Using confirmatory factor analysis to evaluate construct validity of the Brief Pain Inventory (BPI)

Area of Science:

• Psychometric validation
• Patient-reported outcomes
• Pain assessment

Background:

• The Brief Pain Inventory (BPI) is widely used for assessing patient-reported pain outcomes.
• Previous studies suggest a two-factor structure (pain intensity, pain interference) for the BPI.
• Confirmatory factor analysis (CFA) is appropriate for investigating the BPI's construct validity due to its a priori design.

Purpose of the Study:

• To establish the construct validity of the BPI using CFA.
• To demonstrate factorial invariance of the BPI across demographic variables.

Main Methods:

• A retrospective CFA was conducted on a sample of 364 individuals with HIV/AIDS and cancer.
• One-factor, two-factor, and three-factor models were compared.
• The three-factor model included pain intensity, activity interference, and affective interference.

Main Results:

• The three-factor model demonstrated statistically superior fit compared to the one-factor model.
• The three-factor model showed marginally better fit than the two-factor model.
• The three-factor structure was invariant across disease, age, and ethnicity groups.

Conclusions:

• Evidence supports a three-factor structure for the BPI, alongside the original two-factor hypothesis.
• These findings enhance confidence in the BPI's psychometric properties for clinical research.
• The study provides valuable insights for clinical trialists, pharmaceutical sponsors, and regulators.