Software to reduce emergency hospital admissions gets mixed report from family doctors

A software tool meant to help reduce emergency hospital admissions is of limited use and benefit to patients, a new study of family doctors' views and experiences has revealed. 

The study builds on earlier research by the same team showing that emergency admissions had gone up—not down—when the tool was introduced in Wales (UK), which led to its rollout being halted, though it continues to be used widely in England, Scotland and Northern Ireland. 

Researchers say the new study reinforces the need for more evidence and research on the implementation of the software and its effects.  

The software is a risk prediction tool, which in Wales is called PRISM. It identifies people who are most at risk of needing emergency care, based on past use of healthcare, diagnoses and medications. The thinking is that targeted management of these patients can reduce emergency admissions to hospital, improve patient outcomes and experience, and provide better value for money.  

However, researchers at Swansea University Medical School, who evaluated the use of PRISM in Wales, found that there is little evidence to suggest that it meets these objectives.

Now, to shed more light on how the tool is used in practice, the same team have published a second study, exploring the views and experiences of GPs (geenral practitioners) and practice managers who used PRISM. 

They interviewed 22 GPs (general practitioners) and practice managers in 18 practices in south Wales, between three and six months after they began using PRISM and then again 18 months later.

They found: 

• GPs generally judged it unlikely that PRISM had any effect on emergency admissions, with a widespread feeling that admissions initiated by GPs were already low with little scope for further reductions 
• Respondents reported the decision to use PRISM was based mainly on wanting to secure incentives offered by the Welsh Government under its Quality and Outcome Framework for improving care 
• Use of PRISM was inhibited by it not being integrated with practice systems  
• Most doubted any large-scale impact from PRISM, but cited examples of impact on individual patient care  
• The majority of respondents reported that PRISM had made them more aware of high-risk patients, flagging up some patients who had not previously been considered to be in the high-risk category. 

The researchers' overall conclusions were that views of PRISM in general practice were mixed, and that policymakers need more information about how these tools are used in practice, and the effects of these tools on decision-making as well as patient outcomes. 

Professor Helen Snooks from the project team at Swansea University Medical School said that "tools like PRISM are used widely by the NHS in primary and community care, with the aim of reducing emergency hospital admissions. However, there is a lack of evidence to support the view that they enable proactive care and improve patient outcomes." 

"Our research highlighted very mixed views and experiences among GPs and practice managers about use of PRISM. This was often short-term and driven by external factors rather than embedded in new ways of working."

"Decision-makers need more information about the implementation and effects of such tools in primary and community settings to inform future policy on their use."

"Given the current context of rising emergency admissions, and Department of Health incentives in England to use these risk tools in community services, our findings are important and timely." 

The study was published in the British Journal of General Practice.