SPONDYLOARTHRITIS (SpA) classification criteria show variable diagnostic performance in Latin American populations, according to a newly published systematic review and meta-analysis. The study synthesized evidence from six eligible studies conducted in Latin America, with pooled performance metrics raising concerns about the applicability of commonly used criteria in this region.
SpA includes a spectrum of chronic inflammatory disorders affecting the spine and peripheral joints. While classification criteria such as the ASAS, Amor, and ESSG sets have been validated globally, their regional performance in Latin America remains insufficiently established. This review aimed to address that gap using a structured methodology guided by PRISMA standards and registered under PROSPERO (CRD42023464861).
The review found that fewer than 10% of patients in multinational studies used to assess these criteria were from Latin America. The majority of included studies evaluated the ASAS criteria using physician clinical diagnosis as the reference standard. The pooled sensitivity and specificity of the classification tools were as follows:
ASAS criteria: sensitivity 75%, specificity 43%
Amor criteria: sensitivity 71%, specificity 46%
ESSG criteria: sensitivity 83%, specificity 35%
While the ESSG criteria showed the highest sensitivity, it had the lowest specificity, indicating a higher likelihood of false positives. The ASAS criteria demonstrated the most balanced performance, though still with limitations in specificity.
This analysis emphasizes the need for high-quality, homogeneous studies to validate SpA classification tools in Latin American settings. Accurate classification is essential to ensure timely and appropriate management, particularly in regions where the clinical and demographic profile may differ from the populations used in original validation studies.
Reference:
Calixto OJ et al. Performance of classification criteria for spondyloarthritis: where do we stand in many Latin American countries? A systematic literature review and meta-analysis. Clin Rheumatol. 2025. doi: 10.1007/s10067-025-07474-0. [Epub ahead of print].
