RR, 0.97; 95% CI, 0.89–1.06; adjusted RR, 0.99; 95% CI, 0.90–1.08). There was, however, a significant difference in the reduction in acute MI hospitalization rate between those residing within the Pueblo city limits and those in El Paso County (p < 0.001).
The two studies had the same strengths and limitations. They both had pre-implementation and postimplementation information and a concurrent control group, and the authors adjusted for out-of-hospital deaths, season, and county population. The smoking rate in El Paso County, the concurrent control group, however, increased from 17.4% (95% CI, 14.5–20.2%) to 22.3% (95% CI, 19.3–25.4%), whereas the rate in Pueblo County (including the city of Pueblo) decreased from 25.9% (95% CI, 20.2–31.6%) to 20.6% (95% CI, 15.4–25.8%) (CDC, 2009). The trends in the smoking rates could affect the estimated changes in acute MI in comparisons between the two counties. The authors note that the decrease in Pueblo County was not significant but do not comment on the change in El Paso County. Data on changes in smoking rates in Pueblo city itself, the location of the ordinance, were not available. It is unknown to what extent Pueblo County residents who do not live in Pueblo city work or spend time in Pueblo city. If a substantial number of county residents spend time in the city that could affect comparisons by biasing towards the null. The authors did not confirm the definition of acute MI by verifying an ICD-9 code and did not provide retrospective results from Pueblo for trends in acute MI admissions. The studies lacked information on variant risk factors at the patient level, including changes in smoking status. The authors did not quantify exposure or adjust for air-pollutant concentrations, although they noted that the inclusion of a control county may have accounted for fluctuations in air quality. The studies did not account for confounders that could include prevention activities and pollution reduction in Pueblo or migration. The statistical model that accounted for season demonstrated a poor fit with only 1 degree of freedom.
Monroe County, Indiana, implemented a ban on smoking in all restaurants, retail stores, and workplaces effective August 1, 2003; the ban was extended to bars on January 1, 2005. One publication examined the relationship between the smoking ban and acute coronary events (Seo and Torabi, 2007). The committee was unable to find any published information on decreased concentrations of secondhand-smoke components or compliance with the Monroe County ban.