(PS14-87) Confirmatory Factor and Reliability Analyses of the Parent Version Barriers to Treatment Participation Scale - Expectancies: A Replication Study Using a U.S. Community Sample

While an estimated 7% of youth have mental health (MH) impairments, a mere 2.5% of them receive the MH services they need. Of youths initiating MH care, treatment dropout rates range from 28% to 75% (De Haan et al., 2017). Experiencing barriers to treatment participation has been proposed as a stronger predictor of dropout than pre-treatment family or child variables (De Haan et al., 2013), and is also linked to decreased treatment engagement and motivation (Nanninga et al., 2016). Yet despite studies showing that parent barriers to treatment directly impact children’s access to MH care, there are few measures available to reliably assess perceived barriers.

The Barriers to Treatment Participation Scale - Expectancies (BTPS-exp; Murphy, 2005; Nanninga et al., 2016) is one of few measures that assess parent expectations of barriers when seeking MH care for their child. Results from a previous confirmatory factor analysis (CFA) by Nanninga et al. (2016) indicated a relatively good data fit for both a four-factor and a second-order one-factor model in a Dutch population, and the measure showed good reliability and internal consistency. However, the factor structure of the BTPS-exp parent version has yet to be evaluated in a U.S. community sample. Thus, our study had the following aims: (a) evaluate the factor structure of the BTPS-exp through a CFA and (b) assess the internal consistency of the BTPS-exp using a U.S. community sample.

407 participants across the U.S. completed measures in an online survey battery through Qualtrics. To be inclusive of all family structures, eligible participants were any adult who identified as currently or formerly serving as caregivers of children aged 5-18 years. Actual utilization of youth MH services was not a prerequisite for participation. The average participant age was 45.5 years (SD = 14.0), and 55% identified as female. The sample was ethnically diverse (i.e., 64% White, 10% Multiethnic, 10% Black or African American, and 9% Hispanic or Latino).

 All 44 items of the BTPS-exp are included in the analyses. CFA investigated whether a one-factor, four-factor, and second-order one-factor model best captured the measure’s structure. Model fit indices approached adequate values for both the four-factor model (i.e., CFI = .892, TLI  = .886, RMSEA = .052, SRMR = .048, 90% CI [.048, .055]) and the second-order one-factor model (i.e., CFI = .885, TLI  = .879, RMSEA = .053, SRMR = .103, 90% CI [.050, .057]). Overall findings were slightly worse in comparison to those obtained with Nanninga et al.’s (2016) Dutch participants. Our results also indicated good internal consistency (Cronbach’s α ranging from .90 to .97), consistent with values found in Nanninga et al.’s work.

These results suggest that parent perception of barriers may moderately vary across different countries and systems for MH care. Thus, an exploratory factor analysis of the BTPS-exp’s factor structure may be considered as a critical next step in the process of developing a valuable instrument for assessing treatment barriers in U.S. populations. Insights gained from the exploration of modification indices, as well as future directions, will be discussed further.