Prediction of a Positive ANA Result for a Rheumatological Diagnosis in an Outpatient Setting

Abstract

Objective: To study the predictive value of a positive anti-nuclear antibody (ANA) for a rheumatological diagnosis in an outpatient setting. Methods: Individuals who were referred to the rheumatology outpatient clinics because of a positive ANA between July 2014 and June 2015 were retrospectively reviewed. Presenting symptoms in addition to a positive ANA and whether a final rheumatological diagnosis was made were recorded. The positive predictive value of a positive ANA and its titer for a rheumatological diagnosis, with and without accompanying symptoms was evaluated. Results: A total of 230 patients were included (82% women, age 47.7 ± 14.1 years [range 18-84]). Family medicine and the general outpatient clinic were the main sources of referral (32.2%), followed by ophthalmology (13.0%) and otorhinolaryngology (11.7%). A final rheumatological diagnosis was made in 54 (23.5%) patients, with rheumatoid arthritis being the commonest diagnosis (40.7%). In the absence of any associated symptoms, the predictive value of a positive ANA was 0%. The presence of Raynaud’s phenomenon (100%), joint swelling (59.5%), and joint stiffness (48.9%) predicted a better final rheumatological diagnosis along with a positive ANA. ANA titers of 1:80 or less had a low sensitivity for rheumatic diseases. A receiver operating characteristic (ROC) curve analysis showed that an ANA titer of [Formula: see text]1:128 best predicted a rheumatological diagnosis (AUC 0.78 [0.71–0.85]; sensitivity 0.78; specificity 0.64). Conclusions: To improve the prediction for a rheumatological diagnosis, referral for a positive ANA test should be more appropriately done with compatible symptoms.