Background
Obesity is an epidemic disease that threatens to inundate health care resources through the increased incidence of diseases such as Non-Alcoholic Fatty Liver Disease (NAFLD) and Non-Alcoholic SteatoHepatitis (NASH). Type 2 diabetes (T2D) is a common chronic disease, with a worldwide prevalence of currently more than 6%. NASH is a disease where fat, inflammation and scar-tissue builds-up in the liver. If NASH worsens it can lead to severe complications. People with NASH often have obesity and T2D, and many also develop heart disease, which is the primary cause of death in NASH patients.
Rationale
The Clinical Practice Research Datalink (CPRD) offers a platform where we can compare and contrast the clinical differences between people who are overweight/obese, diagnosed with NASH, and/or have a T2D diagnosis. We also intend to determine if certain background characteristics such as age, body mass index (BMI), and medications are linked to an increase or decrease in adverse clinical events such as cardiovascular disease and mortality. There is diagnostic and prognostic importance of understanding the disease history and estimating the risk of clinical outcomes in these patients.
Impact
We aim to understand the differences and overlap between people who are overweight/obese, are diagnosed with NASH, and/or have a T2D diagnosis. If we discover key differences between groups and find associations to clinical events we aim to inform clinical practice to maximise patient benefit through evidence based decision making.
This is retrospective observational cohort study in which patients who are overweight or have been diagnosed with either non-alcoholic steatohepatitis (NASH) and/or type 2 diabetes mellitus (T2DM) will be selected from all available linked data in HES-CPRD databases spanning 1997-2022/3 using validated codes indicative of diagnoses of these conditions. Patients must be aged ≥18 at point of diagnosis. Recorded body mass index (BMI) ≥25 will also constitute an overweight diagnosis and a secondary source of NASH diagnoses will be sought using non-invasive tests including Fibrosis-4 Index, aspartate aminotransferase (AST)-platelet ratio Index, Forns Index, BARD score, non alcoholic fatty liver disease (NAFLD) fibrosis score, and AST/alanine aminotransferase (ALT) ratio, with diagnostic cut-offs for these defined with clinical input. Patients will be assigned to one of three cohorts – overweight, T2DM, or NASH – based on their most recent diagnosis and further classified into subpopulations according to their history of the remaining two comorbidities.
Phase 1 will compare the covariates of interest between overweight, T2DM and NASH patients. Any comparisons between cohorts/subgroups will be made using t-tests for continuous data and chi-squared test for categorical variables, or non-parametric alternatives if required.
Phase 2 will use the covariate data collected in phase 1 to develop and validate risk equations using established methods similar to those used to develop the QRisk3 calculator. Univariable and multivariable Cox proportional hazards models will be used to estimate the coefficients for each risk factor. From the final models we will use the regression coefficients for each variable as weights, which we will combine with the baseline survivor function evaluated up to 10-years to derive risk prediction models over a period of 10-years of follow-up. This will enable us to derive risk estimates for each year of follow-up which, once validated, could help inform clinical management of these conditions.
We will use a combination of primary care, secondary care, and mortality data from CPRD, HES, and ONS respectively to define our outcomes of interest. The outcomes will be defined using Read, SNOMED, International Classification of Disease version 10 (ICD-10), and OPCS Classification of Interventions and Procedures version 4 (OPCS-4) codes.
Phase 1: Comparisons between overweight, T2D, and NASH patients
Phase 1 will investigate the correlation and the underlying characteristics between people who are overweight/obese, have a T2D diagnosis, and/or NASH diagnosis.
Phase 2: Risk prediction models for adverse events
Further details on the outcomes of interest are provided in the “Exposures, Outcomes and Covariates” section and code lists are provided in the appendices. As an overview, we will measure the time-to-event (survival) for the following events in individual cohorts of overweight/obese, T2D, and NASH patients:
1. Mortality (all cause, cardiovascular related mortality, adverse liver event related)
2. CVD
3. T2D (excluding the T2D cohort)
4. Obesity (excluding the overweight/obese cohort)
5. Adverse liver events
6. CKD
Please note that for all events, an event is defined as present by the presence of a code in CPRD, HES or ONS, and assumed to be absent by the absence of a code.
Joe Hollinghurst - Chief Investigator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Joe Hollinghurst - Corresponding Applicant - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Esmé O'Brien - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Iestyn Lloyd - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
John Houghton - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Mike Hale - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Nia Jenkins - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Oksana Kirichek - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Rokas Lukauskis - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Steffan Hill - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
Thomas Padgett - Collaborator - Health Economics & Outcomes Research Ltd ( HEOR Ltd )
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation