A data driven semantic framework for clinical trial eligibility criteria

Abstract

An important step in the discovery of new treatments for medical conditions is the matching of potential subjects with appropriate clinical trials. Eligibility criteria for clinical trials are typically specified in free text as inclusion and exclusion criteria for each study. While this is sufficient for a human to guide a recruitment interview, it cannot be reliably parsed to identify potential subjects computationally. Standardizing the representation of eligibility criteria can help in increasing the efficiency and accuracy of this process. This thesis proposes a semantic framework for intelligent match matching to determine a minimal set of eligibility criteria with maximal coverage of clinical trials. In contrast to top down existing manual standardization efforts, a bottom-up data driven approach is presented that finds the canonical non-redundant representation of an arbitrary collection of clinical trial criteria set to facilitate intelligent match-making. The approach is based on semantic clustering. The methodology been validated on a corpus of 708 clinical trials related to Generalized Anxiety Disorder containing 2760 inclusion and 4871 exclusion eligibility criteria. This corpus is represented by a relatively small number of 126 inclusion clusters and 175 exclusion clusters, each of which represents a semantically distinct criterion. Internal and external validation measures provide an objective evaluation of the method. Based on the clustering, an eligibility criteria ontology has been constructed. The resulting model has been incorporated into the development of the MindTrial clinical trial recruiting system. The prototype for clinical trial recruitment illustrates the real world effectiveness of the methodology in characterizing clinical trials and subjects, and accurate matching between them.