Clinical Utility of Cutoff Index Adjustment in Hepatitis B Surface Antigen Detection

B型肝炎表面抗原臨界值指數修正之臨床應用

快速且正確進行B型肝炎表面抗原 (Hepatitis B surface antigen, HBsAg) 診斷在B型肝炎疾病控制上十分重要。表面抗原及抗體 (Hepatitis B surface antibodies, anti-HBs) 同時陽性時，需加測確認試驗以確認HBsAg結果。在本研究之兩段實驗區間，各分析35,053與49,206例病患，其中HBsAg與Anti-HBs同時陽性有392與438例，同占1%。實驗一，392病患中有110例 (28%) 其HBsAg臨界值指數低於100，經確認試驗，96例 (87%) 為陽性。再經病歷審查為真陽性者共76例 (87%)，其中臨界值指數介於20至100有21例，皆為真陽性(無偽陽性)。於實驗二，HBsAg加測確認試驗之臨界值指數下修至20，438例中65 (15%) 例符合加測標準，55 例 (86%) 確認試驗陽性，病例審查為真陽性者共40例 (83%)。加測標準修正後之陽性預測率，實驗二為83%近於實驗一之87%，但須做確認試驗之檢體比例由28%減至15%，具統計意義。經實驗證明，加測確認試驗之臨界值指數下修，並不影響陽性預測率。我們提供一合理且實用之策略，去除非必要之確認，降低檢驗室工作量並增進報告時效。

Rapid and accurate identification of HBsAg is critical for the management of hepatitis B. Here, the coexistence of hepatitis B surface antigen (HBsAg) and hepatitis B surface antibodies (anti-HBs) was observed in patients, and an additional antibody neutralization test was performed to confirm the presence of HBsAg. In this study, we enrolled 35,053 and 49,206 nonrepeat patients in two study periods, and the coexistence of HBsAg and anti-HBs was confirmed in 392 (1%) and 438 (1%) patients, respectively. In Study 1, 110 (28%) patients with HBsAg cutoff index (COI) values lower than 100 were subjected to confirmatory assays, and 96 (87%) patients were confirmed to be positive. Following medical chart review, 76 (87%) patients were identified as having HBsAg true-positive cases, including 21 patients with COI values of 20-100. Because no false-positive results were found in patients with COI values of 20–100, we adjusted a cutoff criterion from 100 to 20 in Study 2. Among the 438 patients with the coexistence of HBsAg and anti-HBs, 65 (15%) patients with HBsAg COI values of less than 20 were subjected to confirmatory assays. Fifty-five(86%) patients were confirmed to be positive. Medical records indicated that 40 (83%) patients were true-positive cases. The positive predictive values (PPV) of the confirmatory assay with two cutoff criteria (COI < 100 in Study 1 vs. COI < 20 in Study 2) were compared. Similar PPVs, 87% (Study 1) and 83% (Study 2), were observed. A significant reduction in the confirmatory test rates was identified (p <0.05, 15% in Study 2 vs. 28% in Study 1). In conclusion, we provide evidence of no changes in the PPV of the HBsAg confirmatory assay with adjusted HBsAg COI criteria. We reported a reliable and practical strategy to eliminate unnecessary confirmation and thus reduce workloads in the clinical laboratory and provide rapid diagnostic results.