Riley W A, Craven T, Romont A, Furberg C D
Department of Neurology, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC 27157-1078, USA.
Ultrasound Med Biol. 1996;22(4):405-11. doi: 10.1016/0301-5629(96)00027-0.
A randomly selected subset of 100 pairs of baseline and 36-month follow-up carotid B-mode ultrasound examinations from the 919 patients participating in the Asymptomatic Carotid Artery Progression Study (ACAPS) were subjected to a blinded rereading at the conclusion of the trial to assess temporal bias in the measurement of carotid artery intimal-medial thickness (IMT). The original measurements of the primary outcome variable and five secondary outcome variables at baseline and 36 months, respectively, and the 3-year change in each of these variables, were compared with those obtained from the rereadings. For the primary outcome variable, the mean value of 12 IMT measurements obtained from predefined carotid segments, the mean difference (original-rereading) and the 95% confidence interval which resulted from the rereadings were -0.005 (-0.033, 0.023) mm at baseline and -0.009 (-0.031, 0.013) mm at 36 months. The difference in the 3-year change was -0.004 (-0.038, 0.028) mm. The 95% confidence interval for the mean difference between the rereadings and the original readings for the baseline and the 36-month follow-up examinations included zero for all of the six outcome variables as was also the case for the 3-year change in each variable. The magnitude of the mean differences for these 18 variables ranged from 0.004 to 0.034 mm. Intraclass correlation coefficients between the original readings and rereadings ranged from 0.56 to 0.87 with the 3-year changes in outcome variables tending to have lower correlations and the 36-month examinations higher correlations. The carotid IMT measurement process, when combined with uniform reader training, certification and monitoring of reading performance throughout the course of the study, can avoid the temporal bias observed in other studies.
从参与无症状颈动脉进展研究（ACAPS）的919名患者中随机抽取100对基线和36个月随访的颈动脉B型超声检查数据，在试验结束时进行盲法重新解读，以评估颈动脉内膜中层厚度（IMT）测量中的时间偏差。将基线和36个月时主要结局变量和五个次要结局变量的原始测量值，以及这些变量各自的3年变化值，与重新解读获得的值进行比较。对于主要结局变量，从预定义的颈动脉节段获得的12次IMT测量的平均值、平均差值（原始值 - 重新解读值）以及重新解读得出的95%置信区间，在基线时为-0.005（-0.033，0.023）mm，在36个月时为-0.009（-0.031，0.013）mm。3年变化的差值为-0.004（-0.038，0.028）mm。对于基线和36个月随访检查，重新解读值与原始读数之间平均差值的95%置信区间对于所有六个结局变量均包含零，每个变量的3年变化情况也是如此。这18个变量的平均差值幅度在0.004至0.034mm之间。原始读数与重新解读之间的组内相关系数在0.56至0.87之间，结局变量的3年变化往往相关性较低，36个月检查的相关性较高。在整个研究过程中，将颈动脉IMT测量过程与统一的读者培训、认证以及阅读表现监测相结合，可以避免在其他研究中观察到的时间偏差。
