Endometriosis is a long-term condition affecting about ten percent of girls and women from puberty to menopause and is commonly associated with pelvic pain, painful periods, and infertility - symptoms which overlap with other conditions. Endometriosis adversely affects women’s quality of life and productivity constituting a huge socio-economic burden. However, its true prevalence is unknown and underestimated as it requires key-hole surgery to confirm or exclude the diagnosis. Consequently, there is on average an 8–9-year delay from symptom to diagnosis in the UK.
This study aims to assess how common endometriosis-like symptoms as well as surgically confirmed endometriosis are in the population and the geographical spread in the UK. Data from the study will be used to develop a model to predict likelihood of having endometriosis based on pattern of symptom presentation and other characteristics. Lifestyle factors will be investigated to explore their effect on the natural course of the disease. Health resource usage will be compared between women with and without endometriosis among those with common symptoms.
This study will use primary care data linked with hospital data held and controlled by the Clinical Practice Research Datalink, which was established to support public health and clinical studies.
This will be the largest population-based study to estimate how common endometriosis-like symptoms are and what proportion have confirmed endometriosis. It will allow development of a prediction model for endometriosis from a combination of symptoms, demographic characteristics, and pattern of presentation potentially facilitating earlier diagnosis and improving care of women with endometriosis.
Aim: To assess the epidemiology of endometriosis, healthcare burden and develop a prediction model to reduce diagnostic delay.
Objectives:
•Estimate prevalence/geographical distribution of endometriosis-like symptoms and endometriosis
•Develop a prediction model to aid earlier diagnosis of endometriosis
•Explore impact of lifestyle factors on the course of endometriosis
•Estimate healthcare burden associated with endometriosis compared to those without endometriosis but with similar symptoms.
Study design: Cohort study
Study population: Women aged 13-50 years
Data sources: CPRD linked data including primary care, HES, Outpatients, A&E, Imaging data
Methods: Anonymised data for women with endometriosis-like symptoms (e.g. pelvic pain, painful periods) will be obtained from primary care database in CPRD (1987-2020) and linked with HES data to estimate prevalence/incidence of symptoms, as well as surgically confirmed endometriosis, over time and by age group.
Primary care data linked with HES, outpatient, A&E and imaging data will be used to develop and validate a prediction model to predict endometriosis from a combination of symptoms, demographic characteristics, and pattern of presentation. Lifestyle factors, such as pregnancies, contraceptive use (duration and type), will be investigated using inferential models to explore their effect on the natural course of the disease (indicated by number of related GP, gynaecology outpatient, A&E visits, imaging and surgical episodes). Healthcare resource utilisation (no. of health care contacts, repeat surgeries, imaging) and its costs will be compared between women who have endometriosis-like symptoms but with or without a laparoscopic confirmation of endometriosis using generalized linear models.
Public health benefit of research: Findings from this study can be used by policy makers for service and financial planning. The prediction model will have a potential to facilitate earlier diagnosis of endometriosis (a key priority of both Women’s health strategy and plan), institute timely management and improve quality of life of those with endometriosis.
Prevalence and geographical distribution of endometriosis-like symptoms; Prevalence of laparoscopically confirmed endometriosis-like symptoms; laparoscopically confirmed endometriosis; health care resource utilisation including number of related surgeries; number of related GP visits; number of related Accident and Emergency visits, prescription data; prediction model based on symptoms.
Lucky Saraswat - Chief Investigator - University of Aberdeen
Dolapo Ayansina - Corresponding Applicant - University of Aberdeen
Andrew Horne - Collaborator - University of Edinburgh
Dorte Rytter - Collaborator - Aarhus University
Mintu Nath - Collaborator - University of Aberdeen
Philippa Saunders - Collaborator - University of Edinburgh
HES Accident and Emergency;HES Admitted Patient Care;HES Diagnostic Imaging Dataset;HES Outpatient;ONS Death Registration Data;Practice Level Index of Multiple Deprivation;CPRD Aurum Pregnancy Register;CPRD GOLD Pregnancy Register;Rural-Urban Classification