The use of patient-generated health data (PGHD) to support clinical care and research in MSK disease

Background
The increasing uptake of consumer mobile devices, such as smartphones and smartwatches, offers unique opportunities to collect data directly from patients for clinical care and research. This might include digital patient-reported outcomes and sensor data including accelerometers and GPS. In addition, citizens are increasingly leaving behind ‘digital traces’ including web searches, social media discussions and activity tracking, all of which might provide insight about the onset and progression of disease, and adverse reactions related to medication use. There are, however, major challenges for this novel data including understanding how it should best be collected, processed, and analysed. Optimal PGHD collection needs to balance the needs of researchers and clinicians with what is acceptable to patients longitudinally. Data can be voluminous, unstructured and messy, and need to be summarised in a clinically-meaningful way. Lastly, traditional epidemiological methods of examining associations, such as logistic regression, may not suit this novel data.

Aims
The overarching aim is to examine how digital PGHD can advance clinical care and research in long-term MSK conditions, using existing and forthcoming digital health studies including apps and wearables hosted at the ARUK CfE. The specific aim(s) of the PhD will be agreed with the successful candidate and may include:
• Collection: Are mHealth study populations representative? How are the granularity, quality and volume of data desirable for clinical care and research balanced with the burden of data entry and subsequent attrition? Are novel data collection tools valid?
• Processing and description: How can temporally-rich and unstructured data be meaningfully summarised to support patient self-management, clinical care and research?
• Analysis: What methods will usefully identify patterns within large PGHD datasets? This is exemplified by examining the pattern and impact of flares, and the feasibility of accurate flare prediction to inform preventive and self-management interventions. How should case-only designs and n-of-1 studies be best used for studies of digital PGHD to assess impact on longitudinal clinical outcomes?

Entry Requirements
Candidates are expected to hold (or be due to obtain) a minimum upper-second (or equivalent) class undergraduate degree in a clinical discipline, statistics, health informatics, psychology or a relevant subject, and will have strong statistical skills. A Masters degree in a related subject and/or relevant research experience is desirable. Experience of digital health studies or patient-generated health data is desirable.