Primary care Screening Questionnaire for Depression (PQS4D)'s dimensionality, internal consistency, nomological network, and gender differential item functioning

The Primary care Screening Questionnaire for Depression (PQS4D) was recently introduced. It is a four-item instrument based on the International Classification of Diseases (ICD-10). However, psychometric indicators such as dimensionality, nomological validity, and gender differential item functioning are unknown. The study aimed to test the dimensionality, internal consistency, nomological network, and gender differential item functioning (Gender-DIF) among elderly Colombian COPD patients. A psychometric study was designed in which 349 COPD outpatients were accepted to participate. They were aged between 60-98 years (M=75.56, SD=8.41); 61.89% were male, and 80.23% presented low deterioration (A or B COPD combined evaluation). The PQS4D (Spanish version), Brief Zungs Self-rating Depression Scale (BSDS), Brief Zungs Self-rating Anxiety Scale (BSAS), COPD Assessment Test (CAT), and Frail Non-Disabled [FiND] Screening Tool were completed. Dimensionality (exploratory and confirmatory factor analysis, internal consistency (Kuder-Richardson coefficient and McDonalds omega), nomological network (Students test and Spearman correlation [rs]), and Gender-DIF (Kendalls tau b correlation [t]) were done. The PQS4D presented a one-dimension structure that accounted for 55.78% of total variance with adequate fit-of-goodness indicator [Normalized chi-square of 2.84, RMSEA of 0.07 (90%CI 0.00-0.15), CFI of 0.99, TLI of 0.96, and SRMR of 0.02], acceptable internal consistency (Kuder-Richardson coefficient of 0.73 and McDonalds omega of 0.74), excellent nomological validity (rs[≥]0.30 with SDS, SAS, CAT and FiND)] and non-Gender-DIF (t[≤] 0.20). In conclusion, the PSQ4Ds one-dimensional structure shows acceptable dimensionality, good reliability, acceptable nomological network, and non-Gender-DIF among Colombian COPD patients. Establishing the cutoff point with the best sensitivity and specificity is necessary.