To study the failure alarm information displayed on the automatic coagulation analyzer (coagulation method) of thrombin time (TT), and formulate the coping strategies combined with clinical information. Methods: A total of 233 failed TT blood samples [132 males, 101 females, with a median age of 73 (66, 79) years] were selected from 21 359 inpatients in Peking University First Hospital from January to June 2021. The statistical analysis was made and the failure causes and solutions were summarized according to the coagulation curve and the error codes displayed on the coagulation instrument, in combination with the clinical information, sample characteristics, medication status and other reasons. Meanwhile, a total of 96 TT detection failed lipid blood samples [56 males, 40 females, with a median age of 72 (65, 79) years] were analyzed from the inpatients in Peking University First Hospital from July to November 2021. TT results were obtained by artificial coagulation curve interpretation method, magnetic bead method and high-speed centrifugal re-detection method, respectively. The TT results of the three methods were compared. Results: The proportion of 233 failed TT tests from the total number of samples was 1.1% (233/21 359). There were 41.2% (96/233) samples with lipids, 23.2% (54/233) samples with heparin interference, 22.3% (52/233) samples with oral anticoagulant, and 13.3% (31/233) samples with micro-coagulation or insufficient plasma volume among these test failure samples. The classifications for these alarm information of coagulation curves showed on the instrument were as follows: 32.6% (76/233) of samples with higher changes in absorbance at baseline (SD>2 mAbs), 30.5% (71/233) of samples without peak values of second derivative, 25.8% (60/233) of samples with absorbance difference<35 mAbs between baseline and plateau period, 8.6% (20/233) samples with too low starting point or no starting point, and 2.6% (6/233) samples without coagulation curves. Among these 233 samples, there were 55.8% (130/233) samples that could be manually judged according to the reaction principle and standard coagulation curve pattern. Among the 96 samples that failed in coagulation method due to lipemia, there were 78 samples with sufficient blood volume tested by magnetic bead method. The TT results of the high-speed centrifugal redetection method, artificial coagulation curve interpretation method and magnetic bead method were 14.10 (14.80, 13.38) s, 14.30 (14.99, 13.60) s, and 15.65 (17.25, 14.65) s, respectively, but the difference was not statistically significant (P=0.055). For 78 lipid samples, there was a correlation between the results of the artificial coagulation curve interpretation method and the results of magnetic bead method (r=0.99,P=0.001). Conclusions: For those samples failed in TT detection by coagulation method on automatic coagulation instrument, the cause of failure can be analyzed through coagulation curve and alarm information. For the lipid samples, TT results can be obtained by manual interpretation method, high-speed centrifugation method and magnetic bead method.
目的： 探讨全自动凝血分析仪（凝固法）检测凝血酶时间（TT）失败原因，并结合临床制定应对策略。 方法： 选取2021年1至6月北京大学第一医院住院部21 359份TT检测血液样本中检测失败的样本233份，其中男132例，女101例，年龄73（66，79）岁。根据全自动凝血分析仪显示凝固曲线及检测失败错误代码，结合患者临床信息、样本性状、用药情况等原因进行检测失败原因分析，并制定应对措施。选取2021年7至11月北京大学第一医院住院部96份TT检测失败的脂血样本作为验证样本，男56例，女40例，年龄72（65，79）岁；分别采用凝固曲线人工判读法、磁珠法、高速离心后再检测3种方法获得TT结果，比较3种方法检测TT结果的差异。 结果： 233份TT检测失败样本占总样本数的1.1%（233/21 359）；其中脂血样本占41.2%（96/233），肝素干扰样本占23.2%（54/233），患者使用口服抗凝药样本占22.3%（52/233），样本存在微小凝块或血浆量不足现象占13.3%（31/233）。仪器所示凝固曲线报警信息分类：基线期吸光度变化值增大（SD>2 mAbs）样本占32.6%（76/233），二阶导数无峰值样本占30.5%（71/233），反应过程吸光度变化值（基线期与平台期吸光度差值）<35 mAbs样本占25.8%（60/233），起始点过低及无法找到起始点样本占8.6%（20/233），无凝固曲线样本占2.6%（6/233）。233份样本中，依据反应原理及标准凝固曲线图形可以进行人工判读样本占55.8%（130/233）。96份因脂血导致凝固法检测失败的样本中，78份样本量充足，可进行磁珠法检测。高速离心后再测法、凝固曲线人工判读法与磁珠法检测TT结果分别为14.10（14.80，13.38）、14.30（14.99，13.60）、15.65（17.25，14.65）s，差异无统计学意义（P=0.055）。78份脂血样本磁珠法检测结果与凝固曲线人工判读法检测结果具有相关性（r=0.99，P=0.001）。 结论： 全自动凝血分析仪（凝固法）检测TT失败的样本可通过凝固曲线与报警信息分析失败原因；对于脂血样本可通过人工判读、高速离心后再测或磁珠法获得TT结果。.