即时多重免疫层析法诊断伤寒的回顾性准确性评价
该研究在巴基斯坦阿迦汗大学实验室对DPP Typhoid即时检测进行回顾性诊断准确性评估,以血培养为参考标准,使用385份存档血清样本。按制造商阈值,敏感度97.8%、特异度65.3%、AUC 81.5%;按Youden最优阈值,敏感度89.7%、特异度82.2%。在潜类别模型比较中,该检测平衡准确度最高(89.2%)。
要点
- DPP Typhoid检测在制造商阈值下敏感度高(97.8%)但特异度偏低(65.3%)。
- 采用Youden最优阈值后敏感度降至89.7%、特异度升至82.2%。
- 在9种相同队列的快速诊断检测中,该检测平衡准确度最高(89.2%)。
核心结论
- DPP Typhoid检测对伤寒具有较高诊断准确性,但建议调整新阈值以提升性能。
研究设计与样本
- 回顾性、观察性实验室研究,在巴基斯坦阿迦汗大学研究实验室进行,使用先前伤寒诊断准确性研究(NCT04801602)采集的存档冻存血清样本,以血培养为参考标准。
- 共纳入385份样本,评估制造商阈值和Youden最优阈值下的敏感度、特异度及AUC。
主要发现
- 制造商阈值下,敏感度97.8%(95%CI 94.6-99.2)、特异度65.3%(95%CI 58.5-71.6)、AUC 81.5%(95%CI 75.5-85.3)。
- Youden最优阈值下,总体敏感度89.7%、特异度82.2%。
- 与同一队列评估的其他9种快速诊断检测相比,潜类别模型显示DPP Typhoid检测平衡准确度最高(89.2%)。
关键数值与推荐
- 制造商阈值:敏感度97.8%(95%CI 94.6-99.2)、特异度65.3%(95%CI 58.5-71.6)、AUC 81.5%(95%CI 75.5-85.3)。
- Youden最优阈值:总体敏感度89.7%、特异度82.2%。
- 潜类别模型比较:DPP Typhoid检测平衡准确度89.2%,为9种快速诊断检测中最高。
临床怎么用
- 该检测可作为伤寒即时诊断工具,但需注意制造商阈值下特异度偏低可能带来假阳性风险。
- 建议进一步调整阈值以平衡敏感度和特异度,提升临床适用性。
局限性
- 回顾性研究设计,使用存档冻存血清样本,可能影响结果的外推性。
- 仅以血培养为参考标准,血培养敏感度有限可能影响准确性评估。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
UNLABELLED: There is a clear medical need for an accurate diagnostic test for typhoid that can be performed at point of care. Two antigens (lipopolysaccharide [LPS] and hemolysin E [HlyE]) have recently been identified that can distinguish typhoid from other bacterial infections. Here, we present the results of a diagnostic accuracy study of the Dual Path Platform (DPP) Typhoid assay (Chembio) that detects IgA to both LPS and HlyE using blood culture as the reference standard. This was a retrospective, observational, laboratory study conducted at the Aga Khan University research laboratory, Pakistan, to evaluate the sensitivity and specificity of the DPP Typhoid assay, using archived frozen serum samples collected during a previous typhoid diagnostic accuracy study (NCT04801602). The sensitivity, specificity, and accuracy (area under the receptor operating characteristics curve [AUC]) were then assessed using the manufacturer's and Youden's optimal thresholds. In total, 385 samples were included in the analysis. Using the manufacturer's thresholds, the sensitivity, specificity, and AUC were 97.8% (95% confidence interval [CI] 94.6-99.2), 65.3% (95% CI 58.5-71.6), and 81.5% (95% CI 75.5-85.3), respectively. At Youden's optimal threshold, the overall sensitivity of the DPP Typhoid assay was 89.7% and the specificity was 82.2%. In latent class modeling compared with other nine rapid diagnostic tests evaluated from the same cohort sample, the DPP Typhoid assay demonstrated the highest balanced accuracy (89.2%). The DPP Typhoid assay demonstrated a high diagnostic accuracy for typhoid fever. However, further adjustment to new thresholds is recommended to enhance its performance capabilities. IMPORTANCE: Currently available diagnostic tests for typhoid have several limitations, including low sensitivity and specificity. Dual Path Platform Typhoid assay is a multiplex rapid test that detects IgA antibodies to lipopolysaccharide and hemolysin E antigen. It is considered to have high sensitivity and specificity, and its results were found to be highly correlated with ELISA results. However, very few studies have been conducted to evaluate this test and limited information about the accuracy of this test is present. Hence, this study evaluated the new typhoid test.
引用
- Hasan R, Azizullah Z, Shams H, et al. Evaluation of a point-of-care multiplex immunochromatographic assay for the diagnosis of typhoid: results from a retrospective diagnostic accuracy study. J Clin Microbiol. 2024. PMID: 39302169. doi:10.1128/jcm.00428-24
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。