I have already explained why it is not appropriate to conclude that the observed 14% decline in heart attack mortality in New York City from 2003 to 2004 is attributable to the smoking ban. I have argued that the simple observation of this year-to-year change in heart attack deaths does not mean that the smoking ban (implemented in July 2003) was the cause of the decline.
I have also shown that even if there was a decrease in the expected number of heart attack deaths in 2004 in NYC, it was not a 14% decrease from expected, because there was a pre-existing trend of not only declining heart attack deaths in NYC from 2000-2003, but an accelerating trend.
Finally, I have suggested that the decline in heart attack deaths could be at least in part attributable to changes in heart disease treatment, not merely to a change in the incidence of heart disease.
Let us ignore these methodologic concerns, however, and assume that there actually was a 14% reduction in heart attack deaths in New York City compared to what would have been expected and that the change reflects a true decline in actual heart attack incidence, even accounting for secular trends.
Could one now conclude that the observed decline in heart attacks in New York City was due to the smoking ban?
The Rest of the Story
The answer is NO.
And the reason is that there is a much stronger factor that is most certainly operating to decrease cigarette consumption and smoking prevalence in New York City during the time period in question.
And that factor is:
the $1.42 per pack increase in the City cigarette tax which was implemented in July, 2002.
There is simply no way that I think one could credibly conclude that any observed change in cigarette smoking behavior and the resulting change in heart attack mortality (or incidence) in New York City during this time period (2002-2004) was due primarily to the smoking ban, when in fact, there was a major intervention in New York City which is well-documented to substantially affect cigarette consumption and smoking prevalence: namely, a 1700% increase in the NYC cigarette tax.
Does this mean that the smoking ban is not contributing to any observed decline in smoking prevalence and heart attack incidence? No. It probably is making some contribution. But the contribution of the cigarette tax increase is arguably likely to be larger, and the bottom line is that without some sort of sophisticated research methods, one cannot tease out any effect of the smoking ban from an effect of the cigarette tax (and even with sophisticated analytic methods, I'm not sure it could be done).
Thus, it is simply not appropriate to take the observed New York City observation of a decline in heart attack mortality and attribute that effect, in its entirety or in any substantial proportion, to the smoking ban, even if one were to ignore all the other methodologic problems in making such a claim.
Post a Comment