Psychometric evaluation of a four-item scale derived from the PHQ-9 and GAD-7 among community adults based on multivariate generalizability theory and item response theory

The Patient Health Questionnaire-4 (PHQ-4) consists of the first two items of the Patient Health Questionnaire-9 (PHQ-9) and the first two items of the Generalized Anxiety Disorder-7 (GAD-7). To date, no studies have examined the psychometric properties of this PHQ-4 in a Chinese community sample using modern measurement theory. This study aims to conduct a preliminary psychometric evaluation of the PHQ-4 among community adults based on modern measurement theory.

The Chinese version of the combined PHQ-9 and GAD-7 scale (termed the PHQ-16) was used to assess depression and anxiety symptoms among community-dwelling adults residing in rural areas of Yunnan Province, based on the follow up survey of the China Multi-Ethnic Cohort. We selected the items from the PHQ-16 based on the original structure of the PHQ-4 to form a 4-item scale and refer to it as a "PHQ-4-like" scale. The reliability of the PHQ-16 and the PHQ-4-like scale was evaluated using multivariate generalizability theory (MGT). The validity of the PHQ-4-like scale was evaluated using item response theory (IRT).

Based on MGT, reducing the number of items in the PHQ-16 from 16 to 4 decreased the composite generalizability coefficient from 0.906 to 0.719 and the composite reliability index from 0.894 to 0.693. However, when the first two items of the PHQ-9 and the first two items of the GAD-7 were combined to form a four-item scale (the PHQ-4-like scale), the composite generalizability coefficient and composite reliability index were 0.818 and 0.815, respectively. Both the PHQ-9 and the GAD-7 exhibited monotonically increasing threshold parameters. Items 1, 2, 4, and 6 of the PHQ-9 and items 2 and 3 of the GAD-7 showed high discrimination parameters. High average information values were observed for items 1, 2, and 4 of the PHQ-9 and items 2, 3, and 4 of the GAD-7.

The PHQ-4-like scale demonstrated satisfactory reliability and validity in this rural Chinese community sample. However, the PHQ-4 screening performance requires criterion validation. Regarding the anxiety dimension of this scale, it may also be worthwhile to consider combining items 2 and 3 from the GAD-7.