摘要目的:探讨肩关节解剖因素与中小型肩袖修补术后再撕裂率的相关性。方法:回顾性收集2017年6月至2019年6月民航总医院收治的55例中小型肩袖损伤行单排缝合术患者资料，包括年龄、性别、病程、体质指数、吸烟史、糖尿病史，术后6个月及术后1、2、3年的再撕裂率，术后3年肩关节Constant-Murley评分、加州大学洛杉矶分校肩关节评分（University of California，Los Angeles shoulder scale，UCLASS）以及基于CT测量的临界肩关节角（critical shoulder angle，CSA）、肩峰指数（acromial index，AI）。根据随访过程中是否发生再撕裂将55例分为再撕裂组和未撕裂组。采用单因素方差分析分别比较两组CSA、AI各随访时点差异，采用Fisher精确检验比较两组性别、吸烟史、糖尿病史，采用独立样本 t检验比较两组年龄、病程、术后次日CSA、AI、各随访时点UCLA评分、Constant-Murley评分的差异，将术后次日CSA与AI作为术后6个月、1年、2年、3年肩袖再撕裂的易感因素进行二分类变量logistic回归分析。绘制受试者工作特征（receiver operating characteristic，ROC）曲线评估术后次日CSA与AI对术后3年再撕裂的预测效能，采用Pearson相关性分析检验CSA、AI与术后功能恢复的相关性。 结果:55例术后6个月、1年、2年、3年的再撕裂率分别为4%、7%、13%、18%。再撕裂组各随访时点CSA或AI差异均无统计学意义（ P>0.05），未撕裂组各随访时点CSA或AI差异均有统计学意义（ F=14.16， P<0.001； F=4.64， P<0.001）。术后3年再撕裂组Constant-Murley评分、UCLA评分分别为（93.60±2.84）分、（32.30±1.49）分，未撕裂组分别为（92.11±4.10）分、（33.18±1.27）分，差异无统计学意义（ P>0.05）。术后次日CSA是术后1年、2年、3年肩袖再撕裂的危险因素[ OR=4.62，95% CI（1.01，21.06）， P=0.048；OR=7.071，95% CI（1.52，32.87）， P=0.013； OR=3.40，95% CI（1.42，8.12）， P=0.006]。术后次日CSA与AI对肩袖修补术后3年再撕裂有一定预测效能，且CSA特异度较AI更高。术后次日CSA、AI预测术后3年肩袖再撕裂最佳截断值分别为35.3°、0.69，AUC分别为0.87[ OR=3.40，95% CI（1.42，8.12）， P<0.001]、0.77[ OR=1.33，95% CI（0.87，2.02）， P=0.008]。CSA、AI与术后功能恢复无相关性（ P>0.05）。 结论:术后次日CSA与AI对中小型肩袖撕裂单排修补术后1~3年再撕裂均有预测效能，肩关节CSA与AI增大提示术后再撕裂概率增大，且CSA较AI预测特异性更高，两者与术后肩关节功能恢复无关。

abstractsObjective:To explore the correlation between the anatomical features of shoulder joint and the re-tear rate after surgical repair for small and medium-sized rotator cuff tears.Methods:From June 2017 to June 2019, 55 patients who were diagnosed with small or medium-sized rotator cuff tears and treated with arthroscopic single-row repair were enrolled. Demographics including age, sex, disease course, history of smoking and diabetes mellitus, re-tear rates, Constant-Murley score, University of California, Los Angeles score (UCLA) at 6-month, 1-year, 2-year and 3-year after operation were collected. Postoperative critical shoulder angle (CSA) and acromial index (AI) were measured and calculated based on CT scan. The patients were divided into two groups: patients who got re-tear history during follow-up were included into endpoint re-tear (ER) group, and those who got no re-tear history during follow-up were included into endpoint non-tear (EN) group. One-way Anova was used to compare the CSA\AI among different follow-up point. Fisher's exact test was used to compare sex, morbidity of smoking and diabetes between the ER and EN groups. Two independent samples t-test were used to compare age, disease course, CSA and AI at 1-day after operation, functional scores at each follow-up point between the two groups. Binomial logistic regression analysis was performed to test CSA and AI at 1-day after operation as the risk factors of rotator cuff re-tear at 6-month, 1-year, 2-year and 3-year after operation. The predictive efficacy of CSA and AI at 1-day after operation on re-tear rate at 3-year after operation were evaluated by receiver operating characteristic (ROC) curves, Pearson correlation analysis was used to evaluate the correlation between postoperative CSA/AI and postoperative functional recovery. Results:The CSA and AI of ER group were insignificantly different among all follow-up point ( P>0.05), the CSA and AI of EN group were significantly different among all follow-up point ( F=14.163, P<0.001; F=4.635, P<0.001). The re-tear rates at 6-month, 1-year, 2-year and 3-year after operation were 3.6%, 7.3%, 12.7%, 18.2%. The Constant-Murley score and UCLA scores of ER group at 3-year after operation were 93.60±2.84 and 32.30±1.49, respectively while in EN group, they were 92.11±4.10 and 33.18±1.27, respectively, there were no difference of the Constant-Murley score and UCLA score between ER and EN group at 3-year after operation ( P>0.05). CSA at 1-day after operation was the risk factor to re-tear at 1-year, 2-year and 3-year after operation [ OR=4.622, 95% CI (1.01, 21.06), P=0.048; OR=7.071, 95% CI (1.52, 32.87), P=0.013; OR=3.40, 95% CI (1.42, 8.12), P=0.006]. CSA and AI at 1-day after operation had certain predictive efficacy for rotator cuff re-tear at 3-year after rehabilitation, and CSA was more specific than AI, the optimal cutoff values of CSA and AI at 1-day after operation for predicting rotator cuff re-tear at 3-year after operation were 35.3°and 0.69, the AUC were 0.87 [ OR=3.40, 95% CI (1.42, 8.12), P<0.001]、0.77 [ OR=1.33, 95% CI (0.87, 2.02), P=0.008] respectively. CSA and AI had no relationship with postoperative functional recovery. Conclusion:Greater CSA and AI were predictive factors of small and medium-sized rotator cuff re-tear 1-3 years after surgery with CSA being more specific than AI. However, CSA and AI had no relationship with postoperative functional recovery.