Jacob Sullum over at Reason.com's Hit & Run blog has alerted us to a new study, published online ahead of print late last month in the Journal of Community Health, which systematically examines changes in heart attack mortality trends within the first year of implementation of smoking bans in the six states which adopted such bans during the period 1995-2003.
(see: Rodu B, Peiper N, Cole P. Acute myocardial infarction mortality before and after state-wide smoking bans. J Community Health 2011; published online ahead of print on August 30, 2001. doi: 10.1007/s10900-011-9464-5)
The authors examined age-adjusted rates of heart attack mortality during the 3 years before implementation of the smoking ban and during the first year after the smoking ban was implemented. These trends were also compared with those in the 44 other states without smoking bans.
The results were that in four of the six states (California, Utah, Delaware, and South Dakota), the smoking bans were not associated with any significant short-term decline in heart attack mortality. In one of these states - South Dakota - there was an 8.9% increase in heart attack mortality during the first year of the smoking ban which was significantly different from the expected decline of 7.2%.
In two of the states - Florida and New York - there were declines in heart attack mortality during the first smoking ban year that were significantly greater than previous trends. However, these declines were not significantly different from the declines during the same year observed in the other 44 states. Specifically, the heart attack rate in Florida fell by 8.8% in 2004 and in New York it fell by 12% in 2004 (2004 was the year of the smoking ban implementation in both states). However, nationally, the heart attack mortality rate fell by 9.8% during 2004, a value not statistically different from what was observed in New York or Florida).
The paper concludes: "The major finding of this study is that state-wide smoke-free laws resulted in little or no measurable immediate effect on AMI [acute myocardial infarction] death rates."
"Smoke-free ordinances may serve public health objectives by providing non-smokers with indoor environments that are free from irritating and potentially harmful pollutants. However, this study does not provide evidence that these ordinances result in a measurable immediate reduction in AMI mortality of the magnitude claimed by reports based on very small incident numbers."
The Rest of the Story
This study adds to the growing body of literature suggesting that in large, systematic studies of multiple states or large population groups, no significant immediate effects of smoking bans on heart attacks can be detected. Jacob Sullum astutely points out that the studies which have reported dramatic effects of smoking bans on heart attacks have been in small communities, with small sample sizes. The larger studies - conducted on entire states or nations - and the systematic studies which have examined multiple communities or states - have generally failed to find any significant effect.
This phenomenon is a class demonstration of publication bias. When small studies are the ones which tend to report an effect and larger studies do not, one must be very suspicious of publication bias. The reason for this is that small studies which find a negative effect tend not to be published (or communities in which no dramatic decline in heart attacks is observed anecdotally tend not to be chosen for study).
I have argued that there is more than just publication bias occurring here, because even with the small positive studies, many have used inappropriate analyses or incorrectly analyzed or reported their findings. A great case in point is the Ohio Department of Health study released just last week, in which the actual data presented showed no evidence of an effect of the smoking ban on heart attacks, but the report concluded that there was a substantial effect. The discrepancy arose because the report essentially ignored the first six months of baseline data, in which there was a dramatic decline in heart attacks. Thus, the report gives the false impression that the baseline rate of change in heart attacks in Ohio was stable. In this light, the decline in heart attacks after the smoking ban is interpreted as demonstrating a huge effect of the ban. Instead, the truth is that the decline in heart attacks before the ban was actually greater than the decline after the ban.
Jacob Sullum nicely summarizes the phenomenon described above and also calls for a re-examination by the Institute of Medicine of the evidence on this issue: "While a few small jurisdictions, such as Helena, Montana, and Pueblo, Colorado, have seen big drops in AMI rates after implementing their smoking bans, studies that look at multiple jurisdictions and bigger populations (including analyses of nationwide data) find no such effect. Ban boosters focus on the few places that fit the story they want to tell, ignoring the broader picture. This blatant cherry picking has been blessed by the National Academy of Sciences, whose Institute of Medicine issued a 2009 report endorsing the biologically implausible notion that smoking bans have a noticeable impact on heart attack rates within a year or two. In light of the accumulating evidence to the contrary (much of which was available when the report was written), that embarrassing conclusion should be revisited."