Question

Public Health Epidemiology

A case control study of household tobacco smoke and sudden infant death syndrome (SIDS) in the state of IL found an OR=1.7, such the odds of SIDS among children in houses in which at least one household member smoked in the house was 1.7 times that of children who were not exposed to household tobacco smoke. The SIDS cases came from a state registry of all SIDS deaths. The controls were selected by identifying a child born on the same day from state birth records, then inviting that parent/child to participate by telephone and letter.

1. The participation rate among parents of cases was 90%, and the participation rate of parents of controls was 66%. From the birth certificate data, we were able to assess the maternal education of participants and non-participants, and learned that 70% of participants had a college education while only 40% of non-participants did.

     a) Briefly identify and explain the type of bias may have resulted?

b) How might this bias the OR? (e.g. This type of _____ bias may have resulted in bias away/toward the null, and therefore the OR reported may be an over/under estimated of the true OR)

2. The interviewers knew the case status, since it is apparent when doing the interview if it is a grieving parent or one with a living baby.

       a) Briefly identify and explain the type of bias may have resulted?

b) How might this bias the OR? (e.g. This type of _____ bias may have resulted in bias away/toward the null, and therefore the OR reported may be an over/under estimated of the true OR)

Answer 1