Displaced femoral neck fractures in young adults account for 11 % of all femoral neck fractures [1], which typically result from high-energy trauma and are frequently accompanied by posterior medial cortical comminution [2]. Displaced femoral neck fractures have a much greater complication risk than nondisplaced fractures [3], which are characterised by inadequate fixation and challenging treatment; approximately one-third of patients require major reconstructive surgery [4]. According to the literature, the incidence of femoral neck shortening, fracture nonunion, and avascular necrosis of the femoral head after surgery for young displaced femoral neck fractures is 28.9 %−54 %, 14.7 %−30.1 %, and 12.3 %−17.5 % [2,[4], [5], [6], [7]], respectively, of which femoral neck shortening is the most frequent complication after internal fixation of femoral neck fracture, and its incidence is correlated with the type of fracture, comminution of the posterior medial cortex, and type of internal fixation [2]. Slobogean et al. showed that severe femoral neck shortening was associated with worse hip function in a follow-up study involving 102 patients [8].

Internal fixation remains the mainstay of therapy for young patients with displaced femoral neck fractures, despite the high incidence of postoperative complications [9]. Currently, internal fixation devices for the treatment of femoral neck fractures include cannulated cancellous screws (CCS), dynamic hip screws (DHS), and the femoral neck system (FNS); however, the choice of internal fixation remains controversial [10].

CCS is widely used in the treatment of femoral neck fractures in young patients because of its simplicity, minimally invasive nature, and minimal tissue damage [11]. The FNS has been used for femoral neck fractures because it provides good fixation and is minimally invasive. An experimental biomechanical study by Stoffel et al. demonstrated that FNS is a reliable alternative to DHS in unstable Pauwels type III femoral neck fractures and is comparable to DHS in terms of stability [12]. However, owing to the short duration of clinical application of FNS, there is a lack of high-level evidence-based medical evidence to validate its reliability. The occurrence of postoperative femoral neck shortening in young patients with displaced femoral neck fractures undergoing FNS fixation has not yet been reported in the literature. To compare femoral neck shortening after FNS versus CCS fixation in young patients with displaced femoral neck fractures, we retrospectively analysed the data of 225 patients with Garden stage III and IV young femoral neck fractures admitted between September 2019 and January 2022.

We conducted a retrospective analysis of patients aged 18–65 years with displaced femoral neck fractures who underwent surgical treatment with FNS or CCS fixation at our institution between September 2019 and January 2022.

Inclusion criteria: 1) 18 ≤ age ≤ 65 years; 2) unilateral femoral neck fracture; 3) fractures managed surgically using the FNS or CCS; 4) follow-up ≥ 18 months; 5) Garden stage III and IV fractures.

Exclusion criteria: 1) multiple fractures; 2) pathological fractures; 3) injury to admission > 2 weeks; 4) congenital hip dysplasia; 5) incomplete follow-up information.

Setting: This study was conducted in a single centre, which is a tertiary hospital and level I trauma centre providing healthcare to a population of 8.5 million with 240 beds, in the Department of Orthopaedic Trauma. A total of six surgeons were involved in this study, and the surgeries were performed by two trained attending surgeons with more than 10 years of experience in orthopaedic trauma, two deputy chief surgeons with more than 20 years of experience in orthopaedic trauma who co-directed the surgeries, and two resident surgeons who served as surgical assistants.

The procedure was performed with the patient in a hemilithotomy position under general anaesthesia. Two Kirschner wires were pre-positioned in the proximal femur along the axis of the femoral neck, not exceeding the fracture line, and one threaded bone traction screw was inserted 2–3 cm below the greater trochanter, combined with the traction of the lower extremities. If the reduction was unsatisfactory on fluoroscopy, a small incision was made anteriorly in the femoral neck to correct the reduction with the use of a screwdriver, and two Kirschner wires were inserted into the femoral head as a joystick to restore the femoral head rotation if it was still rotated under C-arm fluoroscopy. After satisfactory reduction under C-arm fluoroscopy, two prepositioned Kirschner wires were inserted through the fracture line for temporary stabilisation of the femoral head. When the above closed reduction steps have been attempted and satisfactory reduction has not been achieved, open reduction should be considered.

A 130° guide was placed against the femur, and a Kirschner wire was inserted as a guide pin to a depth of 5 mm below the cartilage. Once the position was satisfactory, the depth was measured, bolt channel was drilled along the direction of the guide pin, the correct bolt length was selected, and the locking plate was assembled and gently hammered along the guide pin's direction to place the plate in the centre of the femur. Through the handle guides, locking and anti-rotation screws were installed.

Under C-arm fluoroscopy, three guide wires were positioned parallel to the femoral head in an inverted triangle along the axis of the femoral neck. After the guide wires were in a satisfactory position, three 7.3 mm titanium cannulated cancellous screws (16 mm or 32 mm thread length) were screwed into the femoral head. The threads of the screws should be completely over the fracture line, the tips of the screws should reach the subchondral bone within 5 mm, and the three cannulated cancellous screws should be as close to the cortex as possible.

None of the patients underwent preoperative skin or bone traction, and they were all permitted to consume a moderate amount of a light beverage two hours before surgery. The Visual Analogue Scale (VAS) was used to rate pre-and postoperative pain, and depending on the intensity of pain, oral or intravenous analgesics were administered to the patients. Cefazolin was administered prophylactically 30 min before and 24 h after surgery; water was initiated after awakening, and a gradual transition to a regular diet was enabled in the absence of gastrointestinal symptoms such as nausea and vomiting; prophylactic anticoagulation with low molecular heparin was initiated on admission and discontinued 12 h before surgery; anticoagulation was continued postoperatively; quadriceps contraction training and ankle pumping exercises were performed to prevent deep vein thrombosis; and patients were allowed to engage in full weight-bearing when the clinical signs and imaging suggested that the fracture had healed. Outpatient follow-ups were carried out at 1 month, 3 months, 6 months, 12 months, and 18 months postoperatively.

Two observers measured the distance of the lateral protrusion of the screws on anteroposterior radiography using the method described by Sepehri et al. [13] (Fig. 1). The true value of femoral neck shortening XR was calculated by measuring the cancellous screw's protruding length XM from the bone surface along the screw axis; the actual screw diameter D is known, and multiplying the lateral protruding length XM by the ratio of the actual screw diameter D to the measured screw diameter DM.XR(mm)=DDM×XM

Similar to the CCS group, both observers measured the distance X0 from the end of the bolt to the lateral plate using immediate postoperative anteroposterior radiography, and the diameter D0 of the bolt was measured simultaneously. The distance Xn from the end of the bolt to the lateral plate on anteroposterior X-rays at the respective follow-up times and the diameter of the bolt, Dn, were measured in the same way, and the actual diameter of the bolt, DR, was known, the distance X'R, which is the distance at which the end of the bolt slips, was the actual value of femoral neck shortening (Fig. 2, Fig. 3).X′R(mm)=DR×(X0D0−XnDn)

The quality of the reduction was evaluated using the garden alignment index according to the criteria described in Table 1 [14,15].

SPSS version 26.0 (IBM, Armonk, NY, USA) was used for the statistical analysis. Initially, the measurement data were screened by conducting Shapiro–Wilk normality tests for normality assumption to determine the choice of parametric or nonparametric methods for statistical analysis. Age, BMI, time from injury to surgery, duration of surgery, intraoperative blood loss, follow-up time, length of hospital stay, Harris hip score (HHS), and femoral neck shortening at each follow-up did not conform to the normality assumption, and the results were reported as the median and interquartile range (IQR: percentile 25 to percentile 75). Because the two groups were independent samples, statistical differences between the two groups were compared using the Mann-Whitney U test. Sex, mechanism of injury, injured side, garden classification, reduction method, and reoperation were considered enumeration data, and the results were reported as numbers and percentages. As the total sample size was > 40 cases, statistical differences between the two groups were compared using the chi-square test. The quality of reduction and extent of femoral neck shortening were ranked data, the results were reported as numbers with percentages, and statistical differences between the two groups were compared using the Mann-Whitney U test. GraphPad Prism software 9.4.0 (GraphPad Sofware, CA, USA) was used to analyse the data and draw the figures. P < 0.05 (2-tailed) was considered statistically significant for all analyses.