowed anteriorly, and it was thought that a stress fracture could occur in the future. Four of the patients had a clinically straight tibia and radiographic evidence of thick cortices following corrective osteotomy, intramedullary fixation with bone-grafting, and prolonged immobilization in a cast. At the most recent follow-up examination, the eleventh patient had a persistent pseudarthrosis despite several operative procedures. While there were too few patients in this series for us to draw definite conclusions, our findings suggest that late-onset pseudarthrosis of a dysplastic tibia has a better prognosis than does congenital pseudarthrosis.

In twenty-five patients, the Pavlik harness failed to obtain or maintain reduction in thirty of thirty-five congenital dislocations of the hip. All of the patients had met the clinical criteria for use of the harness in our institution: they were less than seven months old, the femoral head pointed to the triradiate cartilage on anteroposterior radiographs that were made with the child wearing the harness, and they had no evidence of neuromuscular disease or teratological dislocation. These patients were compared with seventy-one patients (eighty-one dislocations) who had also been treated with the Pavlik harness and in whom a stable reduction was obtained and maintained. Statistically significant risk factors for failure of the harness included an absent Ortolani sign at the initial evaluation, bilateral dislocation, and an age of more than seven weeks before treatment with the harness was begun. All thirty hips in which the harness failed to obtain or maintain reduction had a subsequent attempt at closed reduction after preliminary Bryant traction. Fifteen of these hips were successfully reduced closed, but two later redislocated and needed an open reduction. The remaining fifteen hips needed an open reduction, and two redislocated and needed a second open reduction.

A retrospective review of 38 patients with myelomeningocele scoliosis, focusing on the incidence of eventual fusion, curve correction, and change in pelvic obliquity, was performed. Single stage anterior or posterior fusion was compared to combined anterior and posterior spinal fusion. Single stage anterior or posterior procedures resulted in only a 50% fusion rate compared to rates of 83 to 100% for various combinations of combined anterior and posterior fusions. Similarly, average curve correction and change in pelvic obliquity were much improved with a two-stage approach. No statistical difference in fusion rate, curve correction, or change in pelvic obliquity was noted between the various combinations of two-stage anterior and posterior fusions.

Postoperative decompensation has been reported following Cotrel-Dubousset instrumentation for right thoracic idiopathic scoliosis. The authors examined balance in the frontal and sagittal planes in 53 patients to determine optimal levels for fusion. King et al Type II curves, particularly larger ones, shifted to the left when the thoracic curve was fused to the stable vertebra or just below. Most Type III curves balanced well regardless of the levels fused. One-third of all patients developed mild radiographic junctional kyphosis at the lower level instrumented, more commonly when instrumentation ended at or above T12. The authors recommend fusing one segment short of the stable vertebra in most Type II curves. Large Type II curves need both curves fused for optimal balance. Type III curves can be fused short of the stable vertebra.

We reviewed the results of primary operative treatment in twenty-five patients (thirty-three hips) who were two years or older and had congenital dislocation of the hip. None of the patients had had previous treatment for the dislocation. Preliminary traction was not used in any patient. Femoral shortening and, in twenty-one hips, pelvic osteotomy were performed at the time of open reduction. At the most recent follow-up (average, three years and seven months), according to the radiographic classification system of Severin, there were seven excellent, seventeen good, and eight fair results; one hip had a poor result. Avascular necrosis developed in three of the thirty-three hips. At follow-up, these hips had a radiographic result of excellent, good, and fair, respectively. Twenty-one patients (twenty-eight hips) were reviewed with respect to range of motion and recovery from limb-length discrepancy. According to the rating system of Ferguson and Howorth, there were seventeen excellent, seven good, and three fair results; one hip had a poor result. It was concluded that children who are two years or older and who have a congenital dislocation of the hip can safely be treated with an extensive one-stage operation consisting of open reduction combined with femoral shortening and, often, pelvic osteotomy, without increasing the risk of avascular necrosis. The limb-length discrepancy that is produced by the shortening does not appear to cause a clinical problem.

This is a preliminary report on Cotrel-Dubousset (CD) instrumentation for the surgical management of idiopathic thoracic scoliosis. From September 1985 through April 1986, 37 patients were treated at the authors' hospital, by posterior spinal fusion with CD instrumentation. Twelve patients had surgical treatment of spinal deformity associated with other disorders or had revision surgery. The remaining 25 patients, with no prior surgery, were diagnosed as having juvenile or adolescent idiopathic scoliosis. After operation, this group of patients was routinely noted to have significant improvement in rib deformity. This is associated with the rotational correction achieved with CD instrumentation and contrasts with the minimal rib deformity correction with Harrington instrumentation documented by some workers. In this series, no rib resections have been necessary in conjunction with CD instrumentation. No postoperative external immobilization was used. Ambulation began on the second postoperative day, and patients were discharged five to seven days after operation. Gradual resumption of normal activities was allowed at six weeks, and full activities, other than contact sports, after three months.

Clinically, implant failure is often the result of fatigue from continuous cyclic loading. Because of the inadequacies of long-run cyclic testing, fatigue susceptibility of implants was investigated by means of strain measurements and stress analysis under physiologic loads. The implants were equipped with strain gauges during load-deformation testing, and the tensile stress (the component of stress-producing fatigue failure in metals) was calculated for that site on the implant. For metals most often implanted for spinal surgery, such as stainless steel and chrome-cobalt alloys, a stress exists, known as the endurance limit, below which failure will not occur, even if cycled indefinitely. By calculating the tensile stresses in an implant and relating them to the endurance limit, the implant's susceptibility to fatigue can be determined at the site of stress analysis without formal cyclic load testing.

This study demonstrates that the broad 4.5 mm dynamic compression plate provides superior stiffness in axial and torsional loading compared to the two anterior rod constructs (Zielke-Slot and Kostuik-Harrington) and the thinner ASIF T-plate in this ex-vivo testing. The Armstrong National Research Council (NRC) plate, designed specifically for anterior spinal instrumentation, has essentially the same mechanical performance characteristics as the broad dynamic compression plate. When placed on the lateral aspect of the vertebral body, these latter two plates lie well away from the anterior vascular structures. Screws are placed slightly off set in the plate, two each in the vertebral bodies immediately above and below the fracture, the screws directed transversely across the vertebral body. The broad DCP plate is easily contoured and implanted, and will allow 1.8 mm of compression to be applied to a strut graft between vertebral bodies. Anterior surgery for the treatment of burst fractures with retropulsed bone provides a means for direct decompression of the spinal canal. With an appropriate implant, additional advantages include instrumentation over a shorter distance (one above and one below the fracture) and no need for further operative procedures to implant or remove posterior implants. Due to their superior stiffness in axial and torsional load, the broad dynamic compression plate and the Armstrong NRC plate appear to fulfill most nearly the ideal attributes of an anterior spinal implant for the treatment of burst fractures of the throacic and lumbar spine.

Magnetic resonance imaging (MRI) was performed 49 times in 42 patients with spinal dysraphism. Scoliosis and a changing neurological picture were the primary indications. Spinal cord anomalies included hydromyelia, diastematomyelia, lipoma, thickened filum terminali, and spinal cord atrophy. All but one patient exhibited Arnold-Chiari malformation. Twenty-two of the 42 patients had computed tomography (CT) scans, myelograms, or operations that corroborated the 41 MRI findings. Three false-positive MRI findings of hydromyelia and no false-negative studies were observed. MRI is a noninvasive investigative technique that provides more information than myelography or CT in defining spinal cord anatomy in spinal dysraphism.

Intraoperative somatosensory evoked potentials in 50 patients were reviewed; each waveform was correlated with intraoperative surgical events and conditions. Twenty-two patients maintained reproducible waveforms and awoke without neurological deficit. Two patients had random waveform changes not correlated with a surgical event and awoke without deficit. Twelve patients showed waveform changes temporally related to circumstances that might endanger cord function; two of these awoke with transient neurological abnormality. Satisfactory waveforms were not obtained from 14 patients. Somatosensory evoked potential monitoring in scoliosis surgery appears to be sensitive, but may not be sufficiently specific; waveform changes do not necessarily indicate objective neurological damage.

Summary We reviewed 27 teratologic hip dislocations in 17 patients. Four hips underwent closed reduction, 10 hips had medial adductor open reduction, 9 hips had iliofemoral open reduction, and 4 hips had iliofemoral open reduction and femoral shortening. Avascular necrosis occurred in 48%, redislocation in 19%, and subluxation in 22% of the hips. Results were best in those hips treated by anterior open reduction and femoral shortening and worst in those having closed reduction. Poor results correlated with multiple other involved joints, hypotonia, developmental delay, and poor head control.

Fifty-seven children with cerebral palsy and subluxated hips underwent soft tissue releases to obtain hip reductions. Eight patients in Group I (preoperatively mildly subluxated hips) had postoperative bracing, all with good results. Twenty-two unbraced patients in Group I had 14 good, 5 fair, and 3 poor results. Thirteen patients in Group II (preoperatively severely subluxated hips) received postoperative splinting with nine good, one fair, and three poor results. The 14 unbraced patients in Group II had one fair and 13 poor results. This study indicates that even severely subluxated hips in spastic children less than 6 years of age can eventually achieve stable reductions provided long-term abduction bracing is used following adequate soft tissue releases.

Patients with cerebral palsy and fixed hip extension are often unable to sit and therefore lose social and educational opportunities. At Texas Scottish Rite Hospital (Dallas, TX, U.S.A.), 29 patients had hip extension-abduction contractures. Eleven patients with mild involvement were managed with physical therapy. Six patients had moderate involvement and required wheelchair modification to maintain sitting. Twelve patients demonstrated severe contractures that required major soft tissue releases and often femoral shortening osteotomies to regain flexion. Patients with athetosis or rigidity were identified as those most likely to develop extension-abduction contractures either spontaneously or following adductor releases.

Summary: The three-dimensional relationship between acetabulum and femoral head, known as coverage, was evaluated radiographically using a modified inlet view of the pelvis. The relationship of the anterior edge, the center of the femoral head, and the posterior edge of the acetabulum is a straight line (180°) in normal coverage but less in dysplastic hips. This radiographic method provides anatomical information concerning anterior and posterior coverage. Evaluation of 26 hips delineated problems, including possible danger of anterolateral rotation of the acetabulum in the face of posterior deficiency and inadequacy of posterior coverage after Chiari osteotomy.

A technique of intraarticular hip arthrodesis using a cancellous screw to provide compression across the joint is introduced. We present four patients in whom this method was successful, resulting in solid fusion and pain relief at follow-up.