Implementation of predictive risk stratification to reduce emergency admissions to hospital: experiences of general practitioners and practice managers

Abstract Aim In a trial evaluating the introduction of a predictive risk stratification model (PRISM) into primary care, we reported statistically significant increases in emergency hospital admissions and use of other NHS services without evidence of benefits to patients or the NHS. The aim of this study was to explore the views and experiences of general practitioners (GPs) and practice managers on incorporating PRISM into routine practice. Methods We interviewed 22 GPs and practice managers in 18 participating practices at two timepoints: 3-6 months after PRISM was available in their practice; and at study end, up to 18 months later. We recorded and transcribed interviews and analysed data thematically using Normalisation Process Theory. Results Respondents reported that the decision to use PRISM was based mainly on fulfilling reporting requirements for Quality and Outcome Framework (QOF) incentives. Most applied it to a very small number of patients for a short period. Using PRISM entailed technical tasks, information sharing within practice meetings and changes to patient care. These were diverse and generally small scale. Use was inhibited by PRISM not being integrated with practice systems. Respondents’ evaluation of PRISM was mixed: most doubted it had any large scale impact, but many cited examples of impact on individual patient care. They reported increased awareness of patients in high risk groups. Conclusions Qualitative results suggest mixed views of predictive risk stratification in primary care and raised awareness of highest-risk patient groups, potentially affecting unplanned hospital attendance and admissions. To inform future policy, decision-makers need more information about implementation and effects of emergency admissions risk stratification tools in primary and community settings. Trial registration Controlled Clinical Trials no. ISRCTN55538212 .