[Skip to Navigation]
Editor's Correspondence
June 27, 2005

Diagnostic Accuracy in Systemic Autoimmune Diseases—Reply

Arch Intern Med. 2005;165(12):1439. doi:10.1001/archinte.165.12.1439-a

In reply

We would like to thank the editor for the opportunity to reply to the observations of Goichot and Vinizio regarding our study. We agree that the observations of the study would have been strengthened had the definitive diagnosis been established by an independent panel not involved in the patient’s care. However, we believe that some of that bias has been eliminated because the diagnoses were established after a consensus had been reached by experts in a postclinic conference. Goichot and Vinizio correctly point out that we used classification criteria. Classification criteria are mainly used to “standardize clinical definitions for use in research studies”1 and they work better in groups of patients (such as our cohort) than in an individual patient. At least in the academic practice setting at Johns Hopkins, most patients (93%) with a clinical diagnosis of systemic lupus erythematosus (SLE) met classification criteria.2 The problem is that the diagnosis of individual patients rests on a gold standard, which is usually the rheumatologist’s clinical judgment and “this may or may not match the judgment of others.”3,4 We recognize that not all cases of autoimmune disease follow a particular onset pattern, and there is always a margin of error while using criteria to classify a disease. The referring physicians were given the benefit of the doubt if they had diagnosed a patient as having SLE, and the experts strongly believed that a patient had SLE despite not meeting criteria. However, when medical records of patients who we considered “misdiagnosed” were reviewed, a majority had nothing more than a positive antinuclear antibody test result. Few rheumatologists would consider this part of the spectrum of authentic SLE.

Add or change institution