NHIS - Adult Tobacco Use Information
Major changes in data collection technology
- From 1963 through 1996, the NHIS questionnaire was printed as a paper booklet. The interviewer read the questions and wrote the answers in the booklet. Results were subsequently keyed into a computer in preparation for data processing.
- From 1997 through 2003, the NHIS used CASES software for computer-assisted personal interviewing (CAPI). The interviewer read the questions and entered the responses directly into a laptop computer.
- In 2004 the NHIS questionnaire converted from CASES to BLAISE CAPI software. This technology continues to the present.
Major changes to smoking status questions
- From 1965 through 1991, the basic cigarette smoking status questions consisted of two parts: (1) "Have you smoked at least 100 cigarettes during your entire life?" If yes, "Do you smoke now?"
- In 1991, a transition to a new measure of smoking status was initiated. For the first time, the NHIS distinguished smokers who smoked daily from those who smoked less often than daily. The 1991 Health Promotion and Disease Prevention Supplement included an expanded set of smoking status questions in which respondents were first asked, "Have you smoked at least 100 cigarettes during your entire life?" If yes, "Do you smoke now?" For those who said "yes" to the question about current smoking, the question was asked, "Do you some everyday or some days?" For those who said "no" to the current smoking status question, the question was asked, "Do you smoke some days or not at all?" This additional follow-up resulted in the classification of persons as "someday smokers" who would otherwise have been considered former smokers, since they initially said that they did not smoke now.
- In 1992, the transition to the new set of questions continued. The NHIS Cancer Supplement consisted of two separate questionnaires, each asked of a representative sample of the U.S. civilian, non-institutionalized adult population aged 18 years and over. The Cancer Control questionnaire contained the same set of questions asked the previous year (see above), while the Cancer Epidemiology questionnaires transitioned to the more parsimonious set of questions: "Have you smoked at least 100 cigarettes during your entire life? " If yes, "Do you NOW smoke cigarettes every day, some days, or not at all?" The inclusion of both sets of questions in the same year allowed for evaluation of the impact of the question change on population prevalence. The revised current smoking status question was estimated to have resulted in an increase in smoking prevalence of about 1% as a result to capturing smoking among persons who would otherwise have been classified as nonsmokers with the original question .