ORIGINAL ARTICLES
`
`Mortality After Elective Abdominal Aortic Aneurysm Repair
`
`Felix J. V. Schlo¨sser, MD, PhD,*†‡ Ilonca Vaartjes, MSc,† Geert J. M. G. van der Heijden, PhD,†
`Frans L. Moll, MD, PhD,‡ Hence J. M. Verhagen, MD, PhD,§ Bart E. Muhs, MD, PhD,*¶
`Gert J. de Borst, MD, PhD,‡ Andreas T. Tiel Groenestege, MSc,‡ Jan W. P. F. Kardaun, MD, PhD,储
`Agnes de Bruin, MSc,储 Johannes B. Reitsma, MD, PhD,** Yolanda van der Graaf, MD, PhD,†
`and Michiel L. Bots, MD, PhD†
`
`Objective: Purpose of this study is to provide detailed age- and gender-
`specific mortality risks of patients hospitalized for elective AAA repair.
`Summary Background Data: Whether to perform elective abdominal aortic
`aneurysm (AAA) surgery is balancing the risks of natural history against the
`risks of surgical intervention. Literature is lacking mortality risks after
`elective AAA repair with stratification by both age and gender.
`Methods: Mortality risks for 28 days, 1 year, and 5 years were derived
`from a nationwide cohort of patients hospitalized for elective AAA repair
`in 1997 or 2000. This cohort was formed through linkage of the Hospital
`Discharge Register with the Dutch Population Register. The relations
`between demographics, medical history and mortality were studied by
`Cox regression.
`Results: A total of 3457 patients were identified; 86% males, mean age 72
`⫾ 8.0 years. Mortality risks after elective AAA repair increased with age:
`28-day mortality ranged from 3.3% to 27.1% in men and 3.8% to 54.3%
`in women, 5-year mortality from 12.9% to 78.1% in men and 24.3% to
`91.3% in women. Higher age, congestive heart failure, cerebrovascular
`disease and diabetes mellitus were independent risk factors for 5-years
`mortality.
`Conclusions: Mortality risks after elective AAA repair are strongly age-
`related. Age, gender, and comorbidities should be taken into account when
`deciding on surgery. A general threshold of 55 mm for surgery might not be
`justified for all patients.
`
`(Ann Surg 2010;251: 158 –164)
`
`Abdominal aortic aneurysms (AAA) are each year diagnosed in
`
`200,000 patients in the United States. The total number of
`people with abdominal aortic aneurysms is estimated around 2
`million in the United States, and these aneurysms are a major cause
`of death with approximately 6300 deaths annually.1,2 Aging of the
`
`From the *Department of Surgery, Section of Vascular Surgery, Yale University
`School of Medicine, New Haven, CT; †Julius Center for Health Sciences and
`Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands;
`‡Department of Vascular Surgery, University Medical Center Utrecht, Utre-
`cht, The Netherlands; §Department of Vascular Surgery, Erasmus University
`Medical Center, Rotterdam, The Netherlands; ¶Department of Radiology,
`Yale University School of Medicine, New Haven, CT; 㛳Statistics Netherlands,
`The Hague, The Netherlands; and **Department of Clinical Epidemiology
`and Biostatistics, Academic Medical Center Amsterdam, Amsterdam, The
`Netherlands.
`Supported by the Netherlands Heart Foundation (grant number 31653251). The
`study was part of the project “Cardiovascular disease in the Netherlands:
`figures and facts” of the Netherlands Heart Foundation. The authors were in
`part also supported by a Jahnigen Career Development Scholars Award from
`the American Geriatrics Society (to F.J.V.S. and B.E.M.).
`Reprints: Ilonca Vaartjes, MSc, Julius Center for Health Sciences and Primary
`Care, University Medical Center Utrecht (STR 6.131), PO Box 85500, 3508
`GA Utrecht, The Netherlands. E-mail: c.h.vaartjes@umcutrecht.nl.
`Copyright © 2009 by Lippincott Williams & Wilkins
`ISSN: 0003-4932/10/25101-0158
`DOI: 10.1097/SLA.0b013e3181bc9c4d
`
`158 | www.annalsofsurgery.com
`
`population is likely to further increase these numbers in the future.
`Rupture of an AAA causes death in almost all patients, with reported
`mortality risks ranging from 79% to 94%.3–11 The risk of mortality
`at the moment of rupture appears to be higher in females and
`patients with higher age.12,13
`The maximal AAA diameter is an important risk factor for
`AAA rupture. At present,
`the generally accepted threshold for
`elective AAA repair is an AAA diameter of 55 mm for patients who
`are fit for surgery, because the risk of mortality due to rupture is
`expected to outweigh the risk of mortality due to surgical repair in
`these patients. Watchful waiting by ultrasound surveillance is con-
`sidered safe for small AAAs (⬍55 mm) and surgery does not
`improve survival in these patients.14 –16 The treatment decision is
`therefore based on the difference between the risk of mortality due
`to rupture in patient with watchful waiting and the risk of mortality
`due to surgical repair of the AAA.
`Most available articles present the impact of age and gender
`on mortality as relative risk or hazard ratio, without providing
`stratified and detailed mortality risks for age and gender subgroups.
`Some articles present stratified mortality risks, but all these are
`limited to age or gender, instead of stratification by age and gender.
`This is important since both age and gender are important determi-
`nants of the risk of death. Literature is therefore clearly lacking this
`important information. Therefore, the first objective of this study
`was to provide more insight into the short- and long-term mortality
`risks after first hospitalization for elective AAA repair stratified by
`both age and gender. A cohort of patients hospitalized for elective
`AAA repair was constructed through identification of these patients
`in nationwide registries. As the validity of nationwide registries
`often is questioned,
`the second objective of this study was to
`perform parallelly a validation study with data from the hospital
`discharge registry of the University Medical Center Utrecht
`to
`estimate the number of patients being hospitalized for AAA, that
`actually receive an AAA repair.
`
`METHODS
`Registries and Linkage Procedure
`To construct a cohort of patients who were admitted for the
`first time for any type of elective AAA repair, information from
`the national Hospital Discharge Register (HDR), the Dutch Pop-
`ulation Register (PR) and cause of death statistics Netherlands
`were linked. The registries and linkage procedures have been
`detailed earlier.17 For the present study, cohorts were drawn from
`the years 1997 and 2000 because of pragmatic reasons at the time
`of the initiation of this project in 2001. The total population of
`the Netherlands
`in 1997 and 2000 was 15,567,107 (men
`7,696,803; women 7,870,304) and 15,863,950 (men: 7,846,317;
`women: 8,017, 633), respectively.
`Study Population
`All 5642 hospital admissions with principal diagnosis AAA
`without rupture (ICD-9-CM code 441.4)18 between January 1 and
`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`TMT 2095
`Medtronic v. TMT
`IPR2021-01532
`
`

`

`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`Mortality After Elective AAA Repair
`
`FIGURE 1. Flowchart of the selec-
`tion process of patients hospital-
`ized for the first time for elective
`AAA repair in the Netherlands in
`1997 and 2000. PR indicates popu-
`lation register; HDR, hospital dis-
`charge register; AAA, abdominal
`aortic aneurysm.
`
`December 31, 1997 and January 1 and December 31, 2000, were
`selected from the national HDR. Following linkage with the PR
`(based on date of birth, gender, and 4 digits of postal code) and
`identification of the first admission for an individual of all subse-
`quent admissions occurring during the year 1997 and 2000 yielded
`a total of 4658 patients. Through linkage with the HDR, patients
`with previous admissions from January 1, 1995 until
`the first
`admission in 1997 and patients with previous admissions from
`January 1, 1995 until the first admission in 2000 were identified. No
`data were available about the type of performed surgical procedure
`or on the size of the AAA. Those with a previous hospital admission
`for the same condition or other peripheral arterial disease (ICD-
`9-CM codes 440 – 444, 7854) were excluded. This resulted in a
`cohort consisting of 3457 patients with a first hospitalization for a
`nonruptured AAA, of which we assume that these hospitalizations
`were because of an elective AAA repair in the Netherlands in 1997
`and 2000 (Fig. 1).
`Comorbidity
`The presence of comorbidities was determined on the basis
`of the limited information about
`the discharge diagnosis of
`previous hospital admissions (up to 5 years earlier) or on the
`basis of a secondary diagnosis at
`the moment of the index
`admission and was only available for different cardiovascular
`diseases (ICD-9-CM codes 390 – 459) and diabetes mellitus
`(ICD-9-CM code 250).
`Follow-up
`Information on mortality and causes of death of the patients
`was obtained by linkage of the cohort with the national Cause of
`Death Register. Linkage of the PR (with which the HDR cohort was
`previously linked) with the Cause of Death Register was performed
`by means of a unique identification key. Patients were considered
`non-native if at least one of their parents was not born in The
`Netherlands. Age- and gender-specific mortality results of 32 pa-
`tients (male ⬍50 years or female ⬍60 years) could not be reported
`
`due to Dutch privacy regulations. Follow-up was censored when
`patients moved out of the Netherlands. Cause of death has been
`coded using the tenth revision of the International Classification of
`Diseases (ICD-10).19
`
`Validation Study
`To gain more insight in the number of patients hospitalized
`for AAA that actually received AAA repair, we performed a vali-
`dation study. We selected all patients with a first hospital admission
`with principal diagnosis AAA without rupture (ICD-9-CM code
`441.4) from the hospital discharge registry of the University Medical
`Center Utrecht for the years 1997, 2000, 2004, and 2005. Next, 2
`individual researchers (A.T.T.G., I.V.) manually checked the med-
`ical records of all patients to determine whether these patients
`received an AAA repair.
`
`Design of Data Analysis
`Short-term (28-day), midterm (1-year), and long-term (5-
`year) mortality risks (%) were computed using life time table
`approach. Cox proportional hazard models were used to deter-
`mine whether patient characteristics (age, sex, native/non-native)
`and comorbidities (previous hospitalization because of ischemic
`heart disease, congestive heart failure, cerebrovascular event,
`other cardiovascular disease, and diabetes mellitus) were associ-
`ated with short- and long-term mortality risk. Information about
`life expectancies and 1-year mortality risks of the general Dutch
`population were obtained online from Statline, Statistics Nether-
`lands.20 Statistical analyses were done with Episheet21 and SPSS
`software, version 14.0 (SPSS Inc, Chicago, IL). A P ⬍ 0.05 was
`considered significant. All the linkage procedures and the statis-
`tical analyses were performed in agreement with privacy legis-
`lation in the Netherlands at Statistics Netherlands.22 The authors
`had full access to the data and take responsibility for its integrity.
`
`© 2009 Lippincott Williams & Wilkins
`
`www.annalsofsurgery.com | 159
`
`

`

`Schlo¨sser et al
`
`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`RESULTS
`
`Study Population
`In 1997 and 2000, a total of 3457 patients were identified that
`were hospitalized for the first time for elective AAA repair. The
`overall incidence rate was 128 per million person-years. Of the
`patients, 86% were male and the mean age was 72 ⫾ 8.0 years. Of
`all patients 30% had a history of hospitalization for another cardio-
`vascular disease. Women were older than men (74.3 vs. 71.2 years;
`95% CI of difference 2.2–3.9; P ⬍ 0.0001). A higher percentage of
`men had been previously hospitalized for myocardial infarction
`(6.3% vs. 2.7%; RR: 2.3, 95% CI: 1.3– 4.0) or ischemic heart disease
`(17.0% vs. 11.0%; RR: 1.5, 95% CI: 1.2–2.0) than women. Other
`baseline characteristics were not significantly different between men
`and women (Table 1).
`Length of Hospitalization
`The mean length of hospital stay was 12.5 ⫾ 13.9 days. The
`mean length of stay was significantly longer for women than for men
`(14.4 vs. 12.1 days, difference 2.2 days, 95% CI of difference
`0.7–3.7, P ⫽ 0.0034) in univariate analysis. Length of stay remained
`significantly different between men and women when corrected for
`age, prior cardiovascular diseases, diabetes and survival (difference
`1.8 days, 95% CI of difference 0.5–3.2). Higher age was also
`significantly associated with extended length of stay in the same
`multivariate model (difference 0.09 days per year increase in age,
`95% CI: 0.03– 0.15). When only patients were regarded who sur-
`vived during follow-up (N ⫽ 2064), mean length of hospital stay
`was 11 days for men and 14 days for women.
`Mortality During Follow-up
`Mortality risks clearly increased with age in both men and
`women (Table 2). The 28-day mortality risk was higher in women
`for all age groups of 70 years or older, although this difference was
`only significant for patients aged 70 to 74 years and for patients aged
`
`TABLE 1. Patient Characteristics
`Total
`Female
`Male
`N (% orⴞSD) N (% orⴞSD) N (% orⴞSD)
`
`No. patients
`Age
`Non-native born
`Prior hospitalization in
`academic hospital
`Previous hospital
`admission
`Cardiovascular
`disease
`Acute myocardial
`infarction
`Ischemic heart disease
`Congestive heart
`failure
`Cerebrovascular
`accident
`Other cardiovascular
`disease
`Diabetes mellitus
`
`2983
`71.2 (⫾7.7)
`241 (8.0)
`568 (19.1)
`
`474
`74.3 (⫾8.8)
`49 (10.4)
`92 (19.4)
`
`3457
`71.6 (⫾8.0)
`290 (8.4)
`660 (19.1)
`
`914 (30.6)
`
`130 (27.4)
`
`1044 (30.2)
`
`188 (6.3)
`
`13 (2.7)
`
`201 (5.8)
`
`505 (17.0)
`61 (2.0)
`
`52 (11.0)
`12 (2.5)
`
`557 (16.1)
`73 (2.1)
`
`113 (3.8)
`
`18 (3.8)
`
`131 (3.8)
`
`430 (14.4)
`
`78 (16.5)
`
`508 (14.7)
`
`92 (3.1)
`
`12 (2.5)
`
`104 (3.0)
`
`Values are numbers (N) with percentages or means with standard deviations (SD).
`All hospital admissions for cardiovascular diseases (principal diagnoses) and diabetes
`mellitus (principal and secondary diagnoses) were selected in the period from 1995 to
`1997 for the cohort of patients that were hospitalized in 1997 and in the period from
`1995 to 2000 for the patients that were hospitalized in 2000.
`
`85 years or older. The short-term risk of death in certain groups of
`patients was larger than the natural risks of rupture of 55 mm large
`AAAs reported in the literature. The 5-year mortality risks were
`higher in women than in men in almost all age groups, although this
`difference was only significant in patients aged 85 years or older.
`Causes of Death
`In total, 40% of all patients died during the total follow-up
`period (N ⫽ 1393). Thirty percent of all male patients (N ⫽ 1159)
`and 49% of all female patients died (N ⫽ 234). An overview of
`cardiovascular causes of death stratified by gender is shown in Table
`3. Aneurysm-related death occurred significantly more often in
`women during follow-up, and there was a trend in women for
`increased death due to chronic heart failure and death due to
`cerebrovascular accident, whereas death due to myocardial in-
`farction or other coronary artery diseases occurred more often
`in men.
`Determinants of Survival
`Female gender, increased age and prior hospitalization for
`congestive heart failure were independently and significantly
`associated with higher 28-day and 1-year mortality in patients
`hospitalized for elective AAA repair (Table 4). Higher age,
`diabetes mellitus and previous hospital admission for congestive
`heart failure or cerebrovascular accident were associated with
`higher 5-year mortality.
`Validation Study
`In 1997, 2000, 2004, and 2005, a total of 383 patients were
`identified in the University Medical Center Utrecht. In 374 patients
`(97.7%) the presence of an AAA was confirmed. The percentage
`that received AAA repair decreased by age (92%, 83.9%, 80.8%,
`and 67.8% in patients aged 50 –59, 60 – 69, 70 –79, and 80⫹ years,
`respectively). From the medical records, the reasons indicated for
`not receiving an AAA repair during initial admission were (1) poor
`survival prognosis of the patient (14%) (in particular in the elderly);
`(2) AAA smaller than general threshold for AAA repair and there-
`fore repair postponed for more than a year (15%); or (3) patients
`were hospitalized for clinical (diagnostic) evaluation instead of
`actual AAA repair (5%).
`
`DISCUSSION
`This hospital-based study with nationwide coverage clearly
`highlights the clinically relevant differences in mortality risks after
`elective AAA repair by age and gender. Female gender, increased
`age, and prior hospitalization for congestive heart failure were
`independently and significantly associated with higher 28-day and
`1-year mortality in patients hospitalized for the first time for elective
`AAA repair. At 5-year follow-up, higher age, diabetes mellitus, and
`previous hospital admission for congestive heart failure or cerebro-
`vascular accident were associated with higher mortality.
`The presented mortality risks after elective AAA repair strat-
`ified by both age and gender represent clearly novel information,
`since current medical literature lacks these results. Obtaining these
`results was possible due to a linkage of 2 national registries: the
`Dutch Population Register and the Hospital Discharge Register. A
`high validity of both the HDR and the PR has been demonstrated. In
`a random sample of the HDR, 99% of the personal, admission, and
`discharge data, and 84% of the principal diagnoses were correctly
`registered.23 In a random sample of the PR, over 97% of the addresses
`were correctly registered and only 0.4% of days and months of birth
`were missing.24 Furthermore, over 97% of the uniquely linked
`hospital admissions resulting from linkage of the HDR with the PR
`were shown to be correctly linked and the estimated rate of mis-
`matches (false positive linkages) was approximately 1%.24
`
`160 | www.annalsofsurgery.com
`
`© 2009 Lippincott Williams & Wilkins
`
`

`

`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`Mortality After Elective AAA Repair
`
`TABLE 2. Twenty-Eight Day, 1-Year, and 5-Year Mortality
`28-Day Mortality*
`
`1-Year Mortality†
`
`5-Year Mortality†
`
`Age
`(Year)
`
`No.
`Male
`
`No.
`Female
`
`Male
`(%)
`
`Female
`(%)
`
`RR‡
`
`(95% CI)
`
`Male
`(%)
`
`Female
`(%)
`
`RR‡
`
`(95% CI)
`
`Male
`(%)
`
`Female
`(%)
`
`RR‡
`
`(95% CI)
`
`50–54
`55–59
`60–64
`65–69
`70–74
`75–79
`80–84
`ⱖ85
`
`60
`163
`372
`644
`754
`621
`262
`96
`
`—
`—
`37
`79
`102
`119
`70
`46
`
`3.3
`4.9
`4.0
`5.0
`6.8
`11.3
`20.2
`27.1
`
`—
`—
`—
`3.8
`13.7
`12.6
`25.7
`54.3
`
`—
`—
`—
`0.8
`2.0
`1.1
`1.3
`2.0
`
`—
`—
`—
`(0.2–2.4)
`(1.2–3.5)
`(0.7–1.9)
`(0.8–2.0)
`(1.3–3.1)
`
`3.3
`8.0
`9.1
`10.9
`15.3
`19.8
`32.4
`44.8
`
`—
`—
`8.1
`13.9
`21.6
`19.3
`34.3
`73.9
`
`—
`—
`0.9
`1.3
`1.4
`1.0
`1.1
`1.7
`
`—
`—
`(0.3–2.7)
`(0.7–2.3)
`(0.9–2.1)
`(0.7–1.5)
`(0.7–1.5)
`(1.2–2.2)
`
`13.3
`12.9
`18.8
`25.6
`36.6
`44.1
`58.4
`78.1
`
`—
`—
`24.3
`29.1
`39.2
`47.9
`51.4
`91.3
`
`—
`—
`1.3
`1.1
`1.1
`1.1
`0.9
`1.2
`
`—
`—
`(0.7–2.4)
`(0.8–1.6)
`(0.8–1.4)
`(0.9–1.3)
`(0.7–1.1)
`(1.0–1.3)
`
`N ⫽ 3457.
`*Data about 69 patients (male ⬍50 year or female ⬍65 year) could not be reported due to Dutch privacy regulations.
`†Data about 32 patients (male ⬍50 year or female ⬍60 year) could not be reported due to Dutch privacy regulations.
`‡RR indicates relative risk: risk of death in women compared to men.
`
`TABLE 3. Causes of Death During Follow-up After Elective
`AAA Repair. N ⫽ 3457
`
`Total
`Female
`Male
`N ⴝ 2983 N ⴝ 474 N ⴝ 3457
`% (N) % (N) % (N)
`Aneurysm related death† 8.2 (245) 18.6 (88) 9.6 (333)
`Myocardial infarction,
`5.4 (161)
`4.0 (19) 5.2 (180)
`coronary artery
`disease
`Chronic heart failure
`Cerebrovascular disease
`
`1.0 (36)
`1.7 (8)
`3.6 (17) 2.4 (83)
`
`0.9 (28)
`2.2 (66)
`
`RR* 95% CI
`
`2.3 (1.8–2.8)
`0.7 (0.5–1.2)
`
`1.8 (0.8–3.9)
`1.6 (1.0–2.7)
`
`*RR, Relative Risk: risk of death in women compared to men.
`†Death due to ruptured abdominal aortic aneurysm (ICD-10 code I-713), abdominal
`aortic aneurysm without rupture (ICD-10 code I- 714), or other aneurysm (ICD-10 code
`I-71 or I-72).
`
`The results of the validation study based upon data from the
`University Medical Center Utrecht, indicate that the reported mor-
`tality risks among the 80 plus year olds in our study cannot be
`attributed to the AAA repair alone and therefore these risk estimates
`cannot be used to perform or refrain from AAA repair. With 92% of
`the patients receiving AAA repair one may argue that for the 50 to
`59-year-old group the reported risks are correct while in the age
`group 60 to 79-year-old, mortality risk attributed to AAA repair
`might have been underestimated as 19% of those not receiving AAA
`repair proved to have an AAA too small in diameter to be treated
`surgically yet, and therefore have a lower mortality risk while 10%
`was too sick to receive an AAA repair and have a higher mortality
`risk. In the eldest patient group, 32% did not receive AAA repair.
`Ten percent proved to have a too small aneurysm while a consid-
`erable part (⬎30%) of these patients were not treated because of
`their poor survival prognosis. Therefore, it is likely that the results
`with regard to the mortality risk following AAA repair in patients
`older than 80 years were overestimated.
`Hospital-based studies with nationwide coverage have several
`advantages over other study designs. Randomized trials and hospi-
`tal-based registries often report lower morbidity and mortality risks
`than that are observed in hospital-based studies with nationwide
`coverage.25 Patients included in randomized trials are often younger
`and healthier than average patients seen in daily practice. Hospital-
`based randomized trials and registries often lack complete long-term
`
`follow-up. The principal advantages of the current study design over
`hospital-based registries and randomized trials are as follows: it has
`a nationwide coverage, information is collected about all patients
`who are seen in daily practice, and the follow-up is complete and
`sufficiently long.
`This study has however some limitations. Patients were iden-
`tified on the basis of the ICD-9 diagnosis for hospital admission, and
`we assumed that these hospitalizations were because of a planned
`elective AAA repair. Another disadvantage of our study might be
`the absence of information about the AAA diameter of the patients
`at the moment of surgery. Although the absence of the diameter
`information does not affect our observed estimates of prognosis, we
`cannot compare the age and gender stratified subgroups for differ-
`ences in initial AAA diameter and the subsequent effect of these
`differences on mortality risks after AAA repair. Information on the
`number of patients who underwent endovascular repair was not
`available. However, by selection of patients from years 1997 and
`2000, when conventional, open surgical AAA repair was main-
`stream, and endovascular AAA repair was performed only to a
`minor extend in the Netherlands, endovascular AAA repair was
`probably performed in a small number of all patients.
`Most literature reports overall risks of mortality after AAA
`repair instead of age and gender specific results. Because mortality
`risk clearly differs by age and gender subgroups, mortality risks
`from studies that do not provide these data cannot be compared with
`each other. Seven articles were identified in the available literature
`which presented results with stratification by age or gender (Table
`5). These articles clearly show that age and gender impact mortality
`risks after elective AAA.
`Several studies, that have not provided stratified results, have
`reported significant and independent risk factors for higher mortality
`after elective AAA repair. Older patient age,12,28 –31 female gen-
`der,12,28 –29,32–34 and current smoking32 appear to be associated with
`higher mortality. Comorbidities that have been associated with
`higher mortality risks after elective AAA repair are renal fail-
`ure,12,32,35–37 previous cerebrovascular events,32,38 chronic lung
`disease,39 liver disease,12 and a history of cardiac disease,32 notably
`previous myocardial infarction,29 congestive heart failure,12 and
`coronary heart disease.36,40 – 42 No paper presented results that were
`stratified by both age and gender at the same time. Therefore, in our
`opinion our nationwide study adds novel important information to
`the medical literature.
`The average AAA rupture risk for relatively small AAA with
`diameters between 5 and 6 cm is reported to range between 1.0%
`
`© 2009 Lippincott Williams & Wilkins
`
`www.annalsofsurgery.com | 161
`
`

`

`Schlo¨sser et al
`
`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`TABLE 4. Determinants of Mortality. N ⫽ 3457
`Variables
`28 Days (333 Events)
`1 Year (603 Events)
`
`5 Years (1252 Events)
`
`Female gender
`Age in yr
`Non-native citizen
`Previous
`Ischaemic heart disease
`Congestive heart failure
`Cerebrovascular disease
`Other cardiovascular disease
`Presence of diabetes mellitus
`
`1.33 (1.02–1.70)
`1.08 (1.07–1.10)
`1.29 (0.89–1.86)
`
`0.70 (0.49–1.00)
`2.77 (1.80–4.28)
`1.21 (0.74–2.01)
`0.94 (0.69–1.29)
`1.22 (0.68–2.17)
`
`1.25 (1.01–1.53)
`1.08 (1.07–1.09)
`1.07 (0.80–1.43)
`
`0.87 (0.69–1.11)
`2.64 (1.86–3.76)
`1.38 (0.97–1.96)
`0.98 (0.78–1.22)
`1.40 (0.94–2.09)
`
`1.13 (0.97–1.32)
`1.08 (1.07–1.09)
`1.05 (0.85–1.29)
`
`0.95 (0.81–1.12)
`3.01 (2.29–3.95)
`1.51 (1.18–1.92)
`1.12 (0.96–1.31)
`1.35 (1.01–1.80)
`
`Multivariable Cox regression model. Results are expressed as hazard ratios with 95% confidence intervals.
`
`TABLE 5. Prior Published Mortality Risks After Elective AAA Repair With Stratification by Age or Gender
`Determinant
`Risk of Mortality
`
`Author
`Wanhainen et al26
`
`Year
`
`2008
`
`Domain
`
`Patients undergoing
`surgical repair
`for intact AAA
`
`N
`
`7175
`
`Schermerhorn et al27
`
`2008
`
`22,830
`
`Patients who
`underwent
`elective open
`AAA repair
`
`McPhee et al12
`
`2007
`
`Patients undergoing
`surgery for intact
`AAA
`
`183,387
`
`Rigberg et al28
`
`2006
`
`Urbonavicius et al29
`
`2005
`
`Vemuri et al30
`
`2004
`
`Norman et al31
`
`1998
`
`Patients undergoing
`surgical repair
`for intact AAA
`
`Patients who
`underwent
`surgery for non-
`ruptured AAA
`
`Patients who
`underwent repair
`of intact AAA
`
`Patients undergoing
`surgery for intact
`AAA
`
`9778
`
`69
`
`6339
`
`935
`
`Age (Year) Gender
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M
`F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M
`F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M
`F
`M⫹F
`M⫹F
`M⫹F
`M⫹F
`M
`
`50–59
`60–69
`70–79
`⬎80
`All
`All
`67–69
`70–74
`67–74
`75–79
`80–84
`75–84
`ⱖ85
`⬍60
`60–69
`⬎70
`All
`All
`51–60
`61–70
`71–80
`81–90
`91–100
`ⱕ75
`⬎75
`All
`All
`51–60
`61–70
`71–80
`⬎80
`All
`
`All
`
`F
`
`In-Hospital (%)
`
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`0.73
`1.7
`3.9
`2.6
`4.8
`1.5
`2.4
`4.4
`6.1
`5.3
`5.9
`38.5
`9.6
`17.6
`1.1
`2.9
`5.4
`7.9
`—
`
`—
`
`30 Day (%)
`1.2†
`2.8†
`4.4†
`6.8†
`3.6†
`5.3†
`2.5*
`3.3*
`—
`4.8*
`7.2*
`—
`11.2*
`—
`—
`—
`—
`—
`1.7
`2.6
`4.4
`7.1
`10.5
`—
`—
`—
`—
`—
`—
`—
`—
`4.4
`
`3.6
`
`1 Year (%)
`
`5 Year (%)
`
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`2.9
`5.3
`9.5
`14.7
`26.3
`—
`—
`—
`—
`—
`—
`—
`—
`—
`
`—
`
`—
`—
`—
`—
`—
`—
`—
`—
`26†
`—
`—
`40†
`44‡
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`—
`21.0
`
`20.1
`
`*Percentage of patients that died during index hospital admission or within 30 days after surgery.
`†Extracted from graph, data were not presented in text.
`‡Mortality risk after 4.5 years.
`M indicates male; F, female; M ⫹ F, results from male and female patients together (data were not stratified by gender).
`
`162 | www.annalsofsurgery.com
`
`© 2009 Lippincott Williams & Wilkins
`
`

`

`Annals of Surgery (cid:127) Volume 251, Number 1, January 2010
`
`Mortality After Elective AAA Repair
`
`and 11% per year.43– 46 Our study reports 28-day mortality risks of
`older patients that are relatively high and 1-year mortality risks
`exceeding 10% in women aged 65 years or older and in men aged 70
`years or older. The perioperative mortality risk for certain subgroups
`of patients undergoing elective AAA repair, in particular elderly
`patients, may outweigh the rupture risk of 55 mm large AAAs that
`are known from literature. Although our findings need confirmation
`and especially more accurate estimates of mortality risks among
`those older than 80 years are needed, the results point towards
`reconsideration of an aneurysm diameter of 55 mm as a general
`threshold for elective AAA repair for certain subgroups of patients.
`Moreover, the current results support the notion that future studies
`should explicitly present their results on mortality risks after AAA
`repair by age and gender specific subgroups instead of crude overall
`results only.
`
`CONCLUSIONS
`Mortality risks after elective AAA repair are clearly age- and
`gender-related. The perioperative mortality risk for elderly patients
`undergoing elective AAA repair may outweigh the previously re-
`ported rupture risks of 55 mm large AAAs. Therefore, a general
`threshold of 55 mm for surgery may possibly not be justified for
`certain specific subgroups of patients. Future articles should pay
`more attention to differences by age and gender to increase compa-
`rability between studies and to provide the detailed background
`information that is necessary to make better informed treatment
`decisions in patients with AAA, which may subsequently lead to a
`better prognosis of these patients.
`
`REFERENCES
`1. CDC WONDER–Centers for Disease Control and Prevention. Compressed
`mortality, 1999 –2005 results, abdominal aortic aneurysm. Available at:
`http://wonder.cdc.gov/. Accessed May 27, 2009.
`2. Brewster DC. Presidential address: what would you do if it were your father?
`Reflections on endovascular abdominal aortic aneurysm repair. J Vasc Surg.
`2001;33:1139 –1147.
`3. Acosta S, O¨ gren M, Bengtsson H, et al. Increasing incidence of ruptured
`abdominal aortic aneurysm: a population-based study. J Vasc Surg. 2006;44:
`237–243.
`4. Kantonen I, Lepa¨ntalo M, Brommels M, et al; Finnvasc Study Group.
`Mortality in ruptured abdominal aortic aneurysms. Eur J Vasc Endovasc
`Surg. 1999;17:208 –212.
`5. Semmens JB, Norman PE, Lawrence-Brown MMD, et al. Population-based
`record linkage study of the incidence of abdominal aortic aneurysm in Western
`Australia in 1985–1994. Br J Surg. 1998;85:648–652.
`6. Mealy K, Salman A. The true incidence of ruptured abdominal aortic
`aneurysms. Eur J Vasc Surg. 1988;2:405– 408.
`7. Budd J, Finch D, Carter P. A study of the mortality of ruptured abdominal
`aortic aneurysms in a district community. Eur J Vasc Surg. 1989;3:351–354.
`8. Johansson G, Swedenborg J. Ruptured abdominal aortic aneurysms: a study
`of incidence and mortality. Br J Surg. 1986;73:101–103.
`9. Thomas P, Stewart R. Abdominal aortic aneurysm. Br J Surg. 1988;75:733–
`736.
`10. Drott C, Arfvidsson B, O¨ rtenwall P, et al. Age-standardized incidence of
`ruptured aortic aneurysm in a Swedish population between 1952 and 1988:
`mortality rate and operative results. Br J Surg. 1992;79:175–179.
`11. Ingoldby C, Wujanto R, Mitchell J. Impact of vascular surgery on
`community mortality from ruptured aortic aneurysms. Br J Surg. 1986;
`73:551–553.
`12. McPhee JT, Hill JS, Eslami MH. The impact of gender on presentation,
`therapy, and mortality of abdominal aortic aneurysm in the United States,
`2001–2004. J Vasc Surg. 2007;45:891– 899.
`13. Dueck AD, Kucey DS, Johnston KW, et al. Long-term survival and temporal
`trends in patient and surgeon factors after elective and ruptured abdominal
`aortic aneurysm surgery. J Vasc Surg. 2004;39:1261–1267.
`14. Mortality results for randomised controlled trial of early elective surgery
`or ultrasonographic surveillance for small abdominal aortic aneurysms.
`
`The UK Small Aneurysm Trial Participants. Lancet. 1998;352:1649 –
`1655.
`15. Long-term outcomes of immediate repair compared with surveillance of small
`abdominal aortic aneurysms. N Engl J Med. 2002;346:1445–1452.
`16. Lederle FA, Wilson SE, Johnson GR, et al; Aneurysm Detection and Man-
`agement Veterans Affairs Cooperative Study Group. Immediate repair com-
`pared with surveillance of small abdominal aortic aneurysms. N Engl J Med.
`2002;346:1437–1444.
`17. Vaartjes I, Reitsma JB, de Bruin A, et al. Nationwide incidence of first stroke
`and TIA in the Netherlands. Eur J Neur. 2008;15:1315–1323.
`18. The International Statistical Classification of Diseases, Injuries and Causes
`of Death. 9th Revision. Clinical Modification. Washington DC: US Depart-
`ment of Health, Education, and Welfare; 1979.
`19. World Health Organisation. International Statistical Classification of Disease
`and related Health Problems. Tenth Revision. Geneva, Switzerland: World
`Health Organisation; 1992.
`20. Statistics Netherlands. CBS–Statline. Available at: http://www.cbs.nl/en-GB/
`menu/cijfers/statline/default.htm. Accessed April, 2008.
`21. Andersson T, Ahlbom A. Episheet software: spreadsheets for the Analysis of
`Epidemiologic Data, K. Rothman, 2004. Available at: http://members.aol.
`com/krothman/episheet.xls. Accessed October 10, 2008.
`22. Reitsma JB, Kardaun JW, Gevers E, et al. Possibilities for anonymous
`follow-up studies of patients in Dutch national medical registrations using the
`Municipal Population Register: a pilot study 关in Dutch兴. Ned Tijdschr
`Geneesk. 2003;147:2286 –2290.
`23. Paas GR, Veenhuizen KC. Research on the Validity of the LMR 关in Dutch兴.
`Utrecht, The Netherland: Prismant; 2002.
`24. De Bruin A, Kardaun JWPF, Gast A, et al. Record linkage of hospital
`discharge register with populatio