This website is intended for healthcare professionals only
User log in

Trial log in
November 2021

Predictive software has unexpected results

Software designed to reduce emergency hospital admissions for vulnerable patients may, perversely, increase the numbers of patients attending hospitals, a study in the British Journal of General Practice has found.1

General practice clinicians and practice managers who used the Predictive Risk Stratification Model (PRISM) software to target care to vulnerable patients said it raised their awareness of these patients, and they had more contact with them, and made more ‘small-scale’ changes to their care.

But evaluation of the impact of the introduction of the software in 32 general practices in the Swansea area found that emergency admissions to hospital, emergency department attendances and days spent in hospital increased when the software was implemented – the opposite of the intended effect.

UK policy has been to support initiatives to reduce emergency admissions, and the PRISM software, which predicts each patient’s risk, has been widely implemented – without knowing whether it brings benefits to patients or the NHS.

Most participants in this qualitative analysis said that the increased interaction with high-risk patients was probably beneficial, regardless of any treatment delivered, but doubted that it would have any large-scale effect on admissions. However, the full results of the PRISMATIC trial showed statistically significant increases in emergency admissions and use of other NHS services, and increased NHS costs by £7.59 million per 100,000 population a year.2

As a result, implementation of PRISM has been halted in Wales, although it has continued to be rolled out in other parts of the UK.

1. Evans AB, et al. Br J Gen Practice 2021;

2. Snooks H, et al. BMJ Quality & Safety 2019;28:697-705

Practice Nurse 2021;51(9):10