- Email: [email protected]
Age-Stratified Results from 20,095 Aortoiliac Aneurysm Repairs: Should We Approach Octogenarians and Nonagenarians Differently?
Age-Stratified Results from 20,095 Aortoiliac Aneurysm Repairs: Should We Approach Octogenarians and Nonagenarians Differently? Nikolaos Tsilimparis, MD, Sebastian Perez, MSPH, Anand Dayama, MD, Joseph J Ricotta II, MD, MS, FACS In the endovascular era, elderly patients are offered repair of their aortoiliac aneurysms (AAA) more frequently than in the past. Our objective is to compare age groups and draw inferences for AAA repair outcomes. STUDY DESIGN: We identified 20,095 patients who underwent AAA repair between 2005 and 2010 using the American College of Surgeons NSQIP national database. Preoperative characteristics and outcomes were compared among age groups (group A: 0 to 64 years; B: 65 to 79 years; C: 80 to 89 years; and D: 90 years and older). RESULTS: The age distribution of the cohort was A: 17.1%, B: 57.2%, C: 24%, and D: 1.7%. Nonagenarians presented significantly more often as emergencies in comparison with groups A to C (A: 13.8%, B: 10.8%, C: 12.9%, D: 22.1%; p ⬍ 0.001). Endovascular aneurysm repair was performed more frequently in older patients (A: 55.2%, B: 63.7%, C: 74.6%, D: 77.9%; p ⬍ 0.001). Risk of any complication was significantly different among groups, becoming more prevalent with advanced age (A: 22.8%, B: 23.4%, C: 24.7%, D: 27.8%; p ⫽ 0.041). Nonsurgical complications (A: 14.7%, B: 16.4%, C: 18%, D: 19.8%; p ⬍ 0.001) and cardiovascular complications (A: 3.9%, B: 4.5%, C: 5.5%, D: 5.2%; p ⫽ 0.003) were also higher with advanced age. Overall mortality was 3.1%, 4.9%,7.2%, and 13.2% for groups A to D, respectively (p ⬍ 0.001). Mortality after elective AAA repair was significantly higher for open surgery compared with endovascular aneurysm repair in all age groups (open surgery vs endovascular aneurysm repair, A:1.9% vs 0.5%; p ⫽ 0.001; B: 3.9% vs 1.2%; p ⬍ 0.001; C: 7.4% vs 2%; p ⬍ 0.001; D: 18.8% vs 3.8%; p ⫽ 0.004). After adjusting for confounders in the entire cohort, advanced age persisted as an independent factor for postoperative mortality with a higher risk of death of 1.8 (95% CI, 1.3–2.5), 2.7 (95% CI, 1.9–3.8), and 3.3 (95% CI, 1.8–6.1) times for groups B, C, and D, respectively (group A reference). CONCLUSIONS: Advanced age is independently associated with higher risk of death after AAA repair and indication for surgery should be adjusted for different age groups accordingly. Endovascular aneurysm repair should be preferred for octogenarians and nonagenarians with indication to undergo repair of their AAA. (J Am Coll Surg 2012;215:690–701. © 2012 by the American College of Surgeons) BACKGROUND:
Advanced age has been identified as one of the risk factors that negatively affects the outcomes of aortoiliac aneurysm (AAA) repair.1 The Glasgow Aneurysm Score included age as an independent risk factor for death as far back as 1994.2 Since then, several articles have reported outcomes for patients with AAA older than 80 years that included a mortality rate of 7.5% for open surgery (range 0% to 33%) and 4.6% for endovascular aneurysm repair (EVAR; range 0% to 6%).3 In recent years, numerous studies of abdominal aortic aneurysm repair in nonagenarians have been published.4-7 The preferred use of EVAR in the treatment of older patients has been associated with a substantial decrease in mortality
Disclosure Information: Nothing to disclose. Disclaimer: American College of Surgeons NSQIP and the hospitals participating in the American College of Surgeons NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Received February 13, 2012; Revised May 25, 2012; Accepted June 8, 2012. From the Division of Vascular Surgery and Endovascular Therapy, Emory University School of Medicine, Atlanta, GA. Correspondence address: Joseph J Ricotta II, MD, MS, FACS, Division of Vascular Surgery and Endovascular Therapy, Emory University School of Medicine, 1365 Clifton Rd NE, Clinic Building A, Suite 3200, Atlanta, GA 30322. email: [email protected]
© 2012 by the American College of Surgeons Published by Elsevier Inc.
690
ISSN 1072-7515/12/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2012.06.411
Vol. 215, No. 5, November 2012
Tsilimparis et al
and complications in this fragile group of patients, even in the emergent setting.8 The effect of age on AAA repair in a larger cohort of patients outside of centers of excellence, as well as a contemporary detailed report of age-stratified outcomes, are lacking. In this article, we analyze the association of age with outcomes of AAA repair using real-world nationwide data from the American College of Surgeons (ACS) NSQIP database to present the largest cohort of octogenarians and nonagenarians reported to date.
METHODS Data source
ACS NSQIP is a risk-adjusted data collection mechanism that collects and analyzes clinical outcomes data. Participating hospitals use their collected data to develop quality initiatives that improve surgical care and to identify elements in provided health care that can be improved when compared with other institutions. A site’s surgical clinical nurse reviewer captures data on a variety of clinical variables, including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity outcomes for patients undergoing major surgical procedures in both the inpatient and outpatient setting. Data selection
Using this database, we identified patients undergoing open and endovascular AAA repair between 2005 and 2010. Current Procedural Terminology codes that describe open repair and EVAR for AAA were used. For study purposes, 4 clinically relevant subgroups were generated for comparison: group A comprised patients with AAA repair and age 0 to 64 years, group B comprised patients aged 65 to 79 years, group C comprised patients aged 80 to 89 years, and group D comprised patients 90 years and older. Risk factors and end points
All risk factors available in the ACS NSQIP database were compared between groups. The primary end point of the study was analysis of 30-day mortality. Secondary end points included postoperative morbidity, procedure-related complications, and length of hospital stay. Composite end points were created to facilitate a better understanding of the outcomes. Surgical complications (all surgical site infections, wound disruption, bleeding requiring transfusion, graft/ prosthesis/flap failure, peripheral nerve injury), renal complications (progressive renal failure, acute renal failure), pulmonary complications (pneumonia, unplanned intubation, respiratory insufficiency requiring ventilation for 48 hours),
Age-Stratified Results of Aortoiliac Aneurysm Repairs
691
any sepsis (systematic inflammatory response syndrome, sepsis, septic shock), any nonsurgical complication (any complication except surgical complications), cardiovascular complications (pulmonary embolism, stroke/cerebrovascular event, cardiac arrest, myocardial infarction, deep vein thrombosis/ thrombophlebitis) were the complication end points. Statistical analysis
Categorical data were described using absolute numbers and percent in the study cohort. Continuous variables were summarized as mean (⫾SD). Categorical variables were compared using chi-square test or Fisher exact test, and independent 2-sample t-tests were used for continuous variables or the Wilcoxon rank sum test for nonparametric data. Multivariable logistic regression models were used to assess the association between age groups and 30-day postoperative mortality, as well as the association between age groups and any complication, and control for possible confounders. We created 2 models for 30-day mortality with 1 (model A) using age as a categorical variable with 4 groups and another (model B) in which the categorical variable age was replaced by a continuous age variable (years of age). For the creation of the models, we considered all preoperative variables available in the ACS NSQIP database, including comorbidities (eg, diabetes mellitus, congestive heart failure, chronic obstructive pulmonary disease, renal insufficiency, etc), demographic characteristics (ie, sex, race, body mass index, etc), as well as preoperative blood values. Confounders were identified by running regression models with age and one additional preoperative risk factor or demographic variable at a time as predictors and seeing how the results differed from running a logistic model using age alone. A change of ⬎10% between the crude and adjusted odds ratio of the age variable was used as evidence that the covariate was a possible confounder. A final logistic regression model was run using age groups and all confounders found in this way. A 2-sided p value ⬍0.05 was used to define statistical significance. Patients aged older than 90 years were transformed into 90 in the database. Data were analyzed using SPSS Statistics Version 19.0 (SPSS, IBM Inc). Odds ratios are presented with associated 95% CI.
RESULTS Between 2005 and 2010, there were 20,095 procedures for AAA registered in the ACS NSQIP database. The cohort consisted of 3,430 patients younger than 65 years (group A: 17.1%), 11,499 patients aged 65 to 79 years (group B:
692
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
J Am Coll Surg
Table 1. Demographic Characteristics of 20,095 Patients from the American College of Surgeons NSQIP Database Undergoing Aortoiliac Aneurysm Repair between 2005 and 2010 Variable
n Female, n (%) African American, n (%) ASA, n (%) I/II III IV V IDDM, n (%) NIDDM, n (%) Current smoker (within 1 y), n (%) Severe COPD, n (%) Previous PCI, n (%) Previous cardiac surgery, n (%) Hypertension requiring medication, n (%) Rest pain/gangrene, n (%) On dialysis, n (%) CVA, n (%) Stroke with neurological deficit Stroke without neurological deficit Bleeding disorders, n (%) Pack-years of smoking, mean ⫾ SD Body mass index, mean ⫾ SD eGFR, mL/min/1.73 m2, mean ⫾ SD Albumin, g/dL, mean ⫾ SD Hematocrit, %, mean ⫾ SD Emergency, n (%)
0–64 y
65–79 y
Age group 80–89 y
90 Years or older
p Value
Total cohort
3,430 524 (15.3) 200 (7.1)
11,499 2,313 (20.1) 388 (4.1)
4,817 1,236 (25.7) 162 (4.1)
349 103 (29.5) 7 (2.4)
⬍0.001 ⬍0.001
20,095 4,176 (20.8) 757 (4.6)
307 (9) 2,207 (64.3) 828 (24.1) 87 (2.5) 109 (3.2) 372 (10.9) 2,036 (59.4) 536 (15.6) 704 (20.5) 618 (18) 2,619 (76.4) 68 (2) 58 (1.7)
677 (5.9) 7,657 (66.6) 2,842 (24.8) 308 (2.7) 350 (3) 1,389 (12) 4,154 (36.1) 2,470 (21.5) 2,259 (19.6) 2,719 (23.6) 9,199 (80.0) 141 (1.2) 131 (1.1)
237 (4.91) 3,151 (65.4) 1,251 (26) 176 (3.7) 102 (2.1) 450 (9.3) 616 (12.8) 775 (16.1) 826 (17.1) 1,185 (24.6) 3,902 (81) 40 (0.8) 53 (1.1)
15 (4.3) 188 (53.9) 114 (32.7) 30 (8.6) 6 (1.7) 29 (8.3) 15 (4.3) 37 (10.6) 62 (17.8) 70 (20.01) 270 (77.4) 4 (1.1) 2 (0.6)
⬍0.001 ⬍0.001 ⬍0.001 0.001 ⬍0.001 ⬍0.001 0.001 0.033
1,236 (6.2) 13,203 (65.7) 5,035 (25.1) 601 (3) 567 (2.8) 566 (11.3) 6,821 (33.9) 3,818 (19.0) 3,851 (19.2) 4,592 (22.9) 15,990 (79.6) 253 (1.3) 244 (1.2)
119 (3.5) 113 (3.3) 335 (9.8) 41.39 ⫾ 31.9 29.4 ⫾ 6.3 80.4 ⫾ 26.8 3.8 ⫾ 0.6 40.7 ⫾ 5.5 472 (13.8)
567 (4.9) 582 (5.1) 1,208 (10.5) 43.40 ⫾ 36.6 28 ⫾ 5.8 69.8 ⫾ 25.4 3.8 ⫾ 0.6 39.8 ⫾ 5.4 1,245 (10.8)
269 (5.6) 276 (5.7) 557 (11.6) 28.63 ⫾ 35.2 26.3 ⫾ 4.7 61.7 ⫾ 23.1 3.7 ⫾ 0.5 38.1 ⫾ 5.4 619 (12.9)
14 (4.0) 22 (6.3) 47 (13.5) 12.78 ⫾ 23.9 25.1 ⫾ 5.2 58.3 ⫾ 23.3 3.5 ⫾ 0.6 36.5 ⫾ 5.3 77 (22.1)
0.023 ⬍0.001 0.017 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
969 (4.8) 993 (4.9) 2,147 (10.7) 39.16 ⫾ 36 27.8 ⫾ 5.7 69.5 ⫾ 25.8 3.8 ⫾ 0.6 39.5 ⫾ 5.5 2,413 (12.0)
⬍0.001
ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disorder; CVA, cardiovascular accident; eGFR, estimated glomerular filtration rate; IDDM, insulin-dependent diabetes mellitus; NIDDM, non-insulin–dependent diabetes mellitus; PCI, percutaneous coronary intervention.
57.2%), 4,817 patients aged 80 to 89 years (group C: 24%), and 349 patients 90 years of age or older (group D: 1.7%). Demographic data of the cohort are shown in Table 1. The percentage of female patients treated was higher with advanced age group. African Americans represented a higher proportion of patients in group A than in the other groups. As expected, advanced age was associated with higher American Society of Anesthesiologists IV and V classification and relevant comorbidities. Renal function, body mass index, estimated glomerular filtration rate, albumin, and hematocrit decreased considerably with advancing age. Emergent indication for surgery was significantly more frequent in nonagenarians (group A: 13.8%, group B: 10.8%, group C: 12.9%, group D: 22.1%; p ⬍ 0.001). Endovascular aneurysm repair was performed more frequently in elderly patients (Fig. 1). Higher frequency of
EVAR was significantly different among age groups in elective (p ⬍ 0.01) but not emergent cases (p ⫽ 0.136). The decreasing mean operative time as well as the shorter length of postoperative in-hospital stay associated with advancing age were primarily driven by the higher frequency of endovascular procedures in older patients. Age-stratified outcomes Mortality
Advanced age was associated with higher overall 30-day mortality of 13.2% for nonagenarians (group D) compared with 7.2% for octogenarians (group C), and 4.9% and 3.1% for patients 65 to 79 years and younger than 65 years, respectively. The increased risk of death with advanced age persisted when stratified into elective and emergent repair, as well as when comparing endovascular and open repair in both the elective and emergent setting (p ⬍ 0.01 in all cases; Figs. 2A–C; Tables 2, 3).
Vol. 215, No. 5, November 2012
Tsilimparis et al
Figure 1. Distribution of endovascular and open repair according to age group. Red bar, endovascular repair; blue bar, open repair.
Multivariable analysis to investigate the effect of age group on mortality of the entire patient cohort was generated after adjustment for other relevant preoperative factors. Patients 65 to 79 years had a 1.85-fold higher odds of death, and octogenarians and nonagenarians had a 2.65 and 2.28 times higher risk, respectively, compared with patients aged 0 to 64 years (Table 4, model A). Interaction terms were tested in the model for interaction of age and emergency procedure (p ⫽ 0.44), as well as interaction of age group and EVAR (p ⫽ 0.68), and showed no significance. The same model was also run with age as a continuous variable, but it did not yield a significantly better model (⫺2 Log Likelihood of 3,113.58 vs 3,123.28). The model with age as a continuous variable estimated the odds of mortality for older patients to be even greater than that presented in the categorical model (Table 4, model B). Complications
In elective cases, patients of all age groups had a similar risk for any complication (group A: 18.8%, group B: 19.3%, group C: 19.5%, group D: 19.9%; p ⫽ 0.9), although this risk for any complication was higher for elderly patients in emergent cases (group A: 47.9%, group B: 57.3%, group C: 60.6%, group D: 55.8%; p ⬍ 0.001). Elderly elective patients also had higher rates of nonsurgical complications, despite higher rates of endovascular repair (group A: 11%, group B: 12.6%, group C:13.4%, group D: 14%; p ⫽ 0.022), but interestingly had a lower rate of surgical complications (group A: 10.7%, group B: 10.6%, group C: 9.2%, group D: 7.7%; p ⫽ 0.029), although this could be
Age-Stratified Results of Aortoiliac Aneurysm Repairs
693
attributed to endovascular repair and low rate of wound infections. Graft failure was similar among groups in univariate analysis (group A: 0.9%, group B: 0.8%, group C: 0.9%, group D: 0.4%; p ⫽ 0.83). Cardiovascular complications were significantly higher in elective octogenarians and nonagenarians (group A: 2.8%, group B: 3.3%, group C: 4%, group D: 3.7%; p ⫽ 0.029), but risk for renal complications was not significantly different among groups in elective cases (group A: 2.5%, group B: 2.9%, group C: 3.2%, group D: 2.2%; p ⫽ 0.343). Pulmonary complications were fewer in nonagenarians who also underwent endovascular repair more frequently (group A: 6.2%, group B: 7.7%, group C: 6.9%, group D: 4.8%; p ⫽ 0.013). In emergent cases, older patients had a higher incidence of any complication (group A: 47.9%, group B: 57.3%, group C: 60.6%, group D: 55.8%; p ⬍ 0.01) and nonsurgical complications (group A: 37.9%, group B: 47.8%, group C: 49.6%, group D: 40.4%; p ⬍ 0.01). However, as with elective cases, pulmonary complications (group A: 31.1%, group B: 38.9%, group C: 38.1%, group D: 29.9%; p ⫽ 0.013) and postoperative sepsis (group A: 14%, group B: 17.4%, group C:19.9%, group D: 6.5%; p ⫽ 0.005) were lower in the older groups in which EVAR was more frequent. Logistic regression analysis was run to investigate the association between age group and the odds of “any complication” after adjustment for confounders. Patients 65 to 79 years had 1.12-fold higher odds of any complication, and octogenarians and nonagenarians 1.33 and 1.38 times higher odds compared with patients aged 0 to 64 years (Table 4).
DISCUSSION Aortoiliac aneurysms are most commonly a degenerative disease and have been associated with smoking, age, male sex, hypertension, and other risk factors of atherosclerosis.9-11 The growth of the number of octogenarians in the United States is ⬎160,000/year, although data from the United Kingdom and Sweden suggest that the incidence of clinically relevant AAA has decreased during the past decade.12,13 Anjum and colleagues noted that mortality from AAA decreased ⬎2-fold in patients younger than 75 years, and the decrease was only 25% in those older than 75 years.12 Interestingly, patients older than 75 years was the only age subgroup in which elective admissions for AAA increased during this same time period. In the era of epidemiologic and health care changes, the treating physician faces multiple considerations when dealing with the decision to treat an elderly patient: Does the risk of surgery outweigh the risk of rupture? Is the patient’s
694
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Figure 2. (A) Age-stratified mortality in the entire patient cohort; p ⬍ 0.001 between open and endovascular repair in all age groups. (B) Age-stratified mortality in the elective cases; p ⬍ 0.001 between open and endovascular repair in all age groups. (C) Age-stratified mortality in the emergent cases; p ⬍ 0.001 between open and endovascular repair in all age groups. EVAR, endovascular repair; OS, overall survival.
J Am Coll Surg
Vol. 215, No. 5, November 2012
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
695
Figure 2. Continued
life expectancy and postoperative quality of life such to justify the use of considerable resources in a transforming medical system that tries to reduce costs and become more cost effective? Although the second question cannot be answered in this article, we provide relevant evidence that treating physicians can refer to before making clinical decisions on this potentially fragile group of octogenarians and nonagenarians. This articles shows a substantial difference in 30-day postprocedure survival of patients according to their age. Nonagenarians have an overall 30-day risk of death of 13.2% with a 5.5% risk of death in elective procedures. Octogenarians have a 7.2% overall risk of death with 3.1% risk of death in elective cases. Both groups had substantially lower mortality and complications when undergoing EVAR, making the outcomes of endovascular repair acceptable in elderly patients. A trend toward higher frequency of EVAR with advancing age was observed in our dataset (Fig. 1). The test for interaction between age group and endovascular repair, as well as between age group and emergency procedure, was not statistically significant in the regression model A, so that there is no evidence of EVAR having a different effect across all age groups or of emergent procedures across age groups. Postoperatively, cardiovascular events occurred considerably more often in octogenarians and nonagenarians, but
pulmonary and renal complications were not different between the groups. The composite end point of surgical complications did not differ between age groups and specifically no higher rate of graft-related complications was captured in the elderly patients (p ⫽ 0.789). The odds for mortality were higher in nonagenarians as compared with octogenarians, and the odds for any complication were similar in the 2 groups. This can be explained by the fact that “any complication” includes the minor complications that did not affect mortality, and nonagenarians more frequently had severe complications leading to death. The introduction of EVAR for the treatment of AAA has been associated with an increase in treatment of elderly patients. Biancari and colleagues14 reviewed in their metaanalysis 13,419 patients aged 80 years or older and identified a pooled postoperative mortality of 8.6% and an increased risk of renal, cardiac, and pulmonary complications in the open repair group, similar to our findings. Beck and colleagues15 were also able to demonstrate that advanced age is associated with decreased 1-year survival after open repair. Raval and colleagues8 reported results showing a 2.6% overall mortality and 17.4% morbidity for patients older than 80 years compared with 1.4% mortality and 15.4% morbidity for younger patients after elective aneu-
696
Table 2. Operative Parameters and Outcomes of Open and Endovascular Repair in Elective Cases of Aortoiliac Aneurysm Repair Age group 0–64 y
65–79 y
80–89 y
90 Years or older
Overall
Overall
Overall
Overall
Mortality, n (%)
33 (1.1)
216 (2.1)
129 (3.1)
15 (5.5)
OR EVAR
24 (1.9)
132 (3.9)
9 (0.5)
84 (1.2)
Any complication, n (%)
556 (18.8)
Total cohort p Value
62 (7.4)
6 (18.8)
⬍0.001
67 (2)
9 (3.8)
⬍0.001
1,978 (19.3)
817 (19.5)
1,226 (36.7)
326 (38.8)
15 (46.9)
⬍0.001
EVAR
556 (18.8)
1,978 (19.3)
491 (14.6)
39 (16.3)
⬍0.001
95 (5.5)
Surgical complication, n (%)
1,295 (12.6) 852 (25.5) 443 (6.4)
317 (10.7)
561 (13.4) 246 (29.3)
214 (17.4)
649 (19.4)
146 (17.4)
EVAR
103 (6)
438 (6.3)
240 (7.1)
Pulmonary complication, n (%) OR EVAR
184 (6.2) 142 (11.5) 42 (2.4)
Renal complication, n (%)
788 (7.7) 588 (17.6) 200 (2.9)
75 (2.5)
9 (3.8) 135 (3.2)
EVAR
14 (0.8)
87 (1.3)
71 (2.1)
5 (2.1)
⬍0.001
EVAR
34 (2)
Graft failure, n (%)
141 (2) 27 (0.9)
11 (0.9)
30 (0.9)
EVAR
16 (0.9)
57 (0.8) 97 (3.3)
33 (1) 426 (4.2)
1 (0.4)
0.732
174 (4.1) 6 (18.8)
EVAR
25 (1.4)
121 (1.8)
82 (2.4)
5 (2.1) 234 (5.6) 2 (6.3)
0.269
EVAR
89 (5.2)
307 (4.4)
158 (4.7)
9 (3.8)
0.553
2.6 ⫾ 4
2.9 ⫾ 5
0.182
475 (8.7) 233 (1.9)
4.8 ⫾ 7.1
0.565
563 (4.6) 456 (8.4)
4.1 ⫾ 3.7
4.9 ⫾ 7.2
10.33 ⫾ 10
7.9 ⫾ 5.3
⬍0.001
3.4 ⫾ 5
3.6 ⫾ 3.1
⬍0.001
0.017
9.3 ⫾ 9.3 3⫾5 (continued)
J Am Coll Surg
EVAR
5.1 ⫾ 7.1
107 (0.9)
1,019 (5.8)
76 (9)
9.6 ⫾ 9.6
44 (0.8)
11 (4)
291 (8.7)
4.8 ⫾ 6.4
0.829
0.046
87 (7.1)
8 ⫾ 7.7
289 (2.4)
708 (4) ⬍0.001
OR
OR
316 (5.8)
11 (4)
92 (10.9)
Days from operation to discharge, mean ⫾ SD
0.029
151 (0.9) 0.417
305 (9.1)
598 (5.8)
177 (1.4)
0.002
0
72 (5.8)
176 (5.9)
338 (6.2)
1 (0.4)
OR Return to operating room, n (%)
0.343
605 (3.4) 0.002
6 (2.5) 36 (0.9)
3 (0.4)
376 (3.1)
10 (3.7) 4 (12.5)
108 (3.2) 87 (0.8)
OR Sepsis/SIRS postoperative, n (%)
170 (4) 62 (7.4)
899 (16.5) 515 (2.9)
0.077
343 (3.3)
0.013
0.041
1 (3.1)
202 (6)
796 (6.5)
6 (2.2)
64 (7.6)
82 (2.8)
1,015 (18.6)
1,275 (7.2) ⬍0.001
212 (6.3)
48 (3.9)
0.029
13 (4.8)
61 (4.9)
OR
880 (7.2)
0.329
OR Cardiovascular complication, n (%)
1,339 (24.6) 1,811 (10.2)
0.317
4 (12.5)
125 (3.7)
0.022
⬍0.001
15 (6.3) 290 (6.9)
299 (2.9)
1,464 (12)
21 (7.7) 6 (18.8)
165 (19.6)
1,941 (35.6)
2,219 (12.5) ⬍0.001
27 (11.3) 386 (9.2)
OR
0.900
38 (14) 11 (34.4)
315 (9.4) 1,087 (10.6)
169 (1.4)
Age-Stratified Results of Aortoiliac Aneurysm Repairs
EVAR
325 (11) 230 (18.7)
224 (4.1) 3,405 (19.3)
374 (30.3)
OR
⬍0.001
54 (19.9)
OR Nonsurgical complication, n (%)
Overall
393 (2.2)
Tsilimparis et al
Variable
⬍0.001
158 ⫾ 75 157 ⫾ 75 151 ⫾ 72 EVAR
EVAR, endovascular aneurysm repair; OR, open repair; SIRS, systematic inflammatory response syndrome.
162 ⫾ 78
174 ⫾ 85 237 ⫾ 98 191 ⫾ 99 246 ⫾ 105 OR
183 ⫾ 91
222 ⫾ 93
0.5 ⫾ 1.5 0.3 ⫾ 1.5 0.2 ⫾ 1.2 EVAR
Operation time, mean ⫾ SD, min
OR
No. of RBC U transfused intraoperative, mean ⫾ SD
Variable
161 ⫾ 74
168 ⫾ 80 220 ⫾ 102
0.6 ⫾ 1.3
4.4 ⫾ 4.3 1.05 ⫾ 2.5 3.1 ⫾ 4 1.13 ⫾ 2.6 2.7 ⫾ 3.4 1.10 ⫾ 2.3 2.2 ⫾ 2.8
Overall
⬍0.001
1.09 ⫾ 2.4
⬍0.001
p Value Overall
90 Years or older
Overall 80–89 y Overall
Age group 65–79 y 0–64 y
Table 2. Continued
⬍0.001
⬍0.001
236 ⫾ 99
0.4 ⫾ 1.5
181 ⫾ 91
1.1 ⫾ 2.5
Tsilimparis et al
0.577
2.6 ⫾ 3.4
Overall
Total cohort
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
697
rysm repair. Open repair was associated with considerably worse outcomes when compared with EVAR (mortality 1.8% vs 6.1%). Indes and colleagues16 presented similar findings for patients with occlusive disease and, again, endovascular approach was associated with lower mortality. Recent publications on endovascular repair of aortic aneurysms in octogenarians and nonagenarians has engendered enthusiasm with respect to safety and efficacy in the management of these patients.3-5,7,17-19 Most authors claim that despite the higher perioperative risk, EVAR can be performed safely in these patients cohorts. The obvious advantage of EVAR in the management of elderly patients should not compromise thoughtful patient selection of octogenarians and nonagenarians with AAA. Our study clearly demonstrates how these patients are at higher risk for cardiovascular events, but also for any postoperative complications or death, even with use of endovascular techniques, although the selection process of patients older than 80 years who undergo surgery is more meticulous than that for younger patients. Although elective EVAR should be preferred for anatomically suitable patients, we suggest that octogenarians and nonagenarians who are not suitable for EVAR should also be evaluated for open repair. The option of a fenestrated-branched stent graft could also be considered for aneurysms with short infrarenal neck or pararenal aneurysms. Aggressive treatment of AAA with challenging anatomies can be pursued in elderly patients, given that already 30% of patients in the United States are treated outside instructions for use of the devices.20 However, physicians should be aware of the risk of sac enlargement, as demonstrated by Schanzer and colleagues,20 for patients treated outside the instructions for use, especially for patients older than 80 years. Overall, perioperative risk of death and complications should be evaluated before proceeding with any kind of invasive therapy and risk of aneurysm rupture should exceed the estimated risk. We believe that best medical therapy is an acceptable alternative for selected octogenarians or nonagenarians whose aneurysm size has reached the established repair threshold but is associated with anatomic characteristics that could make it a high-risk repair. Appropriate evaluation of elderly patients is essential, and age older than 80 years should be considered a risk factor that, in the presence of another 2 cardiac risk factors, should trigger additional cardiac stress testing as recommended by the guidelines of vascular societies.21,22 Routine preoperative carotid duplex sonography cannot be supported by current available evidence. The cost effectiveness of using EVAR more frequently in elderly patients is difficult to speculate because of the
698
Table 3. Operative Parameters and Outcomes of Open and Endovascular Repair in Emergent Cases of Aortoiliac Aneurysm Repair Age group 0–64 y
65–79 y
80–89 y
90 Years or older Overall
Overall
Mortality, n (%) OR EVAR Any complication, n (%) OR EVAR Nonsurgical complication, n (%) OR EVAR Surgical complication, n (%) OR EVAR Pulmonary complication, n (%) OR EVAR Renal complication, n (%) OR EVAR Cardiovascular complication, n (%) OR EVAR Graft failure, n (%) OR EVAR Sepsis/SIRS postoperative, n (%) OR EVAR Return to operating room, n (%) OR EVAR
73 (15.5)
345 (27.7)
220 (35.5)
31 (40.3)
58 (19.1) 15 (8.9)
264 (31.8) 81 (19.6) 226 (47.9)
170 (55.9) 56 (33.3)
713 (57.3) 530 (63.8) 183 (44.2)
179 (37.9) 130 (42.8) 49 (29.2)
147 (31.1)
56 (11.9)
51 (10.8)
7 (1.5)
97 (15.7)
10 (0.8)
66 (14)
7 (1.1)
217 (17.4) 171 (20.6) 46 (11.1)
96 (20.3)
123 (19.9)
259 (20.8) 188 (22.6) 71 (17.1)
0.303 0.056
0.260
467 (29.9) 196 (23.1)
0.040 0.113
0.013
688 (44) 202 (23.8)
0.621 0.240
0.343
264 (16.9) 83 (9.8)
0.045 0.965
0.106
233 (14.9) 95 (11.2)
0.229 0.866
0.640
12 (0.8) 13 (1.5)
0.006 0.262
0.005
320 (20.5) 91 (10.7)
0.565 0.573
0.266
337 (21.5) 139 (16.4)
1112 (46.1)
663 (27.5)
890 (36.9)
347 (14.4)
328 (13.6)
25 (1)
5 (6.5) 5 (11.1) 0
110 (17.8) 77 (20.1) 33 (14)
827 (52.9) 285 (33.6)
1 (0.8) 0 1 (3.1)
97 (25.3) 26 (11.1)
0.001 ⬍0.001 0.229
1,357 (56.2)
8 (10.4) 5 (11.1) 3 (9.4)
3 (0.8) 4 (1.7)
989 (63.2) 368 (43.3)
10 (13) 9 (20) 1 (3.1)
69 (18) 28 (11.9)
4 (0.5) 6 (1.4)
47 (15.5) 19 (11.3)
95 (15.3)
172 (13.8)
0.016 ⬍0.001 0.010
23 (29.9) 20 (44.4) 3 (9.4)
67 (17.4) 28 (11.9)
127 (15.3) 45 (10.9)
5 (1.6) 2 (1.2)
236 (38.1)
186 (14.9)
514 (32.9) 155 (18.3)
24 (31.2) 18 (40) 6 (18.8)
181 (47.1) 55 (23.4)
144 (17.3) 42 (10.1)
32 (10.5) 19 (11.3)
174 (28.1)
484 (38.9)
⬍0.001 ⬍0.001 0.005
31 (40.3) 24 (53.3) 7 (21.9)
108 (28.1) 66 (28.1)
375 (45.1) 109 (26.3)
44 (14.5) 12 (7.1)
307 (4.6)
352 (28.3)
Overall
669 (27.7)
43 (55.8) 31 (68.9) 12 (37.5)
223 (58.1) 84 (35.7)
256 (30.8) 96 (23.2)
112 (36.8) 35 (20.8)
375 (60.6)
595 (47.8)
113 (23.9)
24 (53.3) 7 (21.9)
258 (67.2) 117 (49.8)
450 (54.2) 145 (35)
85 (28) 28 (16.7)
65 (2.4) 31 (18.5)
168 (43.8) 52 (22.1)
Total cohort p Value
411 (17)
11 (14.3) 7 (15.6) 4 (12.5)
476 (19.7)
(continued)
J Am Coll Surg
Overall
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Overall
Tsilimparis et al
Variable
EVAR, endovascular aneurysm repair; OR, open repair; SIRS, systematic inflammatory response syndrome.
0.0114 0.135 0.064
211.3 ⫾ 101 197 ⫾ 99 170.4 ⫾ 89.1
7.1 ⫾ 7.8 0.005 0.005 0.229
8.9 ⫾ 8.3 3.3 ⫾ 5.3
11.6 ⫾ 14 0.359 0.846 0.380
13.3 ⫾ 15.3 8.7 ⫾ 11.7
Overall
Tsilimparis et al
Days from operation to discharge, mean ⫾ SD 11.1 ⫾ 14.6 12.1 ⫾ 14.6 11.4 ⫾ 13.6 9.8 ⫾ 12.5 OR 12.9 ⫾ 16.1 13.6 ⫾ 15.4 13.0 ⫾ 14.5 12.6 ⫾ 15.5 EVAR 7.9 ⫾ 10.7 9.1 ⫾ 12.4 8.79 ⫾ 11.63 5.98 ⫾ 3.90 No. of RBC U transfused intraoperative, mean ⫾ SD 5.8 ⫾ 6.6 OR 7.4 ⫾ 6.9 9.3 ⫾ 8.6 7.5 ⫾ 8.3 9.2 ⫾ 8.0 7.2 ⫾ 7.7 11.3 ⫾ 9.3 7.4 ⫾ 8.5 EVAR 2.4 ⫾ 4.3 3.6 ⫾ 5.7 3.5 ⫾ 5.3 2.8 ⫾ 4.3 Operation time, mean ⫾ SD, min 195.9 ⫾ 94.6 OR 218.2 ⫾ 99.3 213.5 ⫾ 102.7 201.2 ⫾ 101.9 201.3 ⫾ 100.1 190.1 ⫾ 98.1 209.5 ⫾ 83.4 188.2 ⫾ 78.2 EVAR 155.6 ⫾ 69.5 176.6 ⫾ 95.6 171.7 ⫾ 92.2 158.2 ⫾ 59.3
Overall Variable
Table 3. Continued
0–64 y
Overall
65–79 y
Overall
Age group
80–89 y
Overall
90 Years or older
p Value
Total cohort
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
699
higher perioperative mortality of open repair in this patient population and their reduced life expectancy, so that data from larger studies suggesting lack of cost effectiveness cannot be directly applied.23 Several studies have confirmed that although the operative mortality of ruptured abdominal aortic aneurysm remains high in elderly patients, the long-term survival of patients who survive the procedure is very satisfactory.24,25 The survival of patients older than 80 years who survived the early postoperative period (90 days) at 1 and 5 years was 99% and 94%, respectively, in the Swedish Vascular Registry.26 Thirty-day survival of octogenarians and nonagenarians in our study was 64.5% and 59.7%, respectively, in emergent cases, which is similar or better than what has been reported previously.24,25 EVAR was associated with improved survival and outcomes in the emergent setting as well, although this could be attributed to a selection bias in favor of patients that were hemodynamically stable and anatomically suitable for EVAR (30-day survival after EVAR in group C: 77.9% and group D: 78.1% vs C: 56.2% and group D: 46.7% after open repair). The survival rates of octogenarians and nonagenarians in our study suggest that age alone should not be a reason to abandon therapy of a patient with AAA presenting as an emergency. The validity of the data of the ACS NSQIP database is much greater when compared with those of other administrative databases.27-29 Dedicated, trained study nurses capture the data in each participating institution. Participation in ACS NSQIP is at a voluntary setting, but procedures in each institution are captured randomly and no selection bias is present. Limitations
This is an observational study from a registry database. Limitations of the study mainly involve missing variables that are not captured in the ACS NSQIP database, such as aneurysm-specific characteristics, as well as the fact that only patients who underwent therapy are included. The ACS NSQIP database was created to report data collected from surgical patients in different specialties to enhance quality of care by providing data for comparison. Limitations due to the database design are expected when a specific surgical entity and issue is addressed. In addition, data beyond 30 days are not available, which is a major limitation of any study looking into outcomes of aneurysmal repair, especially in very elderly patients. Results on longterm survival would be helpful in determining to what extent the most elderly patients benefit from elective repair. The final logistic regression models for mortality included 9,452 patients. The missing data were due to missing laboratory values. The models produced similar conclusions when they were run without the laboratory values (ie, al-
700
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
J Am Coll Surg
Table 4. Logistic Regression Analyses for Mortality and Any Complication for all Patients with No Missing Variables
Variable
Regression model for mortality Model A Model B Odds ratio 95% CI Odds ratio 95% CI
Regression model for any complication Odds ratio 95% CI
Age, y, continuous 0–64 65–79 80–89 90 years or older Endovascular repair Emergency ASA I and II class ASA III class ASA IV class ASA V class eGFR Preoperative albumin Preoperative white blood cell count Preoperative hematocrit Preoperative total bilirubin
— Reference 1.85 2.65 3.28 0.41 2.90 Reference 1.14 2.71 9.16 0.99 0.62 1.03 0.96 1.15
— Reference 1.12 1.33 1.38 0.26 3.62 — — — — 0.99 — — — —
— Reference 1.34–2.54 1.88–3.75 1.77–6.06 0.332–0.50 2.24–3.75 Reference 0.52–2.47 1.25–5.89 4.05–20.70 0.99–0.99 0.52–0.73 1.01–1.05 0.95–0.98 0.99–1.34
1.04 — — — — 0.40 2.88 Reference 1.12 2.66 8.89 0.99 0.61 1.03 0.97 1.15
1.03–1.05 — — — — 0.33–0.50 2.22–3.73 Reference 0.51–2.42 1.22–5.78 3.99–20.10 0.99–0.99 0.52–0.73 1.01–1.05 0.95–0.98 0.98–1.35
— Reference 1.01–1.24 1.18–1.49 1.04–1.84 0.24–0.28 3.28–3.99 — — — — 0.99–0.99 — — — —
For mortality Model A adjusts for age groups and all relevant confounders, and in model B age is presented as a continuous variable with the same confounders. ASA, American Society of Anesthesiologists; eEGF, estimated glomerular filtration rate.
bumin, white blood count, hematocrit, and bilirubin) leading to exclusion of ⬍1,000 patients (data not shown). Data were collected from ⬎400 hospitals in the United States so that a uniform pattern of patient management, size threshold for aneurysm repair, or indication for surgery cannot be identified.
Analysis and interpretation of data: Tsilimparis, Perez, Dayama, Ricotta Drafting of manuscript: Tsilimparis, Dayama, Perez, Ricotta Critical revision: Tsilimparis, Dayama, Perez, Ricotta
CONCLUSIONS The present study is the largest one to date to report agestratified 30-day outcomes of AAA repair. Contrary to what has been frequently described in previous singlecenter studies, the risk of death and major complications in elderly patients is considerable. These data should be considered when treating octogenarians and nonagenarians with AAA. The estimated risk of rupture should outweigh the risk of perioperative death and major complications as described here in for patients receiving elective therapy. Advanced age is independently associated with higher risk of death after AAA repair and indication for surgery should be accordingly adjusted for different age groups. Endovascular aneurysm repair should be preferred for octogenarians and nonagenarians with indication to undergo repair of their AAA.
REFERENCES
Author Contributions
Study conception and design: Tsilimparis, Ricotta Acquisition of data: Tsilimparis, Dayama, Perez
1. Norman PE, Semmens JB, Lawrence-Brown MM, Holman CD. Long term relative survival after surgery for abdominal aortic aneurysm in western australia: population based study. BMJ 1998;317:852–856. 2. Samy AK, Murray G, MacBain G. Glasgow aneurysm score. Cardiovasc Surg 1994;2:41–44. 3. Henebiens M, Vahl A, Koelemay MJ. Elective surgery of abdominal aortic aneurysms in octogenarians: a systematic review. J Vasc Surg 2008;47:676–681. 4. Geisbusch P, Katzen BT, Tsoukas AI, et al. Endovascular repair of infrarenal aortic aneurysms in octogenarians and nonagenarians. J Vasc Surg 2011;54:1605–1613. 5. Roberts WC, Ko JM, Pearl GJ. Abdominal aortic aneurysm in nonagenarians. Am J Geriatr Cardiol 2006;15:319–321. 6. Bouwmeester MF, van Sterkenburg SM, Zeebregts CJ, et al. Endovascular abdominal aortic aneurysm repair in nonagenarians— beyond limits? Tex Heart Inst J 2009;36:345–348. 7. Prenner SB, Turnbull IC, Serrao GW, et al. Outcome of elective endovascular abdominal aortic aneurysm repair in nonagenarians. J Vasc Surg 2011;54:287–294. 8. Raval MV, Eskandari MK. Outcomes of elective abdominal aortic aneurysm repair among the elderly: endovascular versus open repair. Surgery 2011:245–260.
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Tsilimparis et al
9. Forsdahl SH, Singh K, Solberg S, Jacobsen BK. Risk factors for abdominal aortic aneurysms: a 7-year prospective study: the Tromso study, 1994–2001. Circulation 2009;119:2202–2208. 10. Grant SW, Grayson AD, Purkayastha D, et al. Logistic risk model for mortality following elective abdominal aortic aneurysm repair. Br J Surg 2011;98:652–658. 11. Giles KA, Pomposelli F, Hamdan A, et al. Decrease in total aneurysm-related deaths in the era of endovascular aneurysm repair. J Vasc Surg 2009;49:543–550; discussion 550–541. 12. Anjum A, Powell JT. Is the incidence of abdominal aortic aneurysm declining in the 21st century? Mortality and hospital admissions for England & Wales and Scotland. Eur J Vasc Endovasc Surg 2012;43:161–166. 13. Svensjo S, Bjorck M, Gurtelschmid M, et al. Low prevalence of abdominal aortic aneurysm among 65-year-old Swedish men indicates a change in the epidemiology of the disease. Circulation 2011;124:1118–1123. 14. Biancari F, Catania A, D’Andrea V. Elective endovascular vs open repair for abdominal aortic aneurysm in patients aged 80 years and older: systematic review and meta-analysis. Eur J Vasc Endovasc Surg 2011;42:571–576. 15. Beck AW, Goodney PP, Nolan BW, et al. Predicting 1-year mortality after elective abdominal aortic aneurysm repair. J Vasc Surg 2009;49:838–843; discussion 843–834. 16. Indes JE, Tuggle CT, Mandawat A, Sosa JA. Age-stratified outcomes in elderly patients undergoing open and endovascular procedures for aortoiliac occlusive disease. Surgery 2010;148: 420–428. 17. Baril DT, Palchik E, Carroccio A, et al. Experience with endovascular abdominal aortic aneurysm repair in nonagenarians. J Endovasc Ther 2006;13:330–337. 18. Paolini D, Chahwan S, Wojnarowski D, et al. Elective endovascular and open repair of abdominal aortic aneurysms in octogenarians. J Vasc Surg 2008;47:924–927. 19. Ju MH, Keldahl ML, Pearce WH, et al. Stepwise age-related outcomes of elective endovascular abdominal aortic aneurysm
20. 21. 22.
23. 24. 25.
26. 27.
28.
29.
701
repair: 11-year institutional review. Perspect Vasc Surg Endovasc Ther 2011;23:280–290. Schanzer A, Greenberg RK, Hevelone N, et al. Predictors of abdominal aortic aneurysm sac enlargement after endovascular repair. Circulation 2011;123:2848–2855. Chaikof EL, Brewster DC, Dalman RL, et al. The care of patients with an abdominal aortic aneurysm: the Society for Vascular Surgery practice guidelines. J Vasc Surg 2009;50:S2–S49. Moll FL, Powell JT, Fraedrich G, et al. Management of abdominal aortic aneurysms clinical practice guidelines of the European Society for Vascular Surgery. Eur J Vasc Endovasc Surg 2011;41[Suppl 1]:S1–S58. Brown L, Powell J, Thompson S, et al. The UK Endovascular Aneurysm Repair (EVAR) trials: randomised trials of EVAR versus standard therapy. Health Technol Assess 2012;16:1–218. Opfermann P, von Allmen R, Diehm N, et al. Repair of ruptured abdominal aortic aneurysm in octogenarians. Eur J Vasc Endovasc Surg 2011;42:475–483. Biancari F, Mazziotti MA, Paone R, et al. Outcome after open repair of ruptured abdominal aortic aneurysm in patients ⬎80 years old: a systematic review and meta-analysis. World J Surg 2011;35:1662–1670. Mani K, Bjorck M, Lundkvist J, Wanhainen A. Improved longterm survival after abdominal aortic aneurysm repair. Circulation 2009;120:201–211. Shiloach M, Frencher SK Jr, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16. Hall BL, Hamilton BH, Richards K, et al. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: an evaluation of all participating hospitals. Ann Surg 2009;250:363–376. Davis CL, Pierce JR, Henderson W, et al. Assessment of the reliability of data collected for the Department of Veterans Affairs National Surgical Quality Improvement Program. J Am Coll Surg 2007;204:550–560.
Advanced age has been identified as one of the risk factors that negatively affects the outcomes of aortoiliac aneurysm (AAA) repair.1 The Glasgow Aneurysm Score included age as an independent risk factor for death as far back as 1994.2 Since then, several articles have reported outcomes for patients with AAA older than 80 years that included a mortality rate of 7.5% for open surgery (range 0% to 33%) and 4.6% for endovascular aneurysm repair (EVAR; range 0% to 6%).3 In recent years, numerous studies of abdominal aortic aneurysm repair in nonagenarians have been published.4-7 The preferred use of EVAR in the treatment of older patients has been associated with a substantial decrease in mortality
Disclosure Information: Nothing to disclose. Disclaimer: American College of Surgeons NSQIP and the hospitals participating in the American College of Surgeons NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. Received February 13, 2012; Revised May 25, 2012; Accepted June 8, 2012. From the Division of Vascular Surgery and Endovascular Therapy, Emory University School of Medicine, Atlanta, GA. Correspondence address: Joseph J Ricotta II, MD, MS, FACS, Division of Vascular Surgery and Endovascular Therapy, Emory University School of Medicine, 1365 Clifton Rd NE, Clinic Building A, Suite 3200, Atlanta, GA 30322. email: [email protected]
© 2012 by the American College of Surgeons Published by Elsevier Inc.
690
ISSN 1072-7515/12/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2012.06.411
Vol. 215, No. 5, November 2012
Tsilimparis et al
and complications in this fragile group of patients, even in the emergent setting.8 The effect of age on AAA repair in a larger cohort of patients outside of centers of excellence, as well as a contemporary detailed report of age-stratified outcomes, are lacking. In this article, we analyze the association of age with outcomes of AAA repair using real-world nationwide data from the American College of Surgeons (ACS) NSQIP database to present the largest cohort of octogenarians and nonagenarians reported to date.
METHODS Data source
ACS NSQIP is a risk-adjusted data collection mechanism that collects and analyzes clinical outcomes data. Participating hospitals use their collected data to develop quality initiatives that improve surgical care and to identify elements in provided health care that can be improved when compared with other institutions. A site’s surgical clinical nurse reviewer captures data on a variety of clinical variables, including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity outcomes for patients undergoing major surgical procedures in both the inpatient and outpatient setting. Data selection
Using this database, we identified patients undergoing open and endovascular AAA repair between 2005 and 2010. Current Procedural Terminology codes that describe open repair and EVAR for AAA were used. For study purposes, 4 clinically relevant subgroups were generated for comparison: group A comprised patients with AAA repair and age 0 to 64 years, group B comprised patients aged 65 to 79 years, group C comprised patients aged 80 to 89 years, and group D comprised patients 90 years and older. Risk factors and end points
All risk factors available in the ACS NSQIP database were compared between groups. The primary end point of the study was analysis of 30-day mortality. Secondary end points included postoperative morbidity, procedure-related complications, and length of hospital stay. Composite end points were created to facilitate a better understanding of the outcomes. Surgical complications (all surgical site infections, wound disruption, bleeding requiring transfusion, graft/ prosthesis/flap failure, peripheral nerve injury), renal complications (progressive renal failure, acute renal failure), pulmonary complications (pneumonia, unplanned intubation, respiratory insufficiency requiring ventilation for 48 hours),
Age-Stratified Results of Aortoiliac Aneurysm Repairs
691
any sepsis (systematic inflammatory response syndrome, sepsis, septic shock), any nonsurgical complication (any complication except surgical complications), cardiovascular complications (pulmonary embolism, stroke/cerebrovascular event, cardiac arrest, myocardial infarction, deep vein thrombosis/ thrombophlebitis) were the complication end points. Statistical analysis
Categorical data were described using absolute numbers and percent in the study cohort. Continuous variables were summarized as mean (⫾SD). Categorical variables were compared using chi-square test or Fisher exact test, and independent 2-sample t-tests were used for continuous variables or the Wilcoxon rank sum test for nonparametric data. Multivariable logistic regression models were used to assess the association between age groups and 30-day postoperative mortality, as well as the association between age groups and any complication, and control for possible confounders. We created 2 models for 30-day mortality with 1 (model A) using age as a categorical variable with 4 groups and another (model B) in which the categorical variable age was replaced by a continuous age variable (years of age). For the creation of the models, we considered all preoperative variables available in the ACS NSQIP database, including comorbidities (eg, diabetes mellitus, congestive heart failure, chronic obstructive pulmonary disease, renal insufficiency, etc), demographic characteristics (ie, sex, race, body mass index, etc), as well as preoperative blood values. Confounders were identified by running regression models with age and one additional preoperative risk factor or demographic variable at a time as predictors and seeing how the results differed from running a logistic model using age alone. A change of ⬎10% between the crude and adjusted odds ratio of the age variable was used as evidence that the covariate was a possible confounder. A final logistic regression model was run using age groups and all confounders found in this way. A 2-sided p value ⬍0.05 was used to define statistical significance. Patients aged older than 90 years were transformed into 90 in the database. Data were analyzed using SPSS Statistics Version 19.0 (SPSS, IBM Inc). Odds ratios are presented with associated 95% CI.
RESULTS Between 2005 and 2010, there were 20,095 procedures for AAA registered in the ACS NSQIP database. The cohort consisted of 3,430 patients younger than 65 years (group A: 17.1%), 11,499 patients aged 65 to 79 years (group B:
692
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
J Am Coll Surg
Table 1. Demographic Characteristics of 20,095 Patients from the American College of Surgeons NSQIP Database Undergoing Aortoiliac Aneurysm Repair between 2005 and 2010 Variable
n Female, n (%) African American, n (%) ASA, n (%) I/II III IV V IDDM, n (%) NIDDM, n (%) Current smoker (within 1 y), n (%) Severe COPD, n (%) Previous PCI, n (%) Previous cardiac surgery, n (%) Hypertension requiring medication, n (%) Rest pain/gangrene, n (%) On dialysis, n (%) CVA, n (%) Stroke with neurological deficit Stroke without neurological deficit Bleeding disorders, n (%) Pack-years of smoking, mean ⫾ SD Body mass index, mean ⫾ SD eGFR, mL/min/1.73 m2, mean ⫾ SD Albumin, g/dL, mean ⫾ SD Hematocrit, %, mean ⫾ SD Emergency, n (%)
0–64 y
65–79 y
Age group 80–89 y
90 Years or older
p Value
Total cohort
3,430 524 (15.3) 200 (7.1)
11,499 2,313 (20.1) 388 (4.1)
4,817 1,236 (25.7) 162 (4.1)
349 103 (29.5) 7 (2.4)
⬍0.001 ⬍0.001
20,095 4,176 (20.8) 757 (4.6)
307 (9) 2,207 (64.3) 828 (24.1) 87 (2.5) 109 (3.2) 372 (10.9) 2,036 (59.4) 536 (15.6) 704 (20.5) 618 (18) 2,619 (76.4) 68 (2) 58 (1.7)
677 (5.9) 7,657 (66.6) 2,842 (24.8) 308 (2.7) 350 (3) 1,389 (12) 4,154 (36.1) 2,470 (21.5) 2,259 (19.6) 2,719 (23.6) 9,199 (80.0) 141 (1.2) 131 (1.1)
237 (4.91) 3,151 (65.4) 1,251 (26) 176 (3.7) 102 (2.1) 450 (9.3) 616 (12.8) 775 (16.1) 826 (17.1) 1,185 (24.6) 3,902 (81) 40 (0.8) 53 (1.1)
15 (4.3) 188 (53.9) 114 (32.7) 30 (8.6) 6 (1.7) 29 (8.3) 15 (4.3) 37 (10.6) 62 (17.8) 70 (20.01) 270 (77.4) 4 (1.1) 2 (0.6)
⬍0.001 ⬍0.001 ⬍0.001 0.001 ⬍0.001 ⬍0.001 0.001 0.033
1,236 (6.2) 13,203 (65.7) 5,035 (25.1) 601 (3) 567 (2.8) 566 (11.3) 6,821 (33.9) 3,818 (19.0) 3,851 (19.2) 4,592 (22.9) 15,990 (79.6) 253 (1.3) 244 (1.2)
119 (3.5) 113 (3.3) 335 (9.8) 41.39 ⫾ 31.9 29.4 ⫾ 6.3 80.4 ⫾ 26.8 3.8 ⫾ 0.6 40.7 ⫾ 5.5 472 (13.8)
567 (4.9) 582 (5.1) 1,208 (10.5) 43.40 ⫾ 36.6 28 ⫾ 5.8 69.8 ⫾ 25.4 3.8 ⫾ 0.6 39.8 ⫾ 5.4 1,245 (10.8)
269 (5.6) 276 (5.7) 557 (11.6) 28.63 ⫾ 35.2 26.3 ⫾ 4.7 61.7 ⫾ 23.1 3.7 ⫾ 0.5 38.1 ⫾ 5.4 619 (12.9)
14 (4.0) 22 (6.3) 47 (13.5) 12.78 ⫾ 23.9 25.1 ⫾ 5.2 58.3 ⫾ 23.3 3.5 ⫾ 0.6 36.5 ⫾ 5.3 77 (22.1)
0.023 ⬍0.001 0.017 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
969 (4.8) 993 (4.9) 2,147 (10.7) 39.16 ⫾ 36 27.8 ⫾ 5.7 69.5 ⫾ 25.8 3.8 ⫾ 0.6 39.5 ⫾ 5.5 2,413 (12.0)
⬍0.001
ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disorder; CVA, cardiovascular accident; eGFR, estimated glomerular filtration rate; IDDM, insulin-dependent diabetes mellitus; NIDDM, non-insulin–dependent diabetes mellitus; PCI, percutaneous coronary intervention.
57.2%), 4,817 patients aged 80 to 89 years (group C: 24%), and 349 patients 90 years of age or older (group D: 1.7%). Demographic data of the cohort are shown in Table 1. The percentage of female patients treated was higher with advanced age group. African Americans represented a higher proportion of patients in group A than in the other groups. As expected, advanced age was associated with higher American Society of Anesthesiologists IV and V classification and relevant comorbidities. Renal function, body mass index, estimated glomerular filtration rate, albumin, and hematocrit decreased considerably with advancing age. Emergent indication for surgery was significantly more frequent in nonagenarians (group A: 13.8%, group B: 10.8%, group C: 12.9%, group D: 22.1%; p ⬍ 0.001). Endovascular aneurysm repair was performed more frequently in elderly patients (Fig. 1). Higher frequency of
EVAR was significantly different among age groups in elective (p ⬍ 0.01) but not emergent cases (p ⫽ 0.136). The decreasing mean operative time as well as the shorter length of postoperative in-hospital stay associated with advancing age were primarily driven by the higher frequency of endovascular procedures in older patients. Age-stratified outcomes Mortality
Advanced age was associated with higher overall 30-day mortality of 13.2% for nonagenarians (group D) compared with 7.2% for octogenarians (group C), and 4.9% and 3.1% for patients 65 to 79 years and younger than 65 years, respectively. The increased risk of death with advanced age persisted when stratified into elective and emergent repair, as well as when comparing endovascular and open repair in both the elective and emergent setting (p ⬍ 0.01 in all cases; Figs. 2A–C; Tables 2, 3).
Vol. 215, No. 5, November 2012
Tsilimparis et al
Figure 1. Distribution of endovascular and open repair according to age group. Red bar, endovascular repair; blue bar, open repair.
Multivariable analysis to investigate the effect of age group on mortality of the entire patient cohort was generated after adjustment for other relevant preoperative factors. Patients 65 to 79 years had a 1.85-fold higher odds of death, and octogenarians and nonagenarians had a 2.65 and 2.28 times higher risk, respectively, compared with patients aged 0 to 64 years (Table 4, model A). Interaction terms were tested in the model for interaction of age and emergency procedure (p ⫽ 0.44), as well as interaction of age group and EVAR (p ⫽ 0.68), and showed no significance. The same model was also run with age as a continuous variable, but it did not yield a significantly better model (⫺2 Log Likelihood of 3,113.58 vs 3,123.28). The model with age as a continuous variable estimated the odds of mortality for older patients to be even greater than that presented in the categorical model (Table 4, model B). Complications
In elective cases, patients of all age groups had a similar risk for any complication (group A: 18.8%, group B: 19.3%, group C: 19.5%, group D: 19.9%; p ⫽ 0.9), although this risk for any complication was higher for elderly patients in emergent cases (group A: 47.9%, group B: 57.3%, group C: 60.6%, group D: 55.8%; p ⬍ 0.001). Elderly elective patients also had higher rates of nonsurgical complications, despite higher rates of endovascular repair (group A: 11%, group B: 12.6%, group C:13.4%, group D: 14%; p ⫽ 0.022), but interestingly had a lower rate of surgical complications (group A: 10.7%, group B: 10.6%, group C: 9.2%, group D: 7.7%; p ⫽ 0.029), although this could be
Age-Stratified Results of Aortoiliac Aneurysm Repairs
693
attributed to endovascular repair and low rate of wound infections. Graft failure was similar among groups in univariate analysis (group A: 0.9%, group B: 0.8%, group C: 0.9%, group D: 0.4%; p ⫽ 0.83). Cardiovascular complications were significantly higher in elective octogenarians and nonagenarians (group A: 2.8%, group B: 3.3%, group C: 4%, group D: 3.7%; p ⫽ 0.029), but risk for renal complications was not significantly different among groups in elective cases (group A: 2.5%, group B: 2.9%, group C: 3.2%, group D: 2.2%; p ⫽ 0.343). Pulmonary complications were fewer in nonagenarians who also underwent endovascular repair more frequently (group A: 6.2%, group B: 7.7%, group C: 6.9%, group D: 4.8%; p ⫽ 0.013). In emergent cases, older patients had a higher incidence of any complication (group A: 47.9%, group B: 57.3%, group C: 60.6%, group D: 55.8%; p ⬍ 0.01) and nonsurgical complications (group A: 37.9%, group B: 47.8%, group C: 49.6%, group D: 40.4%; p ⬍ 0.01). However, as with elective cases, pulmonary complications (group A: 31.1%, group B: 38.9%, group C: 38.1%, group D: 29.9%; p ⫽ 0.013) and postoperative sepsis (group A: 14%, group B: 17.4%, group C:19.9%, group D: 6.5%; p ⫽ 0.005) were lower in the older groups in which EVAR was more frequent. Logistic regression analysis was run to investigate the association between age group and the odds of “any complication” after adjustment for confounders. Patients 65 to 79 years had 1.12-fold higher odds of any complication, and octogenarians and nonagenarians 1.33 and 1.38 times higher odds compared with patients aged 0 to 64 years (Table 4).
DISCUSSION Aortoiliac aneurysms are most commonly a degenerative disease and have been associated with smoking, age, male sex, hypertension, and other risk factors of atherosclerosis.9-11 The growth of the number of octogenarians in the United States is ⬎160,000/year, although data from the United Kingdom and Sweden suggest that the incidence of clinically relevant AAA has decreased during the past decade.12,13 Anjum and colleagues noted that mortality from AAA decreased ⬎2-fold in patients younger than 75 years, and the decrease was only 25% in those older than 75 years.12 Interestingly, patients older than 75 years was the only age subgroup in which elective admissions for AAA increased during this same time period. In the era of epidemiologic and health care changes, the treating physician faces multiple considerations when dealing with the decision to treat an elderly patient: Does the risk of surgery outweigh the risk of rupture? Is the patient’s
694
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Figure 2. (A) Age-stratified mortality in the entire patient cohort; p ⬍ 0.001 between open and endovascular repair in all age groups. (B) Age-stratified mortality in the elective cases; p ⬍ 0.001 between open and endovascular repair in all age groups. (C) Age-stratified mortality in the emergent cases; p ⬍ 0.001 between open and endovascular repair in all age groups. EVAR, endovascular repair; OS, overall survival.
J Am Coll Surg
Vol. 215, No. 5, November 2012
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
695
Figure 2. Continued
life expectancy and postoperative quality of life such to justify the use of considerable resources in a transforming medical system that tries to reduce costs and become more cost effective? Although the second question cannot be answered in this article, we provide relevant evidence that treating physicians can refer to before making clinical decisions on this potentially fragile group of octogenarians and nonagenarians. This articles shows a substantial difference in 30-day postprocedure survival of patients according to their age. Nonagenarians have an overall 30-day risk of death of 13.2% with a 5.5% risk of death in elective procedures. Octogenarians have a 7.2% overall risk of death with 3.1% risk of death in elective cases. Both groups had substantially lower mortality and complications when undergoing EVAR, making the outcomes of endovascular repair acceptable in elderly patients. A trend toward higher frequency of EVAR with advancing age was observed in our dataset (Fig. 1). The test for interaction between age group and endovascular repair, as well as between age group and emergency procedure, was not statistically significant in the regression model A, so that there is no evidence of EVAR having a different effect across all age groups or of emergent procedures across age groups. Postoperatively, cardiovascular events occurred considerably more often in octogenarians and nonagenarians, but
pulmonary and renal complications were not different between the groups. The composite end point of surgical complications did not differ between age groups and specifically no higher rate of graft-related complications was captured in the elderly patients (p ⫽ 0.789). The odds for mortality were higher in nonagenarians as compared with octogenarians, and the odds for any complication were similar in the 2 groups. This can be explained by the fact that “any complication” includes the minor complications that did not affect mortality, and nonagenarians more frequently had severe complications leading to death. The introduction of EVAR for the treatment of AAA has been associated with an increase in treatment of elderly patients. Biancari and colleagues14 reviewed in their metaanalysis 13,419 patients aged 80 years or older and identified a pooled postoperative mortality of 8.6% and an increased risk of renal, cardiac, and pulmonary complications in the open repair group, similar to our findings. Beck and colleagues15 were also able to demonstrate that advanced age is associated with decreased 1-year survival after open repair. Raval and colleagues8 reported results showing a 2.6% overall mortality and 17.4% morbidity for patients older than 80 years compared with 1.4% mortality and 15.4% morbidity for younger patients after elective aneu-
696
Table 2. Operative Parameters and Outcomes of Open and Endovascular Repair in Elective Cases of Aortoiliac Aneurysm Repair Age group 0–64 y
65–79 y
80–89 y
90 Years or older
Overall
Overall
Overall
Overall
Mortality, n (%)
33 (1.1)
216 (2.1)
129 (3.1)
15 (5.5)
OR EVAR
24 (1.9)
132 (3.9)
9 (0.5)
84 (1.2)
Any complication, n (%)
556 (18.8)
Total cohort p Value
62 (7.4)
6 (18.8)
⬍0.001
67 (2)
9 (3.8)
⬍0.001
1,978 (19.3)
817 (19.5)
1,226 (36.7)
326 (38.8)
15 (46.9)
⬍0.001
EVAR
556 (18.8)
1,978 (19.3)
491 (14.6)
39 (16.3)
⬍0.001
95 (5.5)
Surgical complication, n (%)
1,295 (12.6) 852 (25.5) 443 (6.4)
317 (10.7)
561 (13.4) 246 (29.3)
214 (17.4)
649 (19.4)
146 (17.4)
EVAR
103 (6)
438 (6.3)
240 (7.1)
Pulmonary complication, n (%) OR EVAR
184 (6.2) 142 (11.5) 42 (2.4)
Renal complication, n (%)
788 (7.7) 588 (17.6) 200 (2.9)
75 (2.5)
9 (3.8) 135 (3.2)
EVAR
14 (0.8)
87 (1.3)
71 (2.1)
5 (2.1)
⬍0.001
EVAR
34 (2)
Graft failure, n (%)
141 (2) 27 (0.9)
11 (0.9)
30 (0.9)
EVAR
16 (0.9)
57 (0.8) 97 (3.3)
33 (1) 426 (4.2)
1 (0.4)
0.732
174 (4.1) 6 (18.8)
EVAR
25 (1.4)
121 (1.8)
82 (2.4)
5 (2.1) 234 (5.6) 2 (6.3)
0.269
EVAR
89 (5.2)
307 (4.4)
158 (4.7)
9 (3.8)
0.553
2.6 ⫾ 4
2.9 ⫾ 5
0.182
475 (8.7) 233 (1.9)
4.8 ⫾ 7.1
0.565
563 (4.6) 456 (8.4)
4.1 ⫾ 3.7
4.9 ⫾ 7.2
10.33 ⫾ 10
7.9 ⫾ 5.3
⬍0.001
3.4 ⫾ 5
3.6 ⫾ 3.1
⬍0.001
0.017
9.3 ⫾ 9.3 3⫾5 (continued)
J Am Coll Surg
EVAR
5.1 ⫾ 7.1
107 (0.9)
1,019 (5.8)
76 (9)
9.6 ⫾ 9.6
44 (0.8)
11 (4)
291 (8.7)
4.8 ⫾ 6.4
0.829
0.046
87 (7.1)
8 ⫾ 7.7
289 (2.4)
708 (4) ⬍0.001
OR
OR
316 (5.8)
11 (4)
92 (10.9)
Days from operation to discharge, mean ⫾ SD
0.029
151 (0.9) 0.417
305 (9.1)
598 (5.8)
177 (1.4)
0.002
0
72 (5.8)
176 (5.9)
338 (6.2)
1 (0.4)
OR Return to operating room, n (%)
0.343
605 (3.4) 0.002
6 (2.5) 36 (0.9)
3 (0.4)
376 (3.1)
10 (3.7) 4 (12.5)
108 (3.2) 87 (0.8)
OR Sepsis/SIRS postoperative, n (%)
170 (4) 62 (7.4)
899 (16.5) 515 (2.9)
0.077
343 (3.3)
0.013
0.041
1 (3.1)
202 (6)
796 (6.5)
6 (2.2)
64 (7.6)
82 (2.8)
1,015 (18.6)
1,275 (7.2) ⬍0.001
212 (6.3)
48 (3.9)
0.029
13 (4.8)
61 (4.9)
OR
880 (7.2)
0.329
OR Cardiovascular complication, n (%)
1,339 (24.6) 1,811 (10.2)
0.317
4 (12.5)
125 (3.7)
0.022
⬍0.001
15 (6.3) 290 (6.9)
299 (2.9)
1,464 (12)
21 (7.7) 6 (18.8)
165 (19.6)
1,941 (35.6)
2,219 (12.5) ⬍0.001
27 (11.3) 386 (9.2)
OR
0.900
38 (14) 11 (34.4)
315 (9.4) 1,087 (10.6)
169 (1.4)
Age-Stratified Results of Aortoiliac Aneurysm Repairs
EVAR
325 (11) 230 (18.7)
224 (4.1) 3,405 (19.3)
374 (30.3)
OR
⬍0.001
54 (19.9)
OR Nonsurgical complication, n (%)
Overall
393 (2.2)
Tsilimparis et al
Variable
⬍0.001
158 ⫾ 75 157 ⫾ 75 151 ⫾ 72 EVAR
EVAR, endovascular aneurysm repair; OR, open repair; SIRS, systematic inflammatory response syndrome.
162 ⫾ 78
174 ⫾ 85 237 ⫾ 98 191 ⫾ 99 246 ⫾ 105 OR
183 ⫾ 91
222 ⫾ 93
0.5 ⫾ 1.5 0.3 ⫾ 1.5 0.2 ⫾ 1.2 EVAR
Operation time, mean ⫾ SD, min
OR
No. of RBC U transfused intraoperative, mean ⫾ SD
Variable
161 ⫾ 74
168 ⫾ 80 220 ⫾ 102
0.6 ⫾ 1.3
4.4 ⫾ 4.3 1.05 ⫾ 2.5 3.1 ⫾ 4 1.13 ⫾ 2.6 2.7 ⫾ 3.4 1.10 ⫾ 2.3 2.2 ⫾ 2.8
Overall
⬍0.001
1.09 ⫾ 2.4
⬍0.001
p Value Overall
90 Years or older
Overall 80–89 y Overall
Age group 65–79 y 0–64 y
Table 2. Continued
⬍0.001
⬍0.001
236 ⫾ 99
0.4 ⫾ 1.5
181 ⫾ 91
1.1 ⫾ 2.5
Tsilimparis et al
0.577
2.6 ⫾ 3.4
Overall
Total cohort
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
697
rysm repair. Open repair was associated with considerably worse outcomes when compared with EVAR (mortality 1.8% vs 6.1%). Indes and colleagues16 presented similar findings for patients with occlusive disease and, again, endovascular approach was associated with lower mortality. Recent publications on endovascular repair of aortic aneurysms in octogenarians and nonagenarians has engendered enthusiasm with respect to safety and efficacy in the management of these patients.3-5,7,17-19 Most authors claim that despite the higher perioperative risk, EVAR can be performed safely in these patients cohorts. The obvious advantage of EVAR in the management of elderly patients should not compromise thoughtful patient selection of octogenarians and nonagenarians with AAA. Our study clearly demonstrates how these patients are at higher risk for cardiovascular events, but also for any postoperative complications or death, even with use of endovascular techniques, although the selection process of patients older than 80 years who undergo surgery is more meticulous than that for younger patients. Although elective EVAR should be preferred for anatomically suitable patients, we suggest that octogenarians and nonagenarians who are not suitable for EVAR should also be evaluated for open repair. The option of a fenestrated-branched stent graft could also be considered for aneurysms with short infrarenal neck or pararenal aneurysms. Aggressive treatment of AAA with challenging anatomies can be pursued in elderly patients, given that already 30% of patients in the United States are treated outside instructions for use of the devices.20 However, physicians should be aware of the risk of sac enlargement, as demonstrated by Schanzer and colleagues,20 for patients treated outside the instructions for use, especially for patients older than 80 years. Overall, perioperative risk of death and complications should be evaluated before proceeding with any kind of invasive therapy and risk of aneurysm rupture should exceed the estimated risk. We believe that best medical therapy is an acceptable alternative for selected octogenarians or nonagenarians whose aneurysm size has reached the established repair threshold but is associated with anatomic characteristics that could make it a high-risk repair. Appropriate evaluation of elderly patients is essential, and age older than 80 years should be considered a risk factor that, in the presence of another 2 cardiac risk factors, should trigger additional cardiac stress testing as recommended by the guidelines of vascular societies.21,22 Routine preoperative carotid duplex sonography cannot be supported by current available evidence. The cost effectiveness of using EVAR more frequently in elderly patients is difficult to speculate because of the
698
Table 3. Operative Parameters and Outcomes of Open and Endovascular Repair in Emergent Cases of Aortoiliac Aneurysm Repair Age group 0–64 y
65–79 y
80–89 y
90 Years or older Overall
Overall
Mortality, n (%) OR EVAR Any complication, n (%) OR EVAR Nonsurgical complication, n (%) OR EVAR Surgical complication, n (%) OR EVAR Pulmonary complication, n (%) OR EVAR Renal complication, n (%) OR EVAR Cardiovascular complication, n (%) OR EVAR Graft failure, n (%) OR EVAR Sepsis/SIRS postoperative, n (%) OR EVAR Return to operating room, n (%) OR EVAR
73 (15.5)
345 (27.7)
220 (35.5)
31 (40.3)
58 (19.1) 15 (8.9)
264 (31.8) 81 (19.6) 226 (47.9)
170 (55.9) 56 (33.3)
713 (57.3) 530 (63.8) 183 (44.2)
179 (37.9) 130 (42.8) 49 (29.2)
147 (31.1)
56 (11.9)
51 (10.8)
7 (1.5)
97 (15.7)
10 (0.8)
66 (14)
7 (1.1)
217 (17.4) 171 (20.6) 46 (11.1)
96 (20.3)
123 (19.9)
259 (20.8) 188 (22.6) 71 (17.1)
0.303 0.056
0.260
467 (29.9) 196 (23.1)
0.040 0.113
0.013
688 (44) 202 (23.8)
0.621 0.240
0.343
264 (16.9) 83 (9.8)
0.045 0.965
0.106
233 (14.9) 95 (11.2)
0.229 0.866
0.640
12 (0.8) 13 (1.5)
0.006 0.262
0.005
320 (20.5) 91 (10.7)
0.565 0.573
0.266
337 (21.5) 139 (16.4)
1112 (46.1)
663 (27.5)
890 (36.9)
347 (14.4)
328 (13.6)
25 (1)
5 (6.5) 5 (11.1) 0
110 (17.8) 77 (20.1) 33 (14)
827 (52.9) 285 (33.6)
1 (0.8) 0 1 (3.1)
97 (25.3) 26 (11.1)
0.001 ⬍0.001 0.229
1,357 (56.2)
8 (10.4) 5 (11.1) 3 (9.4)
3 (0.8) 4 (1.7)
989 (63.2) 368 (43.3)
10 (13) 9 (20) 1 (3.1)
69 (18) 28 (11.9)
4 (0.5) 6 (1.4)
47 (15.5) 19 (11.3)
95 (15.3)
172 (13.8)
0.016 ⬍0.001 0.010
23 (29.9) 20 (44.4) 3 (9.4)
67 (17.4) 28 (11.9)
127 (15.3) 45 (10.9)
5 (1.6) 2 (1.2)
236 (38.1)
186 (14.9)
514 (32.9) 155 (18.3)
24 (31.2) 18 (40) 6 (18.8)
181 (47.1) 55 (23.4)
144 (17.3) 42 (10.1)
32 (10.5) 19 (11.3)
174 (28.1)
484 (38.9)
⬍0.001 ⬍0.001 0.005
31 (40.3) 24 (53.3) 7 (21.9)
108 (28.1) 66 (28.1)
375 (45.1) 109 (26.3)
44 (14.5) 12 (7.1)
307 (4.6)
352 (28.3)
Overall
669 (27.7)
43 (55.8) 31 (68.9) 12 (37.5)
223 (58.1) 84 (35.7)
256 (30.8) 96 (23.2)
112 (36.8) 35 (20.8)
375 (60.6)
595 (47.8)
113 (23.9)
24 (53.3) 7 (21.9)
258 (67.2) 117 (49.8)
450 (54.2) 145 (35)
85 (28) 28 (16.7)
65 (2.4) 31 (18.5)
168 (43.8) 52 (22.1)
Total cohort p Value
411 (17)
11 (14.3) 7 (15.6) 4 (12.5)
476 (19.7)
(continued)
J Am Coll Surg
Overall
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Overall
Tsilimparis et al
Variable
EVAR, endovascular aneurysm repair; OR, open repair; SIRS, systematic inflammatory response syndrome.
0.0114 0.135 0.064
211.3 ⫾ 101 197 ⫾ 99 170.4 ⫾ 89.1
7.1 ⫾ 7.8 0.005 0.005 0.229
8.9 ⫾ 8.3 3.3 ⫾ 5.3
11.6 ⫾ 14 0.359 0.846 0.380
13.3 ⫾ 15.3 8.7 ⫾ 11.7
Overall
Tsilimparis et al
Days from operation to discharge, mean ⫾ SD 11.1 ⫾ 14.6 12.1 ⫾ 14.6 11.4 ⫾ 13.6 9.8 ⫾ 12.5 OR 12.9 ⫾ 16.1 13.6 ⫾ 15.4 13.0 ⫾ 14.5 12.6 ⫾ 15.5 EVAR 7.9 ⫾ 10.7 9.1 ⫾ 12.4 8.79 ⫾ 11.63 5.98 ⫾ 3.90 No. of RBC U transfused intraoperative, mean ⫾ SD 5.8 ⫾ 6.6 OR 7.4 ⫾ 6.9 9.3 ⫾ 8.6 7.5 ⫾ 8.3 9.2 ⫾ 8.0 7.2 ⫾ 7.7 11.3 ⫾ 9.3 7.4 ⫾ 8.5 EVAR 2.4 ⫾ 4.3 3.6 ⫾ 5.7 3.5 ⫾ 5.3 2.8 ⫾ 4.3 Operation time, mean ⫾ SD, min 195.9 ⫾ 94.6 OR 218.2 ⫾ 99.3 213.5 ⫾ 102.7 201.2 ⫾ 101.9 201.3 ⫾ 100.1 190.1 ⫾ 98.1 209.5 ⫾ 83.4 188.2 ⫾ 78.2 EVAR 155.6 ⫾ 69.5 176.6 ⫾ 95.6 171.7 ⫾ 92.2 158.2 ⫾ 59.3
Overall Variable
Table 3. Continued
0–64 y
Overall
65–79 y
Overall
Age group
80–89 y
Overall
90 Years or older
p Value
Total cohort
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
699
higher perioperative mortality of open repair in this patient population and their reduced life expectancy, so that data from larger studies suggesting lack of cost effectiveness cannot be directly applied.23 Several studies have confirmed that although the operative mortality of ruptured abdominal aortic aneurysm remains high in elderly patients, the long-term survival of patients who survive the procedure is very satisfactory.24,25 The survival of patients older than 80 years who survived the early postoperative period (90 days) at 1 and 5 years was 99% and 94%, respectively, in the Swedish Vascular Registry.26 Thirty-day survival of octogenarians and nonagenarians in our study was 64.5% and 59.7%, respectively, in emergent cases, which is similar or better than what has been reported previously.24,25 EVAR was associated with improved survival and outcomes in the emergent setting as well, although this could be attributed to a selection bias in favor of patients that were hemodynamically stable and anatomically suitable for EVAR (30-day survival after EVAR in group C: 77.9% and group D: 78.1% vs C: 56.2% and group D: 46.7% after open repair). The survival rates of octogenarians and nonagenarians in our study suggest that age alone should not be a reason to abandon therapy of a patient with AAA presenting as an emergency. The validity of the data of the ACS NSQIP database is much greater when compared with those of other administrative databases.27-29 Dedicated, trained study nurses capture the data in each participating institution. Participation in ACS NSQIP is at a voluntary setting, but procedures in each institution are captured randomly and no selection bias is present. Limitations
This is an observational study from a registry database. Limitations of the study mainly involve missing variables that are not captured in the ACS NSQIP database, such as aneurysm-specific characteristics, as well as the fact that only patients who underwent therapy are included. The ACS NSQIP database was created to report data collected from surgical patients in different specialties to enhance quality of care by providing data for comparison. Limitations due to the database design are expected when a specific surgical entity and issue is addressed. In addition, data beyond 30 days are not available, which is a major limitation of any study looking into outcomes of aneurysmal repair, especially in very elderly patients. Results on longterm survival would be helpful in determining to what extent the most elderly patients benefit from elective repair. The final logistic regression models for mortality included 9,452 patients. The missing data were due to missing laboratory values. The models produced similar conclusions when they were run without the laboratory values (ie, al-
700
Tsilimparis et al
Age-Stratified Results of Aortoiliac Aneurysm Repairs
J Am Coll Surg
Table 4. Logistic Regression Analyses for Mortality and Any Complication for all Patients with No Missing Variables
Variable
Regression model for mortality Model A Model B Odds ratio 95% CI Odds ratio 95% CI
Regression model for any complication Odds ratio 95% CI
Age, y, continuous 0–64 65–79 80–89 90 years or older Endovascular repair Emergency ASA I and II class ASA III class ASA IV class ASA V class eGFR Preoperative albumin Preoperative white blood cell count Preoperative hematocrit Preoperative total bilirubin
— Reference 1.85 2.65 3.28 0.41 2.90 Reference 1.14 2.71 9.16 0.99 0.62 1.03 0.96 1.15
— Reference 1.12 1.33 1.38 0.26 3.62 — — — — 0.99 — — — —
— Reference 1.34–2.54 1.88–3.75 1.77–6.06 0.332–0.50 2.24–3.75 Reference 0.52–2.47 1.25–5.89 4.05–20.70 0.99–0.99 0.52–0.73 1.01–1.05 0.95–0.98 0.99–1.34
1.04 — — — — 0.40 2.88 Reference 1.12 2.66 8.89 0.99 0.61 1.03 0.97 1.15
1.03–1.05 — — — — 0.33–0.50 2.22–3.73 Reference 0.51–2.42 1.22–5.78 3.99–20.10 0.99–0.99 0.52–0.73 1.01–1.05 0.95–0.98 0.98–1.35
— Reference 1.01–1.24 1.18–1.49 1.04–1.84 0.24–0.28 3.28–3.99 — — — — 0.99–0.99 — — — —
For mortality Model A adjusts for age groups and all relevant confounders, and in model B age is presented as a continuous variable with the same confounders. ASA, American Society of Anesthesiologists; eEGF, estimated glomerular filtration rate.
bumin, white blood count, hematocrit, and bilirubin) leading to exclusion of ⬍1,000 patients (data not shown). Data were collected from ⬎400 hospitals in the United States so that a uniform pattern of patient management, size threshold for aneurysm repair, or indication for surgery cannot be identified.
Analysis and interpretation of data: Tsilimparis, Perez, Dayama, Ricotta Drafting of manuscript: Tsilimparis, Dayama, Perez, Ricotta Critical revision: Tsilimparis, Dayama, Perez, Ricotta
CONCLUSIONS The present study is the largest one to date to report agestratified 30-day outcomes of AAA repair. Contrary to what has been frequently described in previous singlecenter studies, the risk of death and major complications in elderly patients is considerable. These data should be considered when treating octogenarians and nonagenarians with AAA. The estimated risk of rupture should outweigh the risk of perioperative death and major complications as described here in for patients receiving elective therapy. Advanced age is independently associated with higher risk of death after AAA repair and indication for surgery should be accordingly adjusted for different age groups. Endovascular aneurysm repair should be preferred for octogenarians and nonagenarians with indication to undergo repair of their AAA.
REFERENCES
Author Contributions
Study conception and design: Tsilimparis, Ricotta Acquisition of data: Tsilimparis, Dayama, Perez
1. Norman PE, Semmens JB, Lawrence-Brown MM, Holman CD. Long term relative survival after surgery for abdominal aortic aneurysm in western australia: population based study. BMJ 1998;317:852–856. 2. Samy AK, Murray G, MacBain G. Glasgow aneurysm score. Cardiovasc Surg 1994;2:41–44. 3. Henebiens M, Vahl A, Koelemay MJ. Elective surgery of abdominal aortic aneurysms in octogenarians: a systematic review. J Vasc Surg 2008;47:676–681. 4. Geisbusch P, Katzen BT, Tsoukas AI, et al. Endovascular repair of infrarenal aortic aneurysms in octogenarians and nonagenarians. J Vasc Surg 2011;54:1605–1613. 5. Roberts WC, Ko JM, Pearl GJ. Abdominal aortic aneurysm in nonagenarians. Am J Geriatr Cardiol 2006;15:319–321. 6. Bouwmeester MF, van Sterkenburg SM, Zeebregts CJ, et al. Endovascular abdominal aortic aneurysm repair in nonagenarians— beyond limits? Tex Heart Inst J 2009;36:345–348. 7. Prenner SB, Turnbull IC, Serrao GW, et al. Outcome of elective endovascular abdominal aortic aneurysm repair in nonagenarians. J Vasc Surg 2011;54:287–294. 8. Raval MV, Eskandari MK. Outcomes of elective abdominal aortic aneurysm repair among the elderly: endovascular versus open repair. Surgery 2011:245–260.
Vol. 215, No. 5, November 2012
Age-Stratified Results of Aortoiliac Aneurysm Repairs
Tsilimparis et al
9. Forsdahl SH, Singh K, Solberg S, Jacobsen BK. Risk factors for abdominal aortic aneurysms: a 7-year prospective study: the Tromso study, 1994–2001. Circulation 2009;119:2202–2208. 10. Grant SW, Grayson AD, Purkayastha D, et al. Logistic risk model for mortality following elective abdominal aortic aneurysm repair. Br J Surg 2011;98:652–658. 11. Giles KA, Pomposelli F, Hamdan A, et al. Decrease in total aneurysm-related deaths in the era of endovascular aneurysm repair. J Vasc Surg 2009;49:543–550; discussion 550–541. 12. Anjum A, Powell JT. Is the incidence of abdominal aortic aneurysm declining in the 21st century? Mortality and hospital admissions for England & Wales and Scotland. Eur J Vasc Endovasc Surg 2012;43:161–166. 13. Svensjo S, Bjorck M, Gurtelschmid M, et al. Low prevalence of abdominal aortic aneurysm among 65-year-old Swedish men indicates a change in the epidemiology of the disease. Circulation 2011;124:1118–1123. 14. Biancari F, Catania A, D’Andrea V. Elective endovascular vs open repair for abdominal aortic aneurysm in patients aged 80 years and older: systematic review and meta-analysis. Eur J Vasc Endovasc Surg 2011;42:571–576. 15. Beck AW, Goodney PP, Nolan BW, et al. Predicting 1-year mortality after elective abdominal aortic aneurysm repair. J Vasc Surg 2009;49:838–843; discussion 843–834. 16. Indes JE, Tuggle CT, Mandawat A, Sosa JA. Age-stratified outcomes in elderly patients undergoing open and endovascular procedures for aortoiliac occlusive disease. Surgery 2010;148: 420–428. 17. Baril DT, Palchik E, Carroccio A, et al. Experience with endovascular abdominal aortic aneurysm repair in nonagenarians. J Endovasc Ther 2006;13:330–337. 18. Paolini D, Chahwan S, Wojnarowski D, et al. Elective endovascular and open repair of abdominal aortic aneurysms in octogenarians. J Vasc Surg 2008;47:924–927. 19. Ju MH, Keldahl ML, Pearce WH, et al. Stepwise age-related outcomes of elective endovascular abdominal aortic aneurysm
20. 21. 22.
23. 24. 25.
26. 27.
28.
29.
701
repair: 11-year institutional review. Perspect Vasc Surg Endovasc Ther 2011;23:280–290. Schanzer A, Greenberg RK, Hevelone N, et al. Predictors of abdominal aortic aneurysm sac enlargement after endovascular repair. Circulation 2011;123:2848–2855. Chaikof EL, Brewster DC, Dalman RL, et al. The care of patients with an abdominal aortic aneurysm: the Society for Vascular Surgery practice guidelines. J Vasc Surg 2009;50:S2–S49. Moll FL, Powell JT, Fraedrich G, et al. Management of abdominal aortic aneurysms clinical practice guidelines of the European Society for Vascular Surgery. Eur J Vasc Endovasc Surg 2011;41[Suppl 1]:S1–S58. Brown L, Powell J, Thompson S, et al. The UK Endovascular Aneurysm Repair (EVAR) trials: randomised trials of EVAR versus standard therapy. Health Technol Assess 2012;16:1–218. Opfermann P, von Allmen R, Diehm N, et al. Repair of ruptured abdominal aortic aneurysm in octogenarians. Eur J Vasc Endovasc Surg 2011;42:475–483. Biancari F, Mazziotti MA, Paone R, et al. Outcome after open repair of ruptured abdominal aortic aneurysm in patients ⬎80 years old: a systematic review and meta-analysis. World J Surg 2011;35:1662–1670. Mani K, Bjorck M, Lundkvist J, Wanhainen A. Improved longterm survival after abdominal aortic aneurysm repair. Circulation 2009;120:201–211. Shiloach M, Frencher SK Jr, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16. Hall BL, Hamilton BH, Richards K, et al. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: an evaluation of all participating hospitals. Ann Surg 2009;250:363–376. Davis CL, Pierce JR, Henderson W, et al. Assessment of the reliability of data collected for the Department of Veterans Affairs National Surgical Quality Improvement Program. J Am Coll Surg 2007;204:550–560.