Prevalence and incidence of behcets syndrome (bs) in the hospital episode statistics (hes) in england between 2011 and 2021; a retrospective validation study nested within the birmingham centre of excellence for behcets

Abstract Background/Aims Behçet’s Syndrome is considered rare in the UK. Prevalence estimates range from 0.64 to14.61 per 100 000 population. Here we explore validity of HES ICD-10 codes for BS using Birmingham Centre of excellence (BCoE) data as reference. Prevalence, incidence, clinical characteristics and mortality were compared for attendees at the three national Centres of Excellence (CoE) versus non attendees Methods We extracted all patients aged >18 who had a diagnosis for Behçet's Disease in HES between April 2011 and March 2019. A retrospective cross-sectional analysis of medical records (2012 to 2019) was undertaken at BCoE for MDT confirmed BS for validation. A multivariable logistic regression model examined mortality within 3 years following BS diagnosis. Kaplan Meier survival analysis was undertaken comparing CoE attenders versus non-attenders. Chi squared tests assessed significance of the relationship between CoE attendance, demography, symptom categories, infliximab use and mortality within 3 years. To determine the incidence and prevalence rates, we obtained ONS mid-year estimates of England’s adult population for each year from 2011 to 2019. Point prevalence of BS per 100,000 population was calculated per year of study while Incidence was calculated per 100,000 person years for each year of the study period Results 4452 cases of BS including 515 BCoE attendees were identified in HES between 2011 and 2019. 559 BS cases were confirmed between 2012 and 2020 in the BCoE database. BS prevalence and incidence increased from 2.2 to 3.8 and 0.7 to 0.9 per 100,000 respectively between 2011 and 2019. 65.5% of BS cases were female, 84.9% were Caucasian. BS prevalence peaked at ages 40-49 years. HES data only identified 1157/4452 patients attending one of the three national COEs. Male patients with BS were more likely to have ocular, vascular and cardiac characteristics of BS (p < 0.001). Caucasian ethnicity was associated with highest prevalence of genital ulceration, ocular, vascular and neurological BS. 3-year all-cause mortality was higher in the non-COE cohort (6.4%) versus COE attenders (2.6%) (p < 0.001). Overall BS associated mortality at 3 years follow-up was 0.83%, with higher rate noted in males (1.3%). In logistic regression analysis, mortality significantly associated with age >40 years, IMD and Comorbidity. Infliximab reduced risk of mortality (OR 0.51; p 0.037) Conclusion Our study strength includes use of unselected national HES data with potential coding for inpatient and outpatient episodes. Increasing prevalence of BS may be multifactorial (misclassification, improved recognition, migration effects, improved survival). Our study confirms differences in phenotype and mortality between those diagnosed with BS in the England versus endemic regions. Future work will examine primary care datasets (CPRD and linked HES) to evaluate the national prevalence, incidence, phenotype clusters and risk factors for BS and will ultimately facilitate the creation of a National BS Register Disclosure P. Chandratre: None. R. Situnayake: None. B. Coupland: None.