Improving Patient-Reported Outcome Measures Through Visual Analytics

Problem

Patient-Reported Outcome Measures (PROMs) are one way of monitoring changes in the natural history and progression of a patients health. They are generally sets of questions that try to capture the subjective perception of the patient about their health, such as "do you feel frequently tired?", or "do you have problems driving a vehicle?". They can be an important instrument specially in understanding the impact of a condition in quality of life, and have been used increasingly in clinical trials, when testing new medication, etc. They can be generic, trying to capture general aspects of health, such as energy and disposition, or specific, focusing for instance or vision or on a specific disease. A common way to apply them is to combine generic and specific questionnaires so that one has a ground truth of comparison across many patients and conditions.

Among the challenges facing doctors that want to use PROMs is how to chose the adequate ones, how to balance between the cost and effort of applying them, and how to properly interpret them. It is not possible to ask patients to answer hundreds of questions, and the more answers a patient has to give the lower their quality will be. From the large amount of available PROMs many end up covering similar aspects through different questions and scoring systems, making it hard to compare between them and understand the relation between populations and cultures.

Aim

This project aims to improve the use of PROMs in medicine, both in clinical trials, treatment, and research with the assistance of Visual Analytics. There are three main dimensions to be considered, with different research questions:

  1. From the side of the patient: How can we assist the patient in efficiently communicating his issues with the minimum input and cognitive effort, therefore improving the quality of the responses?
    • Can we develop a better visualization and interaction with the questionnaires?
    • What visual encodings could be used for answering health related questions?
    • How to guide and onboard patients to provide better responses?
    • How to show patients their past answers to help them provide better feedback over time?
  2. From the side of the Doctor: How can we assist the doctor in better understanding the condition of the patient based on his responses?
    • How to summarize the data from the patients answer in a visually efficient way?
    • How to chose from the many questions the patient answered which are more important for the doctor in each case?
    • How to better help the doctor to understand the evolution of the patient over time?
  3. From the side of the Data: How to extract useful knowledge from a large amount of PROM data among different domains and conditions?
    • How to summarize and compare data from different PROMs?
    • How to use PROM data to improve diagnosis?

Contact

Further information

Topics
Visual Analytics, Medical Informatics, Patient Reported Outcome
Area
Information Extraction (IE) and Transformation
Information Visualization (IV)
Medical Informatics
Visual Analytics (VA)
Previous knowledge
Good knowledge of some web application or visualization development technology such as JavaScript/TypeScript, D3, Frameworks (React, Angular, Vue, etc.)
English
Scope
SE
BA
PR
MA
PhD
Status
open