Rasch Analysis of the Composite Autonomic Symptom Score (COMPASS-31) From Data Collected in an Online Population

Area of Science:

• Psychometrics
• Autonomic Nervous System Function

Background:

• The Composite Autonomic Symptom Score (COMPASS-31) is a patient-reported outcome measure for autonomic dysfunction.
• Previous validation focused on clinical populations; this study assessed a general adult population.

Purpose of the Study:

• To analyze the psychometric properties of the Dutch version of the COMPASS-31 in a general adult population.
• To evaluate the questionnaire's structure using classical test theory and Rasch analysis.
• To compare findings with a previous validation study on a clinical group.

Main Methods:

• A transversal observational study involving 100 adult volunteers completing the COMPASS-31 online.
• Data analyzed using classical test theory (SPSS) and Rasch analysis (Winsteps).
• Assessment of internal consistency (Cronbach's alpha), person and item reliability, and unidimensionality.

Main Results:

• Moderate performance of COMPASS-31 domains; Cronbach's alpha generally below 0.65, except for GI, Bladder, and Pupillomotor domains.
• Person reliability was below the 0.8 threshold, indicating limitations for decision-making.
• Only the Pupillomotor domain differentiated between population groups; unidimensionality was not confirmed for Vasomotor and Gastrointestinal domains.

Conclusions:

• The COMPASS-31 demonstrated limited reliability and did not align with the Rasch model in this general adult population.
• The scale functions as a screening tool or descriptive profile, not a unidimensional measure for a single score.
• Recommend using COMPASS-31 domains as independent ordinal indices and developing domain-specific item banks for computer adaptive testing.