Using structural equation modeling to investigate change and response shift in patient-reported outcomes: practical...

Area of Science:

• Health Services Research
• Psychometrics
• Statistical Modeling

Background:

• Patient-reported outcomes (PROs) are crucial for healthcare evaluation.
• Response shift can alter PRO interpretation, complicating change assessment.
• Structural equation modeling (SEM) is a key statistical method for investigating response shift.

Purpose of the Study:

• To provide accessible descriptions of SEM for response shift.
• To discuss practical considerations for applying SEM in PRO research.
• To offer guidelines for using SEM to interpret PRO change and response shift.

Main Methods:

• Review and synthesis of SEM applications for response shift.
• Explanation of SEM operationalizations of change relevant to response shift.
• Discussion of practical decision-making in SEM application.

Main Results:

• Lack of non-technical SEM descriptions for response shift hinders application.
• SEM application requires specific, often overlooked, sequential decisions.
• Clearer understanding of SEM is needed for accurate PRO interpretation.

Conclusions:

• Proper SEM application enhances understanding of response shift phenomena in PROs.
• Improved interpretation of patient-reported health trajectories leads to better treatment decisions.
• Accurate analysis of PROs ultimately improves patient well-being.