Overview

The "Uncertainty Project" is a multi-center, phase I, survey-based study to investigate the frequency, consequences, and associations of prognostic uncertainty in patients with cancer.

Our objectives

First and foremost, our objective is to improve care and outcomes for our patients. With the study, as appropriate for any pilot study, our objectives were to establish the acceptability of our study design for participants, the frequency of adverse events, and the feasibility of our study design in a multi-center setting that included both academic and private oncology practices.

Our underlying hypothesis

Prognostic uncertainty operates differently in different situations: patients with dismal tumor-related prognoses might derive hope from uncertainty, while patients with favorable prognoses often respond to uncertainty with fear and apprehension.

Our data-exploration application

This website was originally developed to inform our monthly virtual meetings when we reviewed progress with data collection. It has since become an indispensable tool with the potential for much broader impact. We feel the approach we've taken here can be used in virtually any clinical trial, and might inform the design and conduct of clinical trials at all stages. Moreover, by making this application openly accessible to anyone with an Internet connection, we feel that readers of our scientific publication(s) will be empowered to explore our data, verify our results, and perhaps develop their own novel hypotheses. In a sense, we would thereby recruit those readers as virtual co-investigators. We propose that this approach may serve as a model for publishing clinical trials in the twenty-first century.

Technical implementation

This website relies on two remarkable technologies and the organizations that have made them available for almost no cost. We use the Vanderbilt implementation of REDCap for data collection and storage, which has an open interface for data access (an Application Programming Interface or API). And we built the interactive data visualizations and website with Observable Framework, which is open source software based on the open standards that underlie all web technologies. The result is an easy to use, almost no-cost web app that provides collaborating physicians -- clinical oncologists who spend most of their time saving lives -- with an easy-to-use tool for interactive exploration of de-identified, patient-level data. The collaborators use this site whenever they have time to think about the project.