1. During two years eighty-five children with supracondylar fractures of the humerus were admitted to a children's hospital. This paper is a study of the severe fractures which occurred in twenty-three children of whom fifteen had a deformity at review.

2. The only deformity found after treatment by closed manipulation was an alteration in the carrying angle. It is considered that cubitus varus was caused by medial angulation of the distal fragment.

We followed, prospectively, 77 patients (90 hips) with early-stage, non-traumatic osteonecrosis of the femoral head for a mean period of nine years. At the time of final review, 56 (62%) were symptomatic. Using the Cox model, the initial radiological stage, the progression of staging, the reduction in size of the lesions and the percentage of necrotic volume on MRI were identified as risk factors.

Using the Harris score of < 70 or surgery as the endpoint, the cumulative rates of survival were 60.0% at one year, 43.3% at two years, 38.9% at five years and 37.2% at ten years. Survival curves demonstrated that clinical deterioration could occur 90 months after the initial diagnosis, suggesting that asymptomatic patients should be followed carefully for several years. Radiological time-dependent reduction in size without progressive collapse may represent repair even when the collapse is minimal.

We report a postal survey of 59 families of children with osteogenesis imperfecta. From the 51 replies we collected data on developmental milestones and walking ability and related them to the Sillence and the Shapiro classifications of osteogenesis imperfecta. Twenty-four of the patients had been treated by intramedullary rodding.

Both classifications helped to predict eventual walking ability. We found that independent sitting by the age of ten months was a predictor for the use of walking as the main means of mobility with 76% attaining this. Of the patients who did not achieve sitting by ten months, walking became the main means of mobility in only 18%. The developmental pattern of mobility was similar in the rodded and non-rodded patients.

One hundred and twenty-nine unselected club feet were classified at birth into three grades of severity; 123 were followed up. The results of primary treatment were analysed and it is shown that the bad feet did worst. Serial splinting in plasters achieved lasting correction in nine in ten mild club feet, in half of the moderately deformed, but in only one in ten of the severely affected. Surgical correction succeeded in two out of three of the resistant feet, but had to be repeated in the others.

Between March 1969 and May 1978, 36 babies sustained birth injuries of the brachial plexus at the National Women's Hospital, Auckland, New Zealand. This gives an incidence of 0.87 per 1000 live births. Nearly 80 per cent of these children had made a complete recovery by the age of 13 months, while none of those with significant residual defects has severe sensory or motor deficit of the hand.

In recent years, machine learning (ML) and artificial neural networks (ANNs), a particular subset of ML, have been adopted by various areas of healthcare. A number of diagnostic and prognostic algorithms have been designed and implemented across a range of orthopaedic sub-specialties to date, with many positive results. However, the methodology of many of these studies is flawed, and few compare the use of ML with the current approach in clinical practice. Spinal surgery has advanced rapidly over the past three decades, particularly in the areas of implant technology, advanced surgical techniques, biologics, and enhanced recovery protocols. It is therefore regarded an innovative field. Inevitably, spinal surgeons will wish to incorporate ML into their practice should models prove effective in diagnostic or prognostic terms. The purpose of this article is to review published studies that describe the application of neural networks to spinal surgery and which actively compare ANN models to contemporary clinical standards allowing evaluation of their efficacy, accuracy, and relatability. It also explores some of the limitations of the technology, which act to constrain the widespread adoption of neural networks for diagnostic and prognostic use in spinal care. Finally, it describes the necessary considerations should institutions wish to incorporate ANNs into their practices. In doing so, the aim of this review is to provide a practical approach for spinal surgeons to understand the relevant aspects of neural networks.

Cite this article: Bone Joint J 2021;103-B(9):1442–1448.

We have studied 105 patients with 107 acute, primary, dislocations of the glenohumeral joint seen between January 1, 1991 and July 1, 1994. The mean time of follow-up was 71 months (46 to 91). In 34% the injury occurred during a sports activity and in 28% at home. The bias toward sport was even greater in patients less than 40 years of age, and in men. In patients older than 40 years of age, and in women, the dislocation occurred more often at home.

The overall probability of recurrence within four years was 26%. Age was the most significant prognostic factor in recurrence which took place in 64% of patients less than 20 years of age and in 6% of those older than 40 years. Statistically, there was no difference between the rates of recurrence in patients who were active in sport and those who were not. The mean Rowe score for the whole group was 87 (15 to 100). Associated fractures were found in 20 patients (19%) and nerve injuries in 22 (21%). None of those in whom a fracture of the greater tuberosity was seen subsequently suffered a recurrent dislocation.

At follow-up we found that 36 patients (34%) had not returned to their former employment but in only 2% was this owing to the injured shoulder.

We have investigated the significance of local recurrence on survival in 173 patients with localised soft-tissue sarcomas of the limbs and of the trunk.

The overall survival rates at five and ten years were 75.2% and 68.0%, respectively. After definitive surgery at our hospitals, there was local recurrence in 25 patients (14.5%). After inadequate operations elsewhere, there was a higher incidence of late local recurrence (28.3%), in comparison with those with primary tumours treated by us (9.0%), or patients referred to us immediately after inadequate surgery elsewhere (10.2%). Because of small numbers these differences in the survival rates were not statistically significantly different.

Univariate survival analysis showed that local recurrence after definitive surgery (p = 0.006) together with the histological grade (p = 0.0002), the size of the tumour (p = 0.002), its depth in relation to deep fascia (p = 0.003), and the surgical margin (p = 0.0001) were the significant prognostic factors. Local recurrence at the initial presentation did not affect survival. Multivariate analysis showed that local recurrence after definitive surgery also lost its apparent prognostic significance.

1. A family study of sixty-one children with infantile idiopathic scoliosis to determine the relationship of other developmental anomalies to the behaviour of the curve is reported.

2. Of the thirty-nine children with resolving curves only one had another defect.

3. Twelve out of twenty-two children with non-resolving curves had at least one other developmental anomaly.

4. The presence of such defects may indicate that the curve is likely to progress.

We have studied the morphometry of the spinal cord in 50 patients with cervical compression myelopathy. Computed tomographic myelography (CTM) showed that the transverse area of the cord at the site of maximum compression correlated significantly with the results of surgery. In most patients with less than 30 mm2 of spinal cord area, the results were poor; the cord was unable to survive. Several factors, such as chronicity of disease, age at surgery and multiplicity of involvement are said to influence the results of surgery, but the transverse area of the cord at the level of maximum compression provides the most reliable and comprehensive parameter for their prediction.

Aims

To assess the correlation between the histological response to preoperative chemotherapy and event-free survival (EFS) or overall survival (OS) in patients with high-grade localized osteosarcoma.

Aims

The early mortality in patients with hip fractures from bony metastases is unknown. The objectives of this study were to quantify 30- and 90-day mortality in patients with proximal femoral metastases, and to create a mortality prediction tool based on biomarkers associated with early death.

We studied the risk of recurrent dislocation in 121 primary and 39 revision Charnley or Charnley hybrid total hip arthroplasties which had been treated for a primary dislocation between 1979 and 1995. Only 35% of these hips had no further dislocation or a revision for instability within one year. The rates of survival gradually declined with time or if a second, third or fourth dislocation occurred. The risk of recurrence was greater in men, but was not related to age, diagnosis, time of the first dislocation or whether the index operation had been a primary or a revision procedure.

Operative treatment included 15 reoperations leaving intact components, 50 revisions, and permanent removal of the femoral stem in seven patients. The operation was successful in four patients with reoperations and in 36 who had an exchange procedure within two years. Treatment was successful in 35 of 49 hips in which it was possible to correct a technical error compared with 5 out of 16 hips in which malposition of the components was not seen (p = 0.007).

Aims

The aim of this study was to determine the risk of local recurrence and survival in patients with osteosarcoma based on the proximity of the tumour to the major vessels.

We reviewed 26 fractures involving the distal physis of the tibia to identify the patterns of formation and displacement of the subsequent growth disturbance lines. Twenty-one patients showed a regular "normal" pattern of line and healed with no deformity. Three patients had medical physeal arrest revealed by abnormal lines. Two other cases had a minor central physeal arrest without subsequent deformity. The pattern and character of the growth disturbance line can provide an early warning of potential deformity.

Aims

The purpose of this study was to identify prognostic indicators of outcome at presentation to the orthopaedic surgeon, in patients with metastatic prostate cancer. Our aim was to use this information in a pragmatic, clinic-based approach so that surgical decision making could be optimized to benefit the patient in their remaining lifetime.