More than one million people in the UK have heart failure, one of the most common reasons for hospitalisation, with a mortality rate similar to that of some of the worst cancers. Unfortunately, heart failure is often recognised only when it is so severe that the patient needs to be admitted to hospital. Earlier diagnosis of heart failure could lead to better treatment, improve quality of life, avoid hospital admissions, and possibly prolong life. A failing heart causes kidneys to retain salt and water, leading to a build-up of fluids, or “congestion”. Congestion decreases both heart and kidney function, leading to a decline in general health and increased chances of an early death. Congestion aggravates symptoms, such as breathless on exertion, and causes ankles to swell. These symptoms and signs are not specific for heart failure, are often poorly investigated, and are treated with strong water tablets called loop diuretics. We looked at routinely collected electronic from people living around Glasgow, and found that many more people are prescribed loop diuretics than are diagnosed with heart failure. We found that being prescribed loop diuretics is an indicator of poor health. Those who receive loop diuretics have an outlook similar to that of people diagnosed with heart failure. These findings suggest that many people prescribed a loop diuretic might actually have undiagnosed heart failure. We request to repeat this analysis in the CPRD dataset to see if we find the same results in people living across different areas of the UK.
This study has two parts, to:
1. calculate patterns of prevalent and incident cases of heart failure and/or loop diuretic use in the UK.
2. compare characteristics, comorbidities, concurrent medications, hospitalisation and mortality rates in incident cases (defined above) against a sex, age, and geographically matched control group.
1: We will calculate the prevalence and incidence of cases amongst adults (age >18 years) in CPRD Aurum, Hospital Episode Statistics (HES) admitted and patient records. Standardized rates will be calculated using age and sex for the 2013 European Standard Population. Time-dependent covariates and Poisson regression with robust standard errors will be used to estimate overall and category-specific incidence ratios.
2: A sex, age, and geographically matched control group as case-control for our cohort. All-cause and cardiovascular mortality and place of death will be identified from ONS Death Registrations. Patient sex, age, body mass index (BMI), systolic and diastolic blood pressure, serum creatinine, standard haematology and biochemistry profiles (eg:- haemoglobin, sodium, creatinine, albumin), indices of iron deficiency (eg:- ferritin), smoking status, diagnostic care-setting, comorbidities (atrial fibrillation, hypertension, ischaemic heart disease and myocardial infarction, peripheral artery disease, stroke, chronic respiratory disease, anaemia, iron deficiency, hyponatraemia, cancer, chronic kidney disease, dementia, depression, diabetes mellitus, dyslipidaemia, obesity, osteoarthritis, thyroid disorder), concurrent medications (acetazolamide, renin-angiotensin-aldosterone system inhibitors, beta blockers, calcium channel blockers, digoxin, thiazides and related, low dose aspirin, warfarin, insulin, other treatments for diabetes mellitus, bronchodilators, lipid regulating medications, thyroid disorder medications, non-steroidal anti-inflammatory agents, depression medication, oral steroids, gastroprotectants, and opioids/opiates) will be extracted from primary care and HES records as appropriate. Competing risk analysis will be use to study all-cause, cardiovascular and heart failure hospitalisations. Cox proportional hazards regression will be used to study all-cause and cardiovascular mortality.
When applicable, we will adjust the models for age, sex, region and socioeconomic status.
Prevalence rates of heart failure and/or repeat loop diuretic prescriptions 2001-2003; Incident diagnosis of heart failure and/or repeat loop diuretic use and the temporal patterns in diagnosis from 2004-2017; Hospitalisation, all-cause, cardiovascular and for heart failure. All-cause mortality with follow-up through 2019; and cardiovascular and non-cardiovascular mortality with follow-up through 2019. Place of death from the death certificate.
Baseline characteristics, comorbidities and concurrent medications are defined using a priori code lists provided in Appendices B, C, and D.
Baseline characteristics, comorbidities and concurrent medications are defined using a priori code lists provided in Appendices B, C, and D.
John Cleland - Chief Investigator - University of Glasgow
Jocelyn Friday - Corresponding Applicant - University of Glasgow
Fraser Graham - Collaborator - University of Glasgow
Maria Wolters - Collaborator - University of Edinburgh
Pierpaolo Pellicori - Collaborator - University of Glasgow
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation