Latarjet in lateral decubitus vs. beach-chair: effect on screw inclination and coracoid graft positioning—A single-surgeon comparative series Authors: Ali Kalhor Published: 2026 ABSTRACT AbstractBackgroundThe Latarjet procedure has traditionally been performed in the beach-chair position. Optimizing graft position is one of the key factors to reduce failure, prevent redislocation, and minimize the risk of future arthritis.MethodsSeventeen consecutive Latarjet procedures were reviewed (8 beach-chair; 9 lazy lateral decubitus) performed between January 2020 and July 2025. All procedures were carried out with 2 partially threaded screws through the plate (Arthrex) and labrocapsular repair to position the bone block extra-articularly. In the lateral position, the procedure was performed with a modified surgical sequence, with coracoid harvesting undertaken first, followed by glenoid preparation. All patients underwent computed tomography scanning at 4 months to measure screw inclination angle, vertical and medial–lateral position of the graft, as well as early resorption and bony union.ResultsAll patients were male. The average age of the beach-chair and lateral groups was 27.4 and 32 years, respectively. The mean screw-to-glenoid face angle was 40.5° in the beach-chair group compared to 28.0° in the lateral group (P = .008). The average medial/lateral positioning of the graft was +4.6 ± 1.2 in the beach-chair versus 1.0 ± 2.8 mm in the lateral group. Vertical graft position and resorption of the superior aspect of the graft did not show significant differences between the 2 groups.ConclusionThe Latarjet procedure performed in the lateral decubitus position and changing surgical flow were associated with improvement of screw angle and medial/lateral graft positioning compared with the beach-chair position.Keywords: Latarjet procedure, Beach-chair, Lateral decubitus, Screw inclination angle, Shoulder instability, Optimal bone block position Recurrent shoulder instability remains a clinical challenge, predominantly in cases of anterior glenoid bone loss.12 Historically, Bigliani2 proposed the Latarjet procedure for cases with greater than 25% anterior bone loss.21,22 However, the concept of “subcritical bone loss” has expanded the indications for this procedure. Shaha and colleagues18 demonstrated that bone loss between 13.5% and 20% could cause a clinically meaningful decline in quality of life, even in the absence of recurrent dislocation. Similarly, Shin19 identified a threshold of 17.3% bone loss as crucial for recurrent instability, while Yamamoto20 reported that patients with greater than 17% bone loss experienced significantly lower Western Ontario Shoulder Instability Index scores. At the same time, the understanding of bipolar lesions and the “on-track/off-track” concept17 for Hill-Sachs defects has further broadened the role of the Latarjet procedure. Despite its effectiveness, outcomes are closely tied to graft positioning. Mispositioning of the bone block has been reported in up to 20–40% of cases with the open technique.5,24 Optimal graft placement has been described differently: Lafosse9 suggested the ideal position between the 3 and 5 o'clock positions, Boileau3 emphasized placing the coracoid bone block below the glenoid midline, and Zhu24 highlighted the risks of positioning the graft too high or too low relative to the glenoid rim. Together, these evolving insights highlight not only the importance of careful patient selection but also the critical role of precise graft positioning in optimizing outcomes for patients undergoing the Latarjet procedure. This study reports a consecutive case series of Latarjet procedures performed before and after the transition in surgical positioning from the beach-chair to the lateral decubitus position. Methods This series included 17 consecutive patients who underwent the Latarjet procedure for recurrent anterior shoulder instability and anterior bone loss of more than 20% between 2019 and 2025. Of these, 8 procedures were performed in the beach-chair position (mean age: 27.4 years) and 9 in the lazy lateral decubitus position (mean age: 32 years), following a change in my practice. All patients in the series were male. All procedures were performed under general anesthesia using a standard deltopectoral approach. Cephalic vein was mobilized laterally. The subscapularis was split in line with its fibers to approximately 50%, a single-limb blunt Gelpi retractor was placed to maintain the split, and then a vertical incision was continued down on the capsule in a lateral-to-medial fashion to reach the labrum, preserving the labral layer circumferentially and elevating it from the glenoid vault by making an elliptical incision with diathermy perpendicular to the capsular split and parallel to the labrum. Using a Cobb elevator, the labrum mobilized laterally and the glenoid vault exposed. A sharp Bankart retractor was placed on the medial border of the glenoid to maintain the view while a Fukuda retractor was placed on the posterior aspect of the glenoid to keep humerus out of surgical field. Glenoid vault prepared with curette and all scar tissue and remnant of previous bony Bankart was removed. In patients operated on in the beach-chair position, glenoid preparation was carried out prior to harvesting a 2-cm coracoid graft, which was subsequently fashioned with the Arthrex clamp using the congruent arch technique. Maintaining the conjoint tendon allowed the surgeon to use the Kolbel retractor to expose the subscapularis. In contrast, in patients operated on in the lazy lateral decubitus position, the coracoid graft was harvested and fashioned first with the Arthrex clamp, as the author found this step improved subscapularis and glenoid exposure in the lazy lateral position. The subscapularis and labrum management were similar to the beach chair position otherwise. With the exception of one case, a wedge plate was used in all procedures. A zero-offset Arthrex guide was utilized for coracoid fixation, which allowed for consistent placement of the coracoid graft flush with the glenoid. A capsulolabral repair was also performed in every case, securing the capsule to the native glenoid with 2 suture anchors (Arthrex FiberTak anchor, Naples). This was done to restore the soft tissue restraint in addition to the bony augmentation. Patients were kept in a sling for 6 weeks with safe range of motion exercises, then progressed to a full range of motion. At 4 months post-operatively, each patient underwent a computed tomography (CT) scan to evaluate the status of graft healing and positioning. Specific measurements included screw angle and graft position, classified according to Zeng et al.23 In their method, screw trajectory was measured between a line passing through the screw and a line parallel to the face of the glenoid in an axial CT view. The graft position was divided into 3 groups23: high, good, and low positions based on whether the lower aspect of the graft projected below the glenoid (6 o'clock position) (low), stayed in the lower half of the glenoid (ideal), or projected above the 4 o'clock position (high) (Fig. 1A). Medial or lateral placement of the bone block was assessed as described by Kany et al8 (Fig. 1B), using an axial CT scan cut and measuring the highest point of the graft in relation to a line through the face of the glenoid. The screw angle was measured using a line parallel to the glenoid face and another parallel to the angle of the screw (Fig. 1C). The degree of graft resorption was evaluated following the criteria proposed by Zhu et al.24 If a screw head was buried into bone it classified as no resorption, if screw head exposed, grade I resorption, if screw shaft exposed, grade II and whole graft resorption classified grade III. (A) Zeng H-Z method. P represents the lowest point of the coracoid bone block on the glenoid en face three dimensional reconstruction. Oval drawing with 6 o'clock (red line), 4 o'clock (green line), and the midpoint of the glenoid (yellow line). (A) Point P below the 6 o'clock was considered too low, (B) located between 4 and 6 o'clock was ideal, and (C) above the 4 o'clock was considered too high. (B) Axial computed tomography image demonstrating the Latarjet graft. The blue dashed line indicates the linear method used in this study, while the 2 circles represent the circular method. The yellow circle outlines the osseous glenoid; the red circle represents the 3-mm cartilage thickness, which corresponds to the threshold for acceptable lateralization as described by Kany et al6. The yellow arrow indicates a medialized graft, and the red arrow indicates a lateralized graft. (C) Screw angle measured using a line parallel to glenoid face and another parallel to the angle of the screw. Statistical analysis was performed using 2-tailed Student's t-test and chi-square analysis with the assumption of normal distribution of the data. Results The mean screw angle was significantly improved in the lateral decubitus position by 12° (28.0 ± 8.2° vs. 40.5 ± 8.7°). The Student's t-test showed a significant difference between groups (P = .008, Fig. 2). Assessment of graft position using linear measurements demonstrated an average medialization/lateralization of 1.0 ± 2.8 mm in the lateral group compared with +4.6 ± 1.2 mm in the beach-chair group (P < .004; Fig. 3). Comparison of mean screw inclination angles between the lateral decubitus and beach-chair positions. The mean angle in the lateral decubitus group was 28°, compared with 40.5° in the beach-chair group (P < .008). Error bars represent standard deviation. Linear measurement of graft lateralization on axial computed tomography scans. The mean lateralization in the lateral decubitus group was 1.01 mm compared with 4.61 mm in the beach-chair group (P < .004). Error bars represent standard deviation. Chi-square analysis showed no statistically significant difference in graft position distribution between groups (Fig. 4A) (χ2(2) = 3.35, P = .19). However, the effect size was moderate-to-large (Crameˊr's V = 0.44), suggesting a potential association. Given the small sample size, the study was likely underpowered for this categorical comparison, and a type II error cannot be excluded. Bone resorption was observed in both groups, with no significant difference between positions (Fig. 4B). All grafts achieved bony union. (A) Distribution of graft positioning classified as optimal, high, or low on post-operative computed tomography scans in the beach-chair and lateral decubitus groups. (B) Comparison of superior graft bone resorption between the 2 groups, showing no significant difference. Discussion To the best of my knowledge, this is the first study to compare the beach-chair and lateral decubitus positions in a single-surgeon series. Both screw angle and graft medial/lateral positioning improved when changing from the beach-chair to the lateral position. The author believes this is due to the fact that lateral positioning allows for a “lazy lateral” alignment, making the glenoid parallel to the floor, while gravity assists in dispersing the anterior chest wall muscles, thereby creating more space for flush graft placement and better screw angulation. The surgical workflow varied in each positioning. In the lazy lateral position, the coracoid process was harvested first. This approach facilitated improved subsequent exposure of the glenoid, with gravity assisting in retracting the chest wall soft tissues away from the operative field, enhancing visualization and working space. In contrast, in the beach-chair position, a Kolbel retractor was utilized, which relies on an intact conjoint tendon, with one limb engaging the conjoint tendon and the other positioned beneath the deltoid. Consequently, in the beach-chair position, the coracoid was not harvested prior to glenoid exposure, as preservation of the conjoint tendon was necessary to maintain effective retraction. The change in workflow from beach chair to lazy lateral might have had a confounding effect on the author's findings. The optimal method for evaluating coracoid bone block placement after the Latarjet procedure remains debated. While medial and lateral placement is less controversial, vertical positioning continues to be discussed.5 Lafosse et al9 recommended placement between the 3–5 o'clock position, while Boileau et al3 advocated positioning below the horizontal equator. This study followed the technique described by Zhu,24 which is more quantitative than earlier recommendations. Zhu proposed that a graft is too high if the midpoint of its long axis lies above the glenoid midline and too low if it falls below the inferior rim (Fig. 1A). Nourissat et al,14 in a biomechanical study, demonstrated that a 4 o'clock bony-block location yielded the most effective “blocking” function of the graft; these findings correlate with our clinical findings and the absence of redislocations, regardless of some nonideal graft positioning in both the beach-chair and lateral groups. All grafts in this series covered the 4 o'clock position, which is consistent with the clinical finding of no further dislocations among them. Early results in both the beach-chair and lateral groups showed a higher rate of nonideal vertical graft positioning. Although the number of cases was limited, improvement in positioning was observed toward the later part of each group and a better trend across the lateral cases. However, due to the limitation of the sample size in this study, graft position findings need to be interpreted with caution. With respect to screw angle, the consensus is that a lower inclination angle (closer to parallel to the glenoid face) is preferable, as it reduces the risk of graft overhang and the development of future arthritis.10 Lee9 suggested that an angle >25° is associated with poor bone fixation and healing. Boileau3 measured the mean inclination angle between the screw and glenoid surface on axial CT scans, reporting an average of 20.2° ± 8.9° (range, 0°–37°). Similarly, Kany6 measured screw inclination on axial CT, finding a mean angle of 21° (standard deviation, 12; range, 0°–46°), with 22.1% (21 cases) exceeding 30°. Hsu8 performed a biomechanical cadaveric study and concluded screw angle changes from 15 to 30° significantly increase interface displacement of screw and graft. Kany6 reported on medial/lateral graft placement with an average bone-block position of 0 mm (standard deviation 4 mm; range, −10 to +8 mm). Of 95 cases, 68.4% were within the target range (mean, 0 ± 2 mm; range, −5 to +3 mm), 24.2% were lateralized (mean overhang, 5 ± 2 mm; range, +3 to +8 mm), and 7.4% were medialized (mean, −8 ± 1 mm; range, −10 to −6 mm). Interestingly, they found no significant learning-curve effect on positioning or screw angle. The author used a linear technique to measure graft lateralization. As Kany6 mentioned in his study, this technique does not account for the 2 mm cartilage cap. Alternatively, the circle technique has been described9 to calculate a 2–3 mm cartilage gap; therefore, a tolerance of up to 3 mm lateralized graft is acceptable before it starts impinging on the humerus (Fig. 1B). The author contests this concept based on the fact that many of these recurrent dislocation patients have a degree of cartilage thinning and do not necessarily have 2-3 mm of anterior glenoid cartilage cap. In my series, 6 of 8 beach-chair cases showed grafts positioned too laterally (>3 mm overhang), compared with 2 of 9 in the lateral group. Graft overhang has been associated with an increased risk of osteoarthritis after Latarjet procedures. Salem16 performed a systematic review of 16 studies (8 intra-articular and 8 extra-articular grafts), with follow-up classified as short-term (<2 years), mid-term (2–5 years), and long-term (>5 years). Intra-articular grafts showed osteoarthritis (OA) progression rates of 25% and 5.4% in 2 short-term studies,7,15 while extra-articular grafts reported 0% and 1.9%.1,24 At midterm, intra-articular cases showed 21% OA progression,13 versus 9% in extra-articular grafts.3 Long-term results revealed 23.5% OA in intra-articular grafts11 compared with 5.2% in extra-articular grafts.25 Notably, Mizuno et al11 found that 13% of grafts projected laterally, which was a significant risk factor for OA at 20-year follow-up. Conversely, Bouju et al4 reported 12.1% lateralized bone blocks at a 13-year follow-up, without a significant link to OA development. Although this study did not evaluate OA progression, the extra-articular positioning of the grafts may provide some protective effect against future arthritis. No redislocation occurred in the first four months of follow-up, and no perioperative complications were recorded. Limitations This is a retrospective Level 3 study with a small sample size, lacking clinical follow-up and range of motion data. Although these limitations are acknowledged, the finding that simple changes in patient positioning and workflow can improve screw angulation and graft placement is technically relevant and worth sharing with the orthopedic community. Conclusion The Latarjet procedure performed in the lateral decubitus position and the changing of workflow were associated with an improvement of medial/lateral graft and screw position compared with the beach-chair position in a single-surgeon case series. Declaration of generative AI and AI-assisted technologies in the writing process During the preparation of this work, the author used ChatGPT in order to edit English and making sure references are ordered in the requested JSES format. After using this tool/service, the author reviewed and edited the content as needed and takes full responsibility for the content of the published article. Disclaimers: Funding: No funding was disclosed by the author. Conflicts of interest: The authors, their immediate family, and any research foundation with which they are affiliated did not receive any financial payments or other benefits from any commercial entity related to the subject of this article. Footnotes This study has ethical approval of the Mater Misericordiae Ltd Human Research Ethics Committee with reference number: (HREC/MML/111545 [V3]), project ID: 111545. References