Psychometric validation of PROMIS® Anxiety and Depression Item Banks for the Brazilian population

Qual Life Res. 2020 Jan;29(1):201-211. doi: 10.1007/s11136-019-02319-1. Epub 2019 Oct 9.


Purpose: Scientific evidence indicates that depression and anxiety symptoms may be understood as risk factors associated with the incidence and progression of chronic diseases. Considering the lack of mental health assessment tools that meet strict methodological standards, the authors have chosen to validate the psychometric properties of Anxiety and Depression Item Banks - Emotional Distress domain of the Patient-Reported Outcomes Measurement Information System (PROMIS®) for the Brazilian population.

Methods: In this study, 606 adults responded to the self-administered Anxiety and Depression Item Banks, which were calibrated using Factor Analyses (Exploratory and Confirmatory analysis) and adjustment of the Graded Response Model. Transcultural validity was assessed by Differential Item Functioning (DIF).

Results: The two-factor analysis confirmed the unidimensionality of Emotional Distress Items (CFI = 0.96, TLI = 0.96, RMSEA = 0.05). The residual correlation matrix did not identify item pairs with local dependence. Indicators marked with DIF presented a low impact for gender, age, and language variables. The instrument demonstrated greater reliability in the moderate-severe range, indicating that the error reduction is reflected in the - 1.0 to + 3.0 amplitude.

Conclusion: The psychometric measurements of Anxiety and Depression Item Banks in the Brazilian version were equivalent to those in the original version. Additional research contemplating patients with different levels of emotional distress are necessary to better comprehend the results obtained in this study.

Keywords: Adults; Emotional distress; IRT; PRO; PROMIS®.

MeSH terms

  • Adult
  • Anxiety / diagnosis*
  • Brazil
  • Depression / diagnosis*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Psychometrics / methods*
  • Quality of Life / psychology*
  • Reproducibility of Results
  • Surveys and Questionnaires