Analytical performance evaluation of commercially available hepatitis A virus IgM immunoassays: ELISA, ICTs, and chemiluminescence immunoassay

Study population and sample collection

The study was conducted in a tertiary care teaching hospital in central India. A total of 250 residual serum samples were included in the analysis by a convenience sampling strategy. These samples were collected between January 2023 and December 2024 from patients clinically suspected of acute hepatitis A infection and were submitted to the State Virology Laboratory for routine HAV IgM serological testing. Residual sera remaining after completion of routine diagnostic testing were used for the study. Samples with insufficient volume, gross hemolysis, or lipemia were excluded. In addition to the study samples, a separate panel of 48 well-characterized HAV IgM-positive and HAV IgM-negative serum samples was used for initial assay evaluation and verification of performance characteristics. All samples were stored and processed according to manufacturer recommendations. A schematic overview of the study workflow is presented in Figure 1.

Close

Figure 1: Study workflow for the comparative evaluation of hepatitis A virus (HAV) immunoglobulin M (IgM) enzyme-linked immunosorbent assay (ELISA) and rapid immunochromatographic assays using a chemiluminescence immunoassay comparator. Study workflow for the evaluation of HAV IgM immunoassays. An initial assessment was performed using a panel of 48 well-characterized serum samples to compare four ELISA kits and three rapid immunochromatographic test (ICT) kits against the chemiluminescence immunoassay reference comparator. Assays showing concordant performance were subsequently evaluated using 250 clinical serum samples, while rapid ICT kits with discordant results during panel evaluation were excluded from further clinical sample testing.

Export to PPT

Initial kit evaluation using an external quality assurance (EQA) panel

An initial evaluation of the analytical performance of the ELISA and rapid ICT kits was performed using a retrospective panel of 48 pre-characterized serum samples. The reference status of these EQA panel targets was pre-characterized independently of our laboratory’s internal CLIA method. These samples were obtained from the acute hepatitis serology EQA scheme coordinated by the Department of Clinical Virology, Christian Medical College, Vellore, India.^[11] The EQA panel comprised 7 HAV IgM-positive and 41 HAV IgM-negative serum samples, distributed across four different EQA lots. All ELISA and rapid ICT kits were tested using this panel to verify baseline performance before evaluation with routine diagnostic samples. The results obtained for each kit using the EQA panel are summarized in Table 1.

Assay selection

The Architect HAVAb-IgM CLIA (Abbott GmbH, Wiesbaden, Germany), an automated immunoassay platform widely used in routine diagnostic laboratories, was used as the serological reference comparator for this study. The assay was selected due to its established analytical performance and standardized use in high-throughput laboratory settings. Four commercially available HAV IgM ELISA kits were evaluated: Biogenix (Biogenix Inc. Pvt. Ltd., Lucknow, India), Dia.Pro (Diagnostics Bioprobes Srl, Milan, Italy), MBS New S.R.L. (Milan, Italy), and Wantai (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd., Beijing, China). In addition, three commercially available HAV IgM rapid ICT kits were assessed: Biogenix (Biogenix Inc. Pvt. Ltd., Lucknow, India), Biotrol (Biotrol Laboratories Pvt. Ltd., New Delhi, India), and Insight (Tulip Diagnostics [P] Ltd., Goa, India). These specific assays were chosen for evaluation due to their extensive local availability, cost-effectiveness, and routine procurement by healthcare networks across the region, making them highly representative of real-world diagnostic practices in endemic settings. A summary of the CLIA, ELISA, and rapid ICT kits included in the study is provided in Table 2, and the key analytical characteristics of the ELISA kits are summarized in Table 3. All assays were stored, handled, and performed strictly according to the respective manufacturers’ instructions. The manufacturers had no role in study design, data analysis, interpretation of results, or manuscript preparation.

Reference CLIA

The Architect HAVAb-IgM assay (Abbott GmbH, Wiesbaden, Germany) is an automated, two-step CLIA designed for the qualitative detection of IgM antibodies to HAV in human serum and plasma. In the present study, this assay was used as the reference comparator for evaluating the performance of ELISA and rapid immunochromatographic assays. A total of 250 clinical serum samples received for routine diagnosis of suspected acute hepatitis were initially tested using the Architect HAVAb-IgM CLIA, strictly in accordance with the manufacturer’s instructions. Each analytical run included appropriate instrument calibration and internal quality controls, and an external positive control was included to verify assay performance. Following CLIA testing, serum samples were aliquoted into sterile microtubes to avoid repeated freeze-thaw cycles and stored at −20°C until further analysis using the commercially available ELISA and rapid ICT kits.

Comparative ELISA testing

For comparative evaluation, the same 250 clinical serum samples previously tested using the CLIA were retested with the four commercially available HAV IgM ELISA kits (Biogenix, Dia.Pro, MBS New S.R.L., and Wantai). All ELISA procedures were performed strictly in accordance with the respective manufacturers’ instructions, as outlined in Figure 2. Washing steps were automated using a HydroFlex™ 96-well microplate washer (Tecan Life Sciences, Männedorf, Switzerland), and absorbance was measured at 450 nm using an Infinite® F50 microplate reader (Tecan Life Sciences, Männedorf, Switzerland). Incubations were carried out at the temperatures specified in each kit’s Instructions for Use and were monitored using calibrated laboratory thermometers. All assay runs met the manufacturer’s validity criteria, and results were interpreted according to the respective kit-specific cut-off values. To minimize pre-analytical variability, all samples were tested after a single freeze–thaw cycle.

Close

Figure 2: Comparative workflow of the assay protocols used for four commercially available hepatitis A virus (HAV) immunoglobulin M (IgM) enzyme-linked immunosorbent assay (ELISA) kits. Step-by-step comparison of the assay workflows for four commercially available HAV IgM ELISA kits Biogenix, Dia.Pro, MBS New S.R.L., and Wantai are highlighting differences in reagent preparation, sample dilution, incubation conditions, washing steps, substrate development, optical density reading, and cut-off calculation criteria as per manufacturers’ instructions. NC: Negative control, PC: Positive control, OD: Optical density, RT: Room temperature

Export to PPT

Statistical analysis

Statistical analysis was performed using MedCalc Statistical Software (MedCalc Software Ltd., Ostend, Belgium). Diagnostic performance of each ELISA and rapid immunochromatographic assay was evaluated by calculating sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), using the Architect HAVAb-IgM CLIA as the reference comparator. Ninety-five percent confidence intervals (95% CIs) were calculated for all performance estimates. Agreement between each assay and the CLIA comparator was assessed using Cohen’s kappa (κ) coefficient, and agreement strength was interpreted according to standard criteria. A p <0.05 was considered statistically significant.

Assessment using the EQA panel

Initial performance assessment of the ELISA and rapid ICT kits was conducted using the EQA panel comprising 48 pre-characterized serum samples. The CLIA comparator correctly classified all 48 samples. All four ELISA kits demonstrated complete concordance with the EQA panel results. In contrast, variable performance was observed among the rapid ICT kits. The Biogenix rapid ICT kit yielded two false-negative results (one each in lots 22 and 24). The Biotrol rapid ICT kit produced one false-positive result (lot 24). The Insight rapid ICT kit correctly classified all EQA panel samples, demonstrating complete concordance with the comparator assay. Given the suboptimal concordance observed with selected rapid ICT kits during EQA evaluation, and considering the limited availability of clinical sample material, rapid ICT kits were excluded from subsequent analysis using the 250 clinical samples. The detailed performance of each kit using the EQA panel is summarized in Table 1.

Comparative analysis of ELISA kit characteristics

The key operational characteristics of the four commercially available HAV IgM ELISA kits, including reagent preparation requirements, sample volume, incubation conditions, assay duration, and reagent stability, have been summarized in Table 3. Among the evaluated kits, Biogenix and Wantai required preparation of only the wash buffer, whereas Dia. Pro and MBS New S.R.L. required preparation of three components, including wash buffer, calibrator, and HAV antigen-antibody or immunocomplex reagents. Sample volume requirements varied considerably across kits, with Biogenix requiring the highest volume (50 µL), followed by Dia.Pro (10 µL), MBS New S.R.L. (2 µL), and Wantai (1 µL). Two kits required a combination of 37°C and room temperature (18–24°C) incubation steps. Total assay incubation times ranged from ~2 h for Biogenix and Wantai to 3 h and 30 min for Dia.Pro and MBS New S.R.L. With respect to reagent stability at 2–8°C, Biogenix and Dia.Pro kits were stable for 3 months, MBS New S.R.L. for 2 months, and Wantai for 1 month. All kits provided the necessary reagents for assay performance, except the Wantai kit, which required the use of supplemental saline. Result interpretation procedures were straightforward and comparable across all evaluated ELISA kits.

Demographic characteristics of the study population

A total of 250 clinical serum samples were included in the analysis. Based on results obtained using the Architect HAVAb-IgM CLIA comparator, 112 samples (44.8%) were positive for HAV IgM antibodies, and 138 samples (55.2%) were negative. The mean age of the HAV IgM-positive group (11.21 ± 7.36 years) was significantly lower than that of the negative group (29.53 ± 19.80 years) (p <0.001). In addition, significant differences were observed in gender distribution, with a higher proportion of females testing positive (p = 0.003). Age-wise distribution indicated a greater proportion of HAV IgM positivity among younger age groups. The detailed demographic characteristics of the study population, including age and sex distribution, have been presented in Table 4.

Comparative diagnostic performance of the ELISA kits

The diagnostic performance of the four commercially available HAV IgM ELISA kits was evaluated against the CLIA comparator using 250 clinical serum samples. The detailed performance metrics are summarized in Table 5. Among the evaluated kits, Dia.Pro and Wantai demonstrated optimal diagnostic performance, each achieving 100% sensitivity (95% CI: 96.8–100%) and 100% specificity (95% CI: 97.4–100%). Both kits also showed 100% PPV, 100% NPV, and 100% overall diagnostic accuracy (95% CI: 98.5–100%). Perfect agreement with the CLIA comparator was observed for both assays, with a Cohen’s kappa coefficient (κ) of 1.00.

The MBS New S.R.L. ELISA kit demonstrated slightly reduced sensitivity (92.0%, 95% CI: 85.3–96.3%), correctly identifying 103 of 112 HAV IgM-positive samples. Specificity and PPV remained 100% (95% CI: 97.4–100%). The NPV was 93.9% (95% CI: 89.1–96.6%), and the overall diagnostic accuracy was 96.4% (95% CI: 93.3–98.3%). Agreement with the CLIA comparator was strong (κ = 0.93). The Biogenix ELISA kit showed the lowest sensitivity among the four assays (84.8%, 95% CI: 76.8–90.9%), detecting 95/112 positive samples. Specificity and PPV were 100% (95% CI:97.4–100%). The NPV was 89.0% (95% CI: 84.0–92.6%), and the overall diagnostic accuracy was 93.2% (95% CI:89.3–96.0%). Agreement with the CLIA comparator was substantial (κ = 0.86).