A membrane-based test for MAP, the bacterium this assay is designed to detect, showed 90.32% sensitivity and 85.19% specificity when compared with nested PCR in a study of 89 animals. In plain terms, it found most samples that the PCR marked positive and correctly left most PCR-negative samples negative. The result points to close but imperfect agreement with the comparator.
The assay, known as IMS-CNM, uses donkey IgG antibodies fixed to a nitrocellulose membrane to capture MAP, then rabbit anti-MAP IgG linked to rhodamine B hydrazone nanoparticles to signal its presence. The study aimed to simplify detection by placing the capture antibodies directly on the membrane.
The comparison with PCR
Nested PCR classified 62 of the 89 faecal samples as positive and 27 as negative. IMS-CNM classified 60 as positive and 29 as negative. A McNemar test found no statistically significant difference between the paired classifications, while Cohen's kappa, a measure of agreement between two tests, was 0.74, reported as substantial agreement.
Against nested PCR, IMS-CNM's sensitivity was 90.32%, with a reported 95% confidence interval from 82.96% to 97.68%. Its specificity was 85.19%, with a 95% confidence interval from 71.80% to 98.58%. These intervals make clear that the reported percentages are estimates rather than fixed values.
ROC analysis, which tests how well a diagnostic measure separates positive and negative samples at different thresholds, produced an area under the curve of 0.90, with a 95% confidence interval of 0.83 to 0.98. The study selected its diagnostic cut-off through that ROC analysis. This result still needs testing in independent samples before the performance can be judged more broadly.
Additional signals from the assay
In a separate analytical test, the lowest concentration that produced a visible IMS-CNM signal was about 3,000 cells per millilitre, and five of six serial dilutions gave positive reactions. That result describes the assay's signal threshold in a prepared dilution series; it does not by itself establish how the test will perform across field samples.
Laboratory checks also looked at whether the antibodies reacted with other mycobacteria. After adsorption with M. phlei, purified rabbit and donkey antibodies showed no detectable cross-reactivity with the tested species, while the corresponding hyperimmune sera retained measurable cross-reactivity.
Routine ELISA performed less well against PCR in the same evaluation: sensitivity was 75.81%, specificity 59.26%, and overall accuracy 70.79%, with fair agreement by kappa at 0.337. The comparison gives context for the membrane assay, but the stated reference for IMS-CNM remained nested PCR.
Researchers also used ImageJ software to quantify membrane colour rather than relying only on visual scoring. The image-based intensity measure correlated with PCR results at 0.68 and with visual IMS-CNM scores at 0.94. Mean intensity was 2.37 ± 1.11 in PCR-negative samples and 5.16 ± 2.26 in PCR-positive samples; the intraclass correlation coefficient was 0.77, with a 95% confidence interval of 0.64 to 0.85.
What the study does not settle
The clinical evidence came from 89 cattle aged 6 to 24 months in a screened herd. The group included 58 animals classed as ELISA-positive and 31 classed as ELISA-negative, and the population was subclinical. That narrow sample means the findings cannot be assumed to apply to other herds, ages, or stages of disease.
The central comparison was with nested PCR, and the diagnostic cut-off was selected through ROC analysis. The reported sensitivity, specificity, agreement and ROC results should therefore be read as performance relative to that comparator. Independent validation is needed to check whether the reported separation holds outside this dataset.
The authors describe IMS-CNM as a potential screening or complementary platform, not as a settled replacement for nested PCR. They call for larger, more diverse and independently validated studies, a caution that fits the study's single, subclinical herd sample.
Paper data and sources
Original title: Detection of Mycobacterium avium subsp. paratuberculosis in Bovine Faeces Using a Rhodamine B Hydrazone Immunosensor on Antibody‐Coated Nitrocellulose Membranes
Authors: Khosravi M, Gharibi D, Ansari R et al.
Journal/Repository: Veterinary medicine and science
Status: Peer-reviewed
First online: 2026-10-13
DOI: Not available
Original paper · Full text