Psychological distress: Bifactor modeling of the DASS-21 in Chilean healthcare workers

The Depression, Anxiety, and Stress Scale (DASS-21) is widely used to assess mental health symptoms; however, its dimensionality in working populations, particularly healthcare workers, remains under debate.

This study aimed to examine the dimensional structure of the DASS-21 among Chilean healthcare workers to evaluate its suitability as a screening measure of psychological distress in occupational contexts exposed to crises, emergencies, or disasters.

A cross-sectional survey was conducted in 2021 using a non-probabilistic convenience sample of 239 healthcare workers from a hospital in northern Chile. Confirmatory factor analyses were conducted to test four competing models (unidimensional, three correlated factors, second order, and bifactor) using structural equation modeling. Model fit was evaluated using RMSEA, CFI, and TLI, complemented by auxiliary bifactor indices (ECV, ωH, ωHS, PUC, ARPB).

All models showed acceptable fit (RMSEA = 0.038-0.065; CFI = 0.974-0.992; TLI = 0.971-0.990). Although the bifactor model demonstrated the best global fit, auxiliary indices supported the dominance of a general factor (ECV = 0.87; ωH = 0.94), with limited reliable variance attributable to the specific factors (ωHS ≤ 0.16). Measurement invariance across sex was supported.

In this non-clinical sample characterized by low symptom severity, the DASS-21 operates primarily as an essentially unidimensional measure of psychological distress. From a nursing practice perspective, the total score may be used as a parsimonious screening indicator to support early detection, monitoring, and preventive mental health strategies among healthcare workers.