A random-effects mixture model for classifying treatment response in longitudinal clinical trials

Area of Science:

• Biostatistics
• Clinical Trial Methodology
• Psychiatric Research

Background:

• Longitudinal clinical trials often involve heterogeneous patient populations with varying treatment responses.
• Accurately classifying treatment response is crucial for understanding drug efficacy and patient subgroups.
• Existing models may not adequately capture the complexity of diverse patient responses over time.

Purpose of the Study:

• To introduce a novel random-effects regression model for classifying treatment response in longitudinal clinical trials.
• To develop a statistical framework capable of handling longitudinal data from unknown heterogeneous populations.
• To distinguish and assess distinct subgroups of treatment response within clinical trial data.

Main Methods:

• A random-effects regression model incorporating a multivariate normal mixture distribution for random coefficients was developed.
• The model was applied to analyze data from two psychiatric clinical trials involving depression and schizophrenia.
• Longitudinal data with repeated measurements of patient outcomes were analyzed to assess treatment response heterogeneity.

Main Results:

• The proposed model successfully distinguished subgroups of treatment response in both psychiatric clinical trial datasets.
• Analysis revealed ample evidence for a mixture of treatment responses, indicating distinct patient populations.
• Parameter estimates from the mixture model supported the presence of heterogeneity in patient responses to treatment.

Conclusions:

• The developed random-effects mixture model is effective for classifying treatment response in longitudinal clinical trials with heterogeneous populations.
• The model provides valuable insights into distinct patient subgroups and their varied responses to therapeutic interventions.
• This approach enhances the understanding of treatment efficacy by accounting for population heterogeneity in clinical research.