Poster #119 - Moderating Effects of Age on Prime Screen Validity for Early Detection of Psychosis Risk

• In Event: 1-061 - Poster Session 02
  In Poster Session: PS 02 Section - Methods, History, Theory

Integrative Statement

Background: Early detection and treatment of psychosis can have profound effects on quality of life in adolescents and emerging adults. Though valid measures of psychosis risk are available, they are costly and time intensive (SIPS; Miller et al., 2003). The Prime Screen (Miller et al., 2004) is a 12-item self-report measure showing promise for early detection of psychosis and concurrent validity (Kline et al., 2012; 2015). Given the developmental nature of psychotic disorders (Howes & Murray, 2014), the validity of a screening tool for these symptoms should be evaluated across a range of ages. There is evidence to suggest that, when responding to Likert-Type items, younger participants, over-endorse the endpoints of a scale (Kramer, 2009; Mellor & Moore, 2014) (i.e., treating items like yes/no, rather than scaled). Such responding patterns may have deleterious effects on sensitivity and specificity of the Prime Screen for younger individuals, given scoring conventions (1 or more extreme item endorsement) for identifying “at-risk” individuals. Objective: To evaluate age differences in sensitivity and specificity of the Prime Screen (relative to diagnosis on the SIPS). Measures are from a larger study of at-risk youth. The primary goal is to raise awareness of possible false positives for younger participants. Method: The sample of N = 134 help-seeking youth was racially diverse (52% African American/Black, 29% Caucasian/White, 14% Biracial, and 5% otherwise designated). The predominantly female (61%) sample demonstrated a wide range in age (from 12 to 22.5 years, M = 15.41, SD = 2.50); appropriate for the current study. The sample varied in family income, though with 47% reporting income < $40,000. Measures of interest for the current study include the SIPS (the Structured Interview for Psychosis Risk Syndromes) and the Prime Screen (a 12-item self-report inventory developed by the authors of the SIPS). Results: Preliminary analyses suggest younger participants have higher rates of false positives and do appear to over-endorse the extreme items. Reflecting this pattern, exploratory ROC curve analyses suggest the optimal number of items for diagnosis consistent with the SIPS declines with age (from five items for the youngest to one item for the oldest participants). Conclusion: Though not possible with these data to determine whether younger participants differ due to responding styles, due to more intensely feeling the emotions and sensations, or due to misunderstanding the wording of the items, it remains clear that younger participants "use" the Prime Screen differently than older participants. This recognition is important in understanding how use of this screener of emerging psychosis with a more general population (e.g., youth in schools, counseling offices, pediatricians, etc.).

Authors

• Trevor Pitts, University of Maryland, Baltimore County
  Presenting Author

• Pamela Rakhshan, University of Maryland Baltimore County
  Non-Presenting Author

• Jason Schiffman, University of Maryland, Baltimore County
  Non-Presenting Author