Table 2.  Five-year Outcomes of Abdominal Aortic Aneurysm Screening by Smoking History in a Cohort of 100 000 Men 65 to 74 Years of Age.1

Variable Assumptions Ever Smokers Never Smokers Total Cohort
History of smoking, % – – 69
AAA prevalence in men age 65–74 y, %
   Ever smokers – – 6.4
   Never smokers – – 1.8
AAA-related deaths per 1000 person-years in uninvited controls – – 0.72
OR reduction in AAA-related death with screening – – 0.57
U.S. male population age 65–74 y (millions), n – – 8.3
Results
AAAs in cohort, n 4416 558 4974
AAA-related deaths, n
   No screening 320 40 360
   Invited for screening 182 23 205
AAA deaths prevented, n 138 17 155
Estimated 5-y AAA-related deaths in the U.S. male population aged 65-74 y, n
   Not screened 26,521 3351 29,872
   Invited for screening 15,129 1912 17,041
AAA deaths prevented by screening, n 11,392 1439 12,831
AAA-attributable deaths, % 89 11 –

1. AAA= abdominal aortic aneurysm; OR= odds ratio. Approximately 69% of men in the United States age 65 to 74 years have a history of smoking (ever smokers), defined as lifetime consumption of more than 100 cigarettes38. One of the study authors provided the prevalence of AAAs in men age 65 to 74 years from a screening study of 126,696 U.S. veterans1. We estimated AAA-related deaths per 1000 person-years in uninvited controls by summing the number of AAA-related deaths in the control groups across the 4 trials (Figure 1) and dividing by the product of the number of control group participants multiplied by the mean followup for each trial in years (Table 1). We apportioned the expected number of AAA-related deaths without screening in ever smokers and never smokers on the basis of the relative prevalence of AAAs in each group. To model screening benefits, we used the pooled OR 0.57 for reduction of AAA-related mortality from the meta-analysis (Figure 1). We assumed that the screening attendance rate, operative mortality, and other factors were similar to those in the screening trials and that ever smokers and never smokers would receive equal benefit in reduction of AAA-specific mortality if invited to attend screening. The estimated number of men age 65 to 74 years in the U.S. population was obtained from U.S. Census data for the year 200037. The Appendix Table shows the formulas used for calculations.

The following caveats apply to these estimates:

  1. In the veterans screening study cohort, 74% of veterans age 65 to 74 years had a history of smoking compared with 69% in the general population. As a result, the overall prevalence of AAAs in this cohort may also be higher than in the general population.
  2. The key variable in this model is the relative prevalence of AAAs in ever smokers versus never smokers, which is determined by the relative burden of other AAA risk factors in each group. This model assumes that the burden of AAA-risk factors in the general population would not be greater in never smokers in relation to ever smokers than that seen in the veterans screening cohort.
  3. This model assumes that the age-specific AAA-related mortality rate is similar for never smokers compared with ever smokers. This assumption appears to be conservative, since the prevalence of AAAs at a specific age is greater in ever smokers versus never smokers (Figure 2).
  4. Since racial and ethnic data were not reported for the AAA screening trials, we cannot be sure that the AAA-related mortality reduction attributable to screening from the 4 screening trials would be applicable if applied to the racial and ethnic mix of the U.S. population.

Return to Document