Engagement in child psychiatry department appointments: An analysis of electronic medical records in one safety-net hospital in New England, USA.

Objective: This study examined whether being scheduled in a screening clinic versus scheduled directly with a long-term provider to conduct a mental health intake (MHI) is associated with engagement in child psychiatry services in New England, USA.

Method: We used electronic medical record data from one safety-net hospital serving a predominantly low-income and minoritised population. The study sample included 815 youths aged 0 to 25 years, referred or scheduled for a MHI between 1 January 2016 and 31 December 2016. We used chi-square and t-tests to examine the association between referral pathways and engagement, logistic regression to understand the relationship between youth's socio-demographic characteristics and referral pathways, and logistic and Poisson regressions to assess potential moderating effects of socio-demographic characteristics on engagement.

Results: The mean age of the study population was 12 years; 46% were female, and the majority had public health insurance (84%) and lived in high social vulnerability areas (65%). Less than half of the youth attended the first scheduled MHI visit. Those scheduled with the screening clinic were less likely than those scheduled with the provider to ever attend a MHI appointment. Spanish-speakers were more likely to be directly scheduled with a provider (Odds Ratio, OR 0.48; 95% CI: 0.32, 0.73), while those with public health insurance were more likely to be scheduled with the screening clinic (OR 0.56; 95% CI: 0.43, 0.96). Spanish-speaking status and areas social vulnerability scores moderated the relationship between the referral pathway and engagement in psychiatric appointments.

Conclusions: The study highlights the need for psychiatric services to evaluate how MHI referral procedures may mitigate barriers to care and facilitate engagement for youth at high risk of not attending psychiatric service appointments.