Overview/Challenge
The proliferation of large health-related data, their simplified access thanks to Trusted Research Environments (TRE), and the increasing volume of publications showcasing their potential for research have resulted in a rapid rise in the number of mental health researchers interested in making use of them. However, the use of such resources often requires skills such as data managing, data linkage, SQL and programming. These skills have not been typically considered in teaching curricula available to mental health researchers. As a result, there is a recognised lack capacity in the field. DATAMIND is trying to address this lack of capacity not only by training early career mental health researchers in such skills, but by providing clear and accessible definitions of mental health variables for physical health researchers to use in their fields of investigation. However, it is unclear what level of expertise in purely technical skills should be expected from the mental health research community, and how these could impact progress.
Impact and Outcomes
VELA’s lead developer is Dr Marcos Del Pozo Banos. It is currently being tested by Prof. Ann John’s team at Swansea University, whose feedback has resulted in several changes and additions to its first iteration. Testing researchers have provided positive feedback on the current second iteration of VELA:
“VELA allows to smoothly link all the required databases within a project in only a few lines of code… This package is both highly efficient, transparent, and customisable.”
“VELA is a very powerful tool for doing searches of our population health and administrative databases. The use of pre-entered and validated code-lists from primary care, secondary care and emergency department data reduces the possibility of errors, saves repetitive writing of scripts, saves time, and contributes towards increasing reproducibility of research.”
“VELA is a powerful tool allowing me to easily link databases and extract data for projects that would previously have taken thousands of lines of scripts in just a few lines of code. This streamlines the research process allowing research projects to move forward quickly and efficiently.”
What’s next?
- VELA and PHENOMIND will be fully documented in detail, including a set of “demos” and “how to” guides.
- We will further test VELA with other teams outside of DATAMIND to gather further feedback.
- We will publish VELA and its working principles in a methodological journal, using PHENOMIND as an example of an auxiliary package.
- We will make VELA and PHENOMIND packages publicly available.
- We will disseminate VELA and PHENOMIND through DATAMIND, social media and early career researchers groups.
- We will disseminate VELA and PHENOMIND to HDR UK and its Hubs, and participate in the HDR UK Tech Eco meetings.