Selection models for efficient two-phase design of family studies

Area of Science:

• Biostatistics
• Genetic Epidemiology
• Statistical Genetics

Background:

• Family studies often use biased sampling, selecting individuals from disease registries.
• This leads to challenges in efficiently recruiting consenting relatives for genetic and phenotypic data collection.

Purpose of the Study:

• To propose an efficient selection model for recruiting families in two-phase genetic studies.
• To optimize phase II sample selection under budgetary constraints.

Main Methods:

• Comparison of simple random sampling, balanced sampling, and an approximately optimal selection model.
• Utilizing copula models to address within-family dependence for current status disease data.
• Investigating efficiency gains and robustness to model misspecification.

Main Results:

• The optimal selection model demonstrates significant efficiency gains over simple random and balanced sampling.
• The study quantifies the benefits of using an optimized recruitment strategy.
• Robustness of the optimal sampling approach to potential model misspecification was assessed.

Conclusions:

• An efficient selection model can optimize family recruitment in genetic studies.
• This approach is valuable for maximizing information obtained within budget limitations.
• The findings are illustrated with an application to psoriatic arthritis family studies.