In situ pinning in slipped capital femoral epiphysis: long-term follow-up studies.

BACKGROUND: Management of moderately displaced slipped capital femoral epiphysis (SCFE) is debated, mostly because of the risks related to open reduction on one hand, and subsequent evolution toward femoroacetabular impingement (FAI) on the other. METHOD: All SCFE cases treated with in situ fixation (ISF) and a minimum of ten years of follow-up beyond skeletal maturity were analysed in a retrospective multicentre study. Coxometry parameters were measured. Long-term results of ISF were meanwhile analysed at our Institution. RESULTS: A total of 222 patients were included. Patient reported outcome measurements were related to the severity of the initial slip. It suggested a 35.5° threshold for slip angle beyond which FAI was more frequent. Only slight remodelling at the head-neck junction is to be expected. CONCLUSION: Based on these findings, it seems reasonable to perform ISF only in SCFE with a slip angle below 35°.

Arthroscopic treatment of calcaneonavicular coalition in children.

Calcaneonavicular coalition is a common source of pain and more or less severe flat and stiff foot in children. Classically, treatment consists in resecting the coalition using a dorsolateral approach. Good quality resection and interposition can prevent recurrence. The most common complications are infection, hematoma and neuroma. Arthroscopy offers a minimally invasive alternative, but the optimal approach remains undetermined. We describe a surgical technique with an approach based on the anterolateral process of the calcaneus, in three cases with 12 months' follow-up. Arthroscopic resection has certain advantages: recovery is quicker, and the esthetic result is better. For the instrumental portal, skin incision should be superficial, followed by blunt dissection of subcutaneous tissue to avoid superficial peroneal nerve injury. Although longer term follow-up is needed, arthroscopy seems to be an attractive minimally invasive technique in this kind of pathology.

Type IV-D thumb duplication: A new reconstruction method.

INTRODUCTION: In the Wassel type IV classification category, the thumb is duplicated from the metacarpophalangeal joint; this abnormality accounts for approximately 50% of the cases of thumb duplication. Type IV is divided into four subtypes in which the IV-D type, or convergent (9% of cases), is the most complex form because both thumbs are hypoplastic with a divergent metacarpophalangeal joint and a convergent interphalangeal joint. Reconstruction is prone to axis deformity and ligament laxity, whereas the Bilhaut-Cloquet technique's main pitfall is nail dystrophy. We propose a modified Bilhaut-Cloquet procedure to avoid these complications. MATERIAL AND METHOD: Four males (mean age: 11 months; range: 10-12 months) with IV-D thumb duplication were operated on using a modified Bilhaut-Cloquet procedure. SURGICAL TECHNIQUE: The skin is preliminarily marked, taking into account the excision of the central skin and the more hypoplastic nail of the two (most often the radial nail). Both proximal phalanges are split longitudinally and the central halves discarded. An oblique osteotomy is performed at the base of the distal phalanx of the ulnar thumb duplicate (the less hypoplastic) and the radial wedge is excised. The same osteotomy is applied to the distal phalanx of the radial thumb duplicate, but the radial wedge is preserved. The proximal phalanx and the bases of the distal phalanx are joined by bone suture. Axis correction and ligament stability are thus achieved without nail surgery. RESULTS: The patients were examined with a mean 24 months of follow-up (range: 12-36 months). The result was good in all four cases according to the Horii score. DISCUSSION AND CONCLUSION: This procedure combines an excision of the central part of the proximal phalanx and partial excision of the base of the distal phalanx. It provides axis correction and stabilization of the interphalangeal joint while avoiding subsequent nail dystrophy because a single nail is preserved. Preliminary results are encouraging: no axis deformity, instability or nail dystrophy has been noted. Nonetheless, the long-term results need to be evaluated. LEVEL OF EVIDENCE: IV retrospective study.

Spontaneous recovery of experimental valgus deformity in lambs.

INTRODUCTION: Angular deformity in the growing skeleton of animals, especially in the radius and ulna, is occasionally seen in clinical practice. The mechanism of spontaneous correction of these angular deformities however remains to be elucidated. The purpose of our experiment was to explore the ability of a growth plate to correct an induced valgus deformity, and to study the mechanism of correction. METHODS: Before beginning the study, valgus deformity of the distal radius had been induced in lambs by the application of a device that causes asymmetrical compression of the growth plate. The study began after removal of the device and spontaneous correction of the induced deformity was observed weekly for 20 weeks. The angles of the deformity and longitudinal growth on the medial and lateral portions of the growth plate were respectively measured on craniocaudal and mediolateral radiographs. RESULTS AND CONCLUSIONS: Spontaneous correction of the valgus deformity occurred during the first 16 weeks. It resulted from asymmetrical growth characterised by restricted activity of the medial portion of the growth plate (14.8%) in comparison to the lateral portion of the experimental radius, and also in comparison to the medial portion of the control radius.

Tratamiento quirúrgico de la escoliosis congénita por hemivértebra aislada. A propósito de una serie de 27 casos / Surgical treatment of congenital scoliosis resulting from an isolates hemivertebra: 27 cases

Objetivolas escoliosis congénitas evolutivas por hemivértebras pueden ser tratadas quirúrgicamente mediante resección o epifisiodesis convexa. Nuestro objetivo fue evaluar los resultados de una serie de niños operados según una u otra modalidad. Material y métodose realizó una revisión clínica y radiológica de una serie no comparativa de 27 hemivértebras aisladas de las cuales 16 habían sido tratadas por resección y 11 por epifisiodesis. La media de edad en el momento de la intervención fue 50 (24-132) meses y el ángulo de Cobb fue de 33° (20°¿75°). La resección para las hemivértebras lumbares se hizo por una vía de abordaje anterior y posterior y como instrumentación se utilizó el material Baby CD a compresión. La epifisiodesis para las hemivértebras torácicas necesitó de un tiempo anterior de escisión a dos niveles, seguido de un tiempo posterior de descorticación a cuatro niveles: en 6 casos fueron instrumentadas y en 5 casos, escayoladas en posición de reducción. Resultadosno se constató ninguna complicación neurológica. Respecto a las 16 resecciones, la reducción media fue del 75% en postoperatorio y del 73% al final del estudio. Para las 11 epifisiodesis la corrección media obtenida fue del 25% en el postoperatorio y del 39% a los 4 años de seguimiento medio (1¿8 años). Hubo 2 casos de fracaso a largo plazo en ausencia de instrumentación. Conclusionesla resección de las hemivértebras proporciona resultados satisfactorios inmediatos y estables en el tiempo. Sin embargo, en las localizaciones torácicas y a causa del riesgo neurológico, las epifisiodesis convexas con instrumentación garantizan el resultado a largo plazo

Purpose: Congenital scoliosis resulting from hemivertebrae can be treated surgically through resection or convex epiphysiodesis. Our purpose was to assess the results obtained in a series of children operated using both methods. Materials and methodsA clinical and radiological review was carried out of a non-comparative series of 27 isolated hemivertebrae of which 16 had been treated by means of resection and 11 by means of epiphysiodesis. Mean age at surgery was 50 months (24-132) and Cobb's angle was 33° (20°¿75°). In the case of lumbar hemivertebrae, resection was performed through an anterior and posterior approach; compressive CD instruments were used. Epiphysiodesis for the thoracic vertebrae required a prior excision phase at two levels, followed by a subsequent 4-level decortication phase: in 6 cases they were instrumented and in 5 a plaster cast was applied in the reduced position. ResultsNo neurological complication was found. With respect to the 16 resections, mean reduction was 75% postoperatively and 73% at the end of the study. For the 11 epiphysiodeses mean correction obtained was 25% postoperatively and 39% at 4 years¿ mean follow-up (1¿8 years). There were 2 cases of long-term failure in the absence of instrumentation. ConclusionsResection of hemivertebrae provides immediate and stable satisfactory results. However, in the thoracic area, and given the neurological risk involved, instrumented convex epiphysiodesis guarantees a good long-term result (AU)

[Arthroscopic anterior capsular release in medial contracture of the shoulder secondary to brachial plexus birth palsy. Preliminary results]. / Arthrolyse antérieure de l'épaule sous arthroscopie dans les séquelles de paralysie obstétricale du plexus brachial. Résultats préliminaires.

PURPOSE OF THE STUDY: Retraction of the shoulder in internal rotation is observed in 25% of children with brachial plexus birth palsy (C5, C6 +/- C7). Early bone and joint deformities affecting the glenohumeral joint are the consequences. The stiff internal rotation requires surgical release which can involve the capsule and ligaments, muscles, or both. Internal release can be combined with muscle transfer to improve active external rotation. We report the results obtained with arthroscopic anterior capsular release combined with latissimus dorsi transfer. MATERIAL AND METHODS: From 1999 through 2006, fourteen children with a stiff shoulder in internal rotation secondary to brachial plexus birth palsy were managed in our unit. All had recovered biceps function six months after surgery. The glenohumeral dysplasia was analyzed on the preoperative magnetic resonance imaging. Pre- and postoperative passive external rotation (RE) were measured with the arm along the body and at 90 degrees elbow flexion. Internal rotation was measured using the Mallet score (hand-back test). Combined active abduction antepulsion was measured when the child was playing. Mean age at surgery was three years six months. Arthroscopic internal release was performed for eight children. All had an associated latissimus dorsi transfer. RESULTS: Among the 14 children managed in the unit, arthrolysis was not be performed in six, either because of the lack of an adequate electrode (two patients) or because the child presented posterior glenohumeral dislocation making it impossible to introduce the optic channel (four patients). Arthroscopic anterior release was performed for the eight other patients. These eight patients were reviewed at a mean three-year follow-up. Passive external rotation was improved, with a mean gain of 60 degrees with no recovery of passive internal rotation. The abduction antepulsion movement was also improved, mean gain 90 degrees . DISCUSSION: A stiff shoulder in internal rotation can develop during the first two years of life. Several techniques have been proposed for internal release. The origin of the progressive limitation of passive external rotation remains a subject of debate. Is it due to retraction of the internal rotators, or to capsule-ligament retraction, or both? In 1992, Harryman et al. demonstrated the role of the capsule and the coracohumeral ligament in limiting external rotation. Consequently, we have opted for early release (less than two years of age) using an arthroscopic method limited to the capsule and ligaments. Our results for passive external rotation are comparable to those reported by others. However, this technique enables preserved mobility for internal rotation. CONCLUSION: Arthroscopic anterior release limited to the capsule and the ligaments is an effective, minimally invasive technique. Leaving the internal rotator muscles intact preserves internal rotation of the shoulder and reduces the risk of anterior instability.

Comportamiento de la primera cuña en la corrección quirúrgica del metatarso adducto / Behavior of the first cuneiform in the surgical correction of metatarsus adductus

Objetivo. El metatarso adducto (MTA) es una deformación en adducción del antepié. Nuestro objetivo es apoyar la hipótesis por la cual la principal anomalía se encuentra en el lado cuneiforme de la articulación cuneo-metatarsiana, y que el crecimiento de la primera cuña es fundamental en la corrección tras la capsulotomía. Material y método. Se trata de un estudio radiológico realizado con 23 niños con MTA idiopático o secundario a pie zambo, operados entre los años 1982 y 2000, con liberación de la articulación cuneo-metatarsiana. Se realizaron radiografías de los 30 pies operados, y de 12 pies contralaterales utilizados como controles, con las siguientes mediciones: ángulo cuneo-metatarsiano (FMCA), ángulo de inclinación distal de la primera cuña (DCAA), ángulo entre esta superficie y la del primer metatarsiano (PENTE) y ángulo proximal articular del primer metatarsiano (PMAA). Se compararon las medias de estos ángulos tomadas en pre, post-operatorio inmediato y final del seguimiento, mediante los pertinentes tests estadísticos. Resultados. Respecto al MTA idiopático, el FMCA aumentó de 150,4° en el preoperatorio a 170,2° al final, el DCAA pasó de 62° a 81,1°, y el PMAA de 88,4° a 89,1°. Respecto al MTA secundario a pie zambo, el FMCA aumentó de 155,3° a 169,7°, el DCAA pasó de 61,9° a 79,7°, y el PMAA de 88,3° a 90°. En el lado sano los ángulos apenas se modificaron. Conclusiones. La oblicuidad de la articulación cuneo-metatarsiana medial está estrechamente relacionada con el metatarso adducto. La corrección post-quirúrgica se hace también a expensas de este hueso, que tiende a rellenar el espacio creado por la capsulotomía

Purpose. Metatarsus adductus is an adduction deformity of the forefoot. Our purpose is to uphold the claim that the main anomaly is to be found on the cuneiform side of the cuneo-metatarsal joint and that the growth of the medial cuneiform is fundamental for correction further to capsulotomy. Materials and methods. This is a radiological study of 23 children with idiopathic or clubfoot-related metatarsus adductus subjected between 1982 and 2000 to a release of the cuneiform metatarsal joint. X-rays were taken of the 30 operated feet and of 12 contralateral feet used as controls. The following measurements were made: cuneiform-metatarsal angle, distal inclination angle of the medial cuneiform bone, angle formed between the latter surface and the first metatarsal and the proximal articular angle of the first metatarsal. The mean pre-op, immediate post-op and post-op final values of these angles were compared using the relevant statistical tests. Results. As regards idiopathic metatarsus adductus, the cuneiform metatarsal angle went from a preop value of 150.4° to 170.2° at the end of follow-up; the distal inclination angle of the medial cuneiform went from 62° to 81.1°; and the angle formed between the latter surface and the first metatarsal and the proximal articular angle of the first metatarsal went from 88.4° to 89.1°. With respect to clubfoot-related metatarsus adductus, the cuneiform metatarsal angle went from a preop value of 155.3° to 169.7°, the distal inclination angle of the medial cuneiform went from 61.9° to 79.7°, and the angle formed between the latter surface and the first metatarsal and the proximal articular angle of the first metatarsal went from 88.3° to 90°. On the healthy side, the angle values showed virtually no changes. Conclusions. The obliqueness of the medial cuneiform-metatarsal joint is closely related to metatarsus adductus. Post-surgical correction also takes place at the expense of this bone, which tends to fill the space created by the capsulotomy (AU)

[When doors slam, fingers jam!]. / Quand les portes claquent, les doigts craquent!

AIM: Epidemiological analysis in a universitary paediatric emergency unit of children admitted after accidental injuries resulting from fingers crushed in a door. MATERIAL AND METHODS: Prospective, descriptive cohort study from September 6th, 2004 to July 1st, 2005 included all children admitted for finger injuries crushed in a non-automatic door. EXCLUSIONS: included accidents due to automatic doors, toy's or refrigerator doors, families who refused to participate to the study or families who had left the waiting area before medical examination. Collected data were patient and family characteristics, accident characteristics and its management. RESULTS: Three hundred and forty children affected by 427 digital lesions were included. The mean age was 5.5+/-3.8 years (range 4 months - 15.5 years). Male/female ratio was equal to 1.2: 1. Fifty-eight percent of patients belonged to families composed of 3 or more siblings. Ninety-three per cent of families came to hospital within the first 2 hours after the accident (mean delay 99+/-162 min, median range 54 minutes). Location of the accident was: domestic (62%, at home (64%)), at school (17%). Locations within the home were: the bedroom (33%), bathroom and toilets (21%). An adult was present in 75% of cases and responsible for the trauma in 25% of accidents, another child in 44%. The finger or fingers were trapped on the hinge side in 57% of patients. No specific safeguard devices were used by 94% of families. Among victims, 20% had several crushed digits; left and right hand were injured with an equal frequency. The commonest involved digits were: the middle finger (29%), the ring finger (23%). The nail plate was damaged in 60% of digital lesions, associated with a wound (50%), a distal phalanx fracture (P3) (12%). Six children had a partial or complete amputation of P3, 2 children a lesion of the extensor tendon, 1 child had a rupture of the external lateral ligament. Three percent of children required an admission to the paediatric orthopaedic surgery unit. Post-traumatic pain was mainly limited to the first 48 h (64%). Early complications included: 16 cases of infected injuries, 3 cases of pulpar necrosis. The total cost of hospital care was 71,500 euros, the average cost for hospitalised patient equal to 2100 euros and for ambulatory cases equal to 141 euros; the annual cost was estimated at 81,600 euros. CONCLUSION: Associated with potentially serious digital injuries, functional or inesthetic sequelae, this painful experience still remains too frequent in toddlers for a home accident that could be often prevented by the acquisition of specific protective doors devices and for a reasonable cost compared to the cost of hospital care.

Long-term effects on tibial growth after intraosseous infusion: a prospective, radiographic analysis.

BACKGROUND: Evaluate, by radiographic analysis, tibial growth after an intraosseous infusion (IOI) in a pediatric population. METHODS: We performed a prospective simple blind study, between January 1, 1994, and July 1, 2001, which included pediatric patients who needed an intraosseous trocar in emergency situations. During the follow-up, roentgenographs were performed. On each radiologic view, different measurements were carried out: anterior and lateral tibial length, anterior and lateral width at 2 diaphyseal levels. We compared the anterior length values to those published in the Anderson et al tables. When only one tibia was punctured, the mean measurements were compared with the control leg measurements using a paired t test. RESULTS: The initial population included 78 patients. Of these 78 subjects, 42 died, 10 families could not be contacted, and one refused to participate. Two children were excluded because they had other conditions that could influence tibial growth. The study included 23 children. The puncture site was the proximal tibia. The mean age was 18.6 months at the time of IOI, the mean time of infusion was 5 hours, and the mean perfused volume was 225 mL. The mean radiologic follow-up time was 29.2 months. When compared with the Anderson et al tables, all the anterior length values were within the 95% confidence interval. For the other measurements, the statistical analysis showed no significant difference between punctured and control legs. CONCLUSION: There is no long-term effect on tibial growth after an IOI when the IO trocar is properly placed.

Does the meniscus exist in the elbow joint in children?

The authors report a case of meniscus at the elbow joint in a 15-month-old infant causing a limitation of elbow extension. Histological examination demonstrated that this tissue was not a synovial fold or a chondroid metaplasia of the synovial fold. As a meniscus does not appear at any stage of the embryological evolution of the elbow joint, it has been concluded that the presence of the meniscus can be considered as an abnormal condition.

Obstetrical brachial plexus palsy. Prediction of outcome in upper root injuries.

Thirty obstetrical brachial plexus palsies involving the upper roots were retrospectively reviewed. There were 20 C5-C6 palsies and ten C5-C6-C7 palsies in which recovery of C7 occurred by the end of the first month. Recovery of elbow flexion at 3 months, C7 involvement and high birthweight were the best early predictors of outcome, but all were unreliable when used separately. In combination, recovery of elbow flexion and birthweight predicted the final outcome reasonably satisfactorily, particularly when elbow flexion at 9 months, and not 3 months was considered (risk of error = 13%). Brachial plexus reconstruction may therefore be justified when there was initial C7 involvement associated with increased birthweight and poor elbow flexion at 6-9 months.

[Imaging of the hip in healthy children in MRI: evaluation of the acetabular coverage]. / Imagerie de la hanche de l'enfant sain en IRM: évaluation de la couverture cotyloïdienne.

PURPOSE: To define with MR imaging a reference value for acetabular coverage on coronal and sagittal images for a population of healthy children. MATERIALS AND METHODS: 36 children (72 hips) where prospectively studied over 1 year. T1-weighted spin-echo images in both coronal and sagittal planes were obtained. Slice thickness was 4 mm. Measurement of acetabular coverage (AHI: Acetabular-Head Index) was made in both planes with two different landmarks: bone and cartilage. Study of the evolution with age was made. RESULTS: The acetabular coverage is symmetrical and decreases with age. Its measurement is reproducible. CONCLUSION: This study shows that measurement of acetabular coverage (AHI) is easily obtained at MR imaging and is reproducible. We suggest 83 (mean 1SD) as the inferior limit for AHI based on cartilage measurements. We introduce sagittal values, never published. We believe that MR, by its ability to demonstrate the cartilaginous surfaces, will play a major role in the true understanding of the anatomical relationship between the femoral head and the acetabulum.

[Scarf osteotomy of hallux valgus in children and adolescents]. / Traitement de l'hallux valgus de l'enfant et de l'adolescent par ostéotomie scarf.

PURPOSE OF THE STUDY: The aim of this study was to analyze the results of surgical treatment of hallux valgus using scarf osteotomy in children and adolescents. MATERIAL AND METHOD: Twelve children and adolescents (19 feet) operated on with scarf osteotomy were reviewed retrospectively with a mean follow-up of 15 months. Clinical and radiographic results were assessed. RESULTS: We obtained 10 good results (asymptomatic cases), 9 poor with residual symptoms such as pain or cosmetic problems. Two populations could be identified considering the metatarsus varus and distal metatarsal angle. Those with metatarsus varus >/= 10 degrees or normal distal metatarsal angle had good results. Those without metatarsus varus had poor results. There was no disturbance of growth. DISCUSSION: Scarf osteotomy can be used in children without risk of hindering growth. Nevertheless, good results can be achieved only in cases of true metatarsus varus and normal distal metatarsal angle. In other cases, a different osteotomy of the first metatarsal would be preferable.

[Recurrent aggressive chondroblastoma: two cases and a review of the literature]. / Récidive de chondroblastome agressif.

Chondroblastoma is a rare benign bone tumor. Treatment currently consists of curettage and bone graft. Prognosis depends basically on a relatively high rate of recurrence, between 5 and 38%, sometimes with local seeding of soft tissues and joint space. Such recurrences require wide resection with arthrodesis or even amputation. A 13-year-old girl with a humeral head chondroblastoma as treated by curettage and iliac bone graft. Six months later, a recurrence occurred with extension into the rotator cuff and the metaphysis. Remission was achieved by extensive surgical resection and hemiarthroplasty. The second case was a 14-year-old boy with a chondroblastoma of the right talus. He was treated by curettage and packing with bone substitute. After 2 recurrences with soft tissue and intra-articular extension, we performed a wide resection with reconstruction using a vascularized fibular graft. Many other cases in the literature illustrate such complications. We tried to find factors predictive of recurrence. Recurrence is observed when curettage was incomplete or when tumor cells were disseminated during surgery. Chondroblastoma is a benign bone tumor, but prognosis depends on the rapidity and severity of recurrence. Curettage should be as complete as possible and care should be taken to avoid contaminating the operative field.

Repair of lumbar spondylolysis using Morscher material: 14 children followed for 1-5 years.

We evaluated the results of spondylolysis repair in children and adolescents with the Morscher system. 14 patients (8 girls) with symptomatic spondylolysis unresponsive to closed treatment were operated on. Mean age at operation was 12 (7-15) years. The average follow-up was 33 (16-66) months. Results were assessed clinically according to the Henderson classification and radiographically. Clinical results were excellent in 9 patients, good in 4 and poor in 1. Radiographs showed that fusion was obtained in 12 patients, it was doubtful in 1 and 1 patient had non-union. Loosening of the screw on one or both sides was noted in 8 patients and necessitated removal of the osteosynthesis material which, however, did not alter the final outcome. The satisfactory results obtained in our series are similar to those reported in the literature.

Accuracy and precision of in vivo bone mineral measurements in sheep using dual-energy X-ray absorptiometry.

We evaluated the precision and accuracy of in vivo measurements of spine bone mineral density (BMD) and bone mineral content (BMC) in five ewes using dual-energy X-ray absorptiometry (DXA, Lunar DPX-L). The short-term in vivo reproducibility expressed as the coefficient of variation (CV) varied from 0.9 to 1.6% for spine BMD and from 1 to 3.1% for spine BMC. The ex vivo measurements, performed in 20 cm of water to simulate soft tissue thickness, correlated closely with the in vivo measurements, yielding an r value of 0.98 and 0.97 for spine BMD and BMC, respectively. The accuracy was determined by comparing the total BMC of each vertebra measured in vivo with the corresponding ash weight. The correlation coefficient between the two measurements was r = 0.98, with an accuracy error of 5.6%. We concluded that the DXA allows a precise and accurate measurement of spine bone mineral in live ewes using the methodology designed for humans.

Congenital pseudarthrosis of the clavicle and thoracic outlet syndrome in adolescence.

A 15-year-old girl with thoracic outlet syndrome associated with congenital pseudarthrosis of the clavicle was examined. The indication for treatment of congenital pseudarthrosis of the clavicle is discussed.

[Somatosensory evoked potentials during surgery of scoliosis: significance of epidural recording]. / Potentiels évoqués somesthésiques au cours de la chirurgie de la scoliose: intérêt du recueil épidural.

PURPOSE OF THE STUDY: To evaluate the usefulness of the epidural recording in scoliosis surgery in children or young adults. MATERIAL AND METHOD: We used somatosensory-evoked potentials in 60 children or young adults in surgical treatment of scoliosis. Somatosensory-evoked potentials were elicited and recorded using a standard method recording of cortical P40 and subcortical P27, simultaneously with the monitoring of spinal cord function using epidural electrode to record the ascending somatosensory volley (by tibial nerve stimulation) at a high thoracic level. RESULTS: The use of subcortical and epidural recording sites for the somatosensory-evoked potentials indicated that a reliable response could be obtained in 96 p. 100 of the patients. Cortical somatosensory-evoked demonstrated a 48 p. 100 false positive rate. DISCUSSION: Epidural recording in scoliosis surgery is safe and reliable. Combined subcortical and epidural recordings are especially useful in patients with neuromuscular scoliosis or with cerebral palsy that may not have reliable cortical potentials.