Application of statistical and decision-analytic models for evidence synthesis for decision-making in public health and the healthcare sector

Abstract

With the awareness that healthcare is a limited resource, decision-makers are challenged to allocate it rationally and efficiently. Health economic methods of evidence synthesis for decision-making are useful to quantify healthcare resource utilisation, critically evaluate different interventions and ensure the implementation of the most effective or cost-effective strategy. The nine studies included in the present cumulative doctoral thesis aim to demonstrate the capability of statistical and decision-analytic modelling techniques to inform and support rational healthcare decision-making in Germany. Five studies apply statistical modelling in analyses of public health and health economic data. They show that the developed models are valuable instruments for examining patterns in the data and generating knowledge from observable data which can further be used in devising disease management and care programs as well as economic evaluations. Further, two health economic evaluations, which adopt the decision-analytic-modelling approach, show that decision-analytic modelling is a powerful tool to represent the epidemiology of infectious and non-infectious diseases on a population level, quantify the burden of the diseases, generalise the outcomes of clinical trials, and predict how the interventions can change the impact of the diseases on the health of the population. Additionally, two literature reviews examine the application of decision-analytic modelling in health economic evaluations. The first study reviews and empirically analyses health technology assessments by the German Institute for Medical Documentation and Information and demonstrates that the application of decision-analytic models improves the evidence produced for policy-making in the healthcare sector in Germany. The second systematic review focuses on methodological choices made in constructing decision-analytic models and explains how critically the structural and parametrical assumptions can influence the final message of the economic evaluations and shows that building a validated, reliable model as well as the transparent reporting is of high priority in facilitating the communication and implementation of the most cost-effective course of action. Overall, the present thesis shows the relevance and advantage of the application of models in synthesising evidence for decision-making. The included studies contribute to the current and future development of the methods used to address the problems of health economic efficiency. Further advances in the computational modelling techniques and data collection, from one side, will ease the decision-making process, but, from another side, will require increasing competence and understanding within the decision-making bodies.