decline (the value of the parameter representing a change in the series level, ω, was −11.69; p = 0.04)
The results of the study have to be understood in relation to its limitations: the residents of Kent were assumed not to be affected by the restrictions, other risk factors for CHD may have affected admission rates, and smoking status and exposure to secondhand smoke were not accounted for. The study showed a peak in acute MIs in 2002, the year with which postimplementation years are being compared. The smoking ban was implemented in March 2002, but, on the basis of previous studies, the authors “postulated that at least 6 months would be needed to allow for the potential health effects from reduction in exposure to second hand smoke, reduction in smoking prevalence and smokers reducing the quantity of cigarettes smoked.” The authors therefore “waited until October 2002 before assessing the impact of the ordinance.” The sensitivity of the analysis to that choice would have been helpful to see. Annual standardized admission rates varied greatly across years, but the Autoregressive Integrated Moving Average (ARIMA) model used to analyze the data, which estimates the effect of the intervention and accounts for residual correlation, would take that variability into account. The published report provides little information on the fit of the time-series model used to measure the effect of the restrictions. As with Seo and Torabi (2007), a differences-in-differences analysis, as is often used to evaluate the effect of a program (Buckley and Shang, 2003), could have been explored, but it is not clear how it would be done with the information provided in the publication.
On July 24, 2003, New York implemented a statewide ban on smoking in all workplaces, including restaurants, bars, and gaming establishments. Statewide smoking restrictions implemented in 1989 had limited or prohibited smoking in particular public places, such as schools, hospitals, public buildings, and retail stores. By 1995, countywide restrictions had begun to be put into place; by 2002, 75% of residents of New York state were subject to local restrictions more stringent than the statewide restrictions implemented in 1989 (Juster et al., 2007).
Juster et al. (2007) published the only report on the effect of the New York state smoking ban on acute coronary events. The authors did not measure compliance, enforcement, or markers of secondhand-smoke exposure for the report, but they cited a report by RTI International (2004) that showed that 93% of restaurants, bars, and bowling facilities were in compliance in the year after implementation. They took into consideration