鼻漏
医学
脑脊液鼻漏
金标准(测试)
诊断准确性
假阳性悖论
核医学
脑脊液
诊断试验
鼻窦炎
内科学
外科
儿科
机器学习
计算机科学
作者
Jacob G. Eide,William D. Mason,Hussein Mackie,Bernard Cook,Amrita Ray,Karam Asmaro,Adam Robin,Jack Rock,John R. Craig
摘要
Objective Unilateral thin clear rhinorrhea (UTCR) may represent a variety of pathologies including cerebrospinal fluid (CSF) rhinorrhea. Beta‐2 transferrin (B2Tf) gel electrophoresis (GE) has become the preferred testing modality due to reportedly high sensitivity (87%–100%) and specificity (71%–100%). However, there have been relatively few studies assessing its diagnostic accuracy. The purpose of this single‐institution study was to determine the accuracy of B2Tf GE in detecting CSF rhinorrhea. Methods A single‐center retrospective review was conducted from 2016 and 2024 for all patients who presented with UTCR and underwent B2Tf GE. Institutional review board approval was obtained. The gold standard for diagnostic confirmation of true and false positives (TP, FP) as well as false negatives (FN) was endoscopic exploration. The gold standard for true negative (TN) was response to medical therapy. Results A total of 105 patients underwent 149 B2Tf GE tests. 40 (38.1%) patients were diagnosed with CSF rhinorrhea. Of the 149 B2‐Tf GE tests, there were 51 TPs, 72 TNs, 20 FPs, and 6 FNs yielding 89.5% sensitivity, 78.3% specificity, 71.8% positive predictive value, and 92.3% negative predictive value, respectively. Of the false results the most common causes for error were purulent sinusitis ( n = 6, 23.1%), possible mucous contamination from nose‐blowing during collection ( n = 3, 11.5%), patient collection error ( n = 3, 11.5%), and blood contamination ( n = 1, 3.8%). Conclusion Although these single‐institutional data demonstrate test accuracy within ranges previously reported in the literature, they also demonstrate diagnostic limitations. Future studies should explore reasons for erroneous B2Tf GE results and how these may change clinical decision‐making. Level of Evidence IV Laryngoscope , 2024
科研通智能强力驱动
Strongly Powered by AbleSci AI