Table 1

Measures

ConstructInformant (method): assessed at m f-u*InstrumentsDescriptionPsychometricsScoring
Sociodemographic characteristics
Sociodemographic characteristicsYP (I): 0, 9, 15, 24
PC (I): 0, 24		The sociodemographic interview was largely based on the Client Sociodemographic and Service Receipt Inventory EU version (CSSRI-EU).25 Items on medical history were added.Assessing sociodemographic variables, such as living situation, education, and medical history. Within the medical history domain of the interview, the RA also assessed lifetime suicide attempt(s), as indicated by the YP with a ‘yes’ or ‘no’ to the question ‘have you ever tried to commit suicide?’.Psychometric properties of CSSRI-EU for assessing sociodemographic variables are not available, but the instrument has been validated in a large European study on mental health: EPSILON (European Psychiatric Services: Inputs linked to Outcome Domains and Needs).25Categorical answer categories
Family characteristicsPC (I): 0, 24Sociodemographic interview (PC-version)Highest level of PC education of either parent (‘What is your highest completed level of education?’) and (history of) psychopathology in biological parents (‘Were you ever examined or treated for mental, developmental, language, speech or learning problems?’) was assessed in the sociodemographic interview.The item on level of education came from the CSSRI-EU (see psychometrics for sociodemographic characteristics).Categorical answer categories
Clinical characteristics
Clinical classificationsCL (I): 0, 9, 15, 24Clinical classifications (based on the Diagnostic and Statistical Manual of Mental Disorders, version IV or 5 and the International Classification of Diseases, version 10)26 27Official clinical diagnosis classifications registered in the medical records (or, if no official diagnosis was registered: the preliminary/working diagnosis registered) Clinical classifications are dummy coded and indicate presence or absence of a specific clinical classification or category.
Emotional and behavioural problemsYP (OQ: 0, 9, 15, 24
PC (OQ): 0, 9, 15, 24		Youth Self-Report (YSR)
Adult Self-Report (ASR)
Child Behaviour Checklist (CBCL)
Adult Behaviour Checklist (ABCL)		YP (YSR/ASR) and PC reported (CBCL/ABCL) emotional and behavioural problems in the last 6 months in versions for YP under (YSR/CBCL) or over (ASR/ABCL) 18 years old.The Achenbach System of Empirically-Based Assessment28 29 (ASEBA) instruments have been used extensively in different contexts and have shown excellent psychometric properties.Raw scores were converted to t-scores (with a mean of 50 and a SD of 10) to allow comparison between ASEBA measures.
Norm scores were used to differentiate between normal, borderline clinical, and clinically scoring young people.28 29. Higher scores indicate more emotional/ behavioural problems.
Clinician rated severity of psychopathologyCL (I): 0, 9, 15, 24Clinical Global Impression—Severity scale (CGI-S)CL rated severity of psychopathology over the last week relative to other patients with similar problems.The CGI-S30 is extensively used in psychiatric research31 and has proven useful in predicting suicidal ideation and behaviours.32.Single score measuring severity on a 7-point scale (higher scores indicating more severe problems). The CGI-S was used as a categorical variable in the analyses, with the following categories ‘not at all ill’ (score=1), ‘borderline/mildly/moderately ill’ (scores 2–4) and ‘markedly or more severely ill’ (scores 5–7).
Need for careYP (I): 0, 9, 15, 24
PC (I): 0, 9, 15, 24
CL (I): 0, 9, 15, 24		The Health of the Nation Outcome Scale for Children and Adolescents (HoNOSCA)Assesses YP’s health and need for care in the last 2 weeks. In the MILESTONE study, the HoNOSCA is rated by trained research assistants, based on the ‘mental health’-interview with the YP, PCs, the CL and/or medical records.Good interrater reliability cross-nationally33, face validity and sensitivity to change in clinical use15 in adolescent CAMHS patients specifically. Within MILESTONE, research assistants were trained and regular meetings were held to discuss scoring issues and to improve scoring reliability.Total score (ranging 0–52) of 13 health related domains ranging 0–4. Higher scores indicate more severe problems. Domains 14 and 15 are related to lack of information and access of services and not used in computing the HoNOSCA total mental health score.
Psychotic experiencesYP: 0, 24Development and Well-Being Assessment (DAWBA)DAWBA34 assesses a range of psychiatric diagnoses through structured sections of the online questionnaire, among which psychotic experiences. The open sections of the DAWBA were omitted to limit the burden on the participants and to standardise the classification procedure.The DAWBA psychotic experiences section proved valuable as a screening tool in the youth general population (it has not yet been validated in a clinical sample).35.Respondents indicated whether the young person experienced a range of psychotic experiences, with response options ‘no’, ‘a little’, and ‘a lot’. The total number of a total of 10 experiences the young person experienced (either a little or a lot) was calculated.
Service use
Service (& medication) use in the past 6 monthsYP (OQ): 0, 9, 15, 24CSSRI-EU (amended for use in a psychiatric setting)Assesses inpatient and outpatient service use over the last 6 months in different settings (hospital, community and informal) and medication use over the last 6 months.The CSSRI-EU was found to be effective in tracing patterns of service use in an international population and made comparisons between different countries possible.25Dichotomous service use score over different service use types and quantity of service use (number of nights spent or number of visits multiplied by their average duration)
Current mental healthcareYP (I): 0, 9, 15, 24Part of sociodemographic interviewCurrent mental healthcare was assessed with the questions ‘Are you currently using a mental health service?’ and ‘What mental health service are you currently accessing?’. The research assistant administering the interview could help the young person identify what type of care the young person was in care at if necessary. Categorical answer categories
Transition readiness and appropriatenessYP (OQ): 0
PC (OQ): 0
CL (OQ): 0		Transition Readiness and Appropriateness Measure (TRAM)The TRAM assessed the clinician’s transition recommendations and the availability of appropriate services (both in the CL version of the TRAM). The YP-version and PC-version were used to assess young people’s and parents’ need for ongoing treatment.The TRAM has been established to be a reliable instrument for assessing transition readiness and appropriateness.36Categorical answer categories
Impairment and functioning
Quality of lifeYP (OQ): 0, 15, 24World Health Organization Quality of Life Brief Inventory (WHOQOL-BREF)YP reports on quality of life in the last 2 weeks.The WHOQOL-BREF has excellent psychometric properties.37 Internal consistency for assessing quality of life in adolescents is good and the instrument validly discriminated between adolescents with low and high levels of depressive symptoms.38To allow comparison to the WHOQOL-10037, mean domain scores were calculated and multiplied by 4, yielding a 4–20 transformed mean score of quality of life score in 4 domains: psychological, physical, social and environmental quality of life. Higher scores indicate a higher quality of life.
Everyday functional skillsPC (OQ): 0, 15, 24Specific Levels of Functioning (SLOF)Assesses YP’s everyday functional skills, ‘emphasizing patient’s current functioning and observable behaviour, as opposed to inferred mental or emotional states’.39The SLOF domains have acceptable internal consistencies (except for a Cronbach’s alpha of 0.55 for physical functioning) and good concurrent validity.40Average everyday functional skill-scores ranging from 1 to 5 on 6 domains: physical functioning, personal care, interpersonal relationships, social acceptability, activities and work skills, with higher scores indicating more everyday functional skills.
Independent behaviourYP (OQ): 0, 9, 15, 24The Independent Behaviour During Consultations Scale (IBDCS)YPs report on their independent behaviour on a 5-point Likert scale.Independence is a construct sensitive to change at the age of emerging adulthood and closely related to self-efficacy.41Average score of 7 items ranging from 0 to 4 (with higher scores indicating more independence).
Illness perceptionYP (OQ): 0. 24Brief Illness Perception Questionnaire (B-IPQ)Assesses the young person’s perception of their disorder.The B-IPQ has been used extensively in medical research and to a lesser extent in psychiatric research specifically, and has good test-retest reliability and concurrent validity.42 43Average score per item ranging from 0 to 10 (higher scores indicating higher perceived threat).
Experiences
Life eventsYP (OQ): 0, 9, 15, 24Instrument developed specifically for MILESTONE to assess life events13-item scale assessing 13 different life events such as accidents, deaths, separation over the last 9 months. Total score indicating the number of life events experienced (ranging 0 to 13).
BullyingYP (OQ): 0, 24Adapted from Retrospective Bullying and Friendship Interview ScheduleAssesses the YP’s experiences with bullying in different settings (school, at home, college).The Retrospective Bullying and Friendship Interview Schedule has previously been used in various populations and was found to be predictive of mental health.44 45Bullying experiences were classified in four groups: YP who were the victim of bullying (victim), YP who were both the victim of bullying and bullied themselves as well (bully/victim), YP who bullied (bully) and YP who were not involved in bullying (non-involved).

  • *m f-u=months of follow-up.

  • CAMHS, child and adolescent mental health services; CL, clinician; I, interview; OQ, online questionnaire; PC, parent/carer; YP, young person.