Case Report: Proximal femoral valgus osteotomy and dynamic hip screw fixation for post-traumatic coxa vara secondary to femoral neck or intertrochanteric fracture nonunion or malunion: a retrospective case series Authors: Mingyang Cao, Jun Li, Zhiyu Gao Published: 2026 ABSTRACT BackgroundPost-traumatic coxa vara following femoral neck or intertrochanteric fracture nonunion or malunion remains a difficult problem in young and middle-aged patients, in whom salvage total hip arthroplasty carries excess complication risk. Proximal femoral valgus osteotomy is a joint-preserving alternative, but evidence supporting fixation with a dynamic hip screw (DHS) is limited, and is largely confined to younger cohorts.Case summaryWe report a single-center, retrospective, consecutive case series of twelve patients (seven men, five women; mean age 52.2 ± 20.0 years, range 20–73) treated between January 2020 and December 2024. Etiologies included failed internal fixation of intertrochanteric fracture (n = 6), malunion of intertrochanteric fracture (n = 3), failed internal fixation of femoral neck fracture (n = 2), and malunion of femoral neck fracture (n = 1). All patients underwent proximal femoral valgus osteotomy with DHS-based fixation, supplemented as required by cannulated screws or a reconstruction plate. Outcomes were assessed at a minimum follow-up of 12 months.OutcomesBony union was achieved in all twelve patients at a mean of 19.1 ± 4.1 weeks. The mean neck–shaft angle improved from 100.3° to 132.2°, the mean Harris Hip Score from 50.4 to 80.2, and the mean limb shortening reduced from 2.4 cm to 0.7 cm (all p < 0.001). No avascular necrosis, implant failure, clinically significant heterotopic ossification, or conversion to arthroplasty occurred.ConclusionProximal femoral valgus osteotomy with DHS-based fixation reliably restores hip mechanics, achieves union, and produces meaningful functional improvement in post-traumatic coxa vara, including in selected older patients with a viable femoral head. Background Hip fractures remain a major global burden. The annual incidence is projected to nearly double by 2050, with the steepest rise expected in Asia (1, 2). Two anatomical patterns account for most of these injuries: intracapsular femoral neck fractures and extracapsular intertrochanteric fractures. Both are routinely treated by internal fixation, yet failure is far from rare. Pooled data show that around one-third of displaced femoral neck fractures develop nonunion after primary internal fixation, and 16% develop avascular necrosis (AVN) (3). Mechanical failure of intertrochanteric fixation occurs in approximately 16% of cases, most often as cut-out or varus collapse (4). In younger patients, up to one-third require reoperation and 14% are eventually converted to total hip arthroplasty (THA) (5). When these fixation failures heal in a varus position, or when malunion is established without union, the result is post-traumatic coxa vara: a neck–shaft angle (NSA) below 120°. The biomechanical consequence is straightforward. Shortening of the abductor lever arm produces a Trendelenburg gait. Limb shortening (LS) accentuates the lurch. Persistent shear at the fracture line prevents the bone from healing. Pain, fatigue, and progressive arthrosis follow. Salvage THA is the most commonly offered solution. In older patients with degenerative change it works well. In younger and middle-aged patients, however, salvage arthroplasty after failed fixation carries roughly three times the complication rate of primary THA, with higher rates of deep infection, dislocation, and periprosthetic fracture (6). Long-term survivorship is also reduced (7). A joint-preserving alternative is therefore attractive whenever the femoral head remains viable. Valgus intertrochanteric osteotomy, first described by Pauwels in 1935, addresses the problem mechanically (8). Reorientation of the proximal fragment converts shear at the fracture line into compression, simultaneously restoring abductor tension and limb length. Marti and colleagues established the modern clinical template in 1989, reporting union and durable function in the majority of 50 patients (9). Numerous subsequent series have confirmed the procedure's utility in both femoral neck and intertrochanteric nonunion or malunion (10–14). Most published series have used 95° or 120° fixed-angle blade plates. The dynamic hip screw (DHS) offers an attractive alternative. Its sliding mechanism allows controlled compression across the osteotomy, its lateral barrel accommodates supplementary cannulated screws (CS) or a reconstruction plate (RP), and the implant is familiar to most trauma surgeons (15, 16). Evidence supporting DHS-based valgus osteotomy is growing but remains limited to small single-center series (17, 18). The joint-preserving value of valgus osteotomy in young patients is well established. The present study extends this evidence in two respects. First, it focuses specifically on DHS-based fixation rather than the fixed-angle blade plate. Second, it reports outcomes across a deliberately heterogeneous cohort that included older patients and patients with significant comorbidity, a group under-represented in the existing literature. The aim of this study is to report the radiographic and functional outcomes of proximal femoral valgus osteotomy with DHS-based fixation in twelve consecutive patients with post-traumatic coxa vara due to nonunion or malunion of femoral neck or intertrochanteric fractures, and to assess whether this technique offers a reliable joint-preserving option in a heterogeneous clinical population. Methods Study design and setting This study was a single-center, retrospective, consecutive case series conducted at Zhoukou Orthopedic Hospital (Zhoukou, Henan, China). The institutional review board approved the study (approval number: 2025-102), and the requirement for written informed consent for the use of de-identified clinical and radiographic data was waived because of the retrospective design. The study was conducted in accordance with the Declaration of Helsinki. Because the study reports aggregate, group-level outcomes rather than a narration of individual cases, it is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement for observational studies (20); a completed STROBE checklist is provided as a Supplementary File 1. The three illustrative patients in the Representative cases section additionally follow the CARE reporting guideline (21). Participants We reviewed the institutional database for all patients who underwent proximal femoral valgus osteotomy combined with DHS fixation for post-traumatic coxa vara between January 2020 and December 2024. Eligible patients had radiographically confirmed coxa vara (defined as an NSA <120°) secondary to nonunion or malunion of a previous femoral neck or intertrochanteric fracture, together with persistent hip pain, limp, or functional limitation. Etiologies were classified as failure of internal fixation following intertrochanteric fracture (FFIF-ITF), malunion of intertrochanteric fracture (M-ITF), failure of internal fixation following femoral neck fracture (FFIF-FNF), and malunion of femoral neck fracture (M-FNF). Patients with active hip infection, advanced femoral head osteonecrosis with collapse, end-stage hip osteoarthritis, or a follow-up shorter than 12 months were excluded. Twelve consecutive patients met the criteria and constituted the study cohort. Definitions of nonunion and malunion Nonunion was defined as a fracture that had not healed and showed no radiographic progression toward union for at least three consecutive months, with a minimum of six months elapsed since the index fixation (22). Malunion was defined as a healed fracture in a non-anatomical position, in this cohort specifically as union with an NSA below 120°. Plain anteroposterior and lateral radiographs were obtained in all patients. Computed tomography (CT) was used selectively to confirm nonunion when plain radiographs were equivocal, since CT has higher sensitivity than radiography for detecting incomplete bridging (23). Surgery was offered when a symptomatic patient had a confirmed nonunion or a symptomatic malunion with coxa vara and a femoral head judged viable on imaging. Patient selection Patient selection was based on the viability of the femoral head and the physiological status of the patient rather than on chronological age alone. Joint preservation was preferred whenever the femoral head was viable, because salvage arthroplasty after failed fixation carries a higher complication burden than primary arthroplasty (6, 7). Older patients were therefore not excluded; valgus osteotomy has been reported in patients up to the eighth and ninth decades (11, 24), and three patients in this cohort were older than 60 years with substantial comorbidity. The decision to preserve the joint in these patients was made jointly with the patient after discussion of the alternative of salvage arthroplasty. Surgical technique All operations were performed by the same senior orthopedic team. Under general or spinal anesthesia, the patient was placed in the lateral decubitus position on a standard radiolucent table. A lateral approach to the proximal femur was used. Where applicable, previous internal fixation devices were removed. A closing-wedge intertrochanteric valgus osteotomy was performed, with the wedge angle calculated preoperatively from standing anteroposterior pelvic radiographs to restore an NSA of 125°–140° and to correct LS as far as the local anatomy allowed. Osteotomy fixation was achieved with a DHS construct, supplemented when necessary with CS for additional neck fixation or an RP for medial buttress and rotational stability. The choice of construct was individualized on the basis of bone quality, residual neck length, and the pattern of the previous fracture. Autologous bone graft was used at the nonunion site or osteotomy gap when indicated. Postoperatively, all patients followed a standardized rehabilitation protocol of toe-touch weight-bearing for 6 weeks, progressive partial weight-bearing thereafter, and full weight-bearing once radiographic union was confirmed. Outcome measures The primary radiographic outcome was the NSA, measured on standardized anteroposterior pelvic radiographs by two independent observers blinded to the timing of imaging. The primary functional outcome was the Harris Hip Score (HHS) (25), recorded preoperatively and at the final follow-up visit. LS was measured clinically as the leg-length discrepancy in centimeters, using block testing with the pelvis leveled. Time to bony union was defined as the interval from osteotomy to radiographic bridging callus on ≥3 of 4 cortices on anteroposterior and lateral radiographs. Heterotopic ossification was assessed on final follow-up radiographs and graded according to the Brooker classification (26). Final follow-up assessment was carried out at a minimum of 12 months after surgery. Bony union, complications (including infection, implant failure, delayed union, heterotopic ossification, and recurrence of varus deformity), and any reoperation were also recorded. Statistical analysis Given the small sample size, descriptive statistics are presented as mean ± standard deviation (SD) with range for continuous variables and as counts for categorical variables. Preoperative and final follow-up values for NSA, HHS, and LS were compared using the Wilcoxon signed-rank test. A two-sided p value <0.05 was considered statistically significant. Analyses were performed with SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Results Patient characteristics Twelve consecutive patients (7 men, 5 women) with a mean age of 52.2 ± 20.0 years (range, 20–73) were included. The right hip was affected in 6 patients and the left in 6. The underlying cause of coxa vara was failed internal fixation in 8 patients (intertrochanteric fracture, n = 6; femoral neck fracture, n = 2) and fracture malunion in 4 patients (intertrochanteric fracture, n = 3; femoral neck fracture, n = 1). Among the 8 patients with failed prior fixation, the index implant had been a dynamic hip screw in 4, a cephalomedullary nail in 2, and cannulated screws in 2; the 4 patients with malunion had not undergone previous surgery. The osteotomy was stabilized with DHS plus CS in 6 patients, DHS plus RP in 3, DHS plus CS and RP in 2, and isolated DHS in 1. Individual demographic data, index implant, etiology, fixation construct, and outcome measures are summarized in Table 1. Table 1Demographic characteristics, etiologies, index implants, fixation strategies, and preoperative and final follow-up outcomes of the 12 patients with post-traumatic coxa vara treated with proximal femoral valgus osteotomy and dynamic hip screw fixation.Case No.SexAge (y)SideEtiologyIndex implantFixation constructTime to union (wk)NSA (°)HHSLS (cm)PreopFinalPreopFinalPreopFinal1F56RFFIF-ITFDHSDHS + CS + RP179612882972.00.02M72LFFIF-ITFCMNDHS + CS229613436793.01.03F54RFFIF-ITFDHSDHS + RP1811713966831.00.04M73LM-ITFNoneDHS + CS2410812717544.01.55M20RFFIF-FNFCSDHS + CS1410513864961.50.06M20LM-FNFNoneDHS137713052852.51.07F68RM-ITFNoneDHS + CS2010213242762.81.28F55LFFIF-ITFDHSDHS + RP1811013670891.20.09M65RFFIF-ITFCMNDHS + CS + RP269412929673.51.810F62LFFIF-ITFDHSDHS + CS199913158812.20.511M22RFFIF-FNFCSDHS + RP1511213668941.80.012M60LM-ITFNoneDHS + CS238812621613.01.0Mean ± SD—52.2 ± 20.0————19.1 ± 4.1100.3 ± 11.1132.2 ± 4.450.4 ± 21.180.2 ± 13.72.4 ± 0.90.7 ± 0.7DHS, dynamic hip screw; CS, cannulated screw; RP, reconstruction plate; CMN, cephalomedullary nail; NSA, neck–shaft angle; HHS, Harris Hip Score; LS, limb shortening; FFIF-ITF, failure of internal fixation following intertrochanteric fracture; M-ITF, malunion of intertrochanteric fracture; FFIF-FNF, failure of internal fixation following femoral neck fracture; M-FNF, malunion of femoral neck fracture; Preop, preoperative; SD, standard deviation.Continuous variables are presented as mean ± SD; categorical variables are presented as counts. Side: R, right; L, left. Time to union is defined as the interval from osteotomy to radiographic bridging callus on ≥3 of 4 cortices on anteroposterior and lateral radiographs. The index implant column lists the implant used at the original fracture fixation; “None” denotes a primary malunion not previously treated surgically. Demographic characteristics, etiologies, index implants, fixation strategies, and preoperative and final follow-up outcomes of the 12 patients with post-traumatic coxa vara treated with proximal femoral valgus osteotomy and dynamic hip screw fixation. DHS, dynamic hip screw; CS, cannulated screw; RP, reconstruction plate; CMN, cephalomedullary nail; NSA, neck–shaft angle; HHS, Harris Hip Score; LS, limb shortening; FFIF-ITF, failure of internal fixation following intertrochanteric fracture; M-ITF, malunion of intertrochanteric fracture; FFIF-FNF, failure of internal fixation following femoral neck fracture; M-FNF, malunion of femoral neck fracture; Preop, preoperative; SD, standard deviation. Continuous variables are presented as mean ± SD; categorical variables are presented as counts. Side: R, right; L, left. Time to union is defined as the interval from osteotomy to radiographic bridging callus on ≥3 of 4 cortices on anteroposterior and lateral radiographs. The index implant column lists the implant used at the original fracture fixation; “None” denotes a primary malunion not previously treated surgically. Radiographic outcome The mean NSA improved from 100.3° ± 11.1° (range, 77°–117°) preoperatively to 132.2° ± 4.4° (range, 126°–139°) at final follow-up, with a mean correction of 31.8° ± 9.0° (range, 19°–53°; p < 0.001). An NSA within the target range of 125°–140° was achieved in all 12 patients. The mean time to bony union was 19.1 ± 4.1 weeks (range, 13–26), with no nonunion observed. Functional outcome The mean HHS improved from 50.4 ± 21.1 (range, 17–82) preoperatively to 80.2 ± 13.7 (range, 54–97) at final follow-up, representing a mean gain of 29.8 ± 9.5 points (p < 0.001). At final follow-up, hip function was rated as excellent (HHS ≥90) in 3 patients, good (80–89) in 4, fair (70–79) in 2, and poor (<70) in 3. The three patients with a poor result (cases 4, 9, and 12) were among the older patients with longstanding malunion or failed fixation and the most severe preoperative functional impairment, although their NSA correction was comparable to that of the rest of the cohort. One further patient (case 2) achieved a fair score of 79 despite full radiographic correction; this patient had a pre-existing post-stroke hemiparesis of the operated limb, and the residual functional limitation reflected the neurological deficit rather than the surgical result, since the HHS is known to be depressed by pre-existing neurological impairment and comorbidity independent of the hip reconstruction and is an imperfect instrument in this setting (27–29). Limb-length discrepancy The mean LS decreased from 2.4 ± 0.9 cm (range, 1.0–4.0) preoperatively to 0.7 ± 0.7 cm (range, 0–1.8) at final follow-up, with a mean reduction of 1.7 ± 0.4 cm (p < 0.001). Limb length was fully restored (final LS = 0 cm) in 5 patients; the remaining 7 patients had a residual shortening of ≤1.8 cm, which was managed conservatively with a shoe-raise where required. Complications No deep infection, implant failure, recurrence of coxa vara, or femoral head osteonecrosis was observed during the follow-up period. No clinically significant heterotopic ossification was recorded; two patients had Brooker grade I ossification on final radiographs, neither of which restricted hip motion or required treatment. No patient required conversion to THA. Representative cases Four patients are presented in detail to illustrate the range of indications, comorbidities, fixation constructs, and outcomes encountered in this cohort, including one patient with a poor functional result. Case 6 (Figure 1) was a 20-year-old man with longstanding Wilson's disease and hepatic cirrhosis who sustained a left femoral neck fracture in a low-energy fall. He presented one month after injury with hip pain and reduced mobility, and was found to have a malunited femoral neck fracture with severe coxa vara (NSA 77°). Pre-existing hypertonia and choreiform gait were managed perioperatively with continued copper-chelating and hepatoprotective therapy, and a concurrent adductor tenotomy was performed at the time of valgus osteotomy. Fixation was achieved with a DHS alone. Over four years of follow-up the NSA improved to 130°, LS reduced from 2.5 cm to 1.0 cm, and the HHS improved from 52 to 85. Neurological function also improved, with normalization of muscle tone and resolution of gait abnormalities by the final review. Case 6 — A 20-year-old man with Wilson's disease and hepatic cirrhosis presenting with malunion of a left femoral neck fracture and severe coxa vara one month after a low-energy fall. Copper-chelating and hepatoprotective therapy were continued throughout a 4-year follow-up, with progressive resolution of pre-existing hypertonia and choreiform gait. (A,B) Preoperative anteroposterior and lateral pelvic radiographs showing left coxa vara (NSA 77°) with malunion of the femoral neck. (C,D) Immediate postoperative anteroposterior and lateral radiographs after proximal femoral valgus osteotomy with DHS fixation and concurrent adductor tenotomy. (E,F) Anteroposterior and lateral radiographs at 6 weeks, showing early callus formation across the osteotomy. (G,H) One-year follow-up radiographs demonstrating osseous union at the osteotomy and former fracture sites. (I,J) Two-year follow-up radiographs showing maintained correction and progressive remodeling. (K,L) Four-year follow-up radiographs confirming durable union and a preserved NSA. (M,N) Clinical photographs at 4 years demonstrating symmetrical bilateral weight-bearing. (O) Clinical photograph at 4 years demonstrating restored maximal hip flexion. Case 2 (Figure 2) was a 72-year-old man with grade IV residual left lower-limb weakness following an ischemic stroke. He presented nine months after a fall with persistent left hip pain and severe functional impairment (HHS 36) due to failed internal fixation of a previous intertrochanteric fracture, which had originally been treated with a cephalomedullary nail. Valgus osteotomy was performed with DHS and CS fixation. At 18 months postoperatively, the NSA had been restored from 96° to 134°, LS had reduced from 3.0 cm to 1.0 cm, and the HHS had improved to 79. The final functional result was rated fair, which reflected the patient's pre-existing neurological deficit rather than the radiographic correction; the affected limb remained weak from the stroke, and the HHS does not separate this deficit from hip-specific function (27, 28). Case 2 — A 72-year-old man with grade IV residual left lower-limb weakness following an ischemic stroke, presenting nine months after a fall with failed internal fixation of a left intertrochanteric fracture, coxa vara, and a preoperative HHS of 36. (A,B) Preoperative anteroposterior and lateral radiographs of the pelvis and left hip showing failed internal fixation, intertrochanteric nonunion, and coxa vara (NSA 96°). (C,D) Immediate postoperative anteroposterior and lateral radiographs after proximal femoral valgus osteotomy with DHS and CS fixation. (E,F) Anteroposterior and lateral radiographs at 6 weeks, showing early callus across the osteotomy. (G,H) Three-month follow-up radiographs. (I,J) Anteroposterior and lateral radiographs at 18 months showing bony union and a maintained NSA. (K,L) Clinical photographs at 18 months demonstrating symmetrical bilateral standing posture. (M) Clinical photograph at 18 months showing the patient seated comfortably in a standard chair. Case 1 (Figure 3) was a 56-year-old woman who presented two years after internal fixation of a right intertrochanteric fracture, originally treated with a dynamic hip screw, with persistent pain, established nonunion, coxa vara, and a 2 cm LS (preoperative HHS 82). Computed tomography (CT) confirmed incomplete bridging at the fracture site. She underwent valgus osteotomy with a DHS supplemented by CS and an RP. The NSA improved from 96° to 128°, and by six months postoperatively both the original nonunion site and the osteotomy had united, with limb length restored and the HHS increased to 97 — an excellent functional outcome. Case 1 — A 56-year-old woman presenting two years after internal fixation of a right intertrochanteric fracture with persistent nonunion, coxa vara, a 2 cm LS, and a preoperative HHS of 82. (A,B) Preoperative anteroposterior and lateral radiographs showing nonunion of the right intertrochanteric fracture with coxa vara (NSA 96°). (C) Preoperative CT confirming incomplete bridging at the fracture site. (D,E) Immediate postoperative anteroposterior and lateral radiographs after proximal femoral valgus osteotomy with DHS, CS, and RP fixation. (F,G) Anteroposterior and lateral radiographs at 6 weeks, showing early callus across the osteotomy. (H,I) Six-month follow-up radiographs demonstrating union at both the original nonunion site and the osteotomy. Case 4 (Figure 4) was a 73-year-old man who presented with a symptomatic malunion of a previously sustained intertrochanteric fracture that had been treated non-operatively. He had marked coxa vara (NSA 108°), a 4 cm LS, and the most severe preoperative functional impairment in the cohort (HHS 17). Valgus osteotomy was performed with DHS and CS fixation. The osteotomy united at 24 weeks and the NSA was corrected to 127°, but residual shortening of 1.5 cm persisted and the final HHS reached only 54, a poor result. The limited functional gain reflected his advanced age, longstanding deformity, and poor baseline mobility rather than failure of the osteotomy, which united without complication and maintained its correction. This case illustrates that mechanical correction does not fully overcome established disability in elderly patients with longstanding deformity, and underlines the importance of realistic preoperative counseling in this subgroup. Case 4 — A 73-year-old man with a symptomatic malunion of a non-operatively treated intertrochanteric fracture, presenting with severe coxa vara (NSA 108°), 4 cm limb shortening, and a preoperative HHS of 17. Despite uncomplicated union and angular correction, the final functional result was poor (HHS 54), reflecting advanced age and longstanding disability. (A,B) Preoperative anteroposterior and lateral radiographs showing malunion of the right intertrochanteric fracture with severe coxa vara (NSA 108°). (C,D) Immediate postoperative anteroposterior and lateral radiographs after proximal femoral valgus osteotomy with DHS and CS fixation. (E,F) Final follow-up radiographs demonstrating union at the osteotomy with a corrected NSA of 127° and residual limb shortening. Discussion This case series reports the outcomes of twelve consecutive patients who underwent proximal femoral valgus osteotomy with DHS-based fixation for post-traumatic coxa vara. Three findings stand out. All twelve osteotomies united. The mean NSA was restored from 100.3° to 132.2°, within the 125°–140° target band consistent with the Pauwels biomechanical principle (8). The mean HHS improved by 29.8 points, well above the published threshold for a moderate clinical improvement (19). These results compare favorably with the existing literature. Marti and colleagues reported a mean HHS of 91 at 7.1 years in 50 patients with femoral neck nonunion treated by intertrochanteric osteotomy (9). Magu and colleagues reported NSA correction from 100° to 132° and good or excellent outcomes in 32 of 39 patients (10). Bartoníček and colleagues reported uneventful union in 14 of 15 patients with intertrochanteric malunion or nonunion (11). Recent DHS-specific series report similar success rates. Subash reported NSA correction from 95° to 135° and HHS improvement from 72 to 91 in 15 malunited intertrochanteric fractures (17). Xu and colleagues reported NSA correction from 119° to 139° and union in 22 of 23 femoral neck nonunions (18). Our mean time to union of 19.1 weeks fits squarely within the published range of 12–24 weeks for this procedure (10, 12). The choice of fixation for valgus osteotomy and proximal femoral nonunion remains debated. The 95° fixed-angle blade plate has historically been the reference implant and retains an important role; Grünewald and colleagues recently reaffirmed its value in proximal femoral pseudarthrosis (30). Cephalomedullary nails offer load-sharing and a biomechanical advantage in trochanteric and subtrochanteric nonunion, and Rehme-Röhrl and colleagues reported superior six-year function with cephalomedullary nailing compared with revision blade plating in aseptic trochanteric and subtrochanteric nonunion (31). Against this background, the DHS occupies a useful middle ground. Its sliding mechanism delivers dynamic compression across the osteotomy, its lateral plate accepts supplementary CS or an RP for rotational control, and the implant is familiar to most trauma surgeons. The blade plate provides a more rigid fixed-angle construct but is technically demanding, while the nail is well suited to extracapsular nonunion but less so to the controlled angular correction required at the intertrochanteric level. In our cohort, construct choice was individualized: a DHS alone sufficed for a young patient with good bone stock, whereas supplementary screws or a plate were added when rotational stability or bone quality was a concern. We regard implant selection as a decision to be matched to fracture level, bone quality, and surgeon familiarity rather than dictated by a single default. The three patients with a poor functional outcome (cases 4, 9, and 12) shared two features. All were 60 years of age or older. All had severe preoperative functional impairment, with preoperative HHS values below 30. Their time to union was at the upper end of the cohort range. None of them developed nonunion or implant failure. The poor result therefore reflects baseline disability and the limits of pure mechanical correction in this group, rather than failure of the osteotomy itself. Two of these three patients (cases 9) had failed prior fixation with a cephalomedullary nail, and longstanding retention of a load-bearing implant before revision may contribute to localized bone loss through stress shielding, which can compromise the bone stock available for osteotomy fixation (32). This pattern matches the observations of Varghese and colleagues, who identified preoperative neck resorption and pre-existing functional decline as the strongest predictors of suboptimal outcome (13). Three further points deserve specific comment. First, limb length restoration was incomplete in seven patients, with a residual shortening of up to 1.8 cm. This is the expected consequence of a closing-wedge osteotomy and was managed with a shoe-raise where required. Patients should be counseled about this limitation preoperatively. Second, no clinically significant heterotopic ossification occurred; only two patients showed Brooker grade I changes, in keeping with the low rates reported after hip-preserving surgery, and routine prophylaxis was not used (26). Third, no AVN of the femoral head was observed during follow-up. This is consistent with the avascular-sparing nature of the osteotomy, which leaves the head fragment and its blood supply untouched. Several limitations apply. The cohort is small and heterogeneous, with both femoral neck and intertrochanteric pathology pooled together; we justify pooling on biomechanical grounds, since coxa vara is the common mechanical problem being corrected. The study is retrospective, single-center, and not randomized. There is no comparator group. Follow-up was relatively short, with a minimum of 12 months. This is an important limitation, because post-traumatic AVN of the femoral head may not become radiographically apparent for two to three years or longer after the index injury; the absence of AVN at the present follow-up is therefore not definitive, and longer surveillance is required before the joint-preserving result can be considered durable (33, 34). Functional assessment relied on a single observer-administered HHS without independent verification of inter-observer agreement. In summary, proximal femoral valgus osteotomy with DHS-based fixation reliably restores NSA, achieves union, and produces meaningful functional improvement in patients with post-traumatic coxa vara. The technique offers a defensible joint-preserving alternative to salvage THA, particularly in younger and middle-aged patients with a preserved femoral head. Outcomes are most consistent in patients without severe baseline functional impairment.