Feng Xiong ¹ Wenbin Zhang ² Feilong Lu ² Jie Feng ³ Lu Wang ² Yulu Xiang ⁴ Yongtao Wang ² Yimei Hu ^5*

• ¹ Department of Orthopedics, Jiang’an County Traditional Chinese Medicine Hospital, Yibin, China
• ² School of Clinical Medicine, Chengdu University of Traditional Chinses Medicine, Chengdu, Sichuan Province, China
• ³ College of Medicine, Southwest Jiaotong University, Chengdu, Sichuan Province, China
• ⁴ Department of Rehabilitation, Affiliated Hospital of Inner Mongolia Medical University, Hohhot, China
• ⁵ Hospital of Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan Province, China

Background: Due to the scarcity and high cost of MRI in resource-constrained regions, prompt diagnosis and treatment of rotator cuff tears remain problematic in these areas.Therefore, extensive research has been conducted to predict rotator cuff tears using simple and affordable anteroposterior radiographs. It remains unclear whether nonstandard anteroposterior radiographs would have a notable impact on the preciseness of the diagnosis.We analyzed patients treated for shoulder pain at hospitals. These patients underwent shoulder joint MRI and standard anteroposterior radiographs, were categorized into those with rotator cuff tears and a control group. We assessed whether the radiographs were standard anteroposterior radiographs using classification criteria from previous studies. Three assessors independently measured the acromiohumeral interval, upwards migration index, acromion index, critical shoulder angle, and doublecircle radius ratio in radiographic images. The intraclass correlation coefficient and receiver operating characteristic curves were used to assess measurement reliability and predictive capabilities of each predictive method for rotator cuff tears.Results: This study included 102 non-standard radiographs that met the research criteria for the measurement and analysis. The intragroup correlation coefficients for the acromiohumeral interval, upwards migration index, and double-circle radius ratio were above 0.7 (0.77, 0.71, 0.76), while those for the acromion index and critical shoulder angle exceeded 0.8 (0.86 and 0.87). In non-standard radiographs, the doublecircle radius ratio reliably predicted rotator cuff tears (p<0.05), contrary to the other methods (p>0.05). The areas under the receiver operating characteristic curves of the double-circle radius ratio, estimated by the three researchers for rotator cuff tears.This study found that non-standard radiographs significantly impaired the diagnostic performance of the acromiohumeral interval, upwards migration index, acromion index, and critical shoulder angle. Only the double-circle radius ratio maintained its predictive power (although this diminished capability may fall short of clinical relevance) and demonstrated high applicability. These findings indicate the need for researchers to prioritize the quality of radiographs and focus on reducing the sensitivity of the prediction method in relation to radiograph quality. The capability exhibited by the double-circle radius ratio warrants further investigation, to facilitate a simplified diagnosis of rotator cuff tears.