Population Health, AI, predictive analytics, the power of data, and the cutting edge techniques for its use is only increasing. Organisations are rushing to implement these new approaches, in many cases, in a belief that “the new” offers all the answers. But the best AI model, the most advanced predictions, are only as good as the data underpinning them. In rushing to procure these new things, we may put the cart before the horse.
The proliferation of data within health and care has also been accelerating in recent years, thanks to advances in technology and growing amounts of patient-generated data. From electronic health records (EHRs) to wearable devices, the data available about health and care has grown rapidly. In addition, as we move through an increasingly electronic healthcare landscape, the amount of data generated as a by-product of those actions, the “data exhaust” we create, is growing.
Whilst this data exists, the data held is often siloed, held by multiple organisations in primary and secondary care, community and ambulance providers. However, whilst these organisations take care of individual elements of our health and care, we as individuals move through the orbit of these care providers, and to truly understand and improve the provision of care and improve the health of society, we cannot look at a particular element of care in isolation. If we improve A&E care for example, the effects may be seen in the hospital, in primary care, and even in the community. Data is also held by “non-healthcare” organisations; local authorities, the third sector and others who may not consider themselves as providing healthcare, but whose actions are increasingly recognised as playing a crucial role in determining how healthy you or I may be.
Integrated care boards and systems were implemented to resolve some of the problems that occur when health and its determinants are cared for by disparate organisations, however, they each now inherit a data estate covering multiple owners, systems, formats and crucially, of varying coverage and quality.
The whole system view the Integrated Care Boards are required to take will almost certainly improve care and potentially lead to the introduction of more effective population health initiatives, but before all this can be delivered, it is crucial that a clear data strategy is in place.
A data strategy lays out an organisation's objectives, identifying the types of data that the organisation needs to collect, takes stock of the quality and type of existing data, and determines how to collect, store and use data in the future to support the organization's objectives. More than just a determination that data needs to be stored, it involves identifying the sources of data to be preserved, how they should be combined, the access controls that will enable different elements of the system to use the data effectively, and how the metadata and lineage of the data will be recorded and maintained.
Of course, the strategy is only useful if it is followed up by action, but this investment is crucial to monitor and realise the benefits that the (Integrated Care Systems) ICS can bring. This work is necessary for every ICS in the country, as they have each inherited a different estate, but as reported in the Health Service Journal (HSJ), with 40% of ICSs lacking a data strategy, there is still a distance to go before we see the benefits. When a data strategy is designed well, all organisations involved can have trust that their data will be held securely, contributing to more collaboration, and creating the foundations for the delivery of the goal of better healthcare for all.
Neil Mason, Principal Consultant, Methods Analytics.