I am currently into a matched case control study. I have few queries.
1. I have chosen all the patients registered in last two years as my sampling frame from which I will draw the cases and controls. I understand that I recently registered patients if recruited for the study it will reduce recall bias. So in that case how do I select the sample of lets say 100 cases from an available list of 300, should I arrange the cases based on date of registration and choose all recently registered or should I randomly select or should I systematically select the case?
2. When I would like to match the cases for age and sex, how suitable I should match? Is the individual age or age group (5 years) wise to match? Can some show me a data sheet or a snap shot of how the matched case control study master data sheet will look like?
3. What statistical tests can I perform in a matched case control study? I understand odds ratio and regression models. Is there any reading material which can give me an idea of doing analysis in a matched case control study?
The bible on this is Breslow and Day. Statistical Methods in Cancer Research.
very detailed explanations and worked examples. IARC Pubs 32,82 .1980,1987.
Dr. James Leigh, thank you very much for suggesting an important book.
Dr. Miguel Angel Castro Jimenez, Thank you very much for an elaborate answer. It is highly useful. It cleared my few basic confusions.
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Social Media