Creating intelligence for improving value across a system

by Catherine Mitchell

02 September 2020


How do we best allocate resources across a health and care system to improve population health? Most would agree this is a key challenge facing the NHS, but creating the necessary intelligence for value-based decision making is not straightforward.

The roll-out of patient-level costing (PLICS) across England means that the NHS has increasingly rich information about how resources are used within provider trusts. For example, Gloucestershire Health and Care NHS Foundation Trust used PLICS to compare the use of community services by patients with and without diabetes during the Engagement Value Outcome (EVO) pilot. It found that patients with diabetes consume almost six times as much resource as those without. And patients who attend a diabetes education programme have a significantly lower need for community services than those who have not. Already at the provider level, value-based decisions can be made.

However, the increasing emphasis on improving population health and better integrated care means that sustainability and transformation partnerships (STPs) and integrated care systems (ICSs) require information on how resources are used across the organisations in their patch at a patient level. The Healthcare Costing for Value Institute briefing What finance data is required to drive value at a population level? starts to explore what this means in practice.

Continuing with the diabetes example, if Gloucestershire ICS could link community services data with primary care, acute and ambulance data, the system would have a better understanding of the overall pathway for patients with diabetes.  Each part of a local health and care system has a different part of the jigsaw and joining the data up to get a complete picture comes with a number of challenges.

The HFMA briefing Understanding resource consumption across a system describes the approach adopted by Derbyshire STP to join up data across the system in 2019*. As it started to develop integrated care at place-level, the STP wanted to know which patients used most of their services and where they lived in Derbyshire. Focusing on the data that was already held by its trusts, episodic data was linked at the patient level so that the STP had a better understanding of how resources were consumed across the system. 

After gaining agreement across the system for the need to join the data and its proposed use, the biggest (and most time consuming) challenge for those involved was establishing data sharing agreements. These enabled data to be collated at the patient level and fed back in a way that supported decision-making at both the individual organisation and STP level, without contravening data protection rules. If the NHS wants to improve value at the population level, sharing data across the system has to be made simpler.  

Once the data sharing agreements were in place, bringing together the PLICS data from the four providers (acute, mental health and community) was achieved fairly quickly. The STP now has a data cube that identifies the relative resource consumption by GP practice, and demonstrates that the majority of the resources are consumed by less than 20% of the population. The data can be sliced and diced in many ways, providing the STP with a rich source of information to start to consider how services could be provided in a more integrated way in future. 

Derbyshire knows that it is only at the start of the journey in improving value at the population level, but its pragmatic approach is a good first step.




The HFMA has also published High utilisation patients in Derbyshire STP. This is based on the consultancy project submitted by Derbyshire Support and Facilities Services’ Steven Heppinstall as part of his MBA in healthcare finance.

Find out more about HFMA’s Healthcare Costing for Value Institute here.