Introduction: Background and outline of this thesis 17 Outline of this thesis The structure of this thesis will chronologically follow the progression of our research in four parts. Each part contains several chapters. To guide the reader through our research, we will provide further accompanying prologues and epilogues in between Parts. Hence, the outline described below is deliberately kept succinct. We start this thesis by providing the reader first-hand insight into two practical examples of expanded access programs design, conduct, and publication through two case studies of programs of which we presented the results at several conferences. We will cover all challenges and successes that accompany a real-world expanded access data collection program. As a prelude to the official thesis, we will explain (in Dutch) the way expanded access is regulated in the Netherlands as a reference for Dutch physicians. ♣ In Part I will assess the benefits of investigational drugs, as the clinical merit of expanded access is contingent upon the drugs it provides access to. The more expanded access is dedicated to highly effective pharmaceuticals, the higher the value of expanded access itself would be. We attempt to quantify the value of experimental drugs by exploring the likelihood that drugs advance through stages of clinical development and relate these probabilities to the assessed clinical benefit of drugs in development and new drugs on the market. ♦ We dedicate Part II to quantify the usage of expanded access data disseminated through scholarly publications, used by reimbursement bodies, and appraised by regulators. To process this information, we made use of computer algorithms to scan through large bodies of literature. The results of Part II form the basis of our further research and provide a basic understanding of when and how data from expanded access are used. ♥ Part III is dedicated to a novel statistical technique to incorporate expanded access data into clinical trial analyses. We will explore whether previous scholarship on dynamic borrowing of historical control information, together with more recent statistical advances put forth to include patient characteristics to determine the amount of information to borrow, can be adjusted to help
RkJQdWJsaXNoZXIy MTk4NDMw