Elderly adults have high rates of emergency department (ED) visits. Specificities of this population challenge organizations of care in the ED, and older adults are at risk of pejorative outcomes after an ED stay. Numerous interventions have been designed to improve quality of care and outcomes for the older population in these settings, with a specific attention to concerning discharge from the ED. These interventions are interdisciplinary, bridging emergency and geriatric care. The wide range and complexity of these interventions make them difficult to assess and compare, as highlighted by several reviews in the past ten years. Prior analyses helped to categorize different intervention strategies and three main designs: inhospital, community and transitional interventions started in the ED and pursued in collaboration with community primary care professionals . Theses analyses show that the use of multiple strategies and transitional models of care tend to lead to better outcomes, and underline that more robust studies are needed to confirm this hypothesis. In France a majority of EDs collaborate with Geriatric Mobile Teams (GMT) to improve quality of care for older patients. GMTs are dedicated to patients over 75 years old, and interventions in EDs are targeted on patients at risk of worse outcome. When ED physicians detect older patients at risk they may call for the GMT for further assessment and management. GMTs either work in a inhospital standard approach or with a transitional care management. This second strategy, less common in France, is thought to be be efficient and has never been assessed. We have designed a study to compare these methods, with the hypothesis that among at-risk older adults, hospital-community transition care initiated by GMTs during an ED visit with direct discharge home will be associated with a reduction in the risk of early readmission within 30 days, and lower risk of loss of independence at 3 and 6 months. It is a french multicentric study, with a quasi-experimental design, comparing hospitals without transitional care management to hospitals with hospital-community transitional intervention. We aim at enrolling 1322 patients aged 75 and more at risk of pejorative outcomes as determined by the Triage Risk Screening Toll (TRST). The main outcome is a revisit to the ED between day 7 and day 30, secondary outcomes are autonomy, mortality, use of hospital services and caregiving at home at 6 months.

The aim of this study is to evaluate the impact of transitional care initiated by Mobile Geriatrics Team (MGTs) at the time of the ED visit on the incidence of early readmissions, on medical journey and autonomy within 6 months after the ED visit, for adults aged 75 or older and considered at risk for early readmission. This is a multicenter, prospective, controlled, non-randomized quasi-experimental study. Inclusion criteria are age 75 years and more, returning home after an ED visit, with a Triage Risk Screening Tool (TRST) score ≥ 2. Exclusion criteria are living in nursing home, being under legal guardianship or incapacity to consent. All participants will provide a written consent after receiving oral and written information. Participants with language barrier, severe cognitive or psychiatric disorders may be included provided that a relative is physically present at the time of the ED visit. Otherwise they will not be included. The study protocol has been submitted for approval to the ethic committee. Five centers (two university hospitals, three regional hospitals) work on an in-hospital strategy and do not perform transitional care. After identifying the patient at risk, the in-hospital MGT solely provides usual recommendations at the time of the ED visit, without offering home-based intervention or coordination with community actors. The participants included in these centers will form the control group.

Seven centers (three university hospitals, four regional hospitals) work on an transitional care method. These centers have an in and out-hospital MGT that carry out geriatric assessment, offer home-visits if necessary, and have a hospital-community coordination for the management of older patients. This coordination is defined by either regular joint clinical meetings, joint visits, shared professionals, or a shared information system. After identification of the at-risk patient, the first step in management consists in a telephone call by the GMT to the patient or his family and primary care professionals within six days of his return home. Depending on the geriatric assessment performed and the coordination organization specific to each center, multidisciplinary interventions can then be proposed and deployed. The patients included in these centers will form the intervention group.

All 12 centers (intervention and control) meet the following criteria for a standardized in-hospital management during the ED visit, including (i) identification of patients at risk based on clinical characteristics or identification scores with a procedure for reporting to the in-hospital GMT; (ii) a multidisciplinary GMT mobilized in the ED for patients at risk providing a standardized geriatric assessment; (iii) a discharge procedure with at least a medical report and referral to the general practitioner.

The principal investigator and the scientific committee will regularly verify that centers continue to meet the conditions of the group it is allocated.

After inclusion, the participants will benefit from the usual care specific to the inclusion center where they were recruited. In the intervention group, GMT intervention initiation consists in an early, standardized telephone call between D1 and D6 for each participant. To respect a pragmatic approach, the intervention of the MGT and the out-of-hospital management strategy are deliberately not standardized in order to evaluate the hospital-community coordination fitting each territory .

Clinical data will be collected at baseline and during follow-up at three (M3) and six (M6) months (Figure 1). The data collected at baseline during the ED visit (D1) by the MGT. This evaluation will be based on the patients' questioning and the medical record. If deemed necessary, the GMT may also rely on information transmitted by a relative or the general practitioner in order to cross-check the information, as is done routinely. The data relative to socio demographic items are age, sex; autonomy in the week prior to the ED visit using the Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) scales (18,19), self-reported difficulty related to housing, identification of relatives, family and professional caregivers preexisting the ED visit; the nutritional risk using the Mini Nutritional Assessment - Short form (MNA-SF) scale in its alternative version with measurement of calf circumference (20); the screening for cognitive disorders with the Abbreviated Mental Test 4 (AMT4) scale (21) and suggestive clinical diagnosis in the mild, moderate or severe stage, according to the 5th edition of the Diagnostic and Statistical manual of Mental disorders (DSM-V) criteria; the comorbidity level assessed the Charlson index (22) and the number of drugs in the patient's usual treatment; data on the ED visit that including duration the visit in hours and main diagnosis collected from medical records; the history of ED visits or hospitalizations in the previous month and the existence of a regular medical monitoring by the general practitioner (twice a year) or a geriatric team (previously hospitalized in a geriatric wards or appointment with a geriatrician).

One independent investigator will perform follow-up telephone interviews in both groups at M3 and M6 to precise the vital status, place of living (home, nursing home), ADL and IADL, interventions of professional and non-professional caregivers. In addition for participants in the intervention group the independent investigator will identify at M6 the MGT's actions after discharge from ED based on the MGT's medical record including the number and duration of telephone calls and home visits, the coordination with community professionals (medical appointment, rehabilitation care, request for social worker).

The primary endpoint will be the incidence of readmission to an ED between seven (D7) to thirty (D30) days after discharge from the ED visit leading to inclusion in the study, regardless of the reason for admission. This data will be collected from the National Health Data System (NHDS), thus allowing the identification of admissions to an ED on a national level and not exclusively in the centers participating in the research. Very early readmissions (from D1 to D6) will not be taken into account in the primary endpoint, as the time to first intervention in interventional group is a maximum of six days in the framework of the research protocol. Moreover, these very early readmissions are most often motivated by a rapid deterioration of the initial pathology, a diagnostic error or a possible side effect of a therapy (23). Hospital-community transitional care is not intended to avoid this type of very early readmission. The ED visits between D1 and D6 and after D30 will be collected in NHDS as secondary endpoints. The number of ED visits, unplanned hospitalizations and the time duration between inclusion and the first outcome will be also considered.

Inclusion Criteria:

• Patient admitted to the emergency departement over 75 years old and living at home (including independent residence)
• Patient admitted to the Emergency Reception Service (ERS) for less than 48 hours for whom a return home has been decided
• Identified at risk of readmissions to the emergency departement with a Triage Risk Screening Tool (TRST) score > 2.
• Consent to the study possible at the time of his visit to the emergency room by the patient or a caregiver present at the time of inclusion.
• Patient affiliated with a social security (beneficiary or partner)

Exclusion Criteria:

• Person living in an nursing home
• Severe cognitive impairment according to DSM V criteria and absence of a close relative at the time of inclusion
• Unstabilized psychiatric pathology and absence of relatives at the time of inclusion
• Language barrier and absence of relatives at the time of inclusion
• Person under guardianship, under legal safeguard measure, deprived of liberty by judicial or administrative decision, persons subject to psychiatric care without their consent, persons admitted to a health or social establishment for purposes other than those of the research
• Patient under state medical assistance
• Patient already included in the research