Objective:

To evaluate the diagnostic validity of the enzyme-linked immunosorbent assay (ELISA) for human immunodeficiency virus (HIV) in blood banks, based on studies published between 2000 and 2020.

Methodology:

We performed a systematic review and meta-analysis of diagnostic tests, using a random-effects model for the sensitivity, specificity, positive and negative likelihood ratio, diagnostic odds ratio (DOR), and ROC curve, with 95% confidence intervals. Heterogeneity was assessed with the DerSimonianLaird Q(χ2) statistic and the I2 inconsistency statistic. Uncertainty was assessed using percentage study weights.

Results:

15 studies were included. The third-generation ELISA (antibody detection) was applied for 2,992 infected and 4,076 healthy individuals, and the fourth-generation ELISA (simultaneous antigen-antibody detection) was used for 967 infected and 154,264 healthy individuals. Both showed close to 100% sensitivity, but there was an improved specificity in the fourth-generation assays (98% vs. 100%). Both technologies’ likelihood ratios, DOR, and ROC curve aptly distinguished between healthy and infected individuals.

Conclusion:

The third and fourth-generation ELISA were confirmed to have excellent validity and diagnostic utility in blood donors, which is important for HIV control and blood safety policies.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )