Hepatitis E assay comparison in CAR reveals diagnostic limitations across 1,695 samples

A comparative evaluation of hepatitis E virus testing in the Central African Republic found that the routinely used Diapro IgM enzyme-linked immunosorbent assay had limited specificity, while an in-house reverse-transcription polymerase chain reaction had low sensitivity. The study assessed 1,695 samples from people living with HIV, blood donors and jaundiced patients who tested negative for yellow fever. After comparison with reference assays, an adjusted Diapro positivity threshold produced specificity close to 98% while maintaining 100% sensitivity in the evaluated dataset, but the researchers said the cutoff requires prospective validation before clinical implementation.

The investigation was designed to address limited information about hepatitis E disease burden in the Central African Republic despite recurrent epidemics. Hepatitis E virus is not routinely tested in patients with elevated transaminases, and the available serological and molecular tools had not been evaluated together against reference standards in the country. The three cohorts provided contrasting populations for that comparison: 942 samples from people living with HIV, 401 from blood donors and 352 from jaundiced patients who had tested negative for yellow fever.

In the initial Central African Republic testing, Diapro IgM results were positive in 12.7% of samples from people living with HIV, 9.7% from blood donors and 40.3% from jaundiced patients. The in-house RT-PCR detected no hepatitis E virus RNA in the HIV or blood-donor groups. In the jaundiced cohort, however, it detected RNA in 51.4% of the 142 samples classified as IgM-positive by Diapro.

These results showed substantial differences between the routinely used serological and molecular tests, although the initial analysis alone could not establish the underlying infection status of each participant. The researchers then selected 361 samples for which sufficient volumes remained: 226 from people living with HIV, 55 from blood donors and 80 from jaundiced patients who had tested negative for yellow fever. This subset was tested with Wantai serology and the commercial RealStar Altona RT-PCR assay, described in the study as reference methods commonly used in developed countries.

The commercial molecular assay also remained negative in the HIV and blood-donor groups, providing no evidence of detectable hepatitis E virus RNA in those evaluated samples. Serological positivity also changed substantially when the subset was assessed with the reference test: the frequency of positive IgM results decreased from 47.7% with Diapro to 7.6% with Wantai. The comparison therefore identified poor agreement between the routine assays and the reference methods rather than a single consistent pattern across every cohort and test.

Clinical laboratory measurements provided additional evidence for a specificity problem with the routine IgM assay. Among Diapro IgM-positive jaundiced patients, elevated alanine aminotransferase, interpreted in the study as evidence of hepatitis, was observed in 75.0% of those with detectable hepatitis E virus RNA but in only 19.6% of those without detectable viral RNA; the reported p-value was 0.0003. This association supported the investigators’ hypothesis that many routine IgM-positive results were not accompanied by detectable viral RNA or biochemical evidence of hepatitis, although the study did not directly establish why the discordant results occurred.

Threshold analysis showed that changing the Diapro IgM positivity threshold from 1 to 11.5 produced specificity close to 98% and retained 100% sensitivity in the evaluated data. The finding indicates that the numerical interpretation of the routine assay may have contributed to the discrepancy with reference serology. It does not establish that the adjusted threshold is ready for routine clinical use: the investigators specifically called for prospective studies to validate the cutoff before implementation, and the reported performance remains tied to the study’s samples and comparison methods.

Overall, the findings point to two distinct limitations in the diagnostic workflow evaluated in the Central African Republic. Diapro IgM testing produced many results that were not confirmed by the reference serology or by detectable viral RNA, indicating insufficient specificity at the routine threshold. The in-house RT-PCR missed viral RNA that the commercial assay detected in the jaundiced cohort, indicating low sensitivity.

Subject of Research: Comparative evaluation of serological and molecular assays for hepatitis E virus diagnosis in the Central African Republic

Article Title: Limitations of routine hepatitis E virus diagnosis in Central African Republic: a comparative evaluation of serological and molecular assays

Article References: Nombot-Yazenguet, M. P. D. M., Chaghouri, A., Manirakiza, A., Mouna, L., Zemingui-Bembete, P. A., Selekon, B., Roque-Afonso, A. M., & Komas, N. P. J. (2026). Limitations of routine hepatitis E virus diagnosis in Central African Republic: a comparative evaluation of serological and molecular assays. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14244-z

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14244-z

Keywords: hepatitis E virus, Central African Republic, IgM ELISA, RT-PCR, diagnostic specificity, diagnostic sensitivity, jaundiced patients, people living with HIV, blood donors, viral RNA, Limitations, routine