Part II ♦ 80 PROLOGUE In this part, we aim to establish a systematic understanding of the use of data from expanded access programs. Inspired by anecdotal evidence reports, such as the Reagan-Udall Foundation report,1 which mentioned the inclusion of data from expanded access program to support clinical efficacy, we started wondering how often data from expanded access programs were included in regulatory filings. We were primarily interested in the efficacy from these programs – as safety reporting is both obligatory and there has been prior scholarship devoted to exploring the implications of expanded access safety issues on clinical development. Hence, we aimed to first find out how many regulatory filings included data from expanded access programs to support efficacy. Traditional regulatory analysis involves a manual review of each drug approval, a process that would be time-consuming and error-prone for the vast amount of data involved. To give the reader a sense of the magnitude of this issue: the FDA typically employs various types of reviews: a summary review, a medical review, a chemistry review, a pharmacology review, a statistical review, a clinical pharmacology biopharmaceutics review and a microbiology review, to name a few. Scanning these documents for decades of drug approvals would be an onerous, inefficient, and fallible process. Furthermore, given only the handful of examples mentioned in the ReaganUdall Report, we might be looking for a needle in a haystack. Similar issues pertain to the EMA, where thousands of pages of information are available per drug approval. We therefore designed an algorithmic approach that automatically downloaded all available documents from the FDA and EMA websites, and then scanned all documents for any terms relevant to expanded access, ‘Compassionate Use’, ‘Expanded Access’, ‘Early Access’, ‘Single-patient IND’, and all possible spellings thereof. In this way, we could reasonably argue that documents that did not include such a term were unlikely to have employed data from expanded access programs, and we only manually had to search through the documents that appeared in our algorithm. We reused this technique several times. We could not only analyze FDA or EMA documents, but also approvals from other jurisdictions. And there was no reason to limit ourselves to regulators – we extended our work to analyze whether reimbursement decisions employed expanded access data. Lastly, we did a thorough systematic review of all scholar publications that disseminated results of expanded access drugs. Here, we made use of old-fashioned, authentic independent review techniques. Throughout this chapter, we show that the analysis of vast amounts of health policy documents
RkJQdWJsaXNoZXIy MTk4NDMw