Évaluation quantitative de l'effet sur le passage aux urgences et les hospitalisations d'un dispositif départemental innovant de prise en charge des situations de crise durant l'adolescence en Haute-Garonne, France.

Les situations de crise sont fréquentes à l'adolescence. Le Dispositif Départemental Réactif pour Adolescents (DDRA31) a été implanté en 2017, en Haute-Garonne, pour offrir rapidement, aux adolescents en crise, une consultation spécialisée de psychiatrie avec un accompagnement de trois mois. Notre objectif était de mesurer l'effet de l'implantation de ce dispositif sur l'évolution de l'incidence annuelle des consultations aux urgences et hospitalisations des adolescents pour motif psychiatrique au CHU de Toulouse, guichet hospitalier unique de leur prise en charge en Haute-Garonne. Nous avons conduit une étude quasi expérimentale mesurant l'évolution avant-après l'implantation du DDRA31, l'incidence annuelle des passages aux urgences (pédiatriques et adultes) et d'hospitalisation pour motif psychiatrique au CHU de Toulouse parmi la file active annuelle de tous les adolescents âgés de 12 à 17 ans révolus, résidant en Haute-Garonne, de 2014 à 2019. Les facteurs associés au risque d'hospitalisation ont été analysés par régression logistique à effet mixte pour mesurer l'effet du dispositif au cours du temps, avec l'année comme variable explicative principale, ajusté sur le sexe et l'âge. De 2014 à 2019, 6686 passages aux urgences, réalisés par 4245 adolescents ont été observés. Le taux annuel d'incidence de passage aux urgences pour motif psychiatrique était stable, passant de 14,0 (IC 95 % : 13,2–14,9) pour 100 adolescents en 2014, à 13,9 (IC 95 % : 13,2–14,7) pour 100 adolescents en 2019. L'incidence annuelle des hospitalisations a diminué significativement de 38,3 (IC 95 % : 34,1–42,8) pour 100 adolescents en 2014 à 24,2 (IC 95 % : 21,2–27,4) pour 100 adolescents en 2019. La modélisation du risque d'hospitalisation montrait une réduction significative à partir de 2017, avec un rapport de cotes passant de 0,63 (IC 95 % : 0,49–0,80) en 2017 à 0,52 (IC 95 % : 0,41–0,67) en 2019. Nous n'avons pas montré d'effet du dispositif sur l'incidence annuelle du passage aux urgences pour motif psychiatrique mais un effet possible sur la réduction du risque d'hospitalisation en psychiatrie, suite à un passage aux urgences. La mise en place de structures ambulatoires réactives pourrait diminuer la sollicitation de services d'hospitalisation sous tension. Crisis situations are common during adolescence. Current data show that adolescents in crisis are either directly referred to hospital psychiatric emergency rooms, or they make an appointment for a specialized consultation, in a medical-psychological center (CMP), in a medical-psychological-pedagogical center (CMPP), or with a private practitioner. The Departmental Reactive System for Adolescents was implemented in 2017 in the French department of Haute-Garonne to provide adolescents in crisis an immediate specialized psychiatric consultation including on-going support for three months. Our objective was to measure the effect of the implementation of this health system organization (HSO) on the evolution of the annual incidence of emergency room visits and hospitalizations of adolescents for psychiatric reasons at the Toulouse University Hospital, the sole facility for such care in Haute-Garonne. We conducted a quasi-experimental study comparing the evolution, before and after the implementation of the DDRA31, the annual incidence rates of emergency room visits (pediatric and adult) and hospitalizations related to mental health at the Toulouse University Hospital for adolescents aged 12 to 17 years of age, residing in Haute-Garonne, from 2014 to 2019. Factors associated with the risk of hospitalization were analyzed using mixed-effects logistic regression to measure the effect of the system, over time, with the specific year as the primary explanatory variable, adjusted for sex and age. From 2014 to 2019, 6 686 emergency room visits by 4 245 adolescents took placefor psychiatric motives or diagnostics. The annual incidence rate of emergency department visits related to mental health was stable from 14.0 (95 % CI: 13.2–14.9), per 100 adolescents in 2014 to 13.9 (95 % CI: 13.2–14.7) in 2019, with a higher incidence rate for girls and adolescents aged from 15 to 17 years old. The main reasons for emergency room visits were behavioral problems or agitation, drug intoxications or suicide attempts, and drug or alcohol use or abuse. The annual incidence of hospitalizations decreased significantly since 2017, from 38.3 (95 % CI: 34.1–42.8) per 100 adolescents in 2014 to 24.2 (95 % CI: 21.2–27.4) per 100 adolescents in 2019. This decrease was shown for both males and females but only for adolescents aged from 15 to 17 years old, from 21.8 (CI 95 %: 17.9–26.1) in 2014 to 2.6 (CI 95 %: 1.5–4.1) in 2019. Modeling of the hospitalization risk showed a significant reduction from 2017 onward inclusive with an odds ratio ranging from 0.63 (95 % CI: 0.49–0.80) in 2017 to 0.52 (95 % CI: 0.41–0.67) in 2019. The fact that an impact of the HSO was found only for adolescents aged 15 to 17 years of age could indicate a selective effectiveness depending on age possibly caused by differences in physician attitudes depending on the age of the patient. The chronology of changes in hospitalization rates after a visit to the emergency room corresponded to the period when the HSO was implemented, whereas hospitalization rates were stable over the three years prior to its implementation. The main limitations of our study were the absence of data regarding emergency department diagnoses and discharge modes for the year 2019 and the lack of a temporal perspective. Our study did not show any significant effect of the HSO on the annual incidence of emergency room visits related to mental health, but there was a possible effect on reducing the yearly incidence rate and the risk of hospitalization of adolescents in psychiatric services following an emergency room visit. The implementation of reactive ambulatory structures could therefore reduce the demand on hospitalization services that are under pressure. [ABSTRACT FROM AUTHOR]