Outcomes of Surgical Treatment of Femoral Head Fractures with a Minimum of 1-Year Follow-up Authors: Hyun-Chul Shon, Eic-Ju Lim, Kook-Jong Kim, Jae-Young Yang Published: 2026 ABSTRACT BackgroundFemoral head fractures are uncommon injuries, usually resulting from high-energy trauma and often accompanied by other musculoskeletal injuries, which complicate management. Despite advancements in surgical techniques, complications such as posttraumatic osteoarthritis (OA) and heterotopic ossification (HO) remain prevalent, and no standardized treatment protocol has been established. Furthermore, most existing reports are small-scale or multi-center studies with heterogeneous management. The aim of this study was to evaluate the radiological and functional outcomes, as well as complications, of surgically treated femoral head fractures managed by a single surgeon at a single center with a minimum of 1-year follow-up.MethodsThis retrospective study included 48 patients who underwent surgery for femoral head fractures between January 2007 and December 2023. Demographic characteristics, injury mechanisms, associated injuries, surgical methods, conversion to total hip arthroplasty, and complications were obtained from medical records. Fractures were classified using the Pipkin system based on preoperative radiographs and 3-dimensional computed tomography. Postoperative evaluations included radiological assessment for OA, avascular necrosis, and HO, and functional assessments using the modified Postel-Merle d’Aubigné score and Thompson and Epstein clinical criteria. Final outcomes were analyzed according to Pipkin fracture type and surgical approach.ResultsThe rate of excellent or good outcomes was higher in Pipkin types I and II than in type IV according to both the Thompson-Epstein roentgenographic (100%, 100%, and 58.3%) and clinical (90%, 90.9%, and 54.2%) criteria. Merle d’Aubigné-Postel scores were also higher in Pipkin types I (16.4) and II (16.8) than in type IV (15.0). Posttraumatic OA was most frequent in the flip osteotomy group (75%) and in Pipkin type IV fractures (60%). HO occurred most commonly with the anterior and flip osteotomy approaches (33.3%) and in Pipkin type I fractures (40.0%).ConclusionsFemoral head fractures remain uncommon injuries often associated with concomitant trauma and a high incidence of complications. We demonstrated favorable radiological and functional outcomes following surgical treatment, with results analyzed according to Pipkin type and surgical approach. These findings offer clinically meaningful insights that may advance ongoing research on femoral head fractures. Femoral head fractures are rare injuries that primarily result from high-energy trauma and are most frequently associated with posterior hip dislocations.123) Although several surgical methods, such as fragment excision, open reduction internal fixation (ORIF), and arthroplasty, are commonly used, the choice of optimal treatment remains controversial.456) Complications of femoral head fractures include nerve injury, posttraumatic osteoarthritis (OA), avascular necrosis (AVN), and heterotopic ossification (HO). Because femoral head fractures involve intra-articular damage, long-term follow-up is essential to monitor the development of these complications.78) Owing to the rarity of this injury pattern, most published studies are small-scale and lack extended follow-up. Larger studies are generally multi-center and multi-surgeon in design, which introduces variability in surgical approach, fixation technique, and postoperative protocols. In contrast, single-surgeon, single-center series provide greater consistency in surgical management and evaluation.9) Therefore, we aimed to evaluate the radiological and functional outcomes, as well as associated complications, in patients who underwent surgical treatment for femoral head fractures performed by a single surgeon (HCS) at a single center, with a minimum follow-up of 1 year. METHODS The study protocol was reviewed and approved by the Institutional Review Board of Chungbuk National University Hospital (IRB No. CBNUH 2025-02-001-001). The need for informed consent was waived because of the retrospective nature of the study. This retrospective study was conducted at Chungbuk National University Hospital, which is a level 1 trauma center, and included 61 patients who underwent surgery for femoral head fractures between January 2007 and December 2023. The indications for surgery were as follows: (1) incongruent reduction of the hip joint, (2) an articular step-off greater than 1 mm, and (3) fractures involving more than 20% of the posterior acetabular wall. The exclusion criteria were as follows: (1) a follow-up duration of less than 1 year, (2) patients younger than 18 years, (3) inability to perform functional evaluation due to associated injuries (e.g., brain or spinal cord injuries), and (4) preexisting hip joint pathology. Demographic data (age and sex), injury mechanisms, associated injuries, surgical methods, conversion to total hip arthroplasty (THA), and postoperative complications (nerve injury) were obtained from medical records. Preoperative imaging, including pelvis anteroposterior, oblique, inlet, and outlet radiographs, and 3-dimensional computed tomography, was used to classify fractures according to the Pipkin classification.10) Radiological evaluations for posttraumatic OA, AVN, and HO, along with functional assessments using the Thompson and Epstein clinical criteria1) and the modified Postel-Merle d’Aubigné score,1112) were performed at each outpatient follow-up visit. Based on the final follow-up evaluations, radiological and functional outcomes, and complications were analyzed according to the Pipkin classification, and complications associated with each surgical approach were investigated. All radiographic parameters, including the Pipkin classification,10) Thompson and Epstein roentgenographic criteria,1) Association Research Circulation Osseous (ARCO) staging system,13) and Brooker classification,14) were independently assessed by 1 orthopedic surgeon (HCS) and 1 radiologist who were not involved in the surgical procedures. In cases of disagreement, the final assessment was determined through consensus discussion. The Thompson and Epstein roentgenographic criteria were used to evaluate posttraumatic arthritis following femoral head fractures. These criteria assess several parameters, including hip joint congruency (reduced or not), cartilaginous joint space width (normal or diminished, with severity grading), variations in femoral head bone density (mottling, cystic formation, or sequestration), subchondral irregularities, acetabular sclerosis, capsular calcification, and spur formation. Based on these observations, radiographic outcomes were categorized into 4 grades: excellent, good, fair, and poor. AVN was diagnosed radiographically and classified according to the ARCO staging system. HO was identified on radiographs and classified using the Brooker classification. To prevent HO, patients received 100 mg of celecoxib twice daily for 6 weeks in combination with a proton pump inhibitor. Prophylactic radiation therapy was not performed. Nerve injury was diagnosed using electromyography in patients presenting with neurological symptoms. Functional outcomes were evaluated using the modified Postel-Merle d’Aubigné score and the Thompson and Epstein clinical criteria, which are the 2 most commonly used assessment tools for femoral head fractures. The modified Postel-Merle d’Aubigné score was calculated as the sum of pain, walking ability, and range of motion (ROM) sub-scores, yielding a total score ranging from 4 to 18 points. Scores of 18 were classified as excellent, 15–17 as good, 13–14 as fair, and < 13 as poor. The Thompson and Epstein clinical criteria assess functional outcomes based on pain, hip ROM, and the presence of limping, with results categorized into 4 grades: excellent, good, fair, and poor. In patients who underwent THA, the modified Postel-Merle d’Aubigné score was not applied. For these patients, the modified Postel-Merle d’Aubigné classification and the Thompson-Epstein clinical and roentgenographic criteria were all classified as poor. Patient Demographic Data and Characteristics A total of 61 patients who underwent surgery for femoral head fractures were initially included. Among them, the following were excluded: (1) 8 patients with a follow-up duration of less than 1 year, (2) 1 patient younger than 18 years, (3) 3 patients who could not undergo functional evaluation due to associated injuries (e.g., brain or spinal cord injuries), and (4) 1 patient with a preexisting diagnosis of AVN. Consequently, 48 patients were included in the final analysis. The mean follow-up period was 36.5 months (standard deviation [SD], 29.8; range, 12–139 months). The demographic and clinical characteristics of the study population are summarized in Table 1. The mean patient age was 43.1 years (SD, 14.3; range, 18–66 years), and the cohort consisted of 40 males and 8 females. The most common mechanism of injury was automobile traffic accidents (n = 34, 70.8%). According to the Pipkin classification, 10 patients had type I fractures, 11 had type II, 1 had type III, and 24 had type IV fractures. Two patients could not be classified according to the Pipkin system and were categorized as Brumback type 4a injuries.15) Associated injuries included craniofacial injuries in 7 patients, chest injuries in 9 patients, abdominal injuries in 11 patients, and extremity injuries in 16 patients. Isolated hip injuries occurred in 15 patients. Surgery In cases with dislocation, closed reduction was first attempted in the emergency room. If the reduction was unsuccessful, emergency surgery was performed immediately. Even when closed reduction was successful, operative fixation was performed as early as possible, considering the patient’s overall condition, to minimize complications associated with delayed surgery. All surgical procedures were performed by a single surgeon (HCS) with the patients positioned on a radiolucent plain table under general anesthesia. The surgical approach was selected based on the fracture pattern and the presence of associated injuries. Either an anterior (Smith-Petersen) or posterior (Kocher-Langenbeck) approach was selected accordingly. A trochanteric flip osteotomy was performed in cases requiring extensive exposure of the femoral head or when concomitant acetabular fractures required surgical treatment. Fragment excision was performed when the fragment was too small for stable fixation or was located outside the weight-bearing area. Conversely, when the fragment was sufficiently large and involved the weight-bearing area, ORIF was performed using a headless compression screw (Acutrak standard; Acumed) (Figs. 1 and 2). In cases with extensive damage to the weight-bearing area and severe comminution precluding stable fixation, mosaicplasty16) or repositionplasty17) was performed. Considering the high rate of fixation failure reported for ORIF in previous studies,181920) THA was performed for all patients with Pipkin type III fractures, as well as those with Pipkin type IV fractures accompanied by displaced femoral neck fractures. Rehabilitation and Follow-up In cases where simple fragment excision was performed without acetabular surgery, ROM exercises and full weight-bearing (FWB) were allowed immediately after surgery. In patients who underwent ORIF of the acetabulum or femoral head, ROM exercises were started immediately postoperatively. Toe-touch ambulation using crutches began at postoperative week 1, partial weight-bearing (PWB) at week 6, and FWB at week 10. Outpatient follow-up evaluations were conducted at 6 weeks, 12 weeks, 6 months, and 1 year postoperatively, and annually thereafter, unless specific complications occurred. RESULTS Radiological and Functional Outcomes At the final follow-up of the 48 patients, radiological outcomes, measured according to the Thompson-Epstein roentgenographic criteria, were classified as excellent in 24 patients (50.0%), good in 13 (27.1%), fair in 5 (10.4%), and poor in 6 (12.5%). Functional outcomes, measured according to the Merle d’Aubigné score, had an average of 15.8 points. Functional outcomes were classified as excellent in 12 patients (25.0%), good in 22 (45.8%), fair in 5 (10.4%), and poor in 9 (18.8%). According to the Thompson-Epstein functional criteria, the distribution was excellent in 14 patients (29.2%), good in 20 (41.7%), fair in 5 (10.4%), and poor in 9 (18.8%). Ten patients had Pipkin type I fractures; of these, 8 underwent excision and 2 underwent ORIF. Radiographic outcomes were excellent in 8 patients and good in 2 patients. Functional outcomes showed an average modified Postel-Merle d’Aubigné score of 16.4, with 3 patients rated as excellent, 6 as good, and 1 as fair. According to the Thompson-Epstein classification, 3 patients were rated as excellent, 6 as good, and 1 as fair. Eleven patients with Pipkin type II fractures were included; 9 underwent ORIF, and 2 underwent excision. Radiographic outcomes were excellent in 9 patients and good in 2 patients. Functional outcomes demonstrated an average modified Postel-Merle d’Aubigné score of 16.8, with 4 patients rated as excellent, 6 rated as good, and 1 rated as fair. According to the Thompson–Epstein classification, 5 patients were rated as excellent, 5 as good, and 1 as fair. One patient with a Pipkin type III fracture underwent THA. Among the 24 patients with Pipkin type IV fractures, femoral head fractures were treated with ORIF in 4 patients and fragment excision in 14 patients. Two patients with large, crushed defects in the weight-bearing area, precluding ORIF, underwent mosaicplasty and repositionplasty. Seventeen underwent ORIF for associated acetabular fractures, while 3 received conservative treatment. Four patients with displaced femoral neck fractures underwent primary THA. One patient experienced fixation failure 3 months postoperatively and subsequently underwent conversion to THA. Functional outcomes for this group showed an average modified Postel-Merle d’Aubigné score of 15.0, with 4 patients rated as excellent, 8 as good, 3 as fair, and 8 as poor. According to the Thompson-Epstein classification, 5 patients were rated as excellent, 8 as good, 3 as fair, and 8 as poor. Radiographic outcomes were excellent in 6 patients, good in 8 patients, fair in 5 patients, and poor in 5 patients. Two patients could not be classified according to the Pipkin classification: one underwent fragment excision, and the other underwent mosaicplasty. Radiographic outcomes for these patients were excellent in one and good in the other. Functional outcomes showed an average modified Postel-Merle d’Aubigné score of 16.5. Both the modified Postel-Merle d’Aubigné and Thompson-Epstein classifications showed 1 excellent and 1 good outcome each (Table 2). Complications Excluding 5 patients who underwent primary THA (1 patient with a Pipkin type III and 4 patients with Pipkin type IV), complications according to the Pipkin classification and the surgical approach were investigated in the remaining 43 patients (Tables 3 and 4). One patient who underwent a flip osteotomy for a Pipkin type IV fracture experienced fixation failure 3 months postoperatively and subsequently underwent conversion to THA. Posttraumatic OA occurred in 44.2% (19 patients), with an incidence of 75% (9 patients) in the flip osteotomy group, 50% (5 patients) in the Kocher-Langenbeck approach group, and 23.8% (5 patients) in the Smith–Petersen approach group. By Pipkin type, OA occurred most frequently in Pipkin type IV (60%, 12 patients), followed by type I (40%, 4 patients) and type II (18.1%, 2 patients). These patients are currently being followed up. HO developed in 32.6% (14 patients) overall, occurring in 33.3% of patients in both the Smith-Petersen approach group (7 patients) and the flip osteotomy group (4 patients), and 30% (3 patients) in the Kocher-Langenbeck approach group. By Pipkin type, HO was observed in 40.0% of type I cases (4 patients), 35.0% of type IV cases (7 patients), and 18.2% of type II cases (2 patients). Except for mild pain and ROM limitations, no significant symptoms were reported, and all patients were monitored. AVN, corresponding to ARCO stage 2, occurred in 4.7% (2 patients), 1 each in the Smith-Petersen approach group (a Pipkin type II case) and the flip osteotomy group (a Pipkin type IV case). As the symptoms were mild, the patients remain under observation. Nerve injury occurred in 11.6% (5 patients): 4.8% (1 patient) in the Smith–Petersen approach group, 20% (2 patients) in the Kocher-Langenbeck approach group, and 16.7% (2 patients) in the flip osteotomy group. By Pipkin type, nerve injury occurred in 20.0% of type IV cases (4 patients) and 9.1% of type II cases (1 patient). All cases involved sciatic nerve injury, with 1 case in the flip osteotomy group (Pipkin type IV) identified as iatrogenic. All 5 nerve injuries improved spontaneously within 6–9 months postoperatively. Fifteen patients (34.9%) experienced no complications. The lowest complication rate was observed in the Smith–Petersen approach group (52.4%, 11 patients), followed by the Kocher-Langenbeck approach group (30%, 3 patients) and the flip osteotomy group (8.3%, 1 patient). By Pipkin type, no complications were noted in 54.5% of type II cases (6 patients), 50% of type I cases (5 patients), and 20.0% of type IV cases (4 patients). DISCUSSION Femoral head fractures are rare injuries typically resulting from high-energy trauma. Due to the high rates of complications and associated injuries, orthopedic surgeons continue to face significant challenges in management.135) Given the intra-articular nature of these fractures, long-term follow-up is essential to monitor for complications.78) Surgical treatment options include fragment excision, ORIF, and THA; however, a clear treatment protocol has not yet been established.456) Owing to the rarity of femoral head fractures, most studies on their surgical management have been multi-center, involving multiple surgeons. In a recent meta-analysis,21) the sample size per study ranged from 5 to 39 cases. Similarly, in Giannoudis’s meta-analysis,18) which is the largest to date on femoral head fractures, functional outcomes were synthesized from 11 studies, with sample sizes ranging from 6 to 33 patients per study. Given the lack of clear treatment protocols, small-scale studies often produce inconsistent and unreliable outcomes. Therefore, we conducted a single-center, single-surgeon study involving 48 patients with femoral head fractures, all of whom were surgically treated and followed up for a minimum of 1 year. To the best of our knowledge, very few studies from a single center and involving a single surgeon have included outcomes for all Pipkin classifications, underscoring the clinical significance of the present study. According to a meta-analysis by Giannoudis et al.,18) the combined proportion of excellent and good outcomes based on the Thompson-Epstein clinical criteria was reported as 64.7% for Pipkin type I, 73.2% for type II, and 52.6% for type IV fractures. In comparison, our study demonstrated more favorable results, with 90.0% for type I, 90.9% for type II, and 54.2% for type IV fractures. When excluding cases of primary THA due to displaced femoral neck fractures, the type IV group showed an improved rate of 65.0%. Furthermore, the incidence of complications, including AVN and HO, was lower in our study compared to the findings of Giannoudis et al.18) The reported AVN rates in their study were 5.3% for the Smith-Petersen approach, 16.9% for the Kocher-Langenbeck approach, and 8.3% for the flip osteotomy. HO rates were 44.7% for the Smith-Petersen approach, 32.3% for the Kocher-Langenbeck approach, and 47.2% for flip osteotomy. We believe that these differences reflect the contrast between outcomes derived from pooled data of small-scale studies and those achieved by a single experienced surgeon in the treatment of these rare and challenging fractures. In a single-center study by Scolaro et al.,19) which involved 69 patients treated by 2 surgeons and followed for a mean duration of 12.4 months, the reported complication rates included fixation failure in 12%, HO in 40%, and AVN in 9%. In comparison, our study demonstrated a much lower fixation failure rate of 2.3%. This may be attributed to the fact that all cases with concomitant femoral neck fractures were treated with primary THA, resulting in favorable and comparable outcomes. The incidence of HO in our study was also comparable, given the similar follow-up duration. AVN was observed in only 2 patients (4.7%), both corresponding to ARCO stage 2. Given the similarly low incidence in both studies, the results in this regard can be considered comparable. According to a study by Regenbogen,22) clinical outcomes based on the Harris Hip Score were 73 points for Pipkin type I, 79 points for type II, and 68.5 points for type IV fractures, indicating poorer outcomes for Pipkin type IV fractures compared to types I and II. Similarly, our study demonstrated lower clinical scores for Pipkin type IV fractures, with average modified Postel-Merle d’Aubigné scores of 16.4 for type I, 16.8 for type II, and 15.0 for type IV. This trend has also been observed in multiple studies623) and is likely attributable to the fact that Pipkin type IV fractures involve both the femoral head and acetabulum, leading to more complex injury patterns that often require additional surgical intervention and are associated with a higher risk of complications. The original Pipkin classification does not differentiate between cases where femoral head fractures are accompanied by femoral neck and acetabular fractures.10) In our study, 3 patients in the Pipkin type IV group also had associated femoral neck fractures. These patients were treated with primary THA, similar to those with Pipkin type III fractures. This treatment approach may have contributed to the poorer overall outcomes observed in the Pipkin type IV group. Given the distinct injury patterns and clinical implications, future studies should consider separately classifying cases of concomitant femoral head, femoral neck, and acetabular fractures. We propose defining this subgroup as Pipkin type V. In our study, 3 cases (2 Pipkin type IV fractures and 1 unclassified case) involved large, crushed defects in the weight-bearing area of the femoral head, rendering ORIF technically unfeasible. As an alternative, 2 patients underwent mosaicplasty, and 1 underwent repositionplasty. At a mean follow-up of 32.7 months, all 3 cases demonstrated favorable outcomes, with “good” radiological results based on the Thompson-Epstein classification, a mean modified Postel-Merle d’Aubigné score of 15.3, and “good” clinical outcomes according to the Thompson-Epstein criteria. Similarly, a meta-analysis by Athanasiou et al.,24) on mosaicplasty for posttraumatic femoral head lesions, reported excellent outcomes with a mean Harris Hip Score of 94.7. These findings suggest that osteochondral reconstruction techniques, such as mosaicplasty, may serve as valuable surgical options in cases involving large osteochondral defects in the weight-bearing area, where ORIF is not technically feasible. Regarding the mechanism of injury, automobile traffic accidents accounted for 70.8% of cases, which is similar to Regenbogen’s study,22) where 71% of injuries were caused by motorcycles and vehicles. Although differences in the frequency of injury mechanisms may exist due to regional and traffic culture variations, femoral head fractures almost always result from high-energy trauma. According to Park et al.,25) associated injuries were reported in 62% of cases for extremity fractures, 29% for abdominal injuries, and 20% for chest injuries. In our study, extremity fractures occurred in 33.3%, abdominal injuries in 22.9%, and chest injuries in 18.8% of patients, while isolated femoral head fractures accounted for only 31.3%. These findings highlight the importance of recognizing that femoral head fractures are often accompanied by injuries to other regions of the body, necessitating thorough evaluation and management of these concomitant injuries. This study has some limitations. First, the low incidence of femoral head fractures was a major constraint. The small number of patients within each Pipkin subtype limited the statistical power to identify significant risk factors for poor outcomes and complications. Additionally, as all surgeries were performed by a single surgeon, the generalizability of these findings to other settings may be limited. However, the consistency achieved through uniform surgical technique and postoperative management in a single-surgeon series can also be viewed as a methodological strength. This limitation could be addressed in future research by comparing these findings with those from multi-center, multi-surgeon studies. Second, this study did not include outcomes of conservative treatment. According to a study by Giannoudis et al.,18) conservative management accounts for approximately 17% of all femoral head fracture treatments, representing a substantial proportion. Although we initially intended to include both surgically and conservatively treated patients, long-term follow-up beyond 1 year was rarely available for the conservative treatment group. In our cohort, none of the conservatively treated patients continued outpatient follow-up, primarily due to poor early clinical outcomes. Third, the retrospective nature of this study introduces the possibility of selection bias. Additionally, patients with less than 1 year of follow-up were excluded from the analysis, which may have influenced the representativeness of the cohort. Fourth, the outcomes reported in this study reflect the final follow-up assessments rather than evaluations at standardized postoperative intervals, which may limit the accuracy in predicting outcomes progression over time. Finally, unlike radiographic outcomes, clinical outcomes were assessed directly by the operating surgeon during outpatient visits. This may introduce a degree of observer bias, which should be considered when interpreting the findings. Femoral head fractures are rare injuries, often accompanied by associated injuries and high complication rates. In this study, we report favorable radiological and functional outcomes following surgical treatment. Outcomes were stratified according to the Pipkin classification, and complications were analyzed based on the surgical approach and Pipkin classification. We believe these findings provide valuable insights that will contribute to ongoing research on femoral head fractures.