Diagnostic accuracy of ferritin and glycosylated ferritin in adult-onset Still's disease: a meta-analysis.

Objective: This study aimed to ascertain and compare the diagnostic efficacy of ferritin and glycosylated ferritin in adult-onset Still's disease (AOSD).

Methods: A thorough meta-analysis was conducted using data extracted from MEDLINE, Embase, and the Cochrane Library. Separate meta-analyses were performed to assess the diagnostic accuracies of ferritin and glycosylated ferritin in patients with AOSD.

Results: Eight studies encompassing 556 AOSD patients and 763 non-AOSD controls were included in the analysis. Ferritin cut-off values ranged widely, from 400 to 5000 μg/L, with some studies using a relative threshold, such as ≥ 5 times the normal upper limit. Glycosylated ferritin was reported, with thresholds ranging from 16 to 33%. Ferritin displayed a sensitivity of 60.4% and a specificity of 85.7%, accurately identifying AOSD in 60.4% of cases and accurately excluding non-AOSD in 85.7% of cases. The sensitivity in ferritin-specific studies escalated to 66.8%, while specificity was maintained at 84.7%. Glycosylated ferritin exhibited a higher sensitivity of 74% and a slightly lower specificity of 80.5%, indicating superior detection of AOSD cases. The positive likelihood ratio (PLR) and negative likelihood ratio (NLR) for ferritin were 4.179 and 0.399, respectively; for glycosylated ferritin alone, these values were 3.604 and 0.357, demonstrating moderate diagnostic reliability. Diagnostic odds ratios (DORs) were 11.32 for ferritin and 10.14 for glycosylated ferritin. The area under the curve (AUC) was 0.778 for ferritin and 0.835 in ferritin-specific studies, indicating moderate diagnostic precision. The AUC for glycosylated ferritin was not available. The Q* index was 0.717 for ferritin and 0.768 for the ferritin-only group, reflecting a slight improvement when focusing exclusively on ferritin.

Conclusion: Both ferritin and glycosylated ferritin serve as valuable markers for diagnosing AOSD, with ferritin demonstrating superior sensitivity and specificity in targeted studies. Nonetheless, their moderate diagnostic effectiveness suggests that these biomarkers should be used in conjunction with other clinical criteria to ensure a comprehensive diagnosis.