A risk stratfiication tool used by GP practices in Leeds helped to identify where early intervention would benefit patients with long-term conditions - and save money, write James Hoult and Hailey Matheson

Illustration of people walking in a park


The Department of Health estimates approximately 80 per cent of primary care consultations, two-thirds of emergency hospital admissions and two-thirds of healthcare costs in the UK are related to long-term conditions. One in three people in England have a long-term condition.

As the population ages there will be an increase of people affected by long-term conditions creating a strain on existing health and social services. The risk stratification tool allows risk profiling by allocating a risk score dependent on the complexity of their disease type or multiple comorbidities.

‘As the population ages there will be an increase of people affected by long-term conditions’

This risk score indicates those people who are the more frequent users of primary and secondary care services that are likely to increase their level of use. This will support the commissioning of services and enable needs to be flagged up earlier.

Targeted search

Research has shown that patients with multiple comorbidities currently accessing healthcare resources over a 12-month period are likely to continue. Therefore the greatest impact could be seen by targeting those predicted to enter this high-risk category, rather than focusing on those already in this category.

Potentially this will have a further effect, by reducing the chances of these people reaching this high level. Such research prompted GP practices in Leeds to use the risk stratification tool to highlight the patients currently in the medium-risk category, but predicted to be in the high-risk category in 12 months’ time.

Since January 2012, the three clinical commissioning groups in Leeds (Leeds South and East CCG, Leeds North CCG and Leeds West CCG), with support from NHS Leeds, have worked to develop and deploy risk stratification across all GP practices. In doing so each practice was provided with the necessary software and staff training to use the new system.

Key objectives of risk stratification in Leeds

  • Identify people who are more likely to require hospital or long-term care in the future, in order to target them with more intensive support at an earlier stage. This will reduce this risk and so reduce the number of people who go into A&E or hospital unnecessarily, or need long-term social care.
  • When people do go into hospital, our aim is to reduce their stay when possible and ensure they are discharged in a co-ordinated and timely manner, with tailored information and support to help them take more responsibility for their own wellbeing.
  • Reduce the number of patient primary and secondary care interventions and so reduce the increasing cost of patients using primary and secondary care resources where not required. This has the potential to reduce Leeds healthcare costs.
  • To increase the current number of people able to co-manage their own condition with the support of health and social care professionals.

By October 2012 risk stratification had been successfully implemented in 111 out of 112 Leeds GP practices. They were required to sign a data-sharing agreement to provide permission for extraction of the required data from their clinical system to be processed through the risk stratification tool. All data input into the tool is updated monthly.

The tool that practices in Leeds are using is known as the adjusted clinical groups system, developed by Johns Hopkins University. It is based on an algorithm that brings together primary and secondary care patient data in order to calculate an individual’s risk of needing a greater level of support within the following 12 months.

Patients are assigned to unique categories based on patterns of disease and the expected resources that will be needed to treat and support them. These categories are called adjusted clinical groups.

Risk profiling and care management in Leeds will enable GP practices to identify these patients at each individual practice and coordinate and manage the care of high-risk patients, anticipated to be at substantial risk of unscheduled hospital admission. This cohort of patients is considered to become potentially high-risk in future and is predominantly made up of the ageing population.

Assessing the benefits

Following practice feedback, in November 2012 the risk stratification tool was enhanced with the introduction of additional functionality. Enhancements to the risk stratification tool included patient identifiable data including NHS number, forename/surname (this is for direct NHS care only and is protected by appropriate security mechanisms).

New functionality also included a patient search function, quality and outcomes framework indicators, inclusion of body mass index and smoking status data, and a “timeline” that details the patient encounters over a twelve month period with  primary, secondary care and community and social care.

Along with the release of phase 2 of the tool, a risk stratification intranet site and dedicated risk stratification helpdesk was introduced to all practices in order to provide additional information and technical support.

As part of the QP payments for 2012-13 practices are required to review outputs from risk stratification and identify patients who would benefit from a multidisciplinary approach. The multidisciplinary team (MDT) is required to meet twice a year in order to review the health and social care plans/personal care plans for patients selected using the risk stratification tool.

Following their first MDT meeting the majority of practices have recognised the benefits of having this group discussion regarding patient care and have scheduled more than the current required two MDT meetings per year.

The multidisciplinary team comprises of a core team based around the GP practice supplemented with specialist attendance when particular cohorts of people are being focused on, for example respiratory problems. The first phase of looking at the risk stratification outputs enables practices to optimise primary care, and also to identify people who would benefit from an integrated approach.

Co-located district nurses, community matrons and social workers have been co-located in 12 neighborhoods across the city, working with practice populations. We are now working with those core teams to work out how they can begin to work differently and bring in the skills of intermediate care and specialist teams. 

Each team has established a single point of contact to allow a more efficient method of communication to arrange the MDT meetings and discuss the follow-up of patient care post MDT meeting. The intention is for these established teams to proactively provide care for those at risk.

‘Multidisciplinary team meetings are a useful discussion forum to address more complex patients’

Whilst deciding the patients to be discussed at the GP practice based MDT meeting, multiple practices have commented the pre-MDT patient selection process highlighted elements of patient care to be addressed immediately, for example, flu jabs and blood tests.

Feedback from staff is that the meetings are a “very useful discussion forum to address more complex patients” and are of great benefit to patient care, as it allowed the MDT to “holistically look at the patients” but with the attendance of various specialists to provide more “acute care solutions and otherwise unknown available services”.

On occasions MDT meetings have had patient and carer representation and it has been suggested patient attendance could benefit the overall discussion, though this attendance is to be decided on an individual practice level. Where patients and carers have not been present, MDT feedback has indicated the patients views were sought and taken into account during discussions.

Each meeting is approximately two hours, with the majority of the assembled multidisciplinary teams using this time to discuss five to 10 patients in total.

Next steps

In some localities practices have been using the tool to also look at patients grouped into the higher risk bands to investigate what further care can be provided to the patient and also in some cases the carers. Future work will be to incorporate the risk stratification tool into the end of life programme.

NHS Leeds is seeking to incorporate Yorkshire Ambulance Service data into the risk stratification data input to further increase the robustness of the data output. At present, work is being carried out using the risk stratification tool to further assess those who are frequent ambulance callers.

Future work requires the development and implementation of a predictor for future social care usage as the current adjusted clinical groups tool does not include specific social care algorithms.

The risk stratification approach will integrate with the macro commissioning of secondary care contracts. A significant goal of risk stratification is to take advantage of the risk stratification tool data outputs in order to support and benefit future commissioning decisions. The tool has the potential to assess what resources are currently being used to support people and will be used to aggregate resource consumption at any level in the health system, including GP practices and CCGs.

Resource allocation can be made on the basis of actual need based on the data outputs, built up from patient level. For this work, data analysis will include impact ability data, population profile at practice level, including an individual’s use of commissioning resources both currently and predicted usage over the next year. 

Comparative data between practices can be made available using the data outputs and this will support opportunities such as working between practices to manage patients with specific conditions. At the CCG level, we have the potential to carry out the identification of communities, localities and primary care practices that are subject to the most demands given the disease prevalence of their populations, improving targeting of resources.

As the NHS Leeds risk stratification tool is still in the early stages of its usage in practice it not possible to define fully the benefits the tool will have overall across health and social care in Leeds – though we have made a positive start.

James Hoult is project manager, and Hailey Matheson is information and general practice facilitator, risk stratification, at NHS Airedale, Bradford and Leeds integrated health and social care programme.