Background: Anti-SARS-CoV-2 antibodies have been used in the study of the immune response in infected patients. However, differences in sensitivity and specificity have been reported, depending on the method of analysis. The aim of the present study was to evaluate the diagnostic accuracy of an algorithm in which a high-throughput automated assay for total antibodies was used for screening and two semi-automated IgG-specific methods were used to confirm the results, and also to correlate the analytical results with the clinical data and the time elapsed since infection.

Methods: We studied 306 patients, some hospitalized and some outpatients, belonging to a population with a high prevalence of COVID-19. One-hundred and ten patients were classified as SARS-CoV-2 negative and 196 as positive by polymerase chain reaction.

Results: The algorithm and automated assay alone had a specificity and a positive predictive value of 100%, although the sensitivity and negative predictive value of the algorithm was higher. Both methods showed a good sensitivity from day 11 of the onset of symptoms in asymptomatic and symptomatic patients. The absorbance of the total antibodies was significantly higher in severely symptomatic than in asymptomatic or mildly symptomatic patients, which suggests the antibody level was higher. We found 15 patients that did not present seroconversion at 12 days from the onset of symptoms or the first polymerase chain reaction test.

Conclusion: This study highlights the proper functioning of algorithms in the diagnosis of the immune response to COVID-19, which can help to define testing strategies against this disease.

背景：抗 SARS-CoV-2 抗体已用于研究感染患者的免疫反应。然而，根据分析方法的不同，报告了敏感性和特异性的差异。本研究的目的是评估一种算法的诊断准确性，在该算法中，总抗体的高通量自动化检测用于筛选，并使用两种半自动化的 IgG 特异性方法来确认结果，并关联分析结果与临床数据和感染后的时间。

方法：我们研究了 306 名患者，其中一些是住院患者，一些是门诊患者，他们属于 COVID-19 高流行人群。通过聚合酶链反应，110 名患者被归类为 SARS-CoV-2 阴性，196 名患者被归类为阳性。

结果：单独的算法和自动化检测具有100%的特异性和阳性预测值，尽管算法的敏感性和阴性预测值更高。从无症状和有症状患者出现症状的第 11 天起，这两种方法都显示出良好的敏感性。有严重症状的患者总抗体的吸光度明显高于无症状或轻度症状的患者，这表明抗体水平较高。我们发现 15 名患者在症状出现或第一次聚合酶链反应试验后 12 天未出现血清转换。

结论：这项研究强调了算法在诊断 COVID-19 免疫反应方面的正确功能，这有助于确定针对这种疾病的测试策略。