Aims. Using 90% of final height as a benchmark, we sought to develop a quick, quantitative and reproducible method of estimating skeletal maturity based on topographical changes in the distal femoral physis. Patients and Methods. Serial radiographs of the distal femoral physis three years prior to, during, and two years following the chronological age associated with 90% of final height were analyzed in 81 healthy children. The distance from the tip of the central peak of the distal femoral physis to a line drawn across the physis was normalized to the physeal width. Results. A total of 389 radiographs of the distal femur with corresponding Greulich and Pyle bone ages and known chronological ages were measured. Children reached 90% of final height at a mean age of 11.3 years (. sd. 0.8) for girls and 13.2 years (. sd. 0.6) for boys. Linear regression analysis showed higher correlation coefficent in predicting the true age at 90% of final height using chronological age + gender + central peak value (R. 2 . = 0.900) than chronological age + gender (R. 2.  = 0.879) and Greulich and Pyle bone age + gender (R. 2.  = 0.878). Conclusion. Chronological age + gender + central peak value provides more accurate prediction of 90% of final height compared with chronological age + gender and Greulich and Pyle bone age + gender. Cite this article: Bone Joint J 2018;100-B:1106–11

Aims. The aim of this study was to establish the incidence of developmental dysplasia of the hip (DDH) diagnosed after one-year of age in England, stratified by age, gender, year, and region of diagnosis. Patients and Methods. A descriptive observational study was performed by linking primary and secondary care information from two independent national databases of routinely collected data: the United Kingdom Clinical Practice Research Datalink and Hospital Episode Statistics. The study examined all children from 1 January 1990 to 1 January 2016 who had a new first diagnostic code for DDH aged between one and eight years old. Results. The incidence of late-diagnosed DDH was 1.28 per 1000 live births. Within the study population, 754 children were identified with a diagnosis of DDH after one-year of age. Of all late diagnoses, 536 (71.1%) were detected between one to two years of age. There were 608 female patients (80.6%) and 146 male patients (19.4%), giving a female-to-male ratio of 4.2:1. Distribution was evenly spread throughout England. Conclusion. The incidence of late-diagnosed DDH has not been reduced from that reported 35 years ago, prior to the introduction of the national selective screening programme for DDH. Cite this article: Bone Joint J 2019;101-B:281–287

Aims. Delayed diagnosis is a well-known complication of a Monteggia fracture-dislocation. If left untreated, the dislocated radial head later becomes symptomatic. The purposes of this study were firstly, to evaluate the clinical and radiological results of open reduction of the radial head and secondly, to identify the factors that may affect the outcome of this procedure. Materials and Methods. This retrospective study evaluated 30 children with a chronic Monteggia lesion. There were 18 boys and 12 girls with a mean age of 7.4 years (4 to 13) at the time of open reduction. The mean interval to surgery, after the initial fracture, was 23.4 months (6 to 120). Clinical grading used a Kim modified elbow score: radiological outcome was recorded. The effect of the patient’s age, gender, duration from initial injury, Bado classification, and annular ligament reconstruction were analyzed. The mean follow-up was 42.2 months (15 to 20). Results. The Kim elbow scores evaluated at the last clinic visit were excellent in 23 patients, good in three, fair in two, and poor in two. A majority of the patients were found to have significant improvement of elbow flexion (p < 0.001). Six met the criteria of a fair radiological outcome; four of these were operated on more than 24 months after the initial injury, and three had surgery after the age of 11. Univariate analysis failed to find any factor that was significantly associated with a fair or poor outcome. Conclusion. Good clinical and radiological outcomes can be expected in most patients. Osteoarthritic changes were associated with age > 11 years and/or a delay of treatment of > 24 months. However, no statistically significant factor could be identified which correlated with an unfavourable outcome. Cite this article: Bone Joint J 2018;100-B:1117–24

Aims. The aim of this study was to describe the epidemiology of closed isolated fractures of the femoral shaft in children, and to compare the treatment and length of stay (LOS) between major trauma centres (MTCs) and trauma units (TUs) in England. Patients and Methods. National data were obtained from the Trauma and Audit Research Network for all isolated, closed fractures of the femoral shaft in children from birth to 15 years of age, between 2012 and 2015. Age, gender, the season in which the fracture occurred, non-accidental injury, the mechanism of injury, hospital trauma status, LOS and type of treatment were recorded. Results. A total of 1852 fractures were identified. The mean annual incidence was 5.82 per 100 000 children (95% confidence interval (CI) 5.20 to 6.44). The age of peak incidence was two years for both boys and girls; this decreased with increasing age. Children aged four to six years treated in MTCs were more likely to be managed with open reduction and internal fixation compared with those treated in TUs (odds ratio 3.20; 95% CI 1.12 to 9.14; p = 0.03). The median LOS was significantly less in MTCs than in TUs for children aged between 18 months and three years treated in both a spica (p = 0.005) and traction (p = 0.0004). . Conclusion. This study highlights the current national trends in the management of closed isolated fractures of the femoral shaft in children following activation of major trauma networks in 2012. Future studies focusing on the reasons for the differences which have been identified may help to achieve more consistency in the management of these injuries across the trauma networks. Cite this article: Bone Joint J 2018;100-B:109–18

Aims. To explore the of age of onset distribution for Perthes’ disease of the hip, with particular reference to gender, laterality and conformity to the lognormal distribution. Patients and Methods. A total of 1082 patients were identified from the Liverpool Perthes’ Disease Register between 1976 and 2010, of which 992 had the date of diagnosis recorded. In total, 682 patients came from the geographical area exclusively served by Alder Hey Hospital, of which 673 had a date of diagnosis. Age of onset curves were analysed, with respect to the predefined subgroups. Results. The age of onset demonstrated a positive skew with a median of 5.8 years (interquartile range 4.6 to 7.5). Disease onset was a mean five months earlier in girls (p = 0.01) and one year earlier in those who went on to develop bilateral disease (p < 0.001). There was no difference in the age of onset between geographical districts with differing incidence rates. The entire dataset (n = 992) conformed to a lognormal distribution graphically and with the chi-squared test of normality (p = 0.10), but not using the Shapiro-Wilk test (p = 0.01). The distribution for the predefined geographical subgroup (n = 673) conformed well to a lognormal distribution (chi-squared p = 0.16, Shapiro-Wilk p = 0.08). Given the observed lognormal distribution it was assumed that Perthes’ disease followed on incubation period consistent with a point-source disease exposure. The incubation period was further examined using Hirayama’s method, which suggested that the disease exposure may act in the prenatal period. Conclusion. The age of onset in Perthes’ disease conforms to a lognormal distribution, which allows comparisons with infectious disease epidemiology. Earlier onset in girls and those who develop bilateral disease may offer clues to understanding the aetiological determinants of the disease. The analysis suggests that an antenatal aetiological determinant may be responsible for disease. Take home message: Perthes’ disease age of onset conforms to a lognormal model, which is most typical of infectious diseases. The shape of the distribution suggests that an aetiological trigger in the pre-natal period may be an important determinant of disease. Cite this article: Bone Joint J 2016;98-B:710–14

Purpose. The aim of this study was to identify if perioperative outcomes were different in patients with cerebral palsy undergoing unilateral or bilateral hip reconstruction. Method. All consecutive hip reconstructions for cerebral palsy performed by the senior author (FNT) within a tertiary-referral centre were identified between January 2012 and July 2016. Patients were stratified by age, gender, GMFCS and side of procedure. Length of surgery, pre- and post- operative haemoglobin, length of stay and immediate post-operative complications were measured. Results. 158 procedures were performed; 96 bilateral and 62 unilateral (37 left, 25 right). In bilateral procedures, operative time was 14 minutes longer (p= 0.07) and haemoglobin decrease was 1.85g/l greater (p= 0.34) with equivalent length of stay (average 6.4 days) and complication rate when compared to those undergoing unilateral procedures. Irrespective of procedure, those with a GMFCS level V had a longer length of stay (p< 0.05) and post-operative complication rate (p< 0.05) when compared to patients with GMFCS level IV. Conclusion. No significant differences were noted in the outcome measures when comparing unilateral and bilateral hip reconstructions in our group of patients with cerebral palsy. This suggests that concurrent prophylactic surgical management of the radiologically stable hip is a safe surgical option in this high-risk group of paediatric patients in an appropriate tertiary care setting with multidisciplinary team support

We present the long-term results of open surgery for internal shoulder rotational deformity in brachial plexus birth palsy (BPBP). From 1997 to 2005, 207 patients (107 females, 100 males, mean age 6.2 (0.6 to 34)) were operated on with subscapularis elongation and/or latissimus dorsi to infraspinatus transfer. Incongruent shoulder joints were relocated. The early results of these patients has been reported previously. We analysed 118 (64 females, 54 males, mean age 15.1 (7.6 to 34)) of the original patient cohort at a mean of 10.4 years (7.0 to 15.1) post-operatively. A third of patients with relocated joints had undergone secondary internal rotational osteotomy of the humerus. A mixed effects models approach was used to evaluate the effects of surgery on shoulder rotation, abduction, and the Mallet score. Independent factors were time (pre-and post-surgery), gender, age, joint category (congruent, relocated, relocated plus osteotomy) and whether or not a transfer had been performed. Data from a previously published short-term evaluation were reworked in order to obtain pre-operative values. The mean improvement in external rotation from pre-surgery to the long-term follow-up was 66.5° (95% confidence interval (CI) 61.5 to 71.6). The internal rotation had decreased by a mean of 22.6° (95% CI -18.7 to -26.5). The mean improvement in the three-grade aggregate Mallet score was 3.1 (95% CI 2.7 to 3.4), from 8.7 (95% CI 8.4 to 9.0) to 11.8 (11.5 to 12.1). Our results show that open subscapularis elongation achieves good long-term results for patients with BPBP and an internal rotation contracture, providing lasting joint congruency and resolution of the trumpet sign, but with a moderate mean loss of internal rotation. Cite this article: Bone Joint J 2014;96-B:1411–18

Supracondylar humeral fractures are common in children, but there are no classification systems or radiological parameters that predict the likelihood of having to perform an open reduction. In a retrospective case–control study we evaluated the use of the medial spike angle and fracture tip–skin distance to predict the mode of reduction (closed or open) and the operating time in fractures with posterolateral displacement. A total of 21 patients (4.35%) with a small medial spike angle (< 45°) were identified from a total of 494 patients, and 42 patients with a medial spike angle of > 45° were randomly selected as controls. The medial spike group had significantly smaller fracture tip–skin distances (p < 0.001), longer operating times (p = 0.004) and more complications (p = 0.033) than the control group. There was no significant difference in the mode of reduction and a composite outcome measure. After adjustments for age and gender, only fracture tip–skin distance remained significantly associated with the operating time (β = -0.724, p = 0.042) and composite outcome (OR 0.863 (95% confidence interval 0.746 to 0.998); p = 0.048). Paediatric orthopaedic surgeons should have a lower threshold for open reduction when treating patients with a small medial spike angle and a small fracture tip–skin distance. . Cite this article: Bone Joint J 2013;95-B:1290–4

We evaluated results at one year after surgical correction of internal rotation deformities in the shoulders of 270 patients with obstetric brachial plexus palsy. The mean age at surgery was 6.2 years (0.6 to 35). Two techniques were used: open subscapularis elongation and latissimus dorsi to infraspinatus transfer. In addition, open relocation was performed or attempted in all patients with subluxed or dislocated joints. A mixed effects model approach was used to evaluate the effects of surgery on internal and external rotation, abduction, flexion and Mallet score. Independent factors included operative status (pre- or post-operative), gender, age, the condition of the joint, and whether or not transfer was performed. The overall mean improvement in external rotation following surgery was 84.6° (95% confidence interval (CI) 80.2 to 89.1) and the mean Mallet score improved by 4.0 (95% CI 3.7 to 4.2). There was a mean decrease in internal rotation of between 27.6° and 34.4° in the relocated joint groups and 8.6° (95% CI 5.2 to 12.0) in the normal joint group. Abduction and flexion were unchanged following surgery. Adding a latissimus dorsi transfer did not result in greater improvement in the mean external rotation compared with elongation of the subscapularis alone. Cite this article: Bone Joint J 2013;95-B:1432–8

Two types of fracture, early and late, have been reported following limb lengthening in patients with achondroplasia (ACH) and hypochondroplasia (HCH). . We reviewed 25 patients with these conditions who underwent 72 segmental limb lengthening procedures involving the femur and/or tibia, between 2003 and 2011. Gender, age at surgery, lengthened segment, body mass index, the shape of the callus, the amount and percentage of lengthening and the healing index were evaluated to determine predictive factors for the occurrence of early (within three weeks after removal of the fixation pins) and late fracture (> three weeks after removal of the pins). The Mann‑Whitney U test and Pearson’s chi-squared test for univariate analysis and stepwise regression model for multivariate analysis were used to identify the predictive factor for each fracture. Only one patient (two tibiae) was excluded from the analysis due to excessively slow formation of the regenerate, which required supplementary measures. A total of 24 patients with 70 limbs were included in the study. There were 11 early fractures in eight patients. The shape of the callus (lateral or central callus) was the only statistical variable related to the occurrence of early fracture in univariate and multivariate analyses. Late fracture was observed in six limbs and the mean time between removal of the fixation pins and fracture was 18.3 weeks (3.3 to 38.4). Lengthening of the tibia, larger healing index, and lateral or central callus were related to the occurrence of a late fracture in univariate analysis. A multivariate analysis demonstrated that the shape of the callus was the strongest predictor for late fracture (odds ratio: 19.3, 95% confidence interval: 2.91 to 128). Lateral or central callus had a significantly larger risk of fracture than fusiform, cylindrical, or concave callus. Radiological monitoring of the shape of the callus during distraction is important to prevent early and late fracture of lengthened limbs in patients with ACH or HCH. In patients with thin callus formation, some measures to stimulate bone formation should be considered as early as possible. Cite this article: Bone Joint J 2014;96-B:1269–73

We tested the hypothesis that children who sustain a supracondylar fracture have a greater range of elbow hyperextension than those with a fracture of the distal radius. Three observers made 358 measurements in 183 children (114 boys and 69 girls). There were 119 fractures of the distal radius and 64 supracondylar fractures. Initially, the group with a supracondylar fracture appeared to have extension 1.7° greater than that of the group with fracture of the distal radius. On average, there was a maximum variation of 3° between observers. After allowing for age, gender and observer, there was no significant difference between the groups. Our study had greater than 80% power to detect a difference in hyperextension of 2° at the 5% level with the above observer variability. When age and gender are taken into account, any variation in the amount of hyperextension at the elbow is not sufficient to explain the occurrence of a supracondylar fracture

This study was designed to provide normal reference values for the centre-edge angle of Wiberg in the Chinese population by measuring 1494 radiographs according to the criteria of Tönnis. The mean angle was 23.1° (4.7° to 46.4°) in childhood (four to nine years), increasing to 28.9° (6° to 48°) in adolescence (10 to 18 years) and reaching 32.8° (13.7° to 58.8°) in adults. The angle was positively correlated with age, increasing by a mean of 0.78° annually before adulthood and by 0.070° annually in adults. A relative upward tendency after middle age might contribute to degenerative changes. No gender difference was discovered. The change in the angle with age was similar to that observed in other ethnic groups, showing moderate differences in the average of individual age groups, but the ethnic differences are not sufficient to explain the variation of incidence of developmental dysplasia of the hip in different races

We assessed the medium-term outcome of three methods of isolated calf lengthening in cerebral palsy by clinical examination, observational gait analysis and, where appropriate, instrumented gait analysis. The procedures used were percutaneous lengthening of tendo Achillis, open Z-lengthening of tendo Achillis and lengthening of the gastrosoleus aponeurosis (Baker’s procedure). We reviewed 195 procedures in 134 children; 45 had hemiplegia, 65 diplegia and 24 quadriplegia. We established the incidence of calcaneus and recurrent equinus and identified ‘at-risk’ groups for each. At follow-up, 42% had satisfactory calf length, 22% had recurrent equinus and 36% calcaneus. The incidence of calcaneus in girls at follow-up was significantly higher (p = 0.002) while boys had an increased rate of recurrent equinus (p = 0.012). Children with diplegia who had surgery when aged eight years or younger had a 44% risk of calcaneus, while those over eight years had a 19% risk (p = 0.046). Percutaneous lengthening of tendo Achillis in diplegia was the least predictable, only 38% having a satisfactory outcome compared with 50% in the other procedures. The incidence of recurrent equinus in hemiplegic patients was 38%. Only 4% developed calcaneus. The type of surgery did not influence the outcome in patients with hemiplegia or quadriplegia. Severity of involvement, female gender, age at operation of less than eight years and percutaneous lengthening of tendo Achillis were ‘risk factors’ for calcaneus. Hemiplegia, male gender, and an aponeurosis muscle lengthening increased the risk of recurrent equinus

The Ponseti method of clubfoot management requires a period of bracing in order to maintain correction. This study compared the effectiveness of ankle foot orthoses and Denis Browne boots and bar in the prevention of recurrence following successful initial management. Between 2001 and 2003, 45 children (69 feet) with idiopathic clubfeet achieved full correction following Ponseti casting with or without a tenotomy, of whom 17 (30 clubfeet) were braced with an ankle foot orthosis while 28 (39 clubfeet) were prescribed with Denis Browne boots and bar. The groups were similar in age, gender, number of casts and tenotomy rates. The mean follow-up was 60 months (50 to 72) in the ankle foot orthosis group and 47 months (36 to 60) in the group with boots and bars. Recurrence requiring additional treatment occurred in 25 of 30 (83%) of the ankle foot orthosis group and 12 of 39 (31%) of the group with boots and bars (p < 0.001). Additional procedures included repeat tenotomy (four in the ankle foot orthosis group and five in the group treated with boot and bars), limited posterior release with or without tendon transfers (seven in the ankle foot orthosis group and two in the group treated with boots and bars), posteromedial releases (nine in the orthosis group) and midfoot osteotomies (five in the orthosis group, p < 0.001). Following initial correction by the Ponseti method, children managed with boots and bars had far fewer recurrences than those managed with ankle foot orthoses. Foot abduction appears to be important to maintain correction of clubfeet treated by the Ponseti method, and this cannot be achieved with an ankle foot orthosis

We examined differences in the rate of open reduction, operating time, length of hospital stay and outcome between two groups of children with displaced supracondylar fractures of the humerus who underwent surgery either within 12 hours of the injury or later. There were 77 children with type-3 supracondylar fractures. Of these, in 43 the fracture was reduced and pinned within 12 hours and in 34 more than 12 hours after injury. Both groups were similar in regard to gender, age and length of follow-up. Bivariate and logistical regression analysis showed no statistical difference between the groups. The number of peri-operative complications was low and did not affect the outcome regardless of the timing of treatment. Our study confirmed that the treatment of uncomplicated displaced supracondylar fractures of the humerus can be early or delayed. In these circumstances operations at night can be avoided

Aims

The health-related quality of life (HRQoL) of paediatric patients with orthopaedic conditions and spinal deformity is important, but existing generic tools have their shortcomings. We aim to evaluate the use of Paediatric Quality of Life Inventory (PedsQL) 4.0 generic core scales in the paediatric population with specific comparisons between those with spinal and limb pathologies, and to explore the feasibility of using PedsQL for studying scoliosis patients’ HRQoL.

Introduction:. Risk factors for developmental dysplasia of the hip (DDH) in early infancy have never been validated from basic principles; their relevance remains controversial. Purpose:. To determine risk factors for DDH using newly developed diagnostic criteria based on international consensus. Methods:. In this population-based cohort study, 9904 babies born at a secondary care unit (2010–2012) received a standardised examination (usually within 24 hours postpartum) in which we prospectively ascertained the presence of the common risk factors for DDH (breech, family history, etc). Infants exhibiting ≥1 factor were eligible and underwent ultrasound testing within 8 weeks. Alpha angles were measured by surgeon/radiologist in consensus and blinded to risk factors and age. Using multivariable methods we evaluated the association of the risk factors and ultrasonographic DDH using criteria based on international consensus. Results:. 1766 (18%) newborns exhibited ≥1 risk factor for DDH. Of these 1489 (84%) infants participated. To date, 1296 (87%) completed the ultrasound at a mean age of 8±3 weeks. Of the 1296, 55 (4%) patients exhibited alpha <55° and 43 (3%) exhibited alpha <50°. Of all risk factors, only female gender was associated with an alpha <55° (RR=2; 95% CI=1.1, 3.5; p=.01). In contrast, abnormal clinical examination findings of the hip were strongly associated with DDH (p<.0001). Conclusion:. In a prospective study using robust case definitions, commonly known risk factors were not clinically important markers of DDH when DDH was defined by consensus criteria. Given the generally poor and conflicting evidence on risk factors for DDH, our preliminary results suggest a new approach is needed in the risk prediction of DDH. Level of evidence: I

Background. The association between idiopathic scoliosis and intrinsic abnormalities of the spinal cord are well known. MRI is the most sensitive imaging modality available to diagnose an intraspinal anomaly. The use of a routine ‘screening MRI’ prior to scoliosis correction is controversial. Purpose. This study sought to quantify the frequency of previously unidentified cord anomalies identified by a routine pre-operative MRI, in patients planned for surgical scoliosis correction. Methods and results. The patient group comprised 206 patients with idiopathic scoliosis who underwent deformity correction from 1998 to 2008. Of these, 20 (9.7 %) patients were diagnosed with an unexpected intraspinal anomaly (isolated syrinx-7, syrinx with Chiari malformation-4, isolated Chiari malformation-4, syrinx with tethered cord-2, split spinal cord-1, intrinsic spinal cord tumour-1, AV fistula-1) on routine pre-operative MRI. In all cases a neurosurgical opinion was sought prior to further intervention. Of the 20 patients, 11 underwent a neurosurgical procedure (de-tethering of cord, decompression of Chiari malformation, decompression of syrinx). No statistically significant difference was identified between individuals with or without a neurosurgical anomaly with respect to the age of presentation, gender, side and degree of curve. Conclusion. The high frequency of previously unidentified cord abnormalities detected, and the frequent need for subsequent neurosurgical intervention, suggests that MRI assessment prior to deformity correction is important in the management of idiopathic scoliosis

Purpose. To assess the initial rise in inflammatory markers in paediatric patients presenting with long bone osteomyelitis and whether this is comparable with that in septic arthritis, and diagnostic. Methods. All radiologically confirmed cases of long bone osteomyelitis without septic arthritis, joint effusion or abscess, in paediatric patients, presenting to one hospital over an eighteen-month period were included. These patients were compared with all culture positive septic arthritides presenting to the same hospital within the same period. Inflammatory markers taken on the day of admission were studied. Results. Thirty-seven patients with osteomyelitis and thirteen with culture positive septic arthritis were identified. The two groups were comparable with regards to age and gender. At presentation 65% of the osteomyelitis patients had an ESR, (erythrocyte sedimentation rate), less than 40mm/hour, 48% below 20mm/hour and 19% within the normal range. 23% of the septic arthritis patients had an ESR below 40mm/hour, and none had an ESR below 20mm/hour or in the normal range. The CRP, (C reactive protein), was in the normal range in 46% of the osteomyelitis patients and none of the septic arthritis patients. The average ESR and CRP in the osteomyelitis patients were 47mm/hour and 35mg/L respectively, while in the septic arthritis group these were 72mm/hour and 107mg/L, (P<0.028). Conclusion. The initial rise in inflammatory markers due to long bone osteomyelitis is significantly less than that in septic arthritis. A proportion of patients with osteomyelitis have normal inflammatory markers at presentation. A high index of suspicion is required, and inflammatory markers, while a useful adjunct to monitoring treatment, may offer false reassurance regarding the initial diagnosis

Aim. Paediatric fractures are common but those occurring in non-ambulant children are associated with higher rates of Non Accidental Injury (NAI). There is little published on the mechanisms of injury associated with accidental fracture in young children. This study explores the aetiology of long bone fractures in non-ambulant children. Methods. This retrospective multicentre study looked at children aged ≤18 months presenting to three hospitals over 3 years (2009 to 2011). Information was gathered on age, gender, fracture type, injury mechanism, final diagnosis, treatment and details of screening for NAI. Results. 147 children were identified who were ≤18 months old (mean 12 months). There were 32 femoral, 37 tibial, 43 forearm, 17 humeral, 16 clavicular and 3 fibular fractures. We identified 6 confirmed cases of NAI and 7 pathological fractures (osteopenia of prematurity or osteogenesis imperfecta). 5/64 children aged ≤12 months old had NAI compared with 1/83 in those aged >12 months. All 7 pathological fractures occurred in the ≤12 months group. NAI or pathological fracture was more likely in ≤12 months group compared to those >12 months (p=0.0002) Of the 12 children with no clear mechanism of injury, 5 had NAI and 3 had pathological fractures. In 39/147 children NAI was considered in the documentation and 29 had a paediatric review. Falls from beds and change mats were more common in ≤12 months group, as well as transverse femoral fractures; caused when those carrying the child slipped downstairs and applied a sudden bending force to the held leg. In those >12 months falls from chairs, down steps, in playgrounds or on trampolines were more common. 12/147 fractures were caused directly by other children (6 in each group). Conclusion. Our study identified causes of accidental long-bone fracture in non-ambulant children. In cases where there is no clear mechanism of injury, NAI must be carefully excluded