Diagnostic utility of clinical characteristics, laboratory tests, and serum ferritin in diagnosis of adult-onset Still disease.

The diagnosis of adult-onset Still disease (AOSD) is challenging with ambiguous clinical presentation and no specific serological markers. We aim to evaluate the diagnostic utility of clinical, laboratory and serum ferritin features in established AOSD patients. We included all patients >18 years who were admitted to 2 tertiary medical centers (2003-2019) with serum ferritin above 1000 ng/mL. AOSD patients and non-AOSD controls were matched in 1:4 ratio for age and sex. The primary outcomes were sensitivity, specificity, positive/negative likelihood ratio and area under the curve (AUC) using clinical and laboratory characteristics based on the Yamaguchi classification criteria, in addition to serum ferritin. We identified 2658 patients with serum ferritin above 1000 ng/m, of whom 36 diagnosed with AOSD and 144 non-AOSD matched controls. Presence of arthralgia/arthritis showed the highest sensitivity (0.74), specificity (0.93), positive likelihood ratio (10.69), negative likelihood ratio (0.27) and AUC (0.83, 95% confidence interval 0.74-0.92) to the diagnosis of AOSD. On the other hand, serum ferritin showed variation and poorer results, depends on the chosen ferritin cutoff. Joint involvement showed the best diagnostic utility to establish the diagnosis of AOSD. Although clinicians use often elevated ferritin levels as an anchor to AOSD, the final diagnosis should be based on thorough clinical evaluation.