This nationwide register-based cohort follow-up aimed to investigate the psychiatric treatment patterns and treatment outcomes over a five-year period for adolescents who utilized psychiatric services in Finland between 2003 and 2013. Align with previous research [6,7,8,9], we observed a notable shift in treatment patterns, particularly towards outpatient care, which seemed to occur at least partially independently of patient characteristics. Furthermore, there was a significant increase in the proportion of adolescents receiving psychotropic medications, particularly antipsychotics and stimulants.

The cumulative medication exposure to antipsychotics was lower, suggesting that this increase was primarily attributed to the off-label usage of antipsychotics. This aligns with previous studies indicating that antipsychotics are being utilized more frequently for the treatment of other problems instead of psychosis [15,16,17]. While there was a slight decrease in the proportion of adolescents receiving benzodiazepines, the change was modest, and overall, we observed a significant increase in the usage of psychotropic medications. This may indicate a more systematic provision of psychotropics in alignment with current treatment guidelines. For example, it is observed that attention deficit hyperactivity disorders are more frequently diagnosed in recent years, reflecting changes in administrative and clinical practices and likely explaining the increasing rate of stimulant use [18]. In addition to a potentially different threshold for medication, it is possible that the greater overall utilization of psychotropics could also be driven by limited availability of other treatments and services, including nationwide reduction of hospital beds.

The five-year mortality and suicide ratio was significantly lower in adolescents who sought treatment after 2008. The usage of psychotropic medications, particularly antidepressants, has been suggested as a potential explanation for the decline in mortality among individuals with mental health problems [19], and our analysis revealed also a linear increase in the utilization of psychotropics alongside a decrease in mortality ratio. However, in post-hoc analysis adolescents who sought treatment before 2009 and subsequently died were actually more likely to have received antidepressants and benzodiazepines prior to their death, with no other discernible differences in medication treatment patterns. This aligns with recent observations [20, 21], which indicate that the observed declining trend, particularly in suicide rates, is more likely influenced by other time-related factors rather than being a direct consequence of the increased utilization of psychotropic treatments. It is also noteworthy that mortality did not appear to be in a linear relationship with the decrease in hospital care. Although conclusions about causality cannot be drawn from this study, this result may suggest that psychiatric services can be safely developed with a focus on outpatient care without an increase in mortality-ratio.

Adolescents who sought treatment after 2008 were more likely to still be receiving treatment after the five-year follow-up period. Furthermore, there was a significant increase in the annual prevalence of adolescent mental health service users during the inclusion years of the study, indicating that adolescents who sought treatment in later years spent a longer time in services compared to those in previous years. There were no statistically significant differences in the ratio of disability allowances at the end of the follow-up. These findings are particularly noteworthy considering the indications that a higher proportion of adolescents sought psychiatric care, and the proportion of severe psychiatric disorders, such as psychoses and substance abuse disorders, was slightly lower in the latter cohort. Moreover, these results persisted even after adjusting for observable baseline characteristics.

While the observational nature of this study prevents us from drawing causal explanations, there are several potential explanations for this finding. Firstly, the higher usage rates of psychotropic medications, which are often recommended for long-term use, may contribute to the longer treatment periods. Then again, longer usage of services may also reflect systematic efforts to help individuals who are experiencing mental health difficulties, following from the reform of services. It is suggested that the stigma surrounding mental health problems has diminished, leading to better adherence of psychiatric treatment and more individuals seeking help from mental health services [22]. Social network members, including professionals from schools and other services, may also guide adolescents with lower thresholds to psychiatric services.

While increasing awareness and willingness to receive mental health help are positive signs, they also pose a risk of medicalizing mental and social phenomena, leading to an overreliance on medical responses to human life problems. To minimize this, more holistic treatment strategies are suggested [23, 24]. Examples of such strategies, which have preliminary shown reduced cumulative service usage and psychotropic treatments alongside decreased long-term disability ratios, incorporating more contextual and social network-focused approaches [23,24,25,26,27]. Future studies should explore whether these approaches can be more broadly implemented to cost-effectively enhance treatment outcomes for adolescents with mental health difficulties.

Finnish registers are considered as a reliable source of information [28], allowing for the non-selective inclusion of all individuals receiving adolescent psychiatric treatment in Finland. However, registers were not originally designed for research purposes, potentially introducing inaccuracies. Limitations included the lack of information on outpatient primary healthcare prior to 2011 and psychotherapy conducted in the private healthcare sector. Sensitivity analysis focusing on patients with follow-up after 2010 suggested that the missing information on primary-level outpatient care did not bias the main findings. The absence of information on private psychotherapy was also unlikely to impact the main conclusions, given the significant increase in this practice after the 2010s [29].

Since registers were the sole source of information, standardized measures were lacking to estimate the primary outcome. To compensate, strict outcome measures such as survival and non-usage of any services or mental health support at the end of the fixed five-year follow-up were used as proxies for symptomatic recovery, considering that in Finland the healthcare and social services are guaranteed to entire population based on national social insurance. In other words, it is unlikely that there would not be any registered entries of treatments or support in the Finnish system over the long term if the individual’s symptomatology remained disabling. However, this measure does not directly capture symptoms or more existential outcomes, such as subjective experiences of well-being. Disability allowances at follow-up may also be influenced by varying thresholds for receiving such support, which are affected more by policy and financial constraints than individual condition.

Detailed information on symptom severity at onset and time-related factors influencing for example the content of treatment and diagnostic procedures was lacking. Although our approach accounted for evolving and comorbid baseline diagnostic distributions, the potential heterogeneity of diagnostic practices and the possibility of providing treatment without a formal diagnosis necessitate further analysis of different diagnostic combinations and trends during follow-up. Note also that the higher incidence of adolescent patients in the latter cohort suggests an increase in adolescents with less severe symptoms seeking services, which could have particularly impacted outcomes. Due to the limited availability of initial clinical information, the weighting does not fully address this potential confounder.

Due to the aforementioned limitations, the results should be considered as describing rough national trends and changes in adolescent psychiatric treatments, and their associations with long-term outcomes. Note also that the primary objective of this study was not to assess the outcomes of the national mental health plan. Instead, the selection of the 2009 cut-off year was chosen mainly as a historical reference to enable comparison before and after identifiable shifts in mental health regulations and adolescent services. The findings suggest success in this regard, as no extreme weights were observed, and there were noticeable differences in treatment practices regardless of patient characteristics. However, numerous residual factors likely associate with both service intake and treatments, apart from the publication of the national mental health plan. Therefore, a linear causal relationship between the publication of the national plan and treatment practices and outcomes cannot be established.