Smoking, crashing, and burning: Unintentional injury death risks by smoking status in the US

BACKGROUND. Smoking, accidents, and, accidents due to smoking are immense, growing global problems. Accidents represented about 10% of the European and global 2004 burdens of disease in adults and are often forensically linked to cigarettes.

To reduce crash and fire accidents from cigarettes, some bans on smoking while driving, indoors, or in dry lands exist. Recent general population motor vehicle and overall accidental death risks by smoking status might aid deliberations on such bans but are largely unknowns So we estimated recent United States (US) accidental death hazard ratios and rates by smoking status, minimizing selection, follow-up time, and other biases that afflicted previous estimates.

METHODS. Data from the 2002 mortality follow-up of adult respondents to the 1997-2000 US National Health Interview Surveys (NHIS) were analyzed using Cox survival-time methods that adjusted for the complex NHIS survey design. The NHIS enrolls quite representative samples of the US adult noninstitutional population,

RESULTS. 105 motor vehicle and 131 other 1997-2000 NHIS interviewee accident deaths with complete data through 2002 were identified though only 5 and 10, respectively were female ex-smokers. Age-sex-adjusted overall and motor vehicle accident death hazard ratios were 1.62 (95% confidence interval (CI) 1.23-2.13) and 1.41 (CI 0.90-2.22), respectively for current and 0.77 (CI 0.52-1.15) and 0.83 (CI 0.45-1.52) for former smokers.

DISCUSSION. Large accident death excesses are seen in United States smokers, even without adjusting for deaths of nonsmokers in smoking-caused crashes, fires, and other accidents. Reducing or restricting smoking might considerably reduce accident death rates and costs.