Full Program »
115. Symptom fluctuations of attenuated positive symptoms influences predictive power of psychosis-risk screeners
Accurate screening and identification of individuals at clinical high risk (CHR) for psychosis is imperative in early intervention efforts. However, it remains unclear whether symptom fluctuations significantly affect prediction of CHR status as assessed by the Structured Interview for Psychosis-Risk Syndromes. Young adults (N=126) completed the PRIME screen and Prodromal Questionnaire Positive symptom subscale (PQ-P) at two time points within one month (M=22 days). Empirical Bayes Estimates (EBEs) were constructed utilizing Time 1 and Time 2 data, and general linear models (GLMs) were conducted to compare the overall model fit, variance accounted for, and predictive ability of EBEs and Time 1 data. GLM results indicated that utilizing EBEs to predict CHR status provided better overall fit and explained approximately 5% more variance. However, utilizing PQ EBEs decreased the number of correctly predicted CHR individuals by 6.7%, whereas PRIME EBEs increased this estimation by 3.3% as compared to T1; suggesting mixed evidence in regard to utilizing EBEs to increase the ability to accurately predict CHR status. Results also indicate that both EBEs and T1 data for the PRIME and PQ correctly predict CHR status 6.6-10% and 13-20% respectively, suggesting that these measures may not be optimal for predicting CHR risk.