Timely and evidence-based intervention in major psychotic disorders such as schizophrenia improves outcomes and save lives. In the U.S., only a minority of those with psychotic disorders receive appropriate care within the first six months during which treatment has its greatest impact. Timely care is particularly rare in rural communities like Maine’s. One potential strategy for addressing the rate-limiting factor of rapid detection in Maine’s rural communities is to collaborate with the relatively high number of rural primary care practices with integrated mental healthcare. Whereas regular screening through primary care has enabled the rapid detection and treatment of many serious physical and even mental health conditions, universal screening for psychosis, a key marker of the most serious mental health concerns, awaits a suitable screening tool and protocol. This proposal aims to develop and collect initial reliability and validity data on two screening tools that show promise for predicting psychosis in adolescent and young adult primary care patients. The first tool will be a short self-report screen selected from a large pool of items written at a 5th grade reading level and pretested for understanding by adolescents and young adults. The second will be developed from the automated analysis of speech connectedness as captured in recorded speech during a standardized speech protocol, a tool which might be able to identify youth for whom self-report is either infeasible or unreliable. We will evaluate the performance of these tools in a primary care sample aged 12-25 to capture both early and peak psychosis onsets. We will evaluate the test-retest agreement of tool scores measured three days apart. We will use the Mini Structured Interview of Psychosis-Risk Syndromes (Mini SIPS) to estimate the tool’s criterion validity in identifying psychotic spectrum symptoms. Using medical record diagnoses and problem lists, we will explore the degree to which each tool identifies individuals with previously unidentified psychotic spectrum symptoms. To evaluate psychometrics by developmental stage, we will recruit 60 adolescents ages 12-17 and 60 young adults ages 18-25. The ultimate goal of this project is to collect preliminary data necessary for designing a large longitudinal predictive validity study in which we can also test measurement invariance related to important sociodemographic, clinical, and contextual data. With these data on predictive validity and measurement invariance, we expect to be able to embed one or more psychosis screening tools into the electronic health record, a critical step for implementing universal screening. We can then test whether screening plus triage and engagement in the primary care setting (e.g., the SEE ME model) can reliably and accurately identify more youth in time to improve real life outcomes for rural youth at highest risk for serious mental health challenges. This two year proposal will establish a committ...