The healthcare costs of heart failure during the last five years of life: A retrospective cohort study

Abstract

RW is an employee of CPRD, who received funding from the University of Bristol to provide access to the data for this study. CPRD provide research services for, and receive associated funding from, a range of pharmaceutical, academic, governmental, NGO and regulatory organisations. All other authors declare no conflicts of interest.MJD has accepted fees for advisory boards from St Jude who make a range of cardiology devices and also from Bayer for support for attending conferences. 3 Abstract: AimsEvidence on the economic impact of heart failure (HF) is vital in order to predict the costeffectiveness of novel interventions. We estimate the health system costs of HF during the last five years of life. MethodsWe used linked primary care and mortality data accessed through the Clinical Practice Research Datalink (CPRD) to identify 1,555 adults in England who died with HF in 2012/13. We used CPRD and linked Hospital Episode Statistics to estimate the cost of medications, primary and hospital healthcare. Using GLS regression we estimated the relationship between costs, HF diagnosis, proximity to death and patient characteristics. ResultsIn the last 3 months of life, healthcare costs were £8,827 (95% CI £8,357 to £9,296) per patient, more than 90% of which were for inpatient or critical care. In the last 3 months, patients spent on average 17.8 (95% CI 16.8 to 18.8) days in hospital and had 8.8 (95% CI 8.4 to 9.1) primary care consultations. Most (931/1555; 59.9%) patients were in hospital on the day of death. Mean quarterly healthcare costs in quarters after HF diagnosis were higher (£1,439; [95% CI £1,260 to £1,619]) than in quarters preceding diagnosis. Older patients and patients with lower comorbidity scores had lower costs. ConclusionHealthcare costs increase sharply at the end of life and are dominated by hospital care. There is potential to save money by implementation and evaluation of interventions that are known to reduce hospitalisations for HF, particularly at the end of life.

Effectiveness and cost-effectiveness of serum B-type natriuretic peptide testing and monitoring in patients with heart failure in primary and secondary care: an evidence synthesis, cohort study and cost-effectiveness model

“…10 The recent literature 25,33 and our own work in this project have used IPD meta-analysis of RCTs to explore the relative effectiveness of BNP-guided monitoring in various subgroups of the HF population including age (< 75/≥ 75 years) and LVEF (≤ 45%/> 45%) status. The evidence that BNP-guided therapy is effective is strongest in younger patients and in those with HFrEF.…”

“…However, there are no suitable data on the broader economic burden of HF in the UK. Health benefits were quantified in terms of QALYs, 10 calculated by multiplying the utility value for each health state with the time spent in that state. 75 A utility value represents a patient's health-related quality of life …”

“…83 We estimated the costs of managing patients hospitalised with HF and with stable HF in the community, stratified by age group, based on the CPRD-HES-ONS linked data. 10 In brief, we identified 1555 adults in England who died with HF in 2012/13. We used the CPRD-HES linked data to estimate the cost of medications, primary and hospital health care during each 90 day period in the 5 years before death.…”

“…9 Health-care costs increase sharply at the end of life and are dominated by hospital care. 10 The most common causes of HF are ischaemic heart disease and high blood pressure. Other causes include congenital heart defects, genetic disease of the heart muscle, cardiac arrhythmia and alcohol misuse.…”

Methods for estimating costs in patients with hyperlipidemia experiencing their first cardiovascular event in the United Kingdom

“…As a result of these resources, a variety of studies have been conducted that estimate costs based on resource utilization derived from CPRD and HES data, many of which were published recently [7][8][9][10][11][12][13][14][15][16][17][18] . However, it is still challenging to understand the methods used, and to implement them in research 19 .…”