Design and analysis of arm-in-cage experiments: inference for three-state progressive disease models with common...

Area of Science:

• Biostatistics
• Entomology
• Epidemiology

Background:

• Arm-in-cage experiments are crucial for assessing topical treatment efficacy against insect bites.
• Existing methods may not fully capture the nuances of protection over time.
• Efficacy is often evaluated using a three-state model: no landing, landing without biting, and biting.

Purpose of the Study:

• To develop advanced statistical methods for designing and analyzing arm-in-cage experiments.
• To accurately estimate parameters related to insect repellent efficacy.
• To address interval-censored data arising from periodic exposure schedules.

Main Methods:

• Development of a statistical approach combining nonparametric and parametric techniques.
• Modeling dependent sojourn times within a progressive three-state framework.
• Consideration of experimental design factors specific to arm-in-cage studies.

Main Results:

• The proposed methods effectively estimate key parameters for insect repellent efficacy.
• The statistical approach handles interval-censored transition times accurately.
• Illustrative examples using real and simulated arm-in-cage experiment data.

Conclusions:

• The developed statistical methods enhance the design and analysis of arm-in-cage experiments.
• This approach provides a robust framework for evaluating topical treatment efficacy.
• Improved statistical modeling leads to more reliable assessments of insect repellents.