Adequacy of an automatic tool to select and stratify chronic complex patients with multiple unplanned visits to the emergency department

Background: Unplanned emergency department visits are an ongoing challenge, since the patients are increasingly ageing, chronic, complex, frail and multimorbid. The aim is to demonstrate the feasibility of establishing a proactive intervention to detect and distribute complex, chronic, unplanned multiple emergency department patients, tailored to the patient's needs. Methods: Prospective study to evaluate the intervention, which is divided into the following drivers: automated pre-selection sent lists, a specific experienced multiple frequent case-manager nurse, and a 3-level territorial derivation consensus protocol. Chronic complex patients presenting two or more times in the last six months to a tertiary hospital emergency department for COPD or HF. Analysis of the percentage of patients well selected by the three drivers, and their corresponding characterisation by the three levels of care to which they are referred: primary, intermediate and specialist care. Number of visits six months before and after referral. Results: : 368 patients were selected; the 85% were well selected. The patients' age, frailty, home care percentage, Barthel index and cognition are progressively consistent with the three levels of care of referral according to previously defined and consensus criteria. The intervention achieves lowering participant visits: 3.1 before (SD = 2.42) versus 1.6 after (p < 0.001). Conclusions: : An automated selection, plus a territorial protocol consensus of circuits and clinical profiles, together with the autonomous assessment of a multiple case-managing nurse, achieves in a proactive and simple manner, a better and adapted distribution of patients in coherent care levels institutions, regarding their socio-health needs.