目的: 以琼脂稀释法为金标准, 评价Epsilometer试验(Epsilometer test, E-test)法检测幽门螺杆菌(Helicobacter pylori, H. pylori)对甲硝唑敏感性的一致性。方法: 纳入2018年8月至2020年7月因消化不良症状就诊于北京大学第三医院行胃镜检查的H. pylori感染初治患者, 取胃黏膜组织活检行H. pylori培养, 分别采用E-test法和琼脂稀释法检测H. pylori对甲硝唑的敏感性, 比较两种方法检测结果的一致性和相关性。结果: 成功培养105株H. pylori, 将最小抑菌浓度≥ 8 mg/L定义为耐药。琼脂稀释法检测甲硝唑耐药菌株68株, 耐药率64.8%, E-test法检测耐药菌株66株, 耐药率62.9%, 其中, 琼脂稀释法和E-test法检测均为耐药的菌株66株, 均为敏感的菌株37株, 两种方法的一致率为98.1%。2例菌株被琼脂稀释法评价为耐药, 而E-test法评价为敏感, 非常严重错误率为1.9%。没有菌株被琼脂稀释法评价为敏感, 而E-test法评价为耐药(严重错误率为0%)。以琼脂稀释法为金标准, E-test法检测甲硝唑耐药的灵敏度为97.1%(95%CI: 0.888~0.995), 特异度为100%(95%CI: 0.883~1.000)。Cohen’s kappa系数为0.959 (95%CI: 0.902~1.016, P < 0.001), Spearmans相关性检测r=0.807(P < 0.001)。采用Bland-Altman法进行一致性评价, 结果提示较好, 未出现一致性区间外的测值。E-test法比琼脂稀释法的成本更低, 平均完成1例试验两者的成本分别为269.8元和356.6元。结论: E-test法检测H. pylori对甲硝唑的药敏试验与琼脂稀释法相比具有较强的一致性, E-test法省时、省力、价廉, 可以作为H. pylori药敏试验的优选检测方法。

Objective: Agar dilution method (ADM) was used as the golden standard to evaluate the consistency of Epsilometer test (E-test) in detecting the sensitivity of Helicobacter pylori (H. pylori) to metronidazole. Methods: From August 2018 to July 2020, patients with H. pylori infection treated for the first time in Peking University Third Hospital for gastroscopy due to dyspepsia were included in this study. Gastric mucosas were taken from the patients with H. pylori infection. H. pylori culture was performed. Both the ADM and E-test were applied to the antibiotic susceptibility of H. pylori to metro-nidazole, and the consistency and correlation between the two methods were validated. Results: In the study, 105 clinical isolates of H. pylori were successfully cultured, and the minimum inhibitory concentration ≥ 8 mg/L was defined as drug resistance. Both ADM and the E-test showed high resistance rates to metronidazole, 64.8% and 62.9%, respectively. Among them, 66 drug-resistant strains were detected by ADM and E-test, and 37 were sensitive strains, so the consistency rate was 98.1%. Two strains were evaluated as drug resistance by ADM, but sensitive by the E-test, with a very major error rate of 1.9%. There was zero strain sensitive according to ADM but assessed as resistant by the E-test, so the major error rate was 0%. Taking ADM as the gold standard, the sensitivity of E-test in the detection of metronidazole susceptibility was 97.1% (95%CI: 0.888-0.995), and the specificity was 100% (95%CI: 0.883-1.000). Cohen's kappa analysis showed substantial agreement, and kappa coefficient was 0.959 (95%CI: 0.902-1.016, P < 0.001). Spearmans correlation analysis confirmed this correlation was significant (r=0.807, P < 0.001). The consistency evaluation of Bland-Altman method indicated that it was good, and there was no measured value outside the consistency interval. In this study, cost analysis, including materials and labor, showed a 32.2% higher cost per analyte for ADM as compared with the E-test (356.6 yuan vs. 269.8 yuan). Conclusion: The susceptibility test of H. pylori to metronidazole by E-test presents better agreement with ADM. Because it is less expensive, less labor intensive, and more rapid, it is an easy and reliable method for H. pylori susceptibility testing.