Development and validation of an interpretive guide for PROMIS scores

Area of Science:

• Health Informatics
• Medical Statistics
• Patient-Reported Outcomes

Background:

• Accurate interpretation of patient-reported outcome measures (PROMs) is crucial for effective clinical practice.
• The Patient-Reported Outcome Measurement Information System (PROMIS) offers valuable PROMs, but score interpretation can be challenging.

Purpose of the Study:

• To develop and validate graphical tools, termed T-score Maps, for simplifying the interpretation of PROMIS measure scores.
• To enhance the clinical utility of PROMIS measures through improved score interpretability.

Main Methods:

• Item-level data from 21 PROMIS short forms were used to predict item responses across the score range.
• These predicted responses were graphically mapped to create T-score Maps.
• The accuracy of T-score Maps was evaluated by comparing predicted responses to observed responses in a longitudinal study of 1594 adults with chronic conditions.

Main Results:

• T-score Maps were successfully constructed for 21 PROMIS short forms, covering adult and pediatric populations.
• Observed participant responses strongly correlated with predicted responses (r = 0.762 to 0.950).
• T-score Maps demonstrated high accuracy, with predicted responses matching observed responses between 69.5% and 85.3% of the time.

Conclusions:

• The findings support the validity of the predicted responses used in constructing T-score Maps.
• T-score Maps are a promising tool for aiding score interpretation and can be tested in various clinical applications.