Reader response: A multicenter comparison of MOG-IgG cell-based assays

I read with interest the article by Waters et al.¹ In this much-needed case-control study, the authors determined high specificity of cell-based assays (CBAs) for myelin oligodendrocyte glycoprotein immunoglobulin G (MOG-IgG), although the live CBAs had even higher specificity (>99%) than the Euroimmun-fixed CBA (98%). The authors then calculated a positive predictive value (PPV) of only 82% for the fixed CBA compared with >95% for the live CBAs. This calculation is problematic, however, because PPV depends on disease prevalence in the tested population and this is often not accurately represented in a case-control study design.² Instead of calculating PPV in this study, I would have emphasized appropriate patient selection to avoid false-positives if using a fixed CBA with high, but imperfect, specificity for MOG-IgG. The true PPV of the fixed CBA will depend on clinicians' test-ordering practices.³ The authors stated that “MOG-IgGs will likely be commonly ordered in the clinical evaluation of a suspected demyelinating event,” even in patients with clinical diagnoses of multiple sclerosis¹; although this may be true, one should remember that the onus to avoid false-positive results falls not only on the diagnostic assay but also on the clinician who determines which patients undergo testing.