Appendices 210 SUMMARY Medical and technological advances have improved patient care in heart failure, particularly in a subtype known as heart failure with reduced ejection fraction (HFrEF). The evidence which guides treatment standards come from carefully designed and executed clinical trials. A series of successful clinical trials since the 1990s have provided us with at least five drug categories that are effective in reducing deaths and other medications which reduce symptoms and hospitalization in selected patients. On a global scale, this signals availability of effective treatment for over 64 million people living with heart failure. If most of these patients were adequately treated with these medications, we expect to witness progressive gains in population survival. This observation holds true for a period between the 1990s and 2000s, where deaths among people with heart failure declined. However, mortality trends thereafter have stalled. Among the reasons for the lack of improvement in population survival is that only a fraction of people with heart failure receives adequate treatment for the disease. i.e., the full potential of treatment is not yet achieved. Clinical trials are known to recruit a narrow population, usually younger and have few to no background health conditions, which leaves us with an imbalance of data especially among older, sicker patients and women. The patchy evidence makes it challenging for health practitioners to initiate treatment among people who are typically underrepresented in clinical trials. In this thesis, the aim was to understand the mismatch between characteristics of clinical trial participants and usual care patients and subsequently how this mismatch affects disease outcomes between the two populations. This objective will be explored and quantified based on eligibility to participate in trials and how these measures impact representativeness and generalizability of trial data. Understanding the degree of mismatch is necessary to find ways to bridge the imbalance in evidence across all patient types. It is important to realise that prevailing challenges to this first step include limited access to clinical trial datasets particularly for pharmaceutical industry-sponsored trials. In this respect, public-private partnerships such as the BigData@Heart collaboration are instrumental in facilitating the work in this thesis, as the working relationship enabled pooling and sharing of patient data between commercial and academic entities
RkJQdWJsaXNoZXIy MjY0ODMw