Search

In cost-effectiveness analysis, outcomes are typically averaged across large groups to represent a patient population. Implementation and reimbursement decisions based on such analyses often ignore considerable heterogeneity in cost-effectiveness between patients. While good practice guidance for economic evaluations suggest including subgroup analysis, in practice this is frequently overlooked or underutilised. This thesis shows that failing to adequately represent heterogeneity in decision-making leads to an inefficient distribution of healthcare resources. This theory is applied in a study of cholesterol-reducing medication for the primary prevention of cardiovascular disease (CVD). Despite improvements in recent years, CVD remains a significant cause of mortality, morbidity, and health inequality around the world. Rates of the disease have begun to plateau in recent years and novel approaches to its prevention are required. Cholesterol reduction for the primary prevention of cardiovascular disease is a clinical area where better reflection of heterogeneity in cost-effectiveness could significantly improve current practice. Statins are a widely prescribed cholesterol-reducing medication which have recently come off patent. This has led them to become cheaper and cost-effective in a large proportion of CVD-free populations in high-income countries. PCSK9 inhibitors are a more expensive and more effective cholesterol-reducing medication. For both treatments, decision-makers must establish which groups they will prioritise for treatment. Through epidemiologic and health economic analysis, this thesis aims to establish optimal approaches for prioritising patients for cholesterol-reducing therapy. Preventive statin therapy is typically targeted at individuals estimated to have a high ten-year risk of developing CVD. However, individuals with the same ten-year risk may experience different outcomes from preventive treatment. The epidemiologic bases for three alternative approaches to the CVD prevention are discussed. These are: (i) continued use of ten-year risk scoring, (ii) novel decision mechanisms which incorporate ten-year risk, and (iii) direct use of decision-analytic models in clinical practice to guide treatment decisions. Several treatment policies may be characterised by one of the aforementioned approaches to prevention. These include: lowering the risk threshold for treatment initiation, improving the discrimination of risk scores with novel biomarker testing, age-stratified risk thresholds, absolute risk reduction-guided therapy, and outcome maximisation with decision-analytic models. Decision-analytic modelling was employed to assess the long-term effectiveness and cost-effectiveness of these policies. Additional analysis showed how decision-makers can signal demand for PCSK9 inhibitors and achieve welfare gains by reflecting heterogeneity in their decision-making. This thesis demonstrates the importance of reflecting heterogeneity in cost-effectiveness. It shows that standard care regarding the primary prevention of CVD often ignores heterogeneity, leading to suboptimal decision-making. This holds true for long-established, inexpensive treatments like statin therapy and novel, expensive treatments like PCSK9 inhibitors.

Additional follow up study results tables (baseline characteristics, clinical and psychological outcomes, and lifetime economic evaluation). (DOCX 103 kb)

Economic evaluation supplementary tables. (DOCX 57 kb)