Investigating healthcare utilisation for heart failure patients with or without cardiac implantable electronic devices in England: retrospective cohort study.

Date of ISAC Approval: 
Lay Summary: 
Heart failure (HF) is a debilitating, chronic condition with approximately 550,000 UK sufferers. Around 1-2% of the NHS budget is spent on treatment, with 60-70% of the costs related to hospitalisation alone. Though drug therapy has been shown to decrease illness and death, many HF patients will continue to experience poor health. Cardiac implantable electronic devices (CIEDs) such as Cardiac Resynchronization Therapy (CRT), Implantable Cardioverter Defibrillators (ICD) and Left Ventricular Assist Devices (LVAD) have become increasingly important treatments for patients with moderate to severe HF. Implanted within the body, some devices are able to detect abnormal heart rhythms and correct them. Much research on the benefits of CIED treatment has, however, been derived from USA studies/clinical trials. Within the UK, little is known about its use in non-trial ('real world') populations. Our study will investigate the: 1) CIED use for HF patients in England; 2) potential economic impact (costs and/or savings) associated with the devices; 3) adverse outcomes associated with the device implantation and/or hospital stay. Also, comparisons will be made with similar HF patients who do not have a CIED. The study results can support formal economic evaluations (including potential uptake) of such devices within the UK.
Technical Summary: 
The study comprises retrospective cohort analyses examining healthcare utilisation and service outcomes for heart failure (HF) patients with or without a Cardiac Implantable Electronic Device, CIED, in England, diagnosed or with a device implanted between April 2010 and March 2012. For each patient group (those with CRT, ICD, LVAD or none), each kind of NHS contact (primary/secondary care) will be compared, with patient contacts summarised for 5 years after device implantation or first HF admission (diagnosis date will be used instead for HF patients without CIED). Standard national NHS reference costs for the relevant year will be applied for each contact. Analyses will begin with descriptive comparisons of HF patient groups by their socio-demographic factors, LOS and co-morbidities followed by logistic regression. We will attempt to infer, to extent possible, causal effects and other associations between device implantation and subsequent resource usage by use of propensity score matching on socio-demographic factors and co-morbidities. Using the matched groups, logistic regression models will be constructed with device (CRT, defibrillator, LVAD or none) as the treatment of interest; the outcome will be total cost during the five years following device implantation or diagnosis dichotomised into highest quintile vs lowest four quintiles.
Health Outcomes to be Measured: 
- Hospital use (ED visits, admissions, OPD attendances) - GP appointments - Death - Use of heart failure medications

Alex Bottle - Chief Investigator - Imperial College London
Alex Bottle - Corresponding Applicant - Imperial College London
Professor Azeem Majeed - Collaborator - Imperial College London
Professor Martin R Cowie - Collaborator - Imperial College London
Dr Wun Wong - Collaborator - Imperial College London

HES A&E;HES Admitted;HES Outpatient;ONS;Patient IMD