Statistical analysis of Goal Attainment Scaling endpoints in randomised trials

Area of Science:

• Clinical Trials
• Biostatistics
• Patient-Reported Outcomes

Background:

• Goal Attainment Scaling (GAS) is a patient-centered assessment tool evaluating interventions based on individual goals.
• GAS maps goal attainment to a common ordinal scale, accommodating heterogeneous patient symptoms.
• This patient-centric approach is crucial for rare disease clinical trials, enabling larger sample sizes.

Purpose of the Study:

• To investigate the statistical analysis of Goal Attainment Scaling outcomes for comparing two treatments in randomized clinical trials.
• To assess the impact of various model parameters on GAS data using a latent variable approach in a simulation study.
• To provide recommendations for designing clinical trials utilizing GAS as an endpoint.

Main Methods:

• Development of a general statistical model for GAS outcomes.
• Investigation of different hypothesis testing approaches for treatment comparison.
• Proposal of a latent variable model for simulating GAS data to evaluate parameter impacts (e.g., number of goals, correlation, discretization, design type).

Main Results:

• The study analyzes statistical properties of different hypothesis testing methods for GAS data.
• Simulation results reveal the influence of parameters like goal number, correlation, discretization, and trial design (parallel vs. cross-over) on GAS outcomes.
• Findings inform optimal strategies for analyzing GAS endpoints in clinical trials.

Conclusions:

• Statistical analysis of Goal Attainment Scaling outcomes is essential for robust treatment comparisons in clinical trials.
• Recommendations are provided for optimizing clinical trial design when using GAS endpoints, particularly in rare diseases.
• The study highlights the utility of GAS in diverse clinical settings, including mastocytosis trials.