On the Colton model for clinical trials with delayed observations -- normally-distributed responses

Area of Science:

• Medical Decision Making
• Bayesian Statistics
• Biostatistics

Background:

• The Colton model aids in choosing between two medical treatments.
• A critical assumption is the time lag between treatment and response observation.
• This lag creates challenges in patient management during clinical trials.

Purpose of the Study:

• To analyze the Colton model under a time lag assumption.
• To propose and evaluate procedures for patients arriving during the treatment-response lag.
• To compare the performance of these procedures within a Bayesian framework.

Main Methods:

• The study extends the Colton model to incorporate a time lag.
• Two novel procedures are developed for patient management during the lag period.
• Bayesian analysis is employed to assess the relative performance of these procedures.

Main Results:

• The proposed procedures offer practical solutions for managing patient flow during treatment lags.
• The Bayesian framework provides a method for comparing the efficacy of different management strategies.
• Performance is evaluated under assumptions of normally distributed responses with unknown means and known common variance.

Conclusions:

• The developed procedures are effective in managing patients during the time lag in the Colton model.
• Bayesian analysis is a suitable framework for evaluating these procedures.
• The findings contribute to optimizing clinical trial designs involving delayed treatment responses.