DQueST

DQueST is a novel software system designed to help patients find relevant clinical trials more efficiently by reducing information overload. It works in two main steps:

1. Information extraction and criteria library curation: DQueST processes the eligibility criteria from ClinicalTrials.gov, structuring and normalizing the data and storing it in a curated library.

2. Dynamic question generation: When a user searches for trials, DQueST generates questions in real time based on the curated criteria library. It selects the most relevant criteria one at a time, asking the user corresponding questions. Based on the user's responses, DQueST dynamically filters out ineligible trials, iteratively refining the search results.

Simulation experiments showed that DQueST could filter out 60-80% of initial trials after 50 questions, with an estimated accuracy of 63.7%. In a study using mock patient profiles, DQueST retrieved trials with a 1.465 times higher density of eligible trials compared to an existing search engine. Patients found DQueST useful and easy to use and returned relevant results in a usability evaluation.