Real world utilization of the myositis autoantibody panel.

Abstract

Myositis autoantibody panel results can offer diagnostic and prognostic information in patients with concern for idiopathic inflammatory myopathy (IIM). However, there has been widespread utilization of myositis autoantibody testing clinically, often in situations where concern for an IIM is unclear. We sought to determine ordering practices and factors predicting positive results on ordered myositis antibody panels. We included all patients in the Duke University Health System who had a "myositis antibody panel" ordered from October 2014 through December 2016. Retrospective chart review was performed evaluating antibody positivity, provider specialty, ordering location, demographics, medical history, review of systems (ROS), physical examination (PE), and laboratory values. Fisher's exact and t test tests and backward multivariable regression analysis were performed for statistical analysis. There were 642 unique tests obtained with 114 positive autoantibodies (17.7%) over the 26-month period. Myositis-specific autoantibodies (MSAs) were the most common and anti-Mi-2 was the most frequent (40% of MSAs). Pulmonology providers ordered the majority of tests (383; 59.6%). Adult Rheumatology had the highest antibody positivity rate (34.3%, p=0.0001) among specialties with at least 10 panels ordered. In backward multivariable regression analysis, factors independently associated with a positive myositis antibody panel were chronic corticosteroid use (OR: 2.10, 95% CI: 1.30-3.38) and sclerodermoid skin changes (OR: 6.89; 95% CI: 2.02-23.47). The positivity rate of myositis antibody panel testing in this real-world clinical setting was 18%. Anti-Mi-2 antibody was the most frequent autoantibody present. Specific factors associated with positive results can be utilized to identify patients at higher risk for IIM. • Only eighteen percent of all myositis antibody panel tests ordered returned positive. • Anti-Mi-2 antibody was the most frequent autoantibody in our cohort. • Specific factors associated with positive results can help identify patients at higher risk for IIM, particularly for non-rheumatologists.