The Editor
British Medical Journal
Dear Sir,
The recent submission (24 December 2005) by Sargent, Shepard and Glantz (1), suggests that the results of a study in Pueblo, Colorado (2) confirmed the findings of their previous study in Helena, Montana, discussed in this forum (3). Sargent et al. maintain that smoking bans -- in Helena in 2002, and in Pueblo in 2003 -- caused immediate declines (42% and 27% respectively) in hospital admissions for acute myocardial infarctions (AMI) in those communities.
We previously presented data showing that the Helena observations are consistent with random variation because of the small number of observations on which they are based (4). The recent submission from Sargent et al. prompted us to update information for Lewis and Clark County in Montana (which includes Helena) with important data for the year of the ban (2002), and also to present similar data for Pueblo County in Colorado.
The figures present annual mortality rates from 1979 to 2002 for myocardial infarction (ICD-9 codes 410-410.9 for 1979-1998 and ICD-10 codes 121.0-121.9 for 1999-2001), among persons age 35+ years in the respective counties. These rates, expressed as the number of deaths per 100,000 persons per year, are from the National Center for Health Statistics Mortality Database, and are age standardized to the 2000 US population.
Figure 1 shows that in some years AMI death rates in Helena were quite low or high (designated by a * in the figure). The variability was especially striking for two periods (yellow bars). More importantly, we now have evidence that the AMI death rate of 66 in 2002, the year of the ban (** in the figure), is slightly but insignificantly higher than that in 2001 (63). We conclude that the smoking ban certainly had no effect on the number of people in Lewis and Clark county who died from AMI in 2002.
Figure 2 shows a similar overall downward trend in mortality from AMI over the 1979-2002 period in Pueblo county, and with similar variability, especially in specific years (designated by a ! in the figure). Sargent et al. are correct that Pueblo is a larger city than Helena (at risk population in 2002: 75,600 in Pueblo versus 31,600 in Helena) with more cases (AMI deaths in 2002: 59 in Pueblo versus 19 in Helena). But the impact of this larger size on variability is modest. For example, if the Pueblo ban had gone into effect in 2002 (!! in the figure) instead of 2003, smoke-free advocates might have claimed credit for a 28% drop in AMI mortality (compared with 2001).
Sargent et al. believe that the Pueblo report confirms the findings of the Helena study. The AMI mortality data from Pueblo confirms our position that neither report involves anything more than random variation
Brad Rodu
Professor, Department of Medicine and Endowed Chair, Tobacco Harm Reduction Research
Phone
01-502-561-7273
Fax
01-502-561-7280
Email:
brad.rodu@louisville.edu
Philip Cole
Professor Emeritus, Department of Epidemiology
School of Public Health
University of Alabama at Birmingham
Birmingham, AL 35294
USA
Phone 01-205-934-6707
Fax 01-205-934-8665
Email: pcole@uab.edu
References
1. Sargent RP, Shepard RM, Glantz SA. Helena study independently confirmed. 24 December 2005. bmj.com
2. Bartecchi C, et al. A city-wide smoking ordinance reduces the incidence of acute myocardial infarction. American Heart Association Annual Scientific Sessions, 2005. Dallas, TX.
3. Sargent RP, Shepard RM, Glantz SA. Reduced incidence of admissions for myocardial infarction associated with public smoking ban: before and after study. BMJ 2004; 328: 977-980.
4. Rodu B, Cole P. Additional information on acute myocardial infarctions in Helena, Montana. 25 May 2004. bmj.com