• treatment
• disease activity
• rheumatoid arthritis

We appreciate the comments by Pregnolato et al 1 on our original paper2 and subsequent correspondence in the journal.3 4 The data presented are very interesting and suggest that, under certain circumstances, results of antinuclear antibody (ANA) testing with samples from patients with established systemic lupus erythematosus can show both a high frequency of positivity and high concordance among testing laboratories. Such agreement may result from the nature of the samples themselves, including the titre of the antibodies and the array of specificities present, the kits used for the assays and the experience and training of the observers. While we would acknowledge that ANA assays can perform very well, our experience as well as that in the literature indicates that variability occurs significantly in a way that complicates screening of patients for clinical trials. Notwithstanding the demonstrations of the kind reported by Pregnolato et al, we think that it is important to try to understand assay variability and improve serological determinations for both routine care as well as clinical research.