EDITORIAL Burden due to non-radiological axial SpA/S. van der Linden & MA Khan labelled persons for each true axSpA patient. In that case the majority of patients that meet the ASAS criteria do not really have axSpA. Therefore, assessing studies investigating the burden of illness, one should clearly keep in mind that current criteria capture both patients with true axSpA as well as patients incorrectly labelled as having ax-SpA (Table I).

There is now growing evidence that the current ASAS criteria need to be improved (5-7). Misclassification due to low specificity of these criteria was shown in a population based study (8). An unrealistic high proportion (17 of 20 or 85%) of the subset of HLA-B27+ patients with chronic back pain (CBP) and onset before age 45 met the ASAS criteria for axSpA in this study; and the 3 remaining HLA-B27+ patients (15%) had non-specific CBP. Interestingly, prevalence of HLA-B27 among the 278 CBP patients who did not have axSpA was only 1.1%, clearly considerably lower than the expected prevalence of HLA-B27 in that group (8.8%)(5, 8). Therefore, this study shows that many of the HLA-B27+ CBP patients get incorrectly labelled as having axSpA; they are “look-alikes” that will also get captured in a cost-of illness study. Thus presentations of some non-specific non-inflammatory conditions may also get classified as axSpA by the ASAS criteria.