Related Experiment Videos
Design of case-controls studies with unscreened controls
V Moskvina1, P Holmans, K M Schmidt
1Bioinformatics and Biostatistics Unit, Cardiff University, College of Medicine, CF14 4XN, UK.
Annals of Human Genetics
|September 6, 2005
Summary
Using unscreened population controls in genetic studies can be cost-effective. This approach offers similar statistical power to screened controls for many complex disorders, especially with decreasing genotyping costs.
Area of Science:
- Genetics
- Epidemiology
- Biostatistics
Background:
- Traditional genetic case-control studies screen controls to exclude illness, maximizing statistical power but increasing recruitment costs and effort.
- An alternative uses unscreened general population controls, which are abundant and cheaper but may reduce study power due to potential disease presence.
Purpose of the Study:
- To quantify power loss when using unscreened controls.
- To develop formulas for determining the number of unscreened controls needed to match screened control power.
- To evaluate the cost-effectiveness of screened versus unscreened control strategies.
Main Methods:
- Mathematical modeling to quantify power loss and derive sample size formulas.
- Cost-benefit analysis considering variable recruitment and genotyping costs.
- Investigation of the impact of case-control ratios on power.
Main Results:
- Formulas were developed to calculate the number of unscreened controls equivalent to screened controls.
- Cost-benefit analysis identified conditions favoring unscreened controls.
- Unscreened controls are potentially cost-effective for complex disorders with population prevalence Kp < 0.2.
Conclusions:
- Unscreened controls present a cost-effective alternative to screened controls in many genetic studies.
- This approach is particularly viable for complex genetic disorders and is becoming more cost-effective with falling genotyping expenses.
- Study designs can benefit from considering unscreened controls for efficiency and cost savings.