In our study we investigated whether the introduction of youth mental health practice nurses in general practice was associated with outpatient mental health services utilization and costs for outpatient healthcare in one to four years after implementation. Overall, outpatient health care utilization and associated costs increased steadily between 2019 and 2022. After adjustment for relevant confounders, including the COVID-19 pandemic, the presence of a YMHPN in a practice was associated with small, non-significant reductions of outpatient health care utilization and associated costs. Continued evaluation of the introduction of the YMHPN in general practice is needed to confirm the current findings and to assess whether longer-term effects differ from the shorter-term effects.

Worldwide utilization of mental health services has increased substantially among children and adolescents over the past decades [1, 22, 23]. Of the several types of care (e.g., inpatient, school-counselling), the increased utilization of outpatient mental health care is apparent [24]. In our study, we also saw a slow but steady increase in outpatient mental health care utilization in children and adolescents. Furthermore, we noticed a positive correlation between age and the percentage of children using outpatient mental health. Additionally, boys were more likely to receive care than girls. In contrast with our findings, previous research showed inconsistent associations between mental health service use and sex and age [25]. Although adolescent girls tend to experience more mental problems than adolescent boys [26], we noticed that before the age of 15 health utilization in boys is higher than in girls suggesting unmet needs of girls in this age category. In our study on average 5.5% of children attended outpatient mental health services during the study period. Although this seems relatively low considering the high prevalence of child mental health problems, it is comparable to mental care utilization in other high-income countries [27]. It is also in line with regional registry data which show that approximately 7% of children up to 17 years in Rotterdam receive one or more forms of youth care including among others outpatient mental care, help for dyslexia, and developmental problems [28].

Growing up in a socially deprived area has been shown to be an important risk factor for developing mental health problems and is associated with a lower well-being [29]. On the contrary, mental health care utilization is assumed to be lower among children and adolescents living in a socially deprived area than among children living in less deprived areas [30]. This is supported by our findings, where living in a deprived area was associated with less outpatient mental health service use and lower costs. This is worrisome and suggests that although children in socially deprived areas may suffer more from mental health problems, they are less likely to receive care for these issues. One of the goals of integrating mental health services into primary care is to improve accessibility of care for all children regardless of their background [31]. Future research should therefore investigate whether the availability of a YMHPN has differential effects on accessibility of mental health care for children with different socioeconomic, cultural and ethnic backgrounds.

In the past decades, several studies evaluated initiatives to integrate child mental health services into primary care. A meta-analysis of randomized clinical trials showed that integrated mental health care for children and adolescents was associated with improved health outcomes compared with usual care [6]. Additionally, integrated child mental care within primary care may have the potential to reduce disparities for vulnerable children (e.g., children from a minority or social-economically disadvantaged background) [31]. To our knowledge only a limited number of studies evaluated cost effects of such initiatives, using relatively small samples [32, 33]. As far as we are aware, this is the first large-scale study investigating the associations between integrated child mental health care into general practice and the utilization of outpatient mental health care and associated costs. Earlier studies investigating the effects of integrating mental health professionals into general practice in adult populations showed various results, with no clear evidence on whether this form of integrated care changed specialized care utilization and costs [34]. Overall, we found that integrating YMHPNs within general practices did neither lead to short term changes in the number of children receiving outpatient mental health care nor to significant changes in associated costs. It is important to note that we could only investigate overall associations. We were not able to investigate whether the presence of YMHPNs had different effects for specific subgroups (e.g., whether the presence of a YMHPN was associated with specific mental health problems being more or less often treated in outpatient care, or whether it was associated with less unnecessary referrals). Importantly, children registered in a practice with a YMHPN receiving outpatient mental care were not necessarily referred by the YMHPN. They could also be referred directly by their GP or other care providers. Importantly, large variation on both outcomes became apparent between practices which deserves further attention in future research.

YMHPNs can play an important role in lowering barriers to access services. Many barriers (e.g. limited access to mental services and fear of stigmatization) are shown to influence access to mental health services and gaps are still suspected between those who need care and those who receive care [27, 35]. In our study we did not have information on the number of children seen by YMHPNs, and as a result we do not know whether the introduction of the YMHPN led to more children being treated within general practice. However, the monthly rate of children receiving outpatient care did not change following the introduction of YMHPNs. Importantly, an earlier study showed that YMHPNs can successfully manage a substantial part of children without the need for additional referral to specialized care [17].

One of the presumed benefits of integrating child mental health services into primary care is that it improves early detection and treatment [6]. Integrated care may prevent mental health problems from worsening and reduce long-term negative outcomes and associated costs for the affected individual and society [4]. Additionally, integrated care (e.g., YMHPNs) may reduce referrals for non-complex problems and prevent unnecessary medicalization. To determine whether YMHPNs can indeed prevent long-term negative outcomes, studies with longer follow-up periods are necessary. The current study focused on the effects within the first four years of the introduction of the YMHPN. Longer-term effects may differ from the observed shorter-term effects. Therefore, further research is needed to investigate whether the introduction of the YMHPN led to long-term changes in mental health care utilization (e.g., both in absolute rate of mental health services use and on type of service use). Such research should preferably encompass all levels of Dutch mental health care, as well as costs for medication, which will allow stronger and more specific inferences. Besides the prevention of long-term negative outcomes, there might be other effects of introducing the YMHPN into general practice that are not captured in this study and might confound the overall found association, such as an increased interest in children’s mental health by the GP or a lower rejection rate of referrals to specialized mental health care. A study on the long-term effects could be extended with questionnaire data from GP practices, YMHPNs and children and their family members to investigate other effects from the introduction of the YMHPN. This data could also be used to investigate differences between practices and YMHPNs. Moreover, with questionnaire data it is possible to collect information on patient characteristics (e.g., clinical condition) and investigate the effect for different subgroups or certain interactions.

Our study has several limitations. First, it was not possible to investigate on the individual child level whether the introduction of the YMHPN led to changes in the number of children receiving help, because we did not have information on which children were seen by YMHPNs. As such, we were only able to investigate the overall effect of the availability of a YMHPN in a practice on outpatient care utilization. Nevertheless, we believe that this study offers valuable insights for practice and policy as this is the first study to evaluate the effect of the introduction of the YMHPN. Secondly, we had limited information on specific patient characteristics (e.g., no info on ethnic background of children or the clinical condition of the child). Therefore, we could not investigate differences between specific patient populations. Thirdly, we did not have information on costs for general practice care and medication. As such, we only assessed costs for outpatient mental care. We did not take into account the costs for the implementation of the YMHPN. Future research should address the cost-effectiveness of the YMHPN. Fourthly, participation in the YMHPN-project was not random and it is possible that participating practices differed from those that did not participate. For instance, during our study period the number of practices from deprived areas joining the RPCD and the YMHPN-project increased. However, by adjusting for living in a deprived area we minimalized this confounding effect. Fifthly, not all practices provided follow-up information for all months and in some practices that participated in the YMHPN-project only the period with a YMHPN was covered. We addressed the fourth and fifth limitation by conducting two sensitivity analyses in two subsets of practices (i.e., all practices that participated in the YMHPN-project and all practices with > 3 years follow-up). These sensitivity analyses provided comparable results to our main analysis and thus strengthened our results. Lastly, our study period included the COVID-19 pandemic. Although we corrected for the effect of the COVID-19 pandemic in our models, it cannot be ruled out that some remaining confounding effects did affect our results.