While international guidelines for the management of heart failure (HF) largely overlap, important regional differences exist in patient characteristics, prescription patterns, healthcare resource utilisation and consequently, in clinical outcomes, between the Western and Asian countries. To date, comparative data on geographic differences in patient characteristics, management and outcomes has either been based on post-hoc analyses from clinical trials, aggregate patient data or from registries that lacked both quantum and generalisability. I therefore performed an individual patient-level analysis of more than one million non-consecutive heart failure hospitalisations across four countries – UK, US, Japan, and Taiwan – using nationally representative electronic health care records (EHR), which captured routine clinical encounters (Specific Aims 1 and 2).
Although, multiple studies have been undertaken to efficiently identify patients with HF in the real world using EHR, there is a paucity of evidence on identification and validation of algorithms for identifying patients with specific HF phenotypes, including heart failure with reduced ejection fraction (HFrEF) and heart failure with preserved ejection fraction (HFpEF). As part of specific aim 3, I have focussed on improving the ability to identify patients with different HF phenotypes within the EHR, by collecting additional data (including direct questionnaire to General Practitioner’s), and by seeking to understand the impact of different algorithms for identification of HFrEF and HFpEF.
HFpEF patients are widely believed to be a heterogenous and can be broadly categorised into two further sub-phenotypes at the population level: 1) young obese patients and 2) older non-obese patients where hypertension, chronic kidney disease, coronary artery disease (CAD) and atrial fibrillation are the predominant drivers. As part of specific aims 4 and 5, I have evaluated the obese HFpEF sub-phenotype and the impact of coronary revascularisation in patients with HFpEF (CAD-HFpEF sub-phenotype) using nationwide EHR.