Behçet’s disease (BD) is a rare autoimmune condition which is most often characterised by ulceration. The condition is rare, and recent national commissioning decisions have led to the development of the three National Centres of Excellence for BD to improve care for these patients. However, the true extent of the disease is not well understood in the UK. For example, previous research using a UK primary care (General practice) database, The Health Improvement Network database, estimated the prevalence to be 14.61 (95% CI 13.35-15.88) per 100 000 population in 2017. However, this previous work was limited as the coding was not confirmed with secondary care (hospital) coding, the main setting where BD patients are diagnosed and managed. Therefore, the primary aim is to improve the estimation of the coding used primary care data, supported by secondary care which should help support services in planning for the health of these patients.
Secondly, the condition is thought to have originated from the ancient Silk trading route. Hence, historically the condition appears to be more common in countries from these regions. However, there is evidence to suggest that the risk factors and outcomes for those diagnosed with BD outside have different risk factors and outcomes. Therefore, we are proposing to undertake a large study of linked databases of those who attend GPs and hospitals. We will include data on BD diagnoses, potential risk factors and outcomes. Using different approaches we will see what the risk factors and outcomes are for those with BD.
We are planning to undertake a comprehensive exploration of the epidemiology of BD using the CPRD GOLD, AURUM and linked HES datasets.
Current incidence/prevalence estimates derived from The Health Improvement Network (THIN) are limited as: 1) only one clinical code was used to define BD, 2) most diagnoses are made in secondary care so a linked data study would minimise underestimation of BD and 3) there is yet to be exploration of estimations by markers of societal inequalities. Hence, we will describe the burden of BD using CPRD-HES linked data.
BD primarily circulates among Middle Eastern regions. However, the aetiology of patients from these regions may not be generalisable elsewhere. We will undertake a case-control study comparing those with BD against those without and using logistic regression identify the odds-ratio of them having different pre-determined risk factors based on the literature and clinical expertise. Following, using a retrospective cohort study design, we will then examine the risk of patients with BD developing subsequent illnesses compared to those without BD.
As there are numerous crossovers between BD symptomology with other conditions, we will also describe the time to diagnosis from the cohort for patients with typical symptoms. Lastly, it is unclear whether BD is a combination of different disease processes due to the breadth of phenotypical spectrum experienced by patients. Therefore, we will undertake work to ascertain if different phenotypical clusters of patients with BD have different risk factors and outcomes.
Impact: 1) Describing the burden of BD will help with healthcare service planning and provide opportunities to reduce inequalities, 2) Increased knowledge of subsequent morbidity will provide opportunities to tailor preventative advice and therapy, 3) the increased recognition and earlier diagnosis of BD promoted via this study can inform public health policies to improve outcomes by appropriate training/resource allocation.
Objective 1: Epidemiology of BD
• Annual incidence rate per million person years will be calculated stratified by age, sex, ethnicity and patient level/area based deprivation.
• Annual prevalence per million population will be calculated stratified by age, sex, ethnicity and and patient level/area based deprivation.
Objective 2: Risk factors of BD (see appendix 1 for full list)
• The risk factors of interest are: age, smoking, obesity, ethnicity, gender, socio-economic deprivation, presence of MHC class 1 diseases (such as Psoriatic arthritis/Crohns disease/ankylosing spondylitis).
• Time to diagnosis of BD from first key symptom: Ocular lesions, genital apthosis, oral apthosis, skin lesions, neurological manifestations, vascular manifestations and pathergy.
Objective 3: Latent class analysis to examine BD phenotypes (see appendix 1 for full list)
• Phenotypical signs and symptoms explored in objective 2
Objective 4: Outcomes of BD (see appendix 1 for full list)
• Ophthalmic, oro-genital, cardiovascular, neurological, gastroenterological, dermatological outcomes and mortality
Joht Singh Chandan - Chief Investigator - University of Birmingham
Joht Singh Chandan - Corresponding Applicant - University of Birmingham
Krishnarajah Nirantharakumar - Collaborator - University of Birmingham
Nicola Adderley - Collaborator - University of Birmingham
Nigel Trudgill - Collaborator - Sandwell and West Birmingham Hospitals NHS Trust
Priyanka Chandratre - Collaborator - Sandwell and West Birmingham Hospitals NHS Trust
HES Admitted Patient Care;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation