Evaluation of the specificity of the 2023 European Society of Cardiology classification for infective endocarditis

The 2023 Duke-ISCVID and 2023 ESC classifications have recently issued independent diagnostic criteria for infective endocarditis (IE), updating the 2015 ESC criteria. The specificity of the 2023 ESC criteria should be evaluated and compared to the two other classifications in IE suspected patients. We retrospectively collected the characteristics of patients hospitalised in Bichat University Hospital, in 2021, who had been evaluated for suspicion of IE, and in whom IE diagnosis was finally rejected. All were classified by 2015 ESC, 2023 Duke-ISCVID, and 2023 ESC. In total 130 patients were analysed. Mean age was 62 years, 64.6% were male, 30.0% had prosthetic cardiac valve or valve repair, 16.2% had cardiac implanted electronic device, and 23.1% other cardiac conditions. Overall, 2, 5 and 5 patients were falsely classified as definite IE with the 2015 ESC, 2023 Duke-ISCVID and 2023 ESC criteria, respectively. The corresponding specificities were 99% (95% CI [94%; 100%], 96% (95% CI [91%; 99%]), and 96% (95% CI [91%; 99%]). The 2023 ESC and the 2023 Duke-ISCVID criteria are highly specific, although slightly less than the 2015 ESC criteria, for ruling out the diagnosis of definite IE.HIGHLIGHTS2023 Duke-ISCVID and 2023 ESC criteria are recently issued diagnostic classifications2023 ESC criteria have an excellent specificity, equivalent to the 2023 Duke-ISCVID one2023 ESC criteria and the 2023 Duke-ISCVID are less specific than the 2015 ESC criteriaSpecificities were quite similar according to the nature of the cardiac valve (native or prosthetic valve) or the duration of antibiotic therapy 2023 Duke-ISCVID and 2023 ESC criteria are recently issued diagnostic classifications 2023 ESC criteria have an excellent specificity, equivalent to the 2023 Duke-ISCVID one 2023 ESC criteria and the 2023 Duke-ISCVID are less specific than the 2015 ESC criteria Specificities were quite similar according to the nature of the cardiac valve (native or prosthetic valve) or the duration of antibiotic therapy

2023年Duke-ISCVID标准与2023年欧洲心脏病学会（ESC）标准近期分别发布了感染性心内膜炎（infective endocarditis, IE）的独立诊断标准，更新了2015年ESC标准。本研究旨在评估2023年ESC标准的特异性，并在疑似感染性心内膜炎的患者中与另外两套分类标准进行对比。我们回顾性收集了2021年于比沙大学医院住院、接受疑似IE评估但最终排除IE诊断的患者的临床特征。所有患者均采用2015年ESC、2023年Duke-ISCVID及2023年ESC三套标准进行分类。最终共纳入130例患者进行分析，平均年龄为62岁，男性占比64.6%；其中30.0%的患者存在人工心脏瓣膜或瓣膜修复史，16.2%的患者植入了心脏植入式电子装置，23.1%的患者存在其他心脏疾病。整体而言，采用2015年ESC、2023年Duke-ISCVID及2023年ESC标准时，分别有2例、5例及5例患者被误分类为确诊IE。对应的特异性分别为99%（95%置信区间[94%; 100%]）、96%（95%置信区间[91%; 99%]）以及96%（95%置信区间[91%; 99%]）。2023年ESC与2023年Duke-ISCVID标准在排除确诊IE诊断方面具有较高特异性，虽略低于2015年ESC标准。研究亮点：1. 2023年Duke-ISCVID与2023年ESC标准为新近发布的感染性心内膜炎诊断分类体系；2. 2023年ESC标准具有优异的特异性，与2023年Duke-ISCVID标准相当；3. 2023年ESC与2023年Duke-ISCVID标准的特异性均低于2015年ESC标准；4. 按心脏瓣膜类型（自体瓣膜或人工瓣膜）或抗生素治疗时长分层，特异性结果无显著差异。