Construct Validity of the Psychosis Screening Questionnaire in Ugandan Adults

Abstract

Abstract Background: Psychotic disorders are common and contribute significantly to morbidity and mortality of people with psychiatric diseases. Therefore, early screening and detection may facilitate early intervention and reduce adverse outcomes. Screening tools that lay persons can administer are particularly beneficial in low resource settings. However, there is limited research evaluating the validity of psychosis screening instruments in Uganda. We aimed to assess the construct validity and psychometric properties of the Psychosis Screening Questionnaire (PSQ) in Uganda in a population with no history of a psychotic disorder. Methods: The sample consisted of 2101 Ugandan adults participating as controls in a larger multi-country case-control study on psychiatric genetics. We used confirmatory factor analysis (CFA) and item response theory (IRT) to evaluate the factor structure and item properties of the PSQ. Results: The overall prevalence screening positive for psychotic symptoms was 13.9%. “Strange experiences” were the most endorsed symptoms (6.6%). A unidimensional factor was the best fitting model based on the fit indices including the root mean square error of approximation (RMSEA of 0.00), comparative fit index (CFI of 1.000), and Tucker-Lewis Index (TLI of 1.000). The most discriminating items along the latent construct of psychosis were items assessing thought disturbance followed by items assessing paranoia, with a parameter (discrimination) value of 2.53 and 2.40, respectively. Conclusion: The PSQ works well in Uganda as an initial screening tool for moderate to high-level of psychotic symptoms.