By Tom Foley.
Luke Readman, Chief Information Officer, East London Learning Health System
The use of data to drive improvement is central to the Learning Health System (LHS) approach in east London. CCGs, GPs, provider organisations and system suppliers in east London contribute to this work. Data driven collaborative improvement work in primary care locally has a 20 year history. This originated when GP practices in Tower Hamlets secured docklands regeneration funding to employ nurses on a prevention programme and establish electronic health records as early customers of the then new EMIS system. With Hackney and Newham GPs they formed a wider Clinical Effectiveness Group that standardised GP data entry and coding, using in-practice facilitators to support these new IT enhanced quality improvement programmes. All GPs in these CCGs have now joined on the basis of reciprocal sharing and self-funding. This has expanded to cover Waltham Forest, with emergent programmes in Havering and Redbridge. Key datasets have been shared with local authorities for innovative new programmes and research including the Wellcome funded east London Genes & Health study.
From the beginning, this group had a dual focus on analytics and improvement and has achieved notable targeted public health improvements. Despite having one of the most mobile and deprived populations, these CCGs achieved top ranking in 16 of the 65 total clinical measures in the national Quality and Outcomes Framework; including high impact blood pressure and cholesterol control (2016 QOF). The group were early adopters of the LHS philosophy popularised by Prof Charles Friedman. Much of the development, including around IG, is thought to have been possible because of trust established by the central Clinical Effectiveness Group which has established an evolving platform for a LHS.
There are two closely related programmes
1. East London Patient Record
Data sharing is taking place in real time for individual care between, 4 CCGs, 5 acute hospital sites, a mental health trust three sets of community services and almost 200 GP practices, covering a population of almost 1.5 million. Firm plans are in place to connect other care providers such as local councils. The network is used between 9,000 and 11,000 times per week and continues to grow. A variety of technologies are in use, allowing a summary of GP coded data and free text to be shared with secondary care and for data such as, discharge documents, outpatient correspondence and investigation results to be shared in the other direction. Many benefits are emerging and this is now leading to significant potential for wider reform for example of the outpatient pathway supporting the move to ‘advice and book’.
Viewing data between providers has been relatively straight forward. However, sharing it in a way that allows complex analytics requires semantic interoperability – the data has the same meaning within both providers. This interoperability is difficult to achieve, because of different coding and ‘concept’ structures used by each different provider. This is being tackled in two ways within east London. By standardising data entry and coding care, pathway by care pathway, around international SNOMED CT standards. Secondly using Natural Language Processing to derive meaning and codes from free text. This is currently experimental, but has the potential to greatly reduce time taken to achieve semantic interoperability. The group are working on this with a local company. Semantic interoperability is the area where overlap with the potential of the Discovery program is significant.
As well as allowing clinicians to view records from other providers, 18 months ago the group launched a new “Discovery Programme”. This is creating a unique dataservice of linked data from GP, hospital and local authority records. Discovery ensures the ‘Data Controller’ remains with the local GP, hospital and local authority providers and as such is locally responsive whilst being nationally scalable. Discovery has been offered 4 years funding from the CCGs and almost £5m from the Endeavour Health Charity (led by David Stables ex-EMIS founding director). The Discovery Programme is building 3 “engines of improvement”:
· Linked, identifiable real time data to enable predictive analytics and decision support algorithms (e.g. treatment choices or omissions at point of care).
· De-identified data for research and testing the algorithms above.
· Data to support new models of care and determine value (quality and cost). (E.g. identify and follow-up patients missed by vaccination or screening programmes. Patient flows through the system)
Both these programmes are underpinned by three key principles agreed to support informatics development in east London,
1. Develop a single system approach to information sharing that starts locally and builds collaboratively around patient care, rather than being driven top down;
2. Promote connectedness to ensure the information we give to any health or care professional is recorded properly and shared safely so that patients do not have to keep telling people the same thing time and again.
3. Make the best use of the data we have across the partnership for improvement, safely.
 Waltham Forest, Tower Hamlets, Newham and City & Hackney CCGs and GPs. Queen Mary University of London, Homerton University Hospital NHS Foundation Trust, Barts Health NHS Trust, East London NHS Foundation Trust, North East London NHS Foundation Trust, UCLPartners. Cerner, EMIS, TPP, RiO-Servelec