To compare mid- to long-term outcomes of the three major capsular management strategies—capsule preserved (CP), capsule repaired (CR) and capsule unrepaired (CU)—following hip arthroscopy (HAS) for femoroacetabular impingement syndrome (FAIS).
A systematic search was conducted in PubMed, Embase, CENTRAL, and Epistemonikos for studies published up to 31 May 2025. Studies were eligible if they reported ≥2-year outcomes after HAS with clearly described capsular techniques. Outcomes included the modified Harris Hip Score (mHHS), Hip Outcome Score—Activities of Daily Living (HOS-ADL) and Sports Subscale (HOS-SSS) and Visual Analogue Scale (VAS) for pain. Multilevel meta-analyses were performed using random-effects models with Hartung–Knapp adjustment. Failure-related outcomes (complications, reoperations, total hip arthroplasty [THA] conversions) were inconsistently reported across studies and could only be summarised descriptively.
Seven studies with a total of 948 patients were included. Pooled mean mHHS was 81.14 (95% confidence interval [CI]: 77.72–84.56), with CP: 84.90, CR: 80.66 and CU: 80.25 (p = 0.67). HOS-ADL pooled mean was 87.94 (95% CI: 84.79–91.09), with CP: 89.30, CR: 87.36, CU: 88.31 (p = 0.90). HOS-SSS pooled mean was 78.59 (95% CI: 72.69–84.48), with CP: 75.70, CR: 78.89, CU: 80.22 (p = 0.84). VAS pooled mean was 2.42 (95% CI: 2.08–2.77), with CP: 2.46, CR: 2.27, CU: 2.76 (p = 0.51). No statistically significant subgroup differences were detected. CR showed the highest cumulative numbers of reoperations and THA conversions, whereas CP and CU demonstrated lower but heterogeneous event counts.
CP, CR and CU demonstrated comparable mid- to long-term patient-reported outcome measure (PROM) outcomes following HAS for FAIS. Failure-related events varied descriptively across techniques, but inconsistent reporting prevented comparative survivorship assessment.
Level II, systematic review and multilevel meta-analysis of predominantly Level III studies, with additional contributions from Level I and II studies.
Achilles tendon ruptures (ATR) are severe injuries in professional basketball, yet in-game mechanisms and biomechanical patterns remain insufficiently described. The purpose of the study is to characterise biomechanical patterns and basketball-specific contexts of ATR using sequential video analysis.
A systematic search identified 27 confirmed ATR cases between 1970 and 2025 in the NBA league. Contextual factors (court location, playing situation, movement and speed) were analysed for all cases. Frame-by-frame biomechanical assessment was conducted for 17 injuries (63%) with adequate video quality, documenting trunk, hip, knee and ankle angles of the injured limb at initial contact (IC) and the injury frame (IF).
All ruptures occurred during closed-chain movements, mostly with noncontact mechanisms (81.5%), during offensive play actions (85.2%), at low horizontal speeds (63%) and moderate vertical speeds (44.4%). The most frequent action was the take-off/acceleration phase of running (74%), followed by the beginning jump (14.8%) and landing (7.4%). At the IC, players were in forefoot contact, the sagittal plane orientation of the trunk was 18.2° flexed, while the hip was 12.3° extended, the knee was 42.2° flexed and the ankle was 5.6° plantar flexed on the injured side. At IF, players displayed a more flexed torso (27.5°, 9.3° change from IC; p < 0.001), a more extended hip (26.8°, 14.5° change from IC; p = 0.001), a less flexed knee (26.7°, 15.5° change from IC; p = 0.004) and a more dorsiflexed ankle (41.6°, 47.2° change from IC; p < 0.001).
Our study described basketball-specific movement patterns among NBA players who sustained ATRs. Supported the observations of forefoot loading and rapid dorsiflexion of the ankle joint, and introduced new concepts, extension movement of lower limb elements proximal to the ankle joint and proximal elongation of the musculotendinous complex. These findings enhance understanding of basketball-specific ATR mechanisms and may inform targeted prevention strategies such as combining eccentric calf training with functional exercises focusing on lumbopelvic and knee stability, while targeting explosive forefoot contact acceleration manoeuvres.
Level IV, retrospective case series.
The purpose of this study was to compare the accuracy, calibration, reproducibility and operating cost of seven large language models (LLMs)—including four newer models capable of using advanced reasoning techniques to analyse complex medical information and generate accurate responses—on text-only orthopaedic multiple-choice questions (MCQs) and to quantify gains over GPT-4.
From Orthobullets, 702 unique, non-image MCQs (drawn from AAOS Self-Assessment Examinations, Self-Assessment-Based Questions and Orthopaedic In Training Examination-Based Questions banks) were extracted. Each question was submitted to the following LLMs: OpenAI o3, Anthropic Claude Sonnet 4, Claude Opus 4 (with/without ‘Extended Thinking’) and Google Gemini 2.5 Pro. Additionally, OpenAI's GPT-4, GPT-4o and the open-weight Gemma 3 27B served as comparators. The primary outcome was overall accuracy. The secondary outcomes were topic and difficulty-stratified accuracy, calibration (expected calibration error [ECE] and Brier score), reproducibility (flip rate on a retest question subset), latency, token use and cost. Statistical tests included paired McNemar, Cochran Q, ordinal logistic regression and Fleiss κ (Bonferroni-adjusted α = 0.05).
GPT-4 achieved 69.7% accuracy (95% CI = 66.2–72.9). All four reasoning-optimised models scored ≥14 percentage points higher (p < 3.3 × 10−15); OpenAI o3 led with 93.6% (95% CI = 91.5–95.2), which represents a 34% relative error reduction. Accuracy tended to decline with question difficulty, yet the reasoning advantage persisted in every difficulty stratum. Claude Opus 4 showed the best calibration (ECE = 0.023), while GPT-4 exhibited overconfidence (ECE = 0.215). All models except Gemma 3 27B exhibited non-zero flip rates. Median query time: 0.9 s (Gemma) to 15.9 s (Gemini 2.5 Pro). Cost: 0 to 29.9 USD per 1000 queries.
Reasoning-optimised LLMs now answer text-based orthopaedic exam questions with high accuracy and substantially better confidence calibration than earlier models. However, persistent stochasticity and large latency-cost disparities may limit clinical deployment.
N/A.
Tibial deflexion osteotomy (TDO) is performed to correct an increased posterior tibial slope (PTS). Unintended rotation of the osteotomy and the hinge axis (HA) orientation can result in a postoperative deviation of the mechanical medial proximal tibial angle (mMPTA). This study aimed to investigate how internal and external HA rotations affect postoperative mMPTA.
Three-dimensional (3D) bone models of ten patients with increased PTS were used to simulate infratuberositary TDO with different HA orientations and closing distances. Postoperative changes in mMPTA were analysed.
In total, 440 TDOs were simulated. The PTS changed by 0.9 ± 0.0° per mm of closing distance. TDO perpendicular to the coronal plane of the long-leg radiograph showed no significant change in the postoperative mMPTA. Internal and external rotation of the HA resulted in significant changes in postoperative mMPTA, with absolute changes up to 4.5° ± 0.5°.
A TDO oriented perpendicular to the leg's coronal plane preserves the preoperative mMPTA and therefore avoids unintended coronal correction. The mMPTA changed significantly with a rotation of the HA of only 5° and exceeded a postoperative change of ≥2° with 15° of HA rotation.
N/A.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：