Aims. To describe the clinical, radiological, and functional outcomes in patients with isolated congenital thoracolumbar kyphosis who were treated with three-column osteotomy by posterior-only approach. Methods. Hospital records of 27 patients with isolated congenital thoracolumbar kyphosis undergoing surgery at a single centre were retrospectively analyzed. All patients underwent deformity correction which involved a three-column osteotomy by single-stage posterior-only approach. Radiological parameters (local kyphosis angle (KA), thoracic kyphosis (TK), lumbar lordosis (LL), pelvic tilt (PT), sacral slope (SS), C7 sagittal vertical axis (C7 SVA), T1 slope, and pelvic incidence minus lumbar lordosis (PI-LL)), functional scores, and clinical details of complications were recorded. Results. The mean age of the study population was 13.9 years (SD 6.4). The apex of deformity was in thoracic, thoracolumbar, and lumbar spine in five, 14, and eight patients, respectively. The mean operating time was 178.4 minutes (SD 38.5) and the mean operative blood loss was 701.8 ml (SD 194.4). KA (preoperative mean 70.8° (SD 21.6°) vs final follow-up mean 24.7° (SD 18.9°); p < 0.001) and TK (preoperative mean -1.48° (SD 41.23°) vs final follow-up mean 24.28° (SD 17.29°); p = 0.005) underwent a significant change with surgery. Mean Scoliosis Research Society (SRS-22r) score improved after surgical correction (preoperative mean 3.24 (SD 0.37) vs final follow-up mean 4.28 (SD 0.47); p < 0.001) with maximum improvement in self-image and mental health domains. The overall complication rate was 26%, including two neurological and five non-neurological complications. Permanent neurological deficit was noted in one patient. Conclusion. Deformity correction employing three-column osteotomies by a single-stage posterior-only approach is safe and effective in treating isolated congenital thoracolumbar kyphosis. Cite this article: Bone Joint J 2021;103-B(7):1309–1316

Aims. In order to determine whether and for whom serial radiological evaluation is necessary in one-part proximal humerus fractures, we set out to describe the clinical history and predictors of secondary displacement in patients sustaining these injuries. Patients and Methods. Between January 2014 and April 2016, all patients with an isolated, nonoperatively treated one-part proximal humerus fracture were prospectively followed up. Clinical and radiological evaluation took place at less than two, six, 12, and 52 weeks. Fracture configuration, bone quality, and comminution were determined on the initial radiographs. Fracture healing, secondary displacement, and treatment changes were recorded during follow-up. Results. In 100 patients (59 female, 41 male; mean age 57 years), 91 of the fractures (91%) remained stable. In five of nine patients (55%) with secondary displacement, surgery was recommended. Comminution, present in 23 patients (23%), was identified as a predictor of secondary displacement (p < 0.001). Patients’ age, sex, fracture configuration, and bone quality were not associated with secondary displacement (p ≥ 0.438). Nonoperative treatment resulted in a mean absolute Constant score (CS) of 80 (49 to 98), relative CS of 101% (63% to 138%), median subjective shoulder value of 95% (interquartile range (IQR) 90% to 100%), and median EuroQol five-dimensional questionnaire score of 0.89 (IQR 0.80 to 1.00) with bone union in all cases at one-year follow-up. Conclusion. Radiological re-evaluation was only necessary in patients presenting with comminution and may be redundant for 77% of patients with one-part proximal humerus fractures. Nonoperative treatment of one-part proximal humerus fractures remains the mainstay of treatment with a low rate of secondary surgery, a high union rate, and good clinical results. Cite this article: Bone Joint J 2019;101-B:1307–1312

Aims. To clarify the effectiveness of the induced membrane technique (IMT) using beta-tricalcium phosphate (β-TCP) for reconstruction of segmental bone defects by evaluating clinical and radiological outcomes, and the effect of defect size and operated site on surgical outcomes. Methods. A review of the medical records was conducted of consecutive 35 lower limbs (30 males and five females; median age 46 years (interquartile range (IQR) 40 to 61)) treated with IMT using β-TCP between 2014 and 2018. Lower Extremity Functional Score (LEFS) was examined preoperatively and at final follow-up to clarify patient-centered outcomes. Bone healing was assessed radiologically, and time from the second stage to bone healing was also evaluated. Patients were divided into ≥ 50 mm and < 50 mm defect groups and into femoral reconstruction, tibial reconstruction, and ankle arthrodesis groups. Results. There were ten and 25 defects in the femur and tibia, respectively. Median LEFS improved significantly from 8 (IQR 1.5 to 19.3) preoperatively to 63.5 (IQR 57 to 73.3) at final follow-up (p < 0.001). Bone healing was achieved in all limbs, and median time from the second stage to bone healing was six months (IQR 5 to 10). Median time to bone healing, preoperative LEFS, or postoperative LEFS did not differ significantly between the defect size groups or among the treatment groups. Conclusion. IMT using β-TCP provided satisfactory clinical and radiological outcomes for segmental bone defects in the lower limbs; surgical outcomes were not influenced by bone defect size or operated part. Cite this article: Bone Joint J 2021;103-B(3):456–461

Aims. Although knee osteoarthritis (OA) is diagnosed and monitored radiologically, actual full-thickness cartilage loss (FTCL) has rarely been correlated with radiological classification. This study aims to analyze which classification system correlates best with FTCL and to assess their reliability. Methods. A prospective study of 300 consecutive patients undergoing unilateral total knee arthroplasty (TKA) for OA (mean age 69 years (44 to 91; standard deviation (SD) 9.5), 178 (59%) female). Two blinded examiners independently graded preoperative radiographs using five common systems: Kellgren-Lawrence (KL); International Knee Documentation Committee (IKDC); Fairbank; Brandt; and Ahlbäck. Interobserver agreement was assessed using the intraclass correlation coefficient (ICC). Intraoperatively, anterior cruciate ligament (ACL) status and the presence of FTCL in 16 regions of interest were recorded. Radiological classification and FTCL were correlated using the Spearman correlation coefficient. Results. Knees had a mean of 6.8 regions of FTCL (SD 3.1), most common medially. The commonest patterns of FTCL were medial ± patellofemoral (143/300, 48%) and tricompartmental (89/300, 30%). ACL status was associated with pattern of FTCL (p = 0.023). All radiological classification systems demonstrated moderate ICC, but this was highest for the IKDC: whole knee 0.68 (95% confidence interval (CI) 0.60 to 0.74); medial compartment 0.84 (95% CI 0.80 to 0.87); and lateral compartment 0.79 (95% CI 0.73 to 0.83). Correlation with actual FTCL was strongest for Ahlbäck (Spearman rho 0.27 to 0.39) and KL (0.30 to 0.33) systems, although all systems demonstrated medium correlation. The Ahlbäck score was the most discriminating in severe knee OA. Osteophyte presence in the medial compartment had high positive predictive value (PPV) for FTCL, but not in the lateral compartment. Conclusion. The Ahlbäck and KL systems had the highest correlation with confirmed cartilage loss at TKA. However, the IKDC system displayed the best interobserver reliability, with favourable correlation with FTCL in medial and lateral compartments, although it was less discriminating in more severe disease. Cite this article: Bone Joint J 2020;102-B(3):301–309

Aims. Shoulder arthroplasty using short humeral components is becoming increasingly popular. Some such components have been associated with relatively high rates of adverse radiological findings. The aim of this retrospective review was to evaluate the radiological humeral bone changes and mechanical failure rates with implantation of a short cementless humeral component in anatomical (TSA) and reverse shoulder arthroplasty (RSA). Patients and Methods. A total of 100 shoulder arthroplasties (35 TSA and 65 RSA) were evaluated at a mean of 3.8 years (3 to 8.3). The mean age at the time of surgery was 68 years (31 to 90). The mean body mass index was 32.7 kg/m. 2. (17.3 to 66.4). Results. Greater tuberosity stress shielding was noted in 14 shoulders (two TSA and 12 RSA) and was graded as mild in nine, moderate in two, and severe in three. Medial calcar resorption was noted in 23 shoulders (seven TSA and 16 RSA), and was graded as mild in 21 and moderate in two. No humeral components were revised for loosening or considered to be loose radiologically. Nine shoulders underwent reoperation for infection (n = 3), fracture of the humeral tray (n = 2), aseptic glenoid loosening (n = 1), and instability (n = 3). No periprosthetic fractures occurred. Conclusion. Implantation of this particular short cementless humeral component at the time of TSA or RSA was associated with a low rate of adverse radiological findings on the humeral side at mid-term follow-up. Our data do not raise any concerns regarding the use of a short stem in TSA or RSA. Cite this article: Bone Joint J 2019;101-B:610–614

Aims

The preoperative grading of chondrosarcomas of bone that accurately predicts surgical management is difficult for surgeons, radiologists, and pathologists. There are often discrepancies in grade between the initial biopsy and the final histology. Recent advances in the use of imaging methods have shown promise in the ability to predict the final grade. The most important clinical distinction is between grade 1 chondrosarcomas, which are amenable to curettage, and resection-grade chondrosarcomas (grade 2 and 3) which require en bloc resection. The aim of this study was to evaluate the use of a Radiological Aggressiveness Score (RAS) to predict the grade of primary chondrosarcomas in long bones and thus to guide management.

Aims. The diversity of femoral morphology renders femoral component sizing in total hip arthroplasty (THA) challenging. We aimed to determine whether femoral morphology and femoral component filling influence early clinical and radiological outcomes following THA using fully hydroxyapatite (HA)-coated femoral components. Methods. We retrospectively reviewed records of 183 primary uncemented THAs. Femoral morphology, including Dorr classification, canal bone ratio (CBR), canal flare index (CFI), and canal-calcar ratio (CCR), were calculated on preoperative radiographs. The canal fill ratio (CFR) was calculated at different levels relative to the lesser trochanter (LT) using immediate postoperative radiographs: P1, 2 cm above LT; P2, at LT; P3, 2 cm below LT; and D1, 7 cm below LT. At two years, radiological femoral component osseointegration was evaluated using the Engh score, and hip function using the Postel Merle d’Aubigné (PMA) and Oxford Hip Score (OHS). Results. CFR was moderately correlated with CCR at P1 (r = 0.44; p < 0.001), P2 (r = 0.53; p < 0.001), and CFI at P1 (r = − 0.56; p < 0.001). Absence of spot welds (n = 3, 2%) was associated with lower CCR (p = 0.049), greater CFI (p = 0.017), and lower CFR at P3 (p = 0.015). Migration (n = 9, 7%) was associated with lower CFR at P2 (p = 0.028) and P3 (p = 0.007). Varus malalignment (n = 7, 5%), predominantly in Dorr A femurs (p = 0.028), was associated with lower CFR at all levels (p < 0.05). Absence of spot welds was associated with lower PMA gait (p = 0.012) and migration with worse OHS (p = 0.032). Conclusion. This study revealed that femurs with insufficient proximal filling tend to have less favourable radiological outcomes following uncemented THA using a fully HA-coated double-tapered femoral component. Cite this article: Bone Joint Res. 2020;9(4):182–191

Aims

This systematic review aims to identify 3D predictors derived from biplanar reconstruction, and to describe current methods for improving curve prediction in patients with mild adolescent idiopathic scoliosis.

Aim. Nail patella syndrome (NPS) is a skeletal dysplasia with patellofemoral dysfunction as a key symptom. We present the first in-depth radiological evaluation of the knee in a large series of NPS patients and describe the typical malformations. Patients and Methods. Conventional radiological examination of 95 skeletally mature patients with NPS was performed. Patellar morphology was classified according to the Wiberg classification as modified by Baumgartl and Ficat criteria, and trochlear shape was classified according to the Dejour classification. Results. Patellar aplasia was present in 4/90 (4%), and patellar hypoplasia in 77/90 (86%) of patients. The prevailing patellar shapes were type III, type IV and Hunter’s cap. No patellar shape genotype-phenotype association could be found. The malformations of the distal femur comprised shortening of the lateral femoral condyle in 46 out of 84 patients (55%), with a prominent anterior surface of the lateral femoral condyle in 47 out of 84 patients (56%) and a flat anterior surface of the medial femoral condyle in 78 out of 85 patients (92%). The trochlea was type A1 according to the Dejour classification in 79 out of 85 patients (93%). . Conclusion. An easily recognisable characteristic quartet of malformations consisting of patellar aplasiaor hypoplasia and the malformations of the distal femur was found in 22 out of 81 patients (27%), with the majority displaying at least three malformations. Take home message: The distinct malformations of the knee in nail patella syndrome are easily recognisable on conventional radiographs and lead to the correct interpretation of the aberrant morphology which is essential in the treatment of these patellofemoral disorders. Cite this article: Bone Joint J 2016;98-B:483–9

Aims. We conducted a study to determine whether radiological parameters correlate with patient reported functional outcome, health-related quality of life and physical measures of function in patients with a fracture of the distal radius. Patients and Methods. The post-operative palmar tilt and ulnar variance at six weeks and 12 months were correlated with the Patient Rated Wrist Evaluation, Disabilities of the Arm, Shoulder and Hand, and EuroQol scores, grip strength, pinch strength and range of movement at three, six and 12 months for 50 patients (mean age 57 years; 26 to 85) having surgical fixation, with either percutaneous pinning or reconstruction with a volar plate, for a fracture of the distal radius. Results. Radiological parameters were found to correlate poorly with the patient reported outcomes (r = 0.00 to 0.47) and physical measures of function (r = 0.01 to 0.51) at all intervals. Conclusion. This study raises concerns about the use of radiological parameters to determine management, and to act as a surrogates for successful treatment, in patients with a fracture of the distal radius. Restoration of ‘normal’ radiographic parameters may not be necessary to achieve a satisfactory functional outcome for the patient. Cite this article: Bone Joint J 2017;99-B:376–82

Digital templating in hip replacement is commonly performed with radiological markers to determine the magnification. The latter can also be determined by measuring the distance from the x-ray focal spot to the object and the distance from the x-ray focal spot to the radiological cassette or image receptor. We used post-operative radiographs of total hip replacements and hemiarthroplasties from 22 patients to calculate the magnification using both methods. The accuracy of each method was ascertained by measuring the size of the head of the implant projected on to the radiographs and comparing the result with the known size recorded in the medical records. The accuracy was found to be similar with a mean absolute measurement error of 2.6% (. sd. 1.4; 0.0% to 5.2%) for the radiological marker and 2.8% (. sd. 2.2; 0.4% to 10.1%) for the distance method (p = 0.75). The mean radiation dose for templating radiographs (pelvis and lateral of the hip) was similar when taken using a radiological marker (328 mSv . sd. 142) (n = 51) or using the distance measurement method (324 mSv . sd. 39) (n = 39) (p = 0.90). We conclude that the distance measuring method is as accurate as the radiological marker method, but may avoid some of the disadvantages such as misplacement of the marker or placement outside the radiological field. It may also be more acceptable to the patient and radiographer

The incidence of periprosthetic fractures of the ankle is increasing. However, little is known about the outcome of treatment and their management remains controversial. The aim of this study was to assess the impact of periprosthetic fractures on the functional and radiological outcome of patients with a total ankle arthroplasty (TAA). A total of 505 TAAs (488 patients) who underwent TAA were retrospectively evaluated for periprosthetic ankle fracture: these were then classified according to a recent classification which is orientated towards treatment. The outcome was evaluated clinically using the American Orthopedic Foot and Ankle Society (AOFAS) score and a visual analogue scale for pain, and radiologically. . A total of 21 patients with a periprosthetic fracture of the ankle were identified. There were 13 women and eight men. The mean age of the patients was 63 years (48 to 74). Thus, the incidence of fracture was 4.17%. . There were 11 intra-operative and ten post-operative fractures, of which eight were stress fractures and two were traumatic. The prosthesis was stable in all patients. Five stress fractures were treated conservatively and the remaining three were treated operatively. . A total of 17 patients (81%) were examined clinically and radiologically at a mean follow-up of 53.5 months (12 to 112). The mean AOFAS score at follow-up was 79.5 (21 to 100). The mean AOFAS score in those with an intra-operative fracture was 87.6 (80 to 100) and for those with a stress fracture, which were mainly because of varus malpositioning, was 67.3 (21 to 93). Periprosthetic fractures of the ankle do not necessarily adversely affect the clinical outcome, provided that a treatment algorithm is implemented with the help of a new classification system. . Cite this article: Bone Joint J 2015;97-B:950–6

We report the follow-up at 12 years of the use of the Elite Plus total hip replacement (THR). We have previously reported the results at a mean of 6.4 years. Of the 217 patients (234 THRs), 83 had died and nine had been lost to follow-up. The patients were reviewed radiologically and clinically using the Oxford hip score. Of the 234 THRs, 19 (8.1%) had required a revision by the final follow-up in all but one for aseptic loosening. Survivorship analysis for revision showed a survival of 93.9% (95% confidence interval (CI) 89.2 to 96.5) at ten years, and of 88.0% (95% CI 81.8 to 92.3) at 12 years. At the final follow-up survival analysis showed that 37% (95% CI 37.3 to 44.7) of the prostheses had either failed radiologically or had been revised. Patients with a radiologically loose femoral component had a significantly poorer Oxford hip score than those with a well-fixed component (p = 0.03). Radiological loosening at 6.4 years was predictive of failure at 12 years. Medium-term radiographs and clinical scores should be included in the surveillance of THR to give an early indication of the performance of specific implants

Specific radiological features identified by Brandser and Marsh were selected for the analysis of acetabular fractures according to the classification of Letournel and Judet. The method employs a binary approach that requires the observer to allocate each radiological feature to one of two groups. The inter- and intra-observer variances were assessed. The presence of articular displacement, marginal impaction, incongruity, intra-articular fragments and osteochondral injuries to the femoral head were analysed by a similar method. These factors were termed ‘modifiers’ and are generally considered when planning operative intervention and, critically, they may influence prognosis. Six observers independently assessed 30 sets of plain radiographs and CT scans on two separate occasions, 12 weeks apart. They were asked to determine the presence or absence of specific radiological features. This simple binary approach to classification yields an inter- and intra-observer agreement which ranges from moderate to near-perfect (κ = 0.49 to 0.88 and κ = 0.57 to 0.88, respectively). A similar approach to the modifiers yields only slight to fair inter-observer agreement (κ = 0.20 to 0.34) and slight to moderate intra-observer agreement (κ = 0 to 0.55)

We investigated the association between the radiological findings and the symptoms arising from the patellofemoral joint in advanced osteoarthritis (OA) of the knee. Four radiological features, joint space narrowing, osteophyte formation, translation of the patella and focal attrition were assessed in 151 consecutive osteoarthritic knees in 107 patients undergoing total knee replacement. The symptoms which were assessed included anterior knee pain which was scored, the ability to rise from a chair and climb stairs, and quadriceps weakness. Among the radiological features, only patellar translation and obliteration of the joint space had a statistically significant association with anterior knee pain (odds ratio (OR) 4.85; 95% confidence interval (CI) 1.83 to 12.88 and OR 11.23; 95% CI 2.44 to 51.62) respectively. Patellar translation had a statistically significant association with difficulty in rising from a chair (OR 9.06; 95% CI 1.75 to 45.11). Other radiological features, including osteophytes, joint space narrowing, and focal attrition had no significant association. Our study indicates that the radiological findings of patellar translation and significant loss of cartilage are predictive of patellofemoral symptoms and functional limitation in advanced OA of the knee

Aims. Various radiological parameters are used to evaluate a flatfoot deformity and their measurements may differ. The aims of this study were to answer the following questions: 1) Which of the 11 parameters have the best inter- and intraobserver reliability in a standardized radiological setting? 2) Are pre- and postoperative assessments equally reliable? 3) What are the identifiable sources of variation?. Patients and Methods. Measurements of the 11 parameters were recorded on anteroposterior and lateral weight-bearing radiographs of 38 feet before and after surgery for flatfoot, by three observers with different experience in foot surgery (A, ten years; B, three years; C, third-year orthopaedic resident). The inter- and intraobserver reliability was calculated. Results. Preoperative interobserver reliability was high for four, moderate for five, and low for two parameters. Postoperative interobserver reliability was high for four, moderate for five, and low for two parameters. Intraobserver reliability was excellent for all parameters preoperatively as recorded by observer A (PB) and B (MP), and for eight parameters as recorded by observer C (SR). Intraobserver reliability was excellent for ten parameters postoperatively as recorded by observer A and B, and for eight parameters as recorded by observer C. Conclusion. The following parameters can be recommended. For preoperative and postoperative evaluation of flatfoot: anteroposterior, talonavicular coverage angle; lateral, talometatarsal I angle, calcaneal pitch angle, and cuneiform-medial height (high interobserver reliability); and anteroposterior, talometatarsal II angle; lateral, talocalcaneal angle,tibiocalcaneal angle (moderate interobserver reliability). For more experienced observers, we also recommend the anteroposterior talometatarsal I angle (moderate reliability). The inter- and intraobserver reliability for most parameters were similar pre- and postoperatively. The experience of the observer and the definition and ability to measure the parameters themselves were sources of variation. Cite this article: Bone Joint J 2018;100-B:596–602

We have developed a novel method of calculating the radiological magnification of the hip using two separate radio-opaque markers. We recruited 74 patients undergoing radiological assessment following total hip replacement. Both the new double marker and a conventional single marker were used by the radiographer at the time of x-ray. The predicted magnification according to each marker was calculated, as was the true radiological magnification of the components. The correlation between true and predicted magnification was good using the double marker (r = 0.90, n = 74, p < 0.001), but only moderate for the single marker (r = 0.50, n = 63, p < 0.001). The median error was significantly less for the double marker than for the single (1.1% vs 4.8%, p < 0.001). The double marker method demonstrated excellent validity (intraclass correlation coefficient = 0.89), in contrast to the single marker (0.32). The double marker method appears to be superior to the single marker method when used in the clinical environment

Radiological assessment of total and unicompartmental knee replacement remains an essential part of routine care and follow-up. Appreciation of the various measurements that can be identified radiologically is important. It is likely that routine plain radiographs will continue to be used, although there has been a trend towards using newer technologies such as CT, especially in a failing knee, where it provides more detailed information, albeit with a higher radiation exposure. The purpose of this paper is to outline the radiological parameters used to evaluate knee replacements, describe how these are measured or classified, and review the current literature to determine their efficacy where possible

Aims. The aim of this study was to investigate the long-term clinical and radiological outcome of patients who suffer recurrent injuries to the anterior cruciate ligament (ACL) after reconstruction and require revision surgery. Patients and Methods. From a consecutive series of 200 patients who underwent primary reconstruction following rupture of the ACL, we identified 36 who sustained a further rupture, 29 of whom underwent revision surgery. Patients were reviewed prospectively at one, two, seven, 15 and about 20 years after their original surgery. Primary outcome measures were the number of further ruptures, the posterior tibial slope (PTS), and functional and radiological outcomes. These were compared with a gender and age matched cohort of patients who underwent primary ACL reconstruction only. Results. At a mean follow-up of 18.3 years (14.3 to 20.2), 29 patients had undergone revision surgery and within this revision group 11 had sustained more than three ruptures of the ACL (3 to 6). The mean age at the time of revision reconstruction was 26.4 years (14 to 54). The mean PTS was significantly higher in those patients who suffered a further injury to the ACL (11°) compared with the control group (9°) (p < 0.001). The mean PTS in those patients who sustained more than three ruptures was 12°. Conclusion. Patients who suffer recurrent injuries to the ACL after reconstruction have poorer functional and radiological outcomes than those who suffer a single injury. The causes of further injury are likely to be multifactorial but an increased PTS appears to have a significant association with recurrent ACL injuries. Cite this article: Bone Joint J 2017;99-B:337–43

Pelvic obliquity is a common finding in adolescents with cerebral palsy, however, there is little agreement on its measurement or relationship with hip development at different gross motor function classification system (GMFCS) levels. . The purpose of this investigation was to study these issues in a large, population-based cohort of adolescents with cerebral palsy at transition into adult services. . The cohort were a subset of a three year birth cohort (n = 98, 65M: 33F, with a mean age of 18.8 years (14.8 to 23.63) at their last radiological review) with the common features of a migration percentage greater than 30% and a history of adductor release surgery. . Different radiological methods of measuring pelvic obliquity were investigated in 40 patients and the angle between the acetabular tear drops (ITDL) and the horizontal reference frame of the radiograph was found to be reliable, with good face validity. This was selected for further study in all 98 patients. . The median pelvic obliquity was 4° (interquartile range 2° to 8°). There was a strong correlation between hip morphology and the presence of pelvic obliquity (effect of ITDL on Sharpe’s angle in the higher hip; rho 7.20 (5% confidence interval 5.59 to 8.81, p < 0.001). This was particularly true in non-ambulant adolescents (GMFCS IV and V) with severe pelvic obliquity, but was also easily detectable and clinically relevant in ambulant adolescents with mild pelvic obliquity. . The identification of pelvic obliquity and its management deserves closer scrutiny in children and adolescents with cerebral palsy. Cite this article: Bone Joint J 2015;97-B:1435–40