Health system wide “big data” analysis of rheumatologic conditions and scleritis

Abstract Background The development of scleritis in the setting of autoimmune conditions has been well documented. Prior series have assessed the relationship between systemic autoimmune disorders and scleritis only in patients referred for rheumatologic or ocular inflammation. This can lead to a referral bias. We reviewed all charts within the electronic medical record (EMR) of a health system for patients with systemic autoimmune and scleritis diagnoses to determine the prevalence of both and which disorders had the highest relative risk of developing scleritis. Methods The EMR was searched for scleritis and systemic inflammatory diagnoses in the past medical history and diagnosis tabs, and for associated disease specific laboratory values. The intersection of scleritis and systemic inflammatory conditions was assessed through searching both SNOMED Clinical Terminology and ICD-10 codes for diagnoses. The prevalence of each autoimmune disorder, scleritis prevalence, the percentage of patients with an autoimmune condition having scleritis, the percentage of patients with scleritis having an autoimmune condition; the relative risk (RR) of scleritis patients having a specific autoimmune disorder were calculated. Results A total of 5.9 million charts were searched with autoimmune conditions identified in 148,993 patients. The most common autoimmune conditions overall were HLA-B27-associated diseases (n = 26,680; prevalence 0.45%); rheumatoid arthritis (RA)(N = 19,923; prevalence 0.34%). Conversely, 2702 patients were identified with scleritis (prevalence 0.05%), of which 31.4% had an associated autoimmune condition. Patients with RA represented the highest percentage of patients with an autoimmune condition having scleritis. Granulomatosis with polyangiitis (GPA) represented the highest the percentage of patients with scleritis having an autoimmune condition. Sjogrens was the third most common condition associated with scleritis- making up 4.5% of cases. An association with juvenile idiopathic arthritis (JIA) was seen in 0.3% of patients. Conclusions While this is the largest retrospective review examining the association between autoimmune disease and scleritis, the findings are similar to prior studies with nearly a third of scleritis patients having an underlying autoimmune diagnosis. Limitations of the study included accurate chart coding; having laboratory results within the searchable EMR. Future research is needed to delineate associations of systemic disease with the anatomic location of scleritis using EMR.