Date and Time
April 2, 2008
1:30 PM - 3:00 PM

Availability
Open to the public
No RSVP required

Speaker
Kanaka Shetty - Stanford University

Background: In response to evidence on the risks of secondhand smoke, state and local governments in the US have increasingly restricted smoking in public places since the early 1990s. Several prior studies in smaller regions found that local smoking bans decreased acute myocardial infarction incidence by 8-40%, though their generalizability to other areas is uncertain.

Methods: We conduct a retrospective observational study using time trend analysis to estimate short- and medium-term changes in mortality and hospitalization rates due to acute myocardial infarction and other diseases in smoking-restricted regions, after subtracting secular trends observed in control regions in the US. Analysis includes 868 local areas defined by 3-digit zip codes that include nearly 100% of elderly Americans (1997-2004); all admissions from 562 hospitals (taken from the Healthcare Cost and Utilization Project's Nationwide Inpatient Sample - 1993-2004); and all deaths from 468 counties included in nationally representative surveys (1989-2004). Our main outcome measures were the percent changes in mortality or hospitalization rates for acute myocardial infarction and other diseases following the imposition of a workplace smoking ban (exclusive of bars and restaurants) or a smoking ban of any type.

Results: Smoking restrictions are not associated with declines in all-cause hospital admissions or mortality, nor declines in outcomes related to acute myocardial infarction in the working-age population. Restrictions may be associated with declines in mortality in the elderly (-1.4%, p = 0.06) but at the 10% significance level.

Conclusions: US governmental smoking bans are not associated with statistically significant short-term declines in hospital admission rates in the elderly or the working-age population for myocardial infarction or other diseases. There is some evidence that smoking restrictions are associated with declines in all-cause mortality rates in the elderly, though these results are less certain.