Featured Article

Abdominal aortic aneurysm (AAA) differences between men and women are becoming increasingly well-known, with mounting evidence that women rupture at smaller aneurysm sizes, have higher mortality, and may not receive surgery as often as men. In a recent review of the US Medicare database, we found that men have had a 30% decrease in AAA rupture surgery over the past decade, compared to a 12% drop for women (P < .001). Rupture mortality is higher for women (44.2% versus 52.8%, P < .001), as is elective mortality in 2003 (3.2% versus 5.45%, P < .001). Men have a significantly shorter length of stay and are more likely to be discharged home. We conclude that recent improvements in AAA management have decreased aneurysm-related deaths. Women, however, continue to have consistently poorer outcomes after open and ruptured AAA repair.

Introduction
Significant advances in the management of AAA have been made in the last decade, including the introduction of new endovascular treatment modalities (EVAR), screening studies and randomized trials to refine the appropriate timing of intervention.^1,2 There has been work with national small samples3 or geographically–defined populations^4–6 that points to disparity in outcomes between men and women. Additionally, there is evidence that women rupture at a significantly smaller size (5.0 versus 6.0 cm,⁷ 6.0 versus 6.6 cm⁸), and also may have increased rates of aneurysm growth.⁹ At the University of Pittsburgh, we recently reviewed Medicare data from 1994–2003 to assess the impact of the changes in AAA repair practices on men and women.¹⁰

Results
AAA Elective Repair Volume
Over the past decade in the United States, AAA repair rates have remained static at approximately 28,000 repairs per year in patients over 65 years of age. When these totals were normalized for population changes, there has been a small decline in repairs per capita (Figure 1). Endovascular surgery was found to replace, rather than add to the open repairs, with more than 40% of repairs in 2003 being endovascular. Women comprised 22.6% of all elective repairs, and were significantly less likely to have EVAR than men (in 2003, men with 44.3% EVAR versus women with 28.0%, P < .001). The lower rates of EVAR among women have been previously noted, with the differences being attributed to unsuitable anatomy¹¹ and women’s higher incidence of aorto-iliac occlusive disease accompanying AAA.¹²

AAA Rupture
Rupture rates have declined for both men and women, with the overall incidence falling from 22.⁵ ruptures per 100,000 elderly Medicare recipients in 1994 to 16.3/100,000 in 2003 (Figure 2). This represents a 29% decrease for men, but only a 12% decrease for women (P < .001). When presenting with a diagnosis of rupture, 79% of men received surgery versus 62.8% of women (P < .001). The proportion of women as rupture presentations was 25.6% from 1994 to 1999, and 28.2% in 2000–2003 (P < .001). In other series, women have been found to have lower rates of elective surgery after the diagnosis of AAA,¹³ and also a lower rate of surgery with rupture.⁶ The higher rate of women with ruptured AAA versus elective repairs has been noted in previous work^6,7 as has the increasing percentage of women with ruptured AAA as age increases.⁶ This information supports the hypothesis that women are at increased risk of rupture compared to men at the current thresholds for surgery.

Mortality
Mortality for rupture repairs has remained largely unchanged over the past decade, with no significant decreases seen for either men or women. Average rupture mortality for men was 44.2%, with a higher mortality in women, averaging 52.8% (P < .01). Advancing age was clearly associated with increasing mortality (Figure 3). EVAR was used in 10.6% of patients with a ruptured AAA in 2003 with a significantly lower mortality than for open repair; open mortality 52.0% versus EVAR 33.3%, (P < .001). Multivariate analysis demonstrated that age, female gender and open surgery were strong predictors of increased rupture mortality.

Mortality for elective AAA repair has decreased for both men and women over time, although large discrepancies exist between the genders. Overall, male elective mortality has decreased from 5.57% in 1994 to 3.20% in 2003 (P ? .001). For year 2003, male elective open repair mortality (after eliminating endovascular procedures) has decreased to 4.75% (P = .001 versus 1994). For women, overall elective AAA mortality was 7.48% in 1994, decreasing to 5.45% in 2003 (P < .001). Mortality for open repair in women in 2003 was only marginally better than in 1994 at 6.57% (P = .08 versus 1994). As with emergent repairs, elective AAA mortality rose with age (Figure 3). However, there has been a steady decrease in elective mortality for patients over 80 years, from 13.7% in 1994 to 6.0% in 2003 (P < .001). All other age groups have had a more gradual decrease in elective mortality. Endovascular mortality from 2000–2003 was significantly higher in women than men (2.60 versus 1.28%, P = .013). Multivariate analysis again indicated that female gender, advancing age and open repair are predictive of increasing mortality. Women were significantly older than men for both rupture and elective surgery, a finding that has been noted in other series.^12,14

Length of Stay and Discharge Disposition
Length of stay (LOS) after elective and ruptured AAA repair has significantly and steadily decreased from 1994 to 2003. The average length of stay after a ruptured AAA repair has decreased from 14.5 days (standard deviation [STD] 17.3) in 1994 to 10.4 days (STD 11.9) in 2003 (P = .004). Average stay after elective AAA repair has decreased from 11.4 days (STD 10.2) in 1994 to 7.3 days (STD 8.4) in 2003 (P < .001). Again, there were differences in LOS after elective repair when men were compared to women. Women had a significantly longer LOS after elective AAA repair (P < .05) each year of the study with the exception of years 1996 and 1997. LOS for women was on average 1.43 days longer than that of men. In 2003, elective repairs were broken down into open and endovascular procedures. LOS after open repair was 9.6 days (STD 8.4) for men and 10.4 days (STD 11.6) for women (P = .33) while after endovascular repair the average stay was 3.5 days (STD 4.8) for men and 5.3 days (STD 7.0) for women (P = .006). Both EVAR LOS were significantly lower than for open repair. Women were also more likely to be discharged to a care facility rather than home after either elective repair or rupture. The percentage of women discharged to home after elective AAA repair has remained relatively constant from 1994–2003, averaging 74.6%. This was consistently lower than the percentage of men discharged to home, averaging 84.4% over the same period.

Conclusions
Great progress has been made with AAA disease, resulting in fewer aneurysm-related deaths with the same volume of surgery. A decrease in the number of AAA ruptures with stable or declining numbers of elective repairs is gratifying although somewhat unexpected. It clearly shows that the medical community is gradually reaching the desired outcomes for aneurysm treatment, while utilizing fewer resources. This can only be possible if elective repairs are being performed on patients at higher risk of rupture with a shift in practice patterns. This may be the effect of large studies published in the last decade illustrating the lack of benefit of aneurysm surgery at smaller sizes.^1,2

However, this progress has benefited men much more than women, with women having minimal declines in rupture rates and open surgical mortality. Women are less likely to receive EVAR, and for the first time have been found to have higher mortality with this procedure.

The true question is why gender differences in AAA surgical outcomes exist. Advanced age in women has been found to be a consistent finding in virtually all comparisons of elective and ruptured AAA repairs, and was the factor that most consistently predicted mortality in the University of Pittsburgh multivariate analysis. This, however, cannot explain all of the increases in morbidity and mortality seen in women, as mortality differences based on gender were seen within age groups. Advanced age, undoubtedly, has an effect on the lower number of female patients who are discharged to home, as they are likely to have outlived a spouse and may not have another caregiver. Previous studies noted a higher proportion of current smokers in female AAA patients,^6,12 but differences in the prevalence of coronary disease and other predictors of mortality have not been proven to be significantly increased.¹² Unfortunately, the Medicare database is ill-suited to answer these questions and can only highlight their existence and importance to stimulate further research.

Future Directions
Consideration should be given to repair of women’s AAA at a smaller size. In light of the 0.5 to 1 cm smaller size of ruptured aneurysms in women,^7,8 a threshold of 4.5 to 5.0 cm for elective repair would seem appropriate. This recommendation is supported by work that found no ruptures in aneurysms less than 5 cm in either men or women, but did note that the fastest growth of aneurysms was noted in the 4.5 to 5.0 cm size.¹⁵ With women’s higher elective mortality, this recommendation should be tested with prospective study.

The combination of rupture rates at smaller sizes, and women’s disproportionate numbers of ruptures versus elective repairs, should prompt more vigorous screening of women for AAA disease. Continued work to manufacture endografts with smaller delivery systems should also greatly benefit women as it will allow more female EVAR with the demonstrated mortality advantage over open surgery.

Editorial Commentary by Frank Criado, MD

This article carries an important message: women are different from men! —especially as it relates to management of AAA. More vigorous screening for the disease, a higher index of suspicion, and use of a 4.5 cm diameter threshold to recommend elective repair — combined — will go a long way in improving the current unacceptable situation facing female AAA patients.