Outcomes at 1 Year After Total Hip Arthroplasty Combined With Low Femoral Neck Osteotomy vs Subtrochanteric Osteotomy in 73 Patients With Crowe Type IV Developmental Dysplasia of the Hip: A Retrospective Study

Introduction

Developmental dysplasia of the hip (DDH) is a spectrum of disorders involving abnormal development of the acetabulum and proximal femur, which can lead to hip instability and altered biomechanics. The reported incidence of DDH varies widely across populations and depends on screening strategy and diagnostic criteria; in a systematic review and meta-analysis, the pooled incidence of early-detected DDH ranged from 4.4 to 23.0 per 1000 newborns under selective vs universal ultrasonographic screening (clinical screening: 8.4 per 1000) [1]. Although many mild neonatal instabilities resolve spontaneously, late-detected cases still occur (pooled incidence approximately 0.2–0.6 per 1000 newborns), representing a subgroup at risk for persistent dysplasia and dislocation and later reconstructive surgery [1]. Early diagnosis and timely treatment are critical for improving long-term outcomes [2]. The clinical presentation of DDH varies with age and severity; adults may present with chronic hip pain, gait disturbance, functional limitation, and leg length discrepancy (LLD), while radiographic assessment typically demonstrates acetabular dysplasia and varying degrees of femoral head subluxation/dislocation [3]. Diagnosis is commonly established using clinical examination combined with imaging, including anteroposterior pelvic radiographs and, when needed, computed tomography for preoperative assessment of acetabular/femoral morphology and femoral version [4]. Management ranges from observation and hip-preserving procedures in earlier stages to joint replacement in end-stage disease; importantly, untreated hip dysplasia can lead to difficulty walking, hip pain, and early arthritis [5]. When degenerative changes progress to end-stage DDH, total hip arthroplasty (THA) becomes the primary option for pain relief and functional restoration [6]. Since Charnley introduced modern THA, the technique has evolved substantially and has become the cornerstone for treating end-stage hip disease [7]. With increasing clinical experience, THA has also been increasingly applied to complex hip disorders, including DDH. Given the growing number of end-stage DDH cases and the aging population, identifying a safe and effective THA strategy for severe DDH has become increasingly important [8,9]. The Crowe classification, based on the degree of proximal migration of the femoral head on anteroposterior pelvic radiographs, is widely used to grade adult DDH severity. Crowe type IV DDH represents the most severe form and is characterized by complete proximal subluxation/dislocation of the femoral head (high-riding femoral head), a small hypoplastic true acetabulum, and marked proximal femoral deformity (eg, excessive femoral anteversion, a narrow canal, and altered metaphyseal geometry), often accompanied by soft tissue contracture. These anatomic features make anatomic acetabular reconstruction and safe reduction during primary THA particularly challenging [10]. To address these challenges, various surgical techniques have been proposed. Charnley was among the first to use trochanteric osteotomy in complex primary THA, including Crowe type IV DDH [11]. Subsequently, a range of osteotomy and non-osteotomy techniques have been developed to facilitate safer joint reduction. Although favorable outcomes have been reported, each technique carries inherent limitations. For example, severe bone deficiency, extensive soft tissue release, postoperative LLD, and abnormal anatomy may increase the risk of perioperative femoral fractures, osteotomy nonunion, and nerve injury [12,13].

In patients with Crowe type IV DDH, femoral osteotomy is frequently required to facilitate hip reduction and prevent excessive limb lengthening [14]. Among the available techniques, subtrochanteric osteotomy (SO) is widely used because it allows controlled femoral shortening and rotational correction [15,16]. Nevertheless, SO often requires an extended incision and extensive soft tissue dissection, and is associated with risks such as nonunion and prolonged operative time [17]. Low femoral neck osteotomy (LRNO), performed close to the femoral neck and occasionally extending to the level of the lesser trochanter, has been proposed as an alternative strategy that enables hip reduction without a formal subtrochanteric cut [18]. LRNO can be completed through the standard THA incision and may reduce surgical trauma, blood loss, and operative time. Nevertheless, its effects on limb length restoration, early functional recovery, and complication profiles remain incompletely understood. Most existing studies have focused on technical refinements or outcomes of a single osteotomy technique, and direct comparisons of clinical outcomes between different osteotomy strategies remain limited. Therefore, this retrospective study included 73 patients with Crowe type IV DDH and aimed to compare outcomes at 1 year following THA combined with LRNO or SO.

Functional recovery is a primary concern for patients undergoing THA for severe DDH [19]. To comprehensively evaluate postoperative hip function, clinician-reported and patient-reported outcome measures have been used [20–22]. The Harris Hip Score (HHS) assesses pain, function, range of motion, and deformity from a clinical perspective [21], while the Oxford Hip Score (OHS) assesses patient-reported function and satisfaction in daily activities [22]. Given that different osteotomy techniques can influence early soft tissue trauma, limb length restoration, and rehabilitation speed, HHS and OHS were selected as the primary outcome measures. Secondary outcomes included complication rates, perioperative laboratory indicators, and LLD.

Material and Methods

ETHICS STATEMENT:

This retrospective, single-center study was approved by the Institutional Ethics Committee of the Third Hospital of Hebei Medical University (approval No. W2023-002-1). The requirement for informed consent was waived because anonymized data were analyzed. The study complied with the Declaration of Helsinki and Good Clinical Practice guidelines, and patient confidentiality was fully protected.

STUDY DESIGN AND PARTICIPANTS:

Patients with a diagnosis of Crowe type IV DDH who underwent unilateral THA between January 1, 2017, and June 30, 2024, were screened. Of 102 patients initially identified, 29 were excluded: 16 had incomplete records, 7 had prior contralateral THA, and 6 had age <18 years. Ultimately, 73 patients were included: 37 underwent THA combined with LRNO and 36 underwent THA combined with SO (Figure 1).

INCLUSION AND EXCLUSION CRITERIA:

The inclusion criteria were as follows: (1) age <80 years; (2) Crowe type IV DDH; (3) unilateral THA; and (4) follow-up ≥12 months. The exclusion criteria were as follows: (1) age <18 years; (2) history of ipsilateral hip surgery; or (3) incomplete clinical or radiographic data.

SAMPLE SIZE CONSIDERATION:

A sample size estimation was conducted using the expected between-group difference and variability in the primary functional outcomes reported in previous DDH-THA studies comparing osteotomy-related techniques [6]. With a 2-sided alpha of 0.05 and 80% power, the minimum required sample size was 70 patients (approximately 35 per group). Therefore, the final cohort (73 patients: LRNO n=37; SO n=36) met the target and was considered adequate for between-group comparisons.

BLINDING AND OUTCOME ASSESSMENT:

Postoperative functional outcomes and radiographic assessments were performed by independent physicians who were not involved in the surgeries, thereby minimizing surgeon-related assessment bias. Patient blinding was not feasible, because the surgical approaches were identifiable (eg, incision characteristics and osteotomy method). This limitation was considered when interpreting subjective outcome measures.

SURGICAL PROCEDURES:

All THA procedures were performed by a senior orthopedic surgeon with subspecialty training. Participants underwent THA combined with either SO or LRNO. A posterolateral surgical approach was used. Patients were positioned in lateral decubitus position. In the SO group, the posterolateral incision was extended distally, the subcutaneous tissue was dissected layer by layer, the short external rotators and joint capsule were transected, and the hypertrophic capsule was excised to expose the true acetabulum. After sequential reaming, the acetabular component was implanted and secured with screws at an inclination of 45±5° and anteversion of 15±5°. Femoral preparation was then performed, followed by a transverse subtrochanteric osteotomy 2 to 5 cm below the lesser trochanter. A trial stem was inserted and reduction was attempted; after appropriate fit and stability were confirmed, the definitive femoral stem and prosthetic head were implanted, and the wound was closed in layers. In the LRNO group, the acetabular procedure was identical. Femoral osteotomy was performed at a low femoral neck level and extended toward the lesser trochanter when necessary to mitigate sciatic nerve tension.

Intraoperative events were managed as follows: bleeding with electrocautery or ligation; periprosthetic fractures with cerclage wiring or plate fixation; and prosthetic dislocation with closed reduction. All implants were cementless with ceramic-on-ceramic bearings. Femoral stems included S-ROM (Johnson & Johnson, USA), SR, and SL stems (both from Beijing AK Medical). Representative pre- and postoperative radiographs are shown (Figure 2). A schematic illustration (Figure 3) delineates the procedural steps for standard total hip arthroplasty, LRNO, and SO, including osteotomy levels (black lines) corresponding to the operative description.

PERIOPERATIVE MANAGEMENT:

All patients received prophylactic antibiotics intraoperatively, which continued for 24 hours postoperatively [23]. Postoperatively, the operated limb was kept in abduction and neutral rotation using an abduction pillow to prevent excessive adduction or internal rotation. Low-molecular-weight heparin was administered for thromboprophylaxis, followed by oral anticoagulants for up to 35 days postoperatively [24]. From postoperative day 2, patients were allowed partial weight-bearing with crutches, as tolerated [25].

REHABILITATION PROTOCOL:

A standardized rehabilitation protocol was applied to both groups. During hospitalization, rehabilitation was instructed and supervised by physiotherapists and ward nurses under the surgeon’s guidance, following a written protocol. Key milestones (brace use, weight-bearing progression, gait training) were documented in the medical record and discharge education sheet.

After discharge, patients were provided with a standardized home program. Adherence and adverse events were reviewed and reinforced at each scheduled outpatient visit or telephone follow-up. During the early phase (1–6 weeks), patients used an abduction brace, quadriceps isometric training, ankle pumps, ambulation with crutches, and partial weight-bearing less than or equal to 10 kg. During the intermediate phase (6–12 weeks), patients gradually progressed to full weight-bearing by week 12, with active hip flexion and abduction exercises, while avoiding risky movements. During the late phase (≥3 months), patients performed gait training and gradually resumed daily activities, while avoiding strenuous activity or heavy labor.

DATA COLLECTION AND IMAGING:

Baseline demographic and surgical data were collected from hospital electronic records, including age, sex, body mass index, and comorbidities. Preoperative and postoperative erythrocyte sedimentation rate (ESR) and white blood cell (WBC) levels were measured to assess differences in the systemic inflammatory response induced by surgical trauma. In addition, pre- and postoperative data on hemoglobin, albumin, and D-dimer levels were collected to evaluate patients’ nutritional status, together with perioperative data including operative time, intraoperative blood loss, and postoperative complications. All data were obtained from the Donghua medical record system of the Third Hospital of Hebei Medical University. Postoperatively, anteroposterior pelvic radiographs and full-length stitched radiographs of both lower limbs were obtained, with LLD determined by measuring the distance between the line connecting the tuberosity of the ischium and the distal lateral malleoli [26]. On preoperative full-length lower limb radiographs, the bilateral LLD was calculated as |a – b|, representing the absolute difference in the distance from the ischial tuberosity to the lateral femoral condyle on both sides. On postoperative radiographs, the bilateral LLD was calculated as |c – d| using the same landmarks (Figure 4). Figure 5 shows representative anteroposterior pelvic radiographs from before surgery to 12 months postoperatively in 2 patients (LRNO and SO groups), demonstrating the surgical outcomes and limb length restoration over time. Long-term outcomes were assessed using anteroposterior pelvic radiographs at 1, 3, 6, and 12 months postoperatively, as well as additional follow-up radiographic data in some patients. Although the mean follow-up duration was 21.9 months, the 12-month postoperative time point was selected as the primary endpoint because functional recovery after hip arthroplasty typically plateaus within the first year, and follow-up intervals beyond 12 months were less standardized. Prosthetic loosening was defined as a vertical migration greater than 2 mm or radiolucency greater than 2 mm at the bone-implant interface [27]. Nerve injury was classified according to the modified Sunderland grading scale: grade 1, normal limb function; grade 2, mild motor weakness and/or sensory disturbance; grade 3, ambulation requiring orthotic support with mild sensory disturbance; grade 4, restricted ambulation with moderate pain and occupational limitation; and grade 5, severe motor impairment and/or intractable pain [28]. Other complications, including deep vein thrombosis (DVT) of the lower limbs, dislocation, prosthetic loosening, nonunion, periprosthetic fractures (intraoperative and those occurring during postoperative follow-up), and infection, were retrieved from outpatient and medical record data.

CLINICAL FUNCTIONAL ASSESSMENT:

Clinical function was evaluated preoperatively and at 1, 3, 6, and 12 months postoperatively through outpatient visits or telephone follow-up. Patients lost to follow-up within 16 months after surgery were tracked with the assistance of community health centers, and for those with more than 6 months of missed follow-up, the last available follow-up data were recorded along with the reason for loss to follow-up. Hip function was assessed using the HHS, which evaluates pain, function, deformity, and range of motion. Postoperative satisfaction and patient-reported outcomes were further assessed using the OHS. The total HHS is 100 points: <70 indicates poor recovery, 70–90 indicates moderate recovery, and 90–100 indicates good recovery. The OHS has a maximum score of 48 points: 41–48 indicates excellent to good function, 34–40 indicates fair function, 27–33 indicates poor function, and 12–26 indicates very poor function. In this study, HHS and OHS obtained at the 4 follow-up time points within 12 months were used as the primary outcome measures.

STATISTICAL ANALYSIS:

Analyses were conducted in R (version 4.3.1; R Core Team) using the stats package for standard hypothesis tests [29], the tableone package for baseline descriptive summaries [30], the lme4 package for linear mixed-effects modeling (lmer) [31], and the emmeans package for estimated marginal means and contrasts. Multiplicity was controlled using the Holm method implemented in stats:: p.adjust. Continuous variables were first tested for normality using the Shapiro-Wilk test. Data with normal distribution are expressed as mean±standard deviation (mean±SD), and comparisons between groups were conducted using independent-samples t tests. Data not conforming to a normal distribution are expressed as median with interquartile range [M (IQR)], and intergroup comparisons were performed using the Mann-Whitney U test. Categorical variables are presented as counts and percentages [n (%)], and comparisons between groups were performed using the chi-square test or Fisher exact test when expected frequencies were less than 5.

Initially, baseline characteristics of the 2 groups (SO vs LRNO), including age, sex, body mass index, surgical side, and other relevant variables, were compared to test for balance. This was performed to ensure comparability in demographic and clinical characteristics, thereby minimizing potential confounding effects. The primary outcome measures were HHS and OHS measured preoperatively and at postoperative time points of 1, 3, 6, and 12 months. For these longitudinal primary outcomes, linear mixed-effects models were fitted using the lme4 package [32] with group (SO vs LRNO), time, and the group × time interaction as fixed effects, and a subject-specific random intercept to account for within-patient correlation. Model-based estimated marginal means and between-group contrasts at each time point were obtained using the emmeans package [33], with multiplicity controlled using the Holm method. Results are reported as estimated effects with 95% confidence intervals when applicable. Secondary outcomes included perioperative laboratory parameters, including pre- and postoperative hemoglobin, albumin, and D-dimer levels, ESR, and WBC counts; and surgical parameters, including operative time, intraoperative blood loss, transfusion volume, incision length, and lower LLD, as well as the incidence of postoperative complications, including DVT, dislocation, nerve injury, and prosthetic loosening. Group comparisons for secondary continuous outcomes were performed using the aforementioned parametric or nonparametric tests, and complication incidence was compared between groups as categorical data. All statistical tests were 2-sided, with a significance level set at α=0.05. P<0.05 was considered statistically significant. The tableone package was used to automatically generate summary tables and group-comparison results. The resulting figures and tables were exported and saved as electronic files.

Results

JOINT FUNCTION SCORES:

Baseline characteristics of the 2 groups were tested for balance, as shown in Table 1, and no statistically significant differences were observed, indicating comparability between groups. Outcome measures at the completion of the 12-month follow-up are summarized in Tables 2 and 3. At 6 months, HHS and OHS scores had converged; by 12 months, the scores were essentially identical, with no significant differences. Interestingly, at 1 and 3 months postoperatively, the HHS were 57.89±4.02 vs 50.64±2.59 (P<0.001) and 74.16±3.46 vs 65.39±2.27 (P<0.001) for the LRNO and SO groups, respectively, with both differences reaching statistical significance. Similarly, The OHS scores were 29.08±1.53 vs 25.03±1.92 (P<0.001) and 35.16±3.04 vs 32.56±1.73 (P<0.001) for the LRNO and SO groups, respectively. These findings indicate that, among patients with Crowe type IV DDH, THA combined with LRNO facilitates faster early functional recovery (within 3 months), although the advantage diminishes by 6 months as scores converge.

COMPLICATIONS:

Complication data for all 73 patients were collected and are summarized in Table 4. Among the 73 patients, 3 cases of sciatic nerve injury occurred, all within 6 months postoperatively and all in the LRNO group. The incidence of sciatic nerve injury was 3 (8.11%) vs 0 (0.00%) in the LRNO and SO groups, respectively (P=0.240), and all 3 patients had recovered by the 12-month follow-up. Periprosthetic fractures occurred in 3 (8.11%) vs 0 (0.00%) patients in the LRNO and SO groups (P=0.240), with no statistically significant difference. All 3 fractures occurred intraoperatively; 2 were managed with cerclage wiring and 1 was left untreated, and the fracture lines had healed by the 3-month follow-up. Although the LRNO group experienced 3 nerve injuries and 3 periprosthetic fractures, and none occurred in the SO group, the difference was not statistically significant (P=0.240). Although not statistically significant, this trend indicates a potentially higher complication risk associated with LRNO, warranting clinical vigilance.

DVT occurred in 2 (5.41%) vs 2 (5.56%) patients in the LRNO and SO groups, respectively (P=1.000). All 4 events were detected prior to discharge and were located in intramuscular veins, and resolution was confirmed by ultrasound at the 1-month follow-up. The incidence of dislocation was 1 (2.70%) vs 1 (2.78%) in the LRNO and SO groups (P=1.000), with no significant difference observed. Periprosthetic infection occurred in 1 (2.70%) vs 0 (0.00%) patients in the LRNO and SO groups (P=1.000); this case was identified 1 month postoperatively as a superficial infection and was successfully treated with oral antibiotics by the 3-month follow-up.

Nonunion occurred in 0 (0.00%) vs 2 (5.56%) patients in the LRNO and SO groups (P=0.240), and prosthetic loosening occurred in 0 (0.00%) vs 1 (2.78%) patient (P=0.493), with no significant differences between groups. Postoperative LLD was significantly smaller in the LRNO group (0.6 [0.4–0.9] cm vs 1.5±0.6 cm; P<0.001). Among the 73 patients, 1 developed dislocation accompanied by nonunion and prosthetic loosening at the 3-month follow-up and subsequently underwent revision hip arthroplasty.

PERIOPERATIVE OUTCOMES:

Perioperative outcomes are summarized in Table 5. The LRNO group showed significant advantages over the SO group regarding minimally invasive performance and surgical efficiency. Specifically, the LRNO group had shorter operative times (115.00 [100.00–150.00] minutes vs 165.00 [135.00–196.25] minutes; P=0.000), smaller incision lengths (14.00 [12.00–16.00] cm vs 23.00 [21.00–24.25] cm; P=0.000), reduced intraoperative blood loss (350.00 [300.00–600.00] mL vs 700.00 [500.00–900.00] mL P=0.002), and significantly lower transfusion requirements (0.00 [0.00–0.00] mL vs 250.00 [0.00–400.00] mL; P<0.001), as well as plasma usage (0.00 [0.00–0.00] mL vs 0.00 [0.00–200.00] mL; P=0.003).

With regard to laboratory test results, postoperative decreases in hemoglobin and albumin levels were significantly greater in the SO group than in the LRNO group (both P<0.001), suggesting that LRNO was associated with less postoperative anemia and better nutritional preservation.

No significant differences were observed between groups for other laboratory parameters, including ESR, D-dimer levels, and WBC counts (all P>0.05).

In summary, compared with the conventional SO procedure, the LRNO technique reduced surgical trauma, blood loss, and transfusion requirements, shortened operative time, provided superior early functional outcomes, and achieved better restoration of limb alignment without increasing the risk of long-term complications.

Discussion

LIMITATIONS:

This study offers valuable clinical insights into comparing 2 osteotomy techniques in THA for patients with Crowe type IV DDH; however, several limitations should be acknowledged. First, the retrospective study design inevitably introduced bias, and although efforts were made to control confounding factors, they could not be completely eliminated. Second, the relatively small sample size (73 cases) limited the statistical power, particularly for rare complications such as nerve injury and nonunion. Future multicenter, large-scale prospective randomized controlled trials are needed to validate these results. Third, the average follow-up duration was short (21.9 months), making it difficult to assess long-term outcomes, such as prosthesis survival or late loosening. Additionally, all surgeries were conducted by a single senior surgeon; therefore, caution should be exercised when generalizing the findings to surgeons with varying levels of experience. Furthermore, the lack of patient blinding may have introduced bias in subjective functional evaluations. Lastly, this study compared only transverse subtrochanteric osteotomy, without including other approaches such as step-cut or Z-shaped osteotomies, which limits the generalizability of the conclusions.

Conclusions

This retrospective study compared the clinical outcomes of THA combined with LRNO and SO in patients with Crowe type IV DDH. Results demonstrated that patients who underwent LRNO achieved significantly better early functional recovery (at 1 and 3 months), as indicated by higher HHS and OHS scores. LRNO was also associated with shorter operative time, reduced intraoperative blood loss and transfusion requirements, smaller incision sizes, and better restoration of limb length balance. By 6 and 12 months, functional scores had converged between the 2 groups, with no significant differences observed in long-term complication rates. In summary, while functional outcomes and complication rates were similar between groups at 12 months, the minimally invasive nature and early recovery advantages of LRNO suggest it may be an effective clinical option for managing Crowe type IV DDH.