Related Experiment Video
Updated: Jun 3, 2026
Comprehensive & Cost Effective Laboratory Monitoring of HIV/AIDS: an African Role Model
Published on: October 31, 2010
Evaluation of external quality assessment results for HIV testing laboratories in Korea using current analytical
Jin-Sook Wang1, Mee-Kyung Kee, Byeong-Sun Choi
1Division of AIDS, Center for Immunology and Pathology, Korea National Institute of Health, Osong Health Technology Administration Complex, Chungcheongbuk-do, South Korea.
Insights
External Quality Assessment (EQAS) for HIV testing revealed varying laboratory performance. Robust statistical methods and Youden plots improved outlier detection, highlighting the need for reliable HIV panels for quality improvement.
Area of Science:
- Clinical diagnostics
- Immunology
- Public health
Background:
- Presents results of HIV External Quality Assessment (EQAS) from 17 laboratories performing HIV-1 western blot testing.
- Evaluates performance of HIV testing laboratories within the KCDC network.
Purpose of the Study:
- To assess the effectiveness of different statistical methods in analyzing HIV EQAS data.
- To identify outliers and errors in HIV testing laboratory performance.
- To provide recommendations for improving HIV testing quality.
Main Methods:
- Utilized a panel of 10 plasma samples (3 positive, 4 indeterminate, 3 negative) for HIV antibody testing.
- Analyzed results using traditional methods (mean±2 SD), robust statistics (median), and Youden plots.
- Compared the ability of each method to detect outliers and classify errors.
Main Results:
- ELISA sensitivities were 94.1% (antigen) and 98.3% (antibody); specificities were 93.3% (antigen) and 100% (antibody).
- Robust statistics and Youden plots identified more outliers than traditional methods, distinguishing systematic and random errors.
- Identified specific performance issues in participating HIV testing laboratories.
Conclusions:
- Reliable EQAS design using HIV panels with defined titers is crucial for enhancing laboratory quality.
- Improved statistical analysis methods are essential for accurate performance evaluation.
- Findings support the need for continuous quality improvement in HIV diagnostic testing.
Background:
This report presents HIV External Quality Assessment (EQAS) results by the KCDC from the 17 HIV testing laboratories that also performed HIV-1 western blot testing of the 585 laboratories.
Method:
The HIV EQAS panel consisted of 10 undiluted plasma samples from donors who were 3 positive, four indeterminate and 3 negative for HIV antibody. The results were analyzed by traditional methods (mean±2 SD), robust statistics (median), and Youden plots (99% coverage).
Results:
The sensitivities for the HIV antigen and antibody ELISAs were 94.1% and 98.3%, respectively, with the inconsistent 1.7% caused by indeterminate, weakly positive samples. The specificities of HIV antigen and antibody ELISAs were 93.3% and 100%, respectively. The traditional method detected 2 laboratories as outliers for HIV antigen testing and 3 laboratories as outliers for HIV antibody testing. The robust statistics method showed better discrimination of outliers, detecting seven laboratories as outliers for HIV antigen testing and six for HIV antibody testing. The outliers detected by the Youden plot were differentiated as six systematic and 2 random errors.
Conclusion:
Therefore, the design of reliable EQAS based on HIV panels with varied and specified titers is prerequisites for quality improvement of HIV testing laboratories.
