This study aims to develop two statistical models to estimate the propensity to hospitalize patients arriving at the ED with dyspnea or after transient loss of consciousness (TLoC). IT solutions developed in the eCREAM project will be used to obtain high-quality structured data from ED EHRs. A secondary objective is to study the association between the adjusted hospitalization rate and 30-day mortality. It is an observational, multicenter, retrospective study that will last 24 months. 30 EDs from Italy, Poland, Slovakia, Slovenia, Switzerland, and the UK will participate. Data on all patients who attended the ED between 2021 and 2023 will be collected. Three types of information will be collected: 1) structural organizational data, 2) transient organizational data, and 3) data on patient characteristics upon arrival at the ED. Type 1 data will be collected at the individual center level, type 2 data will be collected on all patients who attended the ED, type 3 data will be collected only on patients with dyspnea or TLoC. The latter two will be obtained automatically from the EHRs of the participating EDs.