Do We Need More Criteria for SLE? Is it Just Counting Beans?

Criteria for SLE are not meant to be diagnostic. Instead, they were designed classification of patients for research, i.e. for inclusion into studies. But they should be similar or they would lack validity.

The ACR and SLICC (SLE International Cooperating Clinics) criteria are highly sensitive (>95%) but they both lack specificity (<50%). The criteria are also not inclusive of the complex signs and symptoms of SLE patients.

Today at EULAR, the status of the ACR/EULAR SLE criteria was presented. Items were generated via three cohorts and also a patient survey. Items were reduced through a Delphi technique with the expectation that they would need to be weighted. Previous criteria were not weighted in SLE but this is in contrast to the numerical scoring systems for the ACR/EULAR rheumatoid arthritis and systemic sclerosis criteria, which recognize that individual items to classify a patient with a disease may not have the same weight.

An excellent overview of the Delphi process was presented by Aringer Martin (SP0010). A 1000 Minds exercise that weights the importance of one criterium over another by expert voting was used, which allowed for relative weighting of importance of candidate items. The potential criteria have been clustered into various organs (nephritis), and antibodies and symptoms. The next test will be how the potential criteria perform in 1000 cases and controls, with a comparison of the potential new criteria with the older ACR and SLICC criteria.

So, there is more to be done and the current number of criteria items is too high. I can’t answer if we need criteria for clinical practice but we can use them for research as the specificity of previous criteria is low and this may result in misclassification of patients.

Share This Report

About the Author

Trending Reports From EULAR 2017