Linking Scores with Patient-Reported Health Outcome Instruments:A VALIDATION STUDY AND COMPARISON OF THREE LINKING...

Area of Science:

• Psychometrics
• Health Outcomes Research
• Biostatistics

Background:

• Linking and equating procedures are established psychometric methods.
• Patient-reported outcome (PRO) data increasingly utilizes linking techniques.
• Linking PRO data serves to harmonize data, aggregate scores, and enable score conversion.

Purpose of the Study:

• To evaluate different linking procedures for patient-reported outcome (PRO) instruments.
• To provide guidance on selecting appropriate linking methods for PRO data.
• To address methodological questions in linking health outcome data.

Main Methods:

• Demonstrated linking procedures: equipercentile, unidimensional IRT calibration, and calibrated projection.
• Utilized the Patient-Reported Outcomes Measurement Information System (PROMIS) Depression bank and the Patient Health Questionnaire-9 (PHQ-9).
• Validated linking across two samples and simulated varying instrument correlation levels.

Main Results:

• The study compared the performance of different linking methods under various conditions.
• Findings offer practical guidance for researchers on choosing the most suitable linking strategy.
• The extent of assumption violations impacts the decision to link PRO instruments.

Conclusions:

• Linking PRO instruments is valuable for data harmonization and score interpretation.
• Methodological considerations are crucial when linking PRO data, as equating assumptions are often unmet.
• Further psychometric research is needed for robust linking of PRO instruments.