Psychiatric disorders are a major source of disability. In addition, psychiatric disorders are often chronic and have a profound impact on other chronic medical illnesses. Because psychiatric and medical disorders are so intimately entwined, it is critical to develop methods that account for the influence of both mental and physical health on patient-reported outcomes in chronic disease. The general goal for our independent research project, Psychiatric Symptoms and Social Functioning: IRT and DIP, is to establish with greater precision the magnitude of relationships between psychiatric symptoms (depression, anxiety, anger/hostility, and alcohol and substance abuse) and social functioning. The project will be done with large and diverse samples of psychiatric patients and community residents that display a broad range of psychiatric symptoms, medical conditions, and levels of social functioning. Our plan is to develop comprehensive item banks tapping all levels of severity of the relevant constructs and to administer them on personal computers. Data analysis will rely on the psychometric advantages of item response theory (IRT) models. We will apply such models to the items for psychiatric symptoms and social functioning, first, to calibrate them in graded IRT fashion and, second, to investigate differential item functioning (DIP) among items for social functioning contingent on symptomatic status. Analyses of DIP related to variations in the type and severity of psychiatric symptoms will allow us to create measures of functioning that are less confounded with symptoms and even, more robust than those currently available. After the measures are developed and refined iteratively in three cross-sectional samples (Phases I-III, total N = 2,400), we will use them in a prospective study (Phase IV, N = 500) that relies on longitudinal computerized adaptive testing (CAT). We also propose two projects for the Patient-Reported Outcomes Measurement Information System (PROMIS) research network focused on: 1) assessment of negative affects (depression, anxiety, and anger) and pain and the relationships between these constructs and other core domains likely to be of interest to PROMIS, including physical, social, and occupational functioning; and 2) development and testing of a core battery for sleep-wake function (SWF) and investigation of the relationship between SWF and other core domains relevant to PROMIS.
Showing the most recent 10 out of 103 publications