Evaluation of performance testing of different rapid diagnostic kits in comparison with EIAs to validate detection of hepatitis B virus among high risk group in Nigeria

Background: Hepatitis B virus (HBV) causes chronic liver-associated diseases and its early detection is of high public health importance. Its diagnosis is mainly based on immunological assays among which Enzyme-Linked Immunosorbent Assay (ELISA) and rapid tests are the most common and widespread methods. However, a major challenge is the discordance of results of any two laboratory assays which cannot be easily resolved. Therefore, this study was designed to evaluate the validity and reliability of commercially available five rapid test kits in comparison with two Enzyme Immunoassays (EIAs) in Nigeria using hepatitis B surface antigen as a reference marker.

Methods: A total of 100 sera of previously diagnosed consenting HBV-positive patients from private diagnostic laboratories in Ibadan between March and August, 2011 were tested using two EIA and five rapid commercially available HBV test kits in Nigeria. Data were analyzed by SPSS version 15, while bivariate and multivariate analyses were carried out to identify associations at P < 0.05 considered significant.

Results: Overall, the sensitivity rates of the two EIA kits were 100% and 99.9% (95% confidence interval [CI] = 98.9–99.7) with specificity of 100% and 99.9% (95% CI = 98.9–99.7), respectively. The sensitivity of the five rapid test kits ranged from 97.5% (95% CI = 96.4–97.6) to 98.9% (95% CI = 97.9–99.9) with specificity of 80% (95% CI = 79.3–80.9) to 90% (95% CI = 89.2–91.0). Also, the positive predictive value ranged from 88% (95% CI = 88.2–89.9) to 89% (95% CI = 88.2–89.9), while the negative predictive value ranged from 80% (95% CI = 79.3–80.9) to 90% (95% CI = 89.2–91.0) for the five rapid kits. However, that of the two EIAs ranged from 99.9% (98.9–99.7) to 100%. Further analysis showed significant (P = 0.033) variations in the sensitivity and specificity of the EIAs and rapid test kits.

Conclusions: The results from this study have clearly revealed the challenges of diagnosis of HBV infections in Nigeria. This study has also demonstrated that the sensitivity of most of the rapid test kits may not be adequate when compared with EIA for early detection of HBV infections. The implications of possible misdiagnosis on the various intervention strategies that rely predominantly on correct HBV status of an individual are enormous. Therefore, there is the need to further compliment the use of rapid test kits with EIAs for HBV control in Nigeria.