Performance of existing diagnostic criteria for palindromic rheumatism

期刊论文
发表日期：

2024年5月22日

查看全文

Farhad Soltani,
Amirreza Khalaji,
Rojin Farzaneh,
Aida Malek Mahdavi,
Alireza Khabbazi

摘要

Background

Four criteria have been proposed for the diagnosis of palindromic rheumatism (PR), including those of Hannonen et al., Passero and Barbieri, Guerne and Weisman, and Gonzalez-López. But none of these criteria has been validated. In this research, we investigated the performance of these diagnostic criteria for diagnosing PR.

Methods

In this study, PR and control groups were consecutively recruited from a prospective cohort of intermittent arthritis. Inclusion criteria for PR group were diagnosing PR by an expert rheumatologist, age ≥ 18, having at least 6 months follow-up, and ruling out of other causes of intermittent arthritis. These criteria were applied to both groups. Sensitivity, specificity, positive predictive value, negative predictive value, diagnostic odds ratio (DOR), and Youden’s index were calculated for each criteria.

Results

This study included 197 consecutive subjects diagnosed with PR and 208 subjects with a diagnosis other than PR. The sensitivity of Hannonen et al. criteria was higher than the Gonzalez-Lopez, Guerne and Weisman, and Pasero and Barbieri criteria (96.4% versus 95.4%, 79.2%, and 35.5%, respectively). The specificity of the Pasero and Barbieri criteria was higher than the other criteria. Hannonen al. criteria with a DOR of 325.7, had the highest DOR. In descending order, the best accuracy belonged to Hannonen et al., Gonzalez-Lopez, Guerne and Weisman, and Pasero and Barbieri criteria (94.3%, 94.1%, 86.4%, and 66.9% respectively).

Conclusion

This study showed that the Hannonen et al. and Gonzalez-Lopez criteria have a better performance in diagnosing PR.

Key Points

• The sensitivity of Hannonen et al. criteria and the specifity of Passero and Barbieri criteria are higher than other proposed criteria for diagnosis of palindromic rheumatism.

• Hannonen et al. criteria with a sensitivity of 96.4%, specifity of 92.3% and accuracy of 94.3% has the best performance in diagnosis of palindromic rheumatism between existing diagnostic criteria for palindromic rheumatism.