Management and retrospective analysis of postoperative femoral fractures after proximal femoral osteotomy in children with developmental dysplasia of the hip Authors: Shuhua Deng, Congyang Wang, Qiang Shi Published: 2026 ABSTRACT IntroductionProximal femoral osteotomy (PFO) with locking compression plates (LCPs) is effective for treating developmental dysplasia of the hip (DDH) in children. However, plate insertion or removal has a risk of postoperative femoral fractures. The goal of this study was to identify and analyse the association between postoperative femoral fractures after PFO and potential risk factors for these fractures.MethodsSix patients with DDH and postoperative femoral fractures were retrospectively reviewed at our hospital from June 2014 to October 2020. There were 2 boys and 4 girls whose ages ranged from 2 to 8 years (mean 3 years and 8 months). The timing, circumstances, fracture location, and refixation method were recorded for all patients. Four patients sustained fractures after plate removal: 3 after a fall and 1 spontaneously. Moreover, 2 patients experienced fractures early postoperatively after open reduction and PFO. The mean follow-up time was 31.6 months (25–57 months).ResultsAfter 153 patients underwent open reduction and PFO at our institution, six (3.9%) experienced postoperative femoral fractures. In children who underwent implant removal, all fractures (n = 4/4) occurred at the osteotomy site, whereas in those with retained implants (n = 1/2), fractures occurred at the screw hole. Three fractures were treated with reduction under general anaesthesia and spica casting, and three were treated with ESIN.ConclusionsInjuries from PFO, prolonged immobilization, or LCPs can result in postoperative femoral fractures in children with DDH. After PFO, the implants should be removed when solid union is achieved at the osteotomy site. Moreover, high-intensity activity should not be permitted until moderate or extensive remodelling at the osteotomy site is confirmed after plate removal. Supplementary InformationThe online version contains supplementary material available at 10.1186/s12887-026-07083-z. Background Plate insertion or removal after proximal femoral osteotomy (PFO) may lead to a risk of postoperative femoral fractures, which prolong immobilisation and sometimes necessitate additional surgery [1–3]. In addition, these fractures in children result in surgical risks, medical costs, and patient frustration [4]. The potential risk factors for postoperative femoral fractures (including early postoperative fractures and fractures occurring after plate removal) remain unclear. Therefore, we investigated possible risk factors for postoperative femoral fractures in children with DDH. Patients and methods Informed consent was obtained from all parents and the institutional ethical committee of Changsha Central Hospital, University of South China approved this study. The analyses and procedures performed in the study were conducted in accordance with the Declaration of Helsinki. We retrospectively reviewed data from all 153 children who underwent open reduction and PFO in our hospital from June 2014 to October 2020, and the data presented are primarily descriptive. Six patients with postoperative femoral fractures were retrospectively reviewed, including 2 boys and 4 girls whose ages ranged from 2 to 8 years (mean 3 years and 8 months). The associated conditions of the patients included pain and motion limitations of the hip. Patients with metabolic bone disease, neuromuscular conditions, bone fragility disorders, nutritional deficiencies, and pathologic fractures were excluded from the present study. The timing, circumstances, fracture location, and refixation method were recorded for all patients. Four patients sustained fractures after plate removal: 3 after a fall (falling from standing, twisting or walking) and 1 spontaneously (no falling). Moreover, 2 patients experienced fractures early postoperatively after open reduction and PFO (low-energy trauma such as ground-level falls). The mean follow-up time was 31.6 months (25–57 months) (Table 1). The inclusion criteria were as follows: (a) femoral fracture within 2 years after PFO, regardless of whether the hardware was removed; (b) complete medical data spanning a follow-up period of more than 2 years; and (c) the absence of a history of severe trauma, such as a fall from a height or a car accident. In our hospital, percutaneous adductor tenotomy, open reduction, Salter osteotomy and PFO through the anterolateral approach are always performed in children between 2 and 6 years old with DDH (Pemberton osteotomy for those older than 6 years). For Tönnis grades III or IV, we usually performed capsulotomy and femoral derotation or shortening osteotomy (one or two centimetres according to Tony’s Catterall principles) to decrease the excessive anteversion angle or neck-shaft angle. According to preoperative radiography and CT, the extent of dislocation of the femoral head, neck-shaft angle, and anteversion angle in the proximal femur were carefully evaluated. After open reduction and PFO, a spica cast was used for postoperative immobilisation for 8 weeks. Moreover, when union of the osteotomy site of the femur was confirmed by regular radiography, gradual weight- bearing was permitted. In our institution, hardware removal was performed routinely 12–18 months after open reduction and PFO according to standardised institutional criteria. For children with bilateral DDH, we usually perform osteotomy on each side with an interval of 4–6 months between surgeries. Whether the implant was still in situ was also determined at the endpoint. The remodeling condition at the osteotomy site was also evaluated at the endpoint. According to the methods described by Davids et al., we classified remodeling radiographically as the proportion of the residual trace line located at the osteotomy site to the diameter of the medullary cavity. If the proportion was smaller than 1/2, a moderate or extensive remodeling condition existed at the osteotomy site; if the proportion was larger than 1/2, mild remodeling existed [5]. After the complete union of the osteotomy site, the medial cortex of the proximal segment becomes the residual line in the medullary cavity, as seen in anterior-posterior radiographs. Statistical analysis All measurement data were presented as means ± SD, and Student’s t test was used to examine the differences between groups. Chi-squared test and Fisher’s exact test were applied to analyze the data of the two groups in this study. P-values < 0.05 were considered statistically significant. Results After 153 patients underwent open reduction and PFO at our institution, six (3.9%) experienced postoperative femoral fractures. The 2 groups were similar in terms of age, sex, and weight (Table 1). In children who underwent implant removal, all fractures occurred at the osteotomy site (n = 4/4), whereas in those with retained implants, fractures occurred at the screw hole (n = 1/2). Three fractures were treated with reduction under general anaesthesia and spica casting, and three were treated with ESIN (Table 2). Table 1Comparison of demographic data and characteristics. Values are count (%) or mean (range)VariableAll subjects(N = 153)No fracture(n = 147)Fracture(n = 6)P valueAge (year)5.1 (2–8)5.1 (2–8)4.5 (2–7)0.449Female sex124(81.0)120 (81.6)4 (67.7)0.359Weigh (kg)18.7 (9–55)20.5 (11–55)18.3 (9–47)0.565Side0.475 Unilateral1201164 Bilateral33312 Table 1Comparison of demographic data and characteristics. Values are count (%) or mean (range)VariableAll subjects(N = 153)No fracture(n = 147)Fracture(n = 6)P valueAge (year)5.1 (2–8)5.1 (2–8)4.5 (2–7)0.449Female sex124(81.0)120 (81.6)4 (67.7)0.359Weigh (kg)18.7 (9–55)20.5 (11–55)18.3 (9–47)0.565Side0.475 Unilateral1201164 Bilateral33312 Comparison of demographic data and characteristics. Values are count (%) or mean (range) Table 2Characteristics of the fracture groupAge at timeof surgery (years)SexSideTime from removal to fracture (days)Fracture typeTreatment15FRTransverseESIN24ML14TransverseSpica33FLTransverseSpica47FR10SpiralESIN55FL30TransverseSpica65MR7TransverseESIN Table 2Characteristics of the fracture groupAge at timeof surgery (years)SexSideTime from removal to fracture (days)Fracture typeTreatment15FRTransverseESIN24ML14TransverseSpica33FLTransverseSpica47FR10SpiralESIN55FL30TransverseSpica65MR7TransverseESIN Characteristics of the fracture group Fracture with retained implant (2 patients) Two patients sustained postoperative femoral fractures early postoperatively after PFO. ESIN or a spica cast was used for 12 weeks, and active non-weight-bearing exercises throughout a range of motion were performed for approximately 3–4weeks. Follow-up indicated complete bone healing without further complications (Fig. 1a and 1c). Fig. 1a This patient had a femoral fracture in the screw hole at 24 weeks post-operatively when her femur was injured after a domestic fall. b This patient had a fracture at the osteotomy site at 2 weeks after plate removal when he experienced sudden thigh pain. c Mildly displaced fracture of the distal femur was found 3 weeks after PFO in a 3-year-old girl with DDH. d Ten days after plate removal, a 7-year-old girl had a right femoral fracture with minor injury from a ground-level fall. e Transverse fracture at the osteotomy site at 4 weeks after plate removal in a 4-year-old boy a This patient had a femoral fracture in the screw hole at 24 weeks post-operatively when her femur was injured after a domestic fall. b This patient had a fracture at the osteotomy site at 2 weeks after plate removal when he experienced sudden thigh pain. c Mildly displaced fracture of the distal femur was found 3 weeks after PFO in a 3-year-old girl with DDH. d Ten days after plate removal, a 7-year-old girl had a right femoral fracture with minor injury from a ground-level fall. e Transverse fracture at the osteotomy site at 4 weeks after plate removal in a 4-year-old boy Fracture after implant removal (4 patients) Fractures after implant removal occurred in 4 patients who underwent PFO. Three patients sustained a minimal fall, and 1 patient presented with sudden thigh pain and was unable to bear weight. A spica cast was used for 8–12 weeks for postoperative femoral fractures in patients under the age of 5, while ESIN was performed in patients over 5 years of age (Figs. 1b, 1d, 1e and Fig. 2e). The operative stages for ESIN are shown in Fig. 2. Fig. 2a A 5-year-old girl with bilateral DDH. b This patient underwent PFO, open reduction, and pelvic osteotomy. c The implant was removed at 14 months post-operatively. d This patient had an implant-related fracture at the osteotomy site at 1 week after implant removal when she was walking. e Revision with ESIN. f The ESIN was removed 8 months later, showing complete healing a A 5-year-old girl with bilateral DDH. b This patient underwent PFO, open reduction, and pelvic osteotomy. c The implant was removed at 14 months post-operatively. d This patient had an implant-related fracture at the osteotomy site at 1 week after implant removal when she was walking. e Revision with ESIN. f The ESIN was removed 8 months later, showing complete healing Discussion Implant removal after PFO for DDH is customary. However, postoperative femoral fractures remain challenging because these fractures require additional surgical procedures, long-term hospitalisation and significant economic costs [6, 7]. Recently, the occurrence rate of implant-related femoral fractures has been reported to range from 0.3 to 3.6% [3, 8]. Whether mild remodelling at the osteotomy site weakens the mechanical strength of the proximal femur, making it susceptible to fracture, is still uncertain. Regardless of whether the hardware is removed, close follow-up may be considered for children with mild remodelling at the osteotomy site, and high-intensity activity may not be considered until moderate or extensive remodelling at the osteotomy site is confirmed. In our institution, 2 patients sustained postoperative femoral fractures after PFO, with the implant remaining in situ. LCP and LCP-PHP may reduce the level of stress at the osteotomy site [9]. Fractures occurred after implant removal in 4 patients who underwent PFO, and all fractures occurred at the osteotomy site. All patients had the plates removed after complete osteotomy union and subsequently sustained fractures without any traumatic cause. For postoperative weightbearing instructions after plate removal, we routinely advise patients to begin walking immediately. This may explain why the rate observed in our study was higher than that reported in the literature. On the basis of our results, we have made the following recommendations: (1) during PFO, efforts may be considered to minimise disruption of blood supply and soft tissue damage; (2) LCP may be removed approximately 18 months after PFO, with confirmed and complete healing at the osteotomy site; and (3) After plate removal, we routinely suggest that patients use crutches with partial weight bearing for at least 1 month post–operatively. We also suggest that children may be followed up closely after implant removal and that high-intensity activity may not be considered until moderate or extensive remodelling at the osteotomy site is confirmed. Once postoperative femoral fractures occur in children with DDH, both ESIN and hip spica casting can ultimately yield good radiographic outcomes [10–12]. However, several nonradiographic outcome measures differ between ESIN and hip spica casting. Although the management of postoperative femoral fractures in children with DDH remains controversial, our study reveals that radiological outcomes are generally good when patients are treated with ESIN or hip spica casting. In our study, hip spica casting for eight weeks was suggested for patients under the age of 5 with undisplaced femoral fractures, eight weeks and active nonweight-bearing exercises were performed for four weeks. ESIN is a quick and minimally invasive technique for treating displaced femoral fractures in children older than 5 years, preventing secondary displacement and allowing for early weight-bearing. Nevertheless, the present study has some limitations. First, whether postoperative range of motion in the hip or other clinical factors affects fractures after osteotomy is unknown. Second, the small number of fracture cases (independent risk factor analysis could not be performed), the absence of multivariate and risk-adjusted analyses, the absence of comorbidity assessments, observational and hypothesis-generating single-centre data and a lack of comparative data were also limitations of our study. Therefore, large-scale institutional investigations, preferably prospective in nature, are necessary to corroborate our findings in the future. Third, several potentially relevant confounding variables (such as different pelvic osteotomies, different femoral procedures, different implant types, amount of femoral shortening, screw configuration, hardware dimensions, bone quality, and degree of remodelling at the time of hardware removal) were not analysed in our present study. In addition, no functional outcomes (e.g., range of motion, gait assessment, limb length discrepancy, or validated paediatric hip scores) were provided in the present study. Conclusion In conclusion, many factors such as surgical injury or internal fixation devices, can result in postoperative femoral fractures in children with DDH. High-intensity activity may be avoided until moderate or extensive remodelling at the osteotomy site is confirmed. After PFO, implant removal may be considered after solid union is confirmed at the osteotomy site. We suggest that a protective brace may be considered for 1–2 months after LCP removal and delay exercise as a precaution against secondary postoperative femoral fractures after PFO. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.