This study employed a pragmatic and primarily quantitative design.

Between July 2018 and March 2019, young people aged 15–25 years with CHD attending a specialist outpatient clinic at Starship Hospital (a tertiary hospital in Auckland, New Zealand), who could speak English, use an iPad and provide informed consent (or assent with paired parental consent if under 16 years of age) were invited to participate in the study via flyers disseminated by clinic staff, hence this was a purposive sample. Invited youth participants were approached in clinic by a research assistant before their medical appointments with participant information and consent sheets and a verbal explanation of the study. Clinical staff who reviewed the YouthCHAT results with their young patients were also invited to participate in the study. Those who consented completed an online survey following their consultation with the young person.

Enrolled participants completed YouthCHAT screening on an iPad individually in a consulting room prior to their medical appointments. After the consultation with their cardiac health provider they completed a paper survey to report on YouthCHAT acceptability and utility. YouthCHAT scoring and timing data were collected electronically and securely stored on a central database. Clinicians received YouthCHAT results on their clinic computer at the point of care and discussed them with participants during their medical appointment. Feedback on clinicians’ experience was collected via an electronic survey using SurveyMonkey^®.

The intervention is YouthCHAT, the youth version of the electronic Case-Finding and Help Assessment Tool, a self-reported, tablet-based screener. It consists of up to 72 questions, the exact number answered depending on positive responses to gate-keeper questions in each of 13 domains: smoking, drinking, recreational drug use, problematic gambling, depression, anxiety, sexual health, eating and conduct issues, stress, exposure to abuse, anger management and physical activity [19]. Broadly matched to the areas of HEEADSSS (Home, Education, Eating, Activities, Drugs and Alcohol, Suicide and Depression, Sexuality and Safety) assessment [17].

YouthCHAT includes three well-known and validated screens: for substance abuse (Substances and Choices Scale, SACS), depression (Patient Health Questionnaire—Adolescent Version, PHQ-A) and anxiety (Generalized Anxiety Disorder scale—7 item, GAD-7) with which severity of these issues can be identified. These are validated screening tools. The scores are calculated electronically and provided to the clinician in the summary report including a severity rating for each score range (for example, a PHQ-A score of 15–19 indicates moderately severe depression).

Response options for answers to each question are in the form of yes/no, or Likert scales relevant to the domain (such as ‘not at all’, ‘for several days’, ‘more than half the days’ or ‘nearly every day’ in the past two weeks). For each identified positive domain, a ‘help’ question enables users to communicate whether they would prefer assistance at the time, later or not at all. Previous studies have shown that YouthCHAT is twice as fast as HEEADSSS assessment (mean 8 min 57 s vs. 17 min 22 s; mean difference 8 min 25 s) and that it detects more issues overall on comparable domains [17].

Primary outcomes’ measures were self-reported acceptability and feasibility of YouthCHAT use within an outpatient setting, assessed via feedback from young people and clinicians and the time taken to complete YouthCHAT screening. Young people were asked for overall ratings of the screener and its helpfulness with discussing psychosocial issues with clinicians, specific concerns about included questions, and items regarding YouthCHAT with checkboxes to tick if they agreed with the statements. Clinicians were asked about their usual methods of psychosocial screening, their overall impressions of YouthCHAT and for more detailed open-ended feedback regarding what they liked and did not like about it. Secondary outcome measures were detection rates for each module of YouthCHAT, i.e., mental health or lifestyle issues being ‘present’ or ‘absent’.

All 44 screens were included in data analysis and there were no missing data. Quantitative data were collected through the YouthCHAT database and were processed using Microsoft Excel. Analyses included basic descriptive statistics (e.g., number of youths screened, YouthCHAT summary data and demographic characteristics of the sample). Qualitative data were analyzed using a general inductive approach [20]. Data were independently evaluated, then cross-coded by two researchers (HT and HK) to identify key themes and subthemes, with any disagreements being resolved by consensus.

The study protocol was registered with the Australian and New Zealand Clinical Trials Network Registry ANZCTR in July 2017 (ACTRN12616001243404p.) and received approval as an extension of the previous hospital-based study [18] from the New Zealand Health and Disability Ethics Committee (Reference: 16/CEN/137).

A total of 58 young people with CHD were invited to participate. Of these, 44 provided consent and all (100%) completed YouthCHAT screening and 26 (59%) completed follow-up questionnaires during their congenital cardiology outpatient visit. Although information regarding the age, sex and ethnicity of these young people was collected, their education level, employment experience and health literacy status was not obtained. Out of the non-participants, thirteen said they either were not interested in the study (without giving any specific reason) and one said that they did not have time. Six clinicians participated in the study and four (67%) completed the online feedback survey. Of those surveyed, one was a cardiologist, two were nurse specialists and one was a nurse practitioner. The two clinicians who did not complete surveys were registrars (specialist trainees) who had completed their clinical rotation before survey invitations were dispatched. Information regarding the role and specialty of clinicians was collected, however information regarding their sex and number of years of experience of clinicians was not obtained. Further details of participant characteristics are provided in Tab. 1.

Young people completed YouthCHAT in an average of 9 min and 7 s (range 4 min and 44 s to 44 min and 49 s), and a median of 7 min and 29 s (IQR = 6 min and 40 s–9 min and 36 s).

Young people gave YouthCHAT a rating of 7.0 (range 3–10) on a 10-point scale from ‘lame’ (0) to ‘awesome’ (10) and said that it had been helpful for discussing psychosocial issues with their clinician at a rating of 4.4 (range 1–10) on a 10 point scale from ‘helpful’ (10) to ‘not helpful at all’ (0). Most (N = 25, 96%) young people denied disliking any YouthCHAT questions. One person (4%) out of the young people said they did not like answering questions on worrying and exercise. Optional statements regarding YouthCHAT were acknowledged in a dichotomous manner (agree/disagree) at the rates shown in Tab. 2.

As shown in Tab. 3, the most common issues identified by young people were concerns about life stresses (N = 42; 98%), eating and weight (N = 38; 86%), sexual activity/identity (N = 31; 72%), drinking (N = 27; 61%), depression (N = 18; 41%) and physical inactivity (N = 16; 38%). Positive screens for eating problems were primarily related to dissatisfaction with body shape or weight and serious eating disordered behaviour.

Types of stress experienced by individuals were related to issues at home, school or work, issues with relationships with specific people in their life, being bullied, violence, an upsetting event and money. Eighteen individuals were identified as having symptoms of depression via the PHQ-A. Of these, 12 were rated subthreshold, one was rated mild, one was rated moderate, three were rated moderate-severe and one was rated severe in intensity. Sixteen individuals were identified as having anxiety via the GAD-7. Of these, three were rated subthreshold, 12 were rated mild and one was rated moderate in intensity. Twenty-nine individuals were identified as having substance use problems via the SACS. Of these, 17 were rated subthreshold and 12 were rated as having a low level of problems that required further assessment. Despite YouthCHAT identifying a range of psychosocial issues, only four (9%) young people said that they wanted help that day, 11 (25%) said that they wanted help later and the remainder (66%) said they did not want help at all. The latter category included two young people with moderately severe to severe symptoms of depression who reported suicidal ideation and were flagged as being of immediate concern. They were provided with face to face follow-up by the in-hospital pediatric consult liaison.

Clinicians reported they rarely (N = 2; 50%) or sometimes (N = 2; 50%) conducted intermittent HEEADSSS assessments in an outpatient clinic setting depending to perceived need. Feedback regarding YouthCHAT was generally positive, with all clinicians rating the discussion of psychosocial issues raised by YouthCHAT as ‘very acceptable’ and denying that young people overtly objected to any of the questions. Three clinicians rated YouthCHAT as a ‘very useful’ screener, ‘very useful’ for identifying whether young people were ready to make changes in their behavior and ‘very useful’ for initiating conversation with young people about psychosocial issues. One clinician rated YouthCHAT as ‘neither useful nor un-useful’ for identifying psychosocial issues, ‘neither useful for un-useful’ for identifying whether young people were ready to make changes in their behavior and ‘useful’ for initiating conversation with young people about these issues. Three clinicians said that they would be ‘very likely’ and one clinician said they would be ‘likely’ to use YouthCHAT regularly if it was available in their clinic. Key themes, subthemes and examples of clinician feedback are presented in Tab. 4.