Similar (re-)hospitalization patterns were identified and their sociodemographic and clinical characteristics were examined. Five distinct clusters emerged from the analysis: A: Brief hospitalization; B: Repetitive short hospitalizations; C: Repetitive medium hospitalizations; D: Long hospitalizations; E: Chronic hospitalizations. Although the duration and frequency of hospitalizations varied across clusters, the age of patients did not vary significantly among the clusters. Moreover, there were noticeable differences in terms of gender, diagnoses, and the extent of symptoms severity and psychosocial challenges faced by the patients.

Patterns of (re-)hospitalization and service organization

The identified clusters in this study are largely consistent with those from previous research examining similar patterns or trajectories in adults psychiatry [21, 31, 32]. Indeed, the ratio between the number of patients and the hospitalization days (a small number of patients necessitate a large number of hospitalization is similar to what was observed in adult psychiatry, where this pattern were observed to be even more pronounced in adult psychiatry where roughly 5% of patients consume a total of 30% of the resources) [21]. Thus, the most severe and chronic cases which represent few adolescents need the most hospitalization days (or resources).

This may be understood in the light of the “stepped care” approach [33,34,35]. It tailors treatment responses to the patient's clinical stage and needs [36, 37] and may guide the organization of the hospitalization. In that sense, the efficacy of repetitive short stays is often debated, especially when considering the potential adverse effects of hospitalizations [15]. If CAMHS organizations were structured differently, there could be a bolstering of intermediate or outpatient structures (e.g., intensive community treatment, short-stay day care centers, and specialized short-term inpatient crisis units) to prevent such re-hospitalizations. Such structures aim to decrease the need for hospitalizations for adolescents while supporting continuity of care and offering them the most appropriate treatment to sustain their ongoing development. According to this approach, only patients with the most severe difficulties should be hospitalized. Thus, it emphasizes the relevant or “smarter” utilization or “smarter medicine” [38] of inpatient facilities, underscoring the importance of comprehending trends.

Sociodemographic variables related to the different patterns of (re-)hospitalization

In the analyses of the five clusters, results revealed no age-based differences, a finding consistent with a study on the adult population using analog procedure [21]. More generally, a recent systematic review of the literature [39] yielded inconsistent results. Some studies within this review [39] associated a higher risk of re-hospitalization with younger age, while others found the opposite trend. In another study with a mixed-age population (ranging from 15 to 64 y.o [40]), younger age was found to be predictive of an increased risk of the repetitive medium hospitalization phenomenon.

In the current study, females were overrepresented in the typical user clusters and underrepresented the atypical user. This is somewhat aligned with findings from a systematic review [39], where a majority of studies found an association between being female and an increased risk of re-hospitalization. As for clusters D (long hosp.) and E (chronic hosp.), the underrepresentation of females might be partially due to a significant proportion of patients in these clusters being diagnosed with schizophrenia spectrum disorders. This diagnosis is more prevalent in males than in females, with the difference being particularly marked up to the age of 25 [41,42,43]. Moreover, it should be noted that clusters D (long hosp.) and E (chronic hosp.) represent only 14 patients, which could influence these findings. In addition, this discrepancy can be partly attributed to the overrepresentation of females in the sample.

Distinct clinical profiles of hospitalized patients

To summarize, results revealed three main profiles of hospitalized patients. The first profile, represented by cluster A (brief hospitalization with an average duration of 7.71 days), comprised patients exhibiting acute symptomatology and with mild impairment in psychosocial functioning. For these individuals, a brief hospitalization, typically singular, serves to address the immediate crisis and possibly alleviate the external factors contributing to it. This is supported by the fact that re-hospitalization rarely re-occurred within a year. These individuals, who show clinical benefits from a single hospitalization, are best treated during their acute phases in short-stay hospitalization units which is consistent with the hospitalization’s goal to alleviate symptoms [8, 10] and promote functional level [11,12,13].

The second profile encompasses patients from clusters B (repetitive short hospitalizations with an average duration across hospitalizations of 19.90 days) and C (repetitive medium hospitalizations with an average of 41.33 days). These patients are marked by their propensity for re-hospitalization, which differs in frequency and length, with the repetitive medium hospitalizations group experiencing more prolonged and frequent stays. Predominantly, this group was diagnosed with mood disorders and displayed subpar psychosocial functioning. These patients together accounted for 70% of the patients and 70% of the hospitalization days. Given the high re-hospitalization rate, the authors assumed that a different management strategy needs to be developed for these patients [44]. Indeed, while traditional hospitalization proved beneficial, these patients consumed a significant portion of the available hospital days.

The third profile, represented by clusters D (long hospitalizations with an average duration of 99.36 days) and E (chronic hospitalizations with an average of 138.67 days), consisted of patients with the most severe diagnoses and the most compromised psychosocial functioning which is consistent with previous findings [45, 46]. Although this group generally benefited from traditional hospitalization, an extended duration was necessary to achieve symptom remission that aligned with outpatient treatment standards. Yet, prolonged hospital stays can introduce adverse consequences such as stigmatization and de-socialization [15]. Given these considerations, re-evaluating the efficacy and suitability of extended hospitalizations, particularly for clusters D and E, becomes paramount. While a hospital environment is undeniably beneficial during acute phases, longer-term treatment might find greater success in social-health facilities aimed at fostering reintegration into the patient's natural social setting. Adopting this approach could also free up critical resources within acute inpatient units for more urgent cases.

Limitations and future perspectives

This study has several limitations. Firstly, owing to its retrospective design based on clinical data, it offers limited insights into the clinical condition of the patients. While future studies might benefit from a prospective design, the essential attributes (due to retrospective study of data collected in routine clinical care) for clinical characterization of the clinical and socio-demographic variables, such as age, sex, and diagnoses, were the only available information in this research. Additionally, the analysis was confined to a 1-year period. A more extended study might yield additional insights, such as patterns observable over a longer timeframe. It is also worth noting that the data might have been influenced by the impact of the sanitary crisis on youth mental health. Moreover, the authors did not have access to the type of familial, social and/or community support that the youths may benefit after the hospitalization. This may impact the youths’ ability to adhere to treatment recommendations upon discharge and ultimately their ability to access the outpatient services and resources. In future research endeavors, adopting a more comprehensive approach that encompasses various treatment modalities, such as outpatient care, daycare, and assertive community treatment, could be beneficial (see for instance [47]).