Background Often, only basic nuclear patterns, like homogenous, speckled, nucleolar, centromere or cytoplasmic, are reported by laboratories. The detection of other patterns requires well trained readers. As a result, different systems have been developed which automate part of or the complete IFA method and reading process.
Objectives This study compares 2 commercially available HEp-2 antinuclear antibody (ANA) indirect fluorescent antibody (IFA) assays using a sensitivity panel (120 clinically determined patients) and a specificity panel consisting of 80 clinically confirmed negative patients.
Methods We compared two different automated systems (A and B) to assess their capability for screening and titration of these samples. The automated method was directly compared to manual reading of the same processed slides on respective microscopes and also compared with the known clinical information.
Results The results of the two automated methods were in good agreement. System A detected 188 samples correctly from negative and positive samples, versus 187 samples detected by the system B. The falsely detected positive samples were all of low titer (1:80). System A found 157 and System B 156 patterns in agreement to the target pattern. From 80 negative samples system A detected 73 and system B 71 samples correctly.
Conclusions Both systems resulted in an overall sensitivity >95% and a specificity of 91.25% and 88.75%. The pattern recognition also showed only minor aberrant findings resulting in a slightly better detection of cytoplasmic and nuclear membrane patterns by system A while system B detected slightly better the centromeric pattern.
Disclosure of Interest None declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.