Emergency Admission Predictive Risk Stratification Models: Assessment of Implementation Consequences (PRISMATIC 2): A protocol for a mixed-methods study

Background: Emergency admissions are costly, increasingly numerous, and associated with adverse patient outcomes. Policy responses have included the widespread introduction of emergency admission risk stratification (EARS) tools in primary care. These tools generate scores that predict patients’ risk of emergency hospital admission and can be used to support targeted approaches to improve care and reduce admissions. However, the impact of EARS is poorly understood and there may be unintended consequences. Aim: To assess effects, mechanisms, costs, and patient and healthcare professionals’ views related to the introduction of EARS tools in England. Design & setting: Quasi-experimental mixed-methods design using anonymised routine data and qualitative methods. Method: We will apply multiple interrupted time-series analysis to data, aggregated at former clinical commissioning group (CCG) level, to look at changes in emergency admission and other healthcare use following EARS introduction across England. We will investigate GP decision making at practice level using linked general practice and secondary care data to compare case-mix, demographics, indicators of condition severity, and frailty associated with emergency admissions before and after EARS introduction. We will undertake interviews (approximately 48) with GPs and healthcare staff to understand how patient care may have changed. We will conduct focus groups (n = 2) and interviews (approximately 16) with patients to explore how they perceive that communication of individual risk scores might affect their experiences and health-seeking behaviours. Conclusion: Findings will provide policymakers, healthcare professionals, and patients, with a better understanding of the effects, costs, and stakeholder perspectives related to the introduction of EARS tools.