15581-m-nanninga

CHAPTER 6 110 METHODS Study design We used one-year follow-up data from TakeCare, a large prospective cohort study of children aged 4-18 in the northeast of the Netherlands [8]. TakeCare has been designed to investigate the trajectories and outcomes of children receiving psychosocial care and consists of a care cohort of children enroling in psychosocial care, and a reference cohort of children not in care. Between April 2011 and April 2013, parents/caregivers of children aged 4 to18 years, along with children aged 12 years and over, were invited to participate in TakeCare. The Medical Ethical Committee of the University Medical Center Groningen evaluated the design of the study, and approved it without requiring full assessment. Informed consent was obtained from all participating respondents [8]. Sample and procedure For the care cohort, 2,664 children and their parents/caregivers were recruited via PCH, CASC and CAMH, the main providers of psychosocial care for children and adolescents in the Netherlands. At the time of our study, children entered psychosocial care via either their general practitioner, the youth care office, or PCH [15]. General practitioners and doctors and nurses in PCH provide light psychosocial support to children and their families. In case of more severe problems they refer children to specialized care, either to CASC, primarily staffed by child (social) workers, or to CAMH, primarily staffed by child psychologists and psychiatrists [3]. See Table 1 for a detailed description of the care types. Children with insufficient understanding of Dutch, living outside the northern region, or following special education because of intellectual disability, were excluded ( N =223). Of the eligible (either child and/or parent) 2,441 respondents, 1,382 participated (response 56.6%). Differences between respondents and non-respondents were small regarding characteristics known to influence enrolment and outcomes and available for non-respondents: age, gender, degree of urbanization ( i.e. rural versus urban based on density of living addresses per zip code), and psychosocial problems. For these Cohen’s effect sizes ranging from 0.01 (age) to 0.12 (degree of urbanization) [8,17,18]. For the reference cohort a stratified random sample of 1,025 school children and their parents was approached. The sample was obtained via five primary schools, two secondary schools, and one school for intermediate vocational education. Thereby the distribution of children across the study region according to age, gender, socioeconomic position, and degree of urbanization was taken into account. Of these children, 77 were excluded using the same exclusion criteria as with the care cohort. Of the eligible 948