This group is one of the UK’s leading gastrointestinal epidemiology research groups describing the epidemiology, aetiology and consequences of malignant and non-malignant acute and chronic gastrointestinal disease.
The group makes extensive use of e-health databases (CPRD, THIN, HES).
Major research contributions include:
- evidence on the risk of venous thromboembolism in inflammatory bowel disease
- the occurrence and consequences of diverticular disease
- work in the field of liver disease, coeliac disease and gastrointestinal bleeding
Our work has informed national (British Society of Gastroenterology, NICE) and international (American Association for the Study of Liver Disease) guidelines on management of coeliac disease, inflammatory bowel disease, acute upper GI bleeding and primary biliary cirrhosis.
Gastrointestinal Epidemiology Studies
Inflammatory bowel disease:
The occurrence, natural history and consequences of Inflammatory Bowel Disease
- Funder: National Association for Colitis and Crohn’s Disease (£100,000)
- Investigator: Tim Card
- Study dates: 2010-2012
Though much is known of the epidemiology of Inflammatory Bowel Disease (IBD), there are still large gaps in our knowledge, and some aspects including the most basic studies of prevalence and incidence need to be conducted repeatedly to be useful in the planning of health services.
Some of the outstanding questions can only be adequately addressed through large and expensive studies collecting new data. However with increasingly large quantities of ever richer data available in anonymised databases; database epidemiology provides a relatively fast and reliable method to answer many others.
- To determine the prevalence and incidence of IBD and its variation by type, sex, age, socioeconomic status and geographical distribution in the UK.
- To describe the natural history and progression of disease in terms of the rates of acute flare, burden of steroid use, and rates of surgical intervention.
- To quantify the risk of venous thromboembolism in people with IBD compared with the general population and how this is related to episodes of disease activity, hospitalisation, surgical intervention and comorbidity.
- To determine the fertility rate of women with IBD and the occurrence of complications of pregnancy and congenital anomalies and how this is related to severity of disease and medication use.
Using the General Practice Research Database, linked to Hospital Episodes Statistics and the Office for National Statistics data on births, cancer and deaths, population-based cohorts will be obtained to enable aims 1-3 to be examined addressed. A further cohort from the mother-baby linked dataset of The Health Improvement Network will be obtained to address aim 4.
Upper gastrointestinal haemorrhage:
Predicting the occurrence and mortality of upper gastrointestinal haemorrhage
- Funder: Medical Research Council Population Health Scientist Fellowship (£260,000)
- Investigator: Colin Crooks
- Study dates: 2009-2013
Can we use hospital records to understand trends in upper gastrointestinal haemorrhage?
Information on hospital admissions is collected routinely from all NHS institutions within England. We have been investigating whether using these data can provide reliable information on the trends and patterns of the mortality, occurrence and management of upper gastrointestinal haemorrhage. This could provide a more comprehensive and cheaper alternative (or at least a useful addition) to specifically collecting clinical audit data, so we have worked on better validating this data in these areas.
Using these data we have been able to show that changes in mortality have been partly obscured by changes in underlying illness and age. We have also shown steep gradients in occurrence by deprivation and region that are not explained by differences in severity.
Does the overall burden of co-morbidity predict bleeding?
We are using a general practitioner database linked to a hospital database to model statistically known predictors (medications and diagnoses) of upper gastrointestinal haemorrhage. If overall illness predicts bleeding in addition to this model, an episode of bleeding might indicate a general deterioration in health rather than a specific disease that requires intensive investigation and treatment.
Does a bleeding episode contribute to an increased long-term mortality?
Additionally, we are investigating the long term mortality of upper gastrointestinal haemorrhage in these databases linked to the Office for National Statistics’ death register. Determining whether long-term mortality is predicted wholly by coexisting illness or whether an episode of bleeding itself contributes to long term mortality helps explain why the mortality rate has proved so difficult to reduce.
What are the risks of combinations of medications?
Finally, we are looking at the risk of upper gastrointestinal haemorrhage associated with different combinations of medications after adequately controlling for the risks associated with the underlying illnesses. This allows better informed prescribing decisions in patients who are particularly at risk of bleeding.
- CROOKS, COLIN, CARD, TIM and WEST, JOE, 2011. Reductions in 28-Day Mortality Following Hospital Admission for Upper Gastrointestinal Hemorrhage. Gastroenterology.141(1):62-70
Irritable bowel syndrome:
The burden, consequences and economic impact of a diagnosis of Irritable Bowel Syndrome
- Funder: Medical Research Council Population Health Scientist Fellowship
- Investigator: Caroline Canavan
- Study dates: 2011-2015
Irritable bowel syndrome (IBS) affects up to 20% of the population at any time yet relatively little is known about characteristics of individuals and their prognosis or the financial cost to the health service of managing IBS.
- What is the contemporary epidemiology of IBS in UK primary care?
- How does health care utilization change with secondary care referral?
- What impact does secondary care referral have on quality of life in IBS?
- What is the economic burden of IBS in the UK and the cost benefit of secondary care referral?
Using the General Practice Research Database a cohort of patients with IBS will be identified and used to describe current epidemiology of IBS in the UK. Using the same cohort a self-controlled case series analysis will assess changes in health care utilisation before and after secondary care referral.
A self-controlled questionnaire-based case study of new referrals with asses the changes in the quality of life associated with consultation in patients with IBS. A cost-effectiveness model will then be constructed using the above results based on a hypothetical cohort of IBS patients compared to the general population.
Chronic liver disease:
What is the health care burden of chronic liver disease in the UK?
- Funder: School of Community Health Sciences PhD Studentship
- Investigator: Sonia Ratib
- Study dates: 2011-2014
- What is the incidence and prevalence of cirrhosis of the liver in the UK over the previous 2 decades and how does this vary by age, sex, socioeconomic status, geographical region and aetiology of disease?
- What is the rate of hepatocellular carcinoma among people with cirrhosis and how does it compare to that of the general population?
- What are the rates and reasons for hospital admission among people with chronic liver disease compared to the general population?
- Can prognostic scores for end-stage liver disease designed in secondary care be adapted and used in primary care, using routinely available data?
We will identify cohorts of people with chronic liver disease and a matched cohort of people without chronic liver disease from the General Practice Research Database. Secondary care and national statistics data linked to these cohorts will be obtained and together provide all the necessary information to characterise the populations in great detail alongside the natural history of the disease.
This will include data describing the primary care attendances, hospital admissions, cancer diagnosis date, type, site and histology where available and cause of death.
What is the occurrence and consequence of coeliac disease in the UK?
- Funder: Coeliac UK / core (£78,997)
- Investigators: Joe West, Tim Card, Kate Fleming, Laila Tata, Colin Crooks
- Study dates: 2012-2014
- To provide contemporary and population based estimates of the occurrence of coeliac disease and dermatitis herpetiformis (DH) in the UK in the general population, among family members and the rate of misdiagnosis in irritable bowel syndrome.
- To determine the consequences of the disease in terms of pneumococcal disease and cause-specific mortality compared to the general population.
- To quantify the risk of a diagnosis of coeliac disease and/or dermatitis herpetiformis among offspring of mothers with coeliac disease and/or DH compared to offspring of non-coeliac/DH mothers.
We will identify cohorts of people with coeliac disease and DH within the General Practice Research Database (GPRD), Hospital Episodes Statistics (HES) and the Office for National Statistics (ONS) mortality data which have been linked together. We will also use information from The Health Improvement Network (THIN) mother-baby linked data developed within the University of Nottingham.
The work will inform clinicians and policy makers of the occurrence of disease and it’s natural history for the purposes of health care planning, the prognosis of disease for informing patients and relatives of the real world risks of coeliac disease and will show potential areas for improvement in quality of care and reduction in morbidity and mortality that could be achieved.