Fonseca K, Anand C M
Provincial Laboratory of Public Health for Southern Alberta, Calgary, Canada.
J Clin Microbiol. 1991 Nov;29(11):2507-12. doi: 10.1128/jcm.29.11.2507-2512.1991.
Two algorithms for screening sera for antibody to human immunodeficiency virus type 1 were compared for their efficiency in identifying a true-positive sample in a population with heterogeneous risk factors, using the criteria of specificity and positive predictive value (PPV). In the first algorithm, all sera were screened by using a single enzyme immunoassay (EIA) kit, and a specificity of 98.6% and a PPV of 69.3% was calculated for true-positive sera. The second algorithm employed two different EIA kits in parallel to screen each sample. In the first instance, a specificity and a PPV of 100% was calculated if a positive sample was defined as reactive by both EIA kits; in the second, a specificity of 99.97% and a PPV of 99.4% was obtained if this criterion was extended to include a combination of one reactive and one equivocal result obtained with the two EIA kits.
比较了两种用于筛查血清中1型人类免疫缺陷病毒抗体的算法,根据特异性和阳性预测值(PPV)标准,评估它们在识别具有异质风险因素人群中真阳性样本的效率。在第一种算法中,所有血清均使用单一酶免疫测定(EIA)试剂盒进行筛查,计算出真阳性血清的特异性为98.6%,PPV为69.3%。第二种算法并行使用两种不同的EIA试剂盒对每个样本进行筛查。首先,如果将阳性样本定义为两种EIA试剂盒均呈反应性,则计算出特异性和PPV均为100%;其次,如果将该标准扩展为包括两种EIA试剂盒获得的一个反应性结果和一个可疑结果的组合,则特异性为99.97%,PPV为99.4%。