Alleviate: Hub for Pain

Lead Research Organisation: University of Dundee
Department Name: Population Health and Genomics

Abstract

Our vision for the Alleviate Hub for Pain is to:
- Transform existing UK pain datasets to be Findable, Accessible, Interoperable and Reusable (FAIR)
- Link these datasets with expert data engineering, integrated into the Health Data Research (HDR) UK Gateway
- Catalyse responsible and trustworthy analysis by international researchers and innovators

Our UK-wide consortium will deliver world class health data infrastructure and services for pain research, guided by leading experts in pain research and in partnership with the NHS, Advanced Pain Discovery Platform (APDP) consortia, people with lived experience of chronic pain (PWLE) and industry. This consortium has a common desire and a commitment for change to a "data as infrastructure" approach. The fact that custodians of diverse datasets have chosen to join our consortium is testament to the need to develop new improved, more efficient methods of bringing everyone together under a common approach. Alleviate will engage directly with PWLE to maximise impact and patient benefit. We have already held our first 'In-Principle' meeting with 9 representatives, from diverse backgrounds, with 2 PWLE co-leading the Public and Patient Involvement and Engagement Work Package.

The Dundee Health Informatics Centre (HIC) has considerable data engineering expertise and has provided leadership in health data science and research services for over a decade. Using a secure ISO27001 certified Trusted Research Environment (TRE)/Safe-Haven (where data security and safety is guaranteed) managed by HIC, Alleviate will:
- Leverage existing open-source software and infrastructure to curate and manage data (research cohorts and routinely collected health (and social) care data), streamline access and support governance
- Provide a modern, secure, and flexible design, benefitting from HIC's secure, scalable hybrid cloud 'Next-Generation' infrastructure; supported by cloud computing expertise from Amazon Web Services (AWS), enhancing an existing successful partnership
- Support analysis of multi-modal data using artificial intelligence and machine learning (e.g. epidemiological, clinical, population health, genomics, imaging, psychological, social, biological)
- Deliver a hybrid model of data access and management, i.e. supporting UK-wide data federation combined with a centralised model (Figure.1). The model will also support international datasets in the future
- Enable pain research datasets and registries, facilitating their curation and further collaboration, linking with other HDR UK Hubs, related clinical datasets (e.g. mental health, cancer) and boosting their power and discovery potential for patient benefit

Our consortium has a considerable track record:
- Specific expertise in multi-centre pain-related data
- Leading academics in pain research covering the range of data modalities across biological, psychological and social factors that influence pain. Our Co-I's are leading applicants in 4 separate APDP consortia bids
- Experience in collaborating between clinical researchers, PWLE, laboratory scientists and data scientists at local, national and global level
- Experience as a data processor for several health boards, Scottish Government, and many UK and international research cohorts
- Existing HDR UK Alliance members and strong links with other hubs e.g. Breathe
- Custodians of a TRE for secure analysis of health data at scale
- Leading many HDR UK initiatives: Phenomics portal, Scottish Data Federation, Multiomics project, Biorepository management software, Imaging, open Connector, ATLAS
- Enabling c.900 national and international research projects as a service provider (c.5years)
- 'Big-data' pipelines for efficient operation including omics to explore the biopsychosocial model of chronic pain
- Experience of building research networks and with ready access to the UKCRC Tissue Directory and Coordination Centre of over 200 biobanks

Technical Summary

The Alleviate will adopt a hybrid architecture: each group donating a pain dataset will choose whether to use the centralised hub or act as a federated node based upon data governance, size and complexity of the data and technical feasibility. The Research Data Management Platform (RDMP) open-source software tool will be utilised to curate, clean and transform data and automatically upload metadata and data dictionaries into the HDR Gateway.

Within each federated node, a pseudonymised version of the data will be made available within a secure VM so that it can be queried by federation software integrated into the HDR Gateway. This will enable researchers and innovators to discover what data is available for pain research across the UK and run high level meta-analysis across the datasets to find relevant cohorts. Such capability will support the research community to find relevant datasets which can be efficiently re-used and support recruitment to clinical trials. Alleviate will automate the process to extract research project specific data from across the federated network into a single TRE, streamlining access to data once data governance approvals have been granted.

Alleviate will upload pain specific phenotypes into the HDR phenotype portal.

The Alleviate hybrid secure cloud Trusted Research Environment (TRE) will be delivered in partnership with Amazon Web Services, building on an existing successful collaboration over the past year. The environment will provide Next Generation capabilities including:
- A suite of regularly updated, out-the-box software tools supporting machine learning (ML) and artificial intelligence (AI) development, statistical analysis and data management
- Improved support for multi-modal data cohorts including imaging & genomics integration
- Rapid scalability / elasticity / flexibility - as project demands require
- Reduced capital investment in new physical infrastructure
- High-availability
- On-Demand Self-Service

Publications

10 25 50