Validation of a five-question survey to assess a child's exposure to environmental tobacco smoke.

PURPOSE: To study the potentially adverse health effects of environmental tobacco smoke (ETS) exposure in young children, a short five-question survey was developed to identify routine exposure to ETS in a large epidemiological study.
METHODS: The survey is administered to parents of a healthy cohort of children starting at age 3 months. To validate the survey, urinary cotinine levels were measured on 50 children from this cohort who were selected based on ETS exposure as reported in the survey: 24 with no exposure and 26 with exposure. Cotinine was adjusted for creatinine.
RESULTS: Overall, children with some form of reported ETS exposure had urinary cotinine levels 7.5 times higher than those who were not exposed. Analysis of variance shows that mean levels of log transformed cotinine in children whose parent(s) smoke in the home, parent(s) who smoke but not in the home, and non-smoking parents are 137.13, 75.60, and 43.28 respectively (p = 0.0009), indicating decreasing levels of cotinine as reported exposure decreases. Using a cut-point of 30 ng/mg of cotinine to differentiate unexposed and exposed to ETS, we found 80% agreement with our survey. A Spearman's ranked correlation coefficient of 0.62 indicates a direct relationship between cotinine and an ETS exposure intensity score (p < 0.0001).
CONCLUSIONS: These results suggest that the 5-question survey reflects the child's exposure to passive smoke and that the survey is sensitive to varying levels of exposure.