Screening for Problematic Worry in Adults With a Single Item From the Penn State Worry Questionnaire

Abstract

As the rapid assessment of mental health is a growing need, a quick and valid tool for the early detection of symptoms that can be flexibly deployed across a range of contexts may be especially beneficial. This is particularly true of anxiety problems, which when undetected contribute to health care costs and lost work productivity. Data from more than 10,000 respondents (primarily female undergraduates) were used to test whether a single item from the popular Penn State Worry Questionnaire could serve as a screening tool in settings where administration of the full scale is undesirable. Items were evaluated by examining item response theory models, screening capabilities, stability over time, convergence with other anxiety and depression measures, and a response time analysis that assessed how quickly participants responded to each item. Item 15 (“I worry all the time”) emerged as the strongest item: It was the most discriminating and reliable item, had sensitivity and specificity similar to the full scale, had the highest 1-month and 1-year retest coefficients, the highest convergent correlations with measures of anxiety and depression, and was responded to significantly faster than any other item. We suggest that in time-limited contexts, this item is suitable for screening.