Volume 34, Issue 3 , Pages 260-266, September 2007
The Relationship Between Volume and Outcome Following Elective Open Repair of Abdominal Aortic Aneurysms (AAA) in 131 German Hospitals☆
Article Outline
Objectives
Several studies indicate that high-volume hospitals have better results in open repair of unruptured abdominal aortic aneurysms (AAA). Up to now no studies had addressed this question in German hospitals.
Design
Post-hoc-analysis from a prospective physician-led registry.
Material and methods
Since 1999, the German Society for Vascular Surgery has conducted a prospective registry for open and endovascular repair of AAAs. This study includes 131 hospitals who conducted n
=10163 elective open repairs for unruptured AAA between 1999 to 2004. All perioperative variables including annual volume as a continuous variable were analysed in a step-wise logistic regression model. In order to define a threshold annual volume an additional logistic regression analysis was performed by use of annual volume groups (0–9, 10–19, 20–29, 30–39, 40–49, 50 or more). The relationship between annual volume and further outcome parameters (length of procedure, blood transfusion, length of stay) were also analyzed.
Results
The overall mortality rate was 3.2%. The stepwise logistic regression model identified the following predictors of an increased perioperative mortality: age (OR 1.084, 95% CI 1.066–1.102), AAA diameter (OR 1.008, 95% CI 1.001–1.016), length of procedure (OR 1.008, 95% CI 1.006–1.009), ASA-Score (OR 2.636, 95% CI 2.129–3.264), suprarenal clamping (OR 1.447, 95% CI 1.008–2,078), blood transfusion (OR 1.786, 95% CI 1.268–2.514). Annual volume was moderately predictive (OR 1.003, 95% CI 1–1.006) but failed to reach statistical significance (p=0.07). The analysis of volume groups identified a significantly higher risk for hospitals with an annual volume of 1-9 AAA-repairs by comparison to hospitals with an annual volume of 50 or more AAA-repairs (OR 1.903, 95% CI 1.124–3.222). Operations at low volume hospitals were also longer (p<0.001), with an extended postoperative stay (p<0.001) and a higher transfusion rate (p<0.001).
Conclusions
Patient's age, ASA classification, AAA diameter, length of procedure, suprarenal clamping and blood transfusion are predictive variables for an increased perioperative mortality in elective open AAA repair. Mortality is also increased by a low annual volume. Further studies are needed to examine whether these data are applicable to all German hospitals.
Keywords: AAA repair, Workload, Outcomes, Perioperative mortality, Perioperative complications
Background
Elective open or endovascular repair of an Abdominal Aortic Aneurysm (AAA) is the only effective treatment for the prevention of a rupture, which is fatal in >80% of all cases. Aside from the maximal AAA diameter, and accompanying co-morbidity, the patient's life expectancy and individual wishes need to be taken into consideration when contemplating elective repair.1 The standard treatment is open replacement with a tube or bifurcation prosthesis. Endovascular aortic repair (EVAR) has become more popular since randomized studies have shown low perioperative mortality rates for this procedure.2, 3, 4 Due to morphological limitations and the need for regular follow-up visits, EVAR is only possible for 30–50% of all AAA patients.
Several studies have shown that there is a significant relationship between the annual volume of open AAA repair and the perioperative mortality rate.5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 The respective thresholds, however, vary significantly. To date, there has been no information regarding German hospitals in this respect.
Material and Methods
In 1997, the German Society for Vascular Surgery implemented a prospective surgeon-led registry to document representative data about indications, treatment modalities and complications of open and endovascular repair for ruptured and non-ruptured AAAs in Germany. Participation was voluntary, all data were collected prospectively and were submitted on a machine-readable form including the following variables:
This study evaluated all AAA repairs from 1/1/1999–12/31/2004. During these 6 years, 16603 infrarenal aneurysm repairs were registered (unruptured AAA n=14667, ruptured AAA n=1981). 81.8% of all procedures were open repairs and 18.2% were endovascular repairs (EVAR). This study included all elective open repairs of unruptured AAAs. We excluded hospitals that had not been part of the registry for at least 4 years and patients who underwent another procedure at the same time as the AAA repair (such as carotid endarterectomy or peripheral bypass). As a result, 10163 elective open repairs for unruptured AAAs, performed in 131 different hospitals fulfilled the inclusion criteria for the study.
Data was statistically summarized by quantitive compilation of all preoperative, intraoperative, and postoperative variables. The following variables were considered to be potential factors for influencing perioperative mortality and were analyzed in a step-wise logistic regression model: patient's age, AAA-diameter, risk score according to the American Society of Anesthesiology (ASA), inflammatory AAA, coexistence of iliac artery aneurysms, length of the surgical procedure, suprarenal clamping, blood transfusion, replantation of the inferior mesenteric artery, tube grafting or aortobiiliac grafting, graft material (Dacron or PTFE) and annual volume as continous variable. Perioperative mortality was set as the primary outcome variable. The threshold for exclusion of factors was set at 0.2.
In order to find a significant and also clinical applicable threshold for annual volume we then divided the hospitals into groups by the number of open AAA-repairs per year: 0–9, 10–19, 20–29, 30–39, 40–49 and >/= 50. These thresholds were analysed together with the significant predictors of the step-wise regression analysis in a further regression model. To identify a relationship between annual volume and preoperative and/or intraoperative parameters and further outcome parameters we analyzed the different volume groups descriptively. All statistical calculations were conducted with SAS Version 8.2. Statistics were performed by us of the chi-sqare-test and Odds-Ratios (OR) with a confidence interval (CI) of 95%. Conspicuous parameters in the descriptive analysis of volume groups were subjected to a statistical trend analysis (Cochran Armitage Trend Test).
Results
The patients' average age was 67.5 years (median 68 years). 37.6% of the patients were categorized as ASA category I/II. The median duration of the surgical procedure was 155 minutes, and 54.2% of patients received an aortic tube graft. 71.6% of all patients required a blood transfusion, either during or after the surgery. The median length of the hospital stay was 16 days (2 days of which on the intensive or intermediate care unit).
The quantitative analysis of the different volume groups is included in Table 1. 17.6% of all participating hospitals conducted <10 open repairs per year (low volume) and 15.3% conducted more than 50 procedures per year (high volume). However, only 3.6% of all patients underwent repair in low volume hospitals, whilst 39.3% underwent repair in hospitals with a large volume.
Table 1. Distribution of patients/operations and hospitals regarding 10163 conventional open repairs on unruptured AAAs in 131 hospitals
| Hospitals | Patients | |||
|---|---|---|---|---|
| Number | Percent | Number | Percent | |
| 1–9 cases | 23 | 17.6% | 367 | 3.6% |
| 10–19 cases | 45 | 34.6% | 1881 | 18.5% |
| 20–29 cases | 22 | 16.8% | 1634 | 16.1% |
| 30–39 cases | 14 | 10.7% | 1551 | 15.3% |
| 40–49 cases | 7 | 5.3% | 739 | 7.3% |
| ≥50 cases | 20 | 15.3% | 3991 | 39.3% |
| Total | 131 | 100% | 10163 | 100% |
The overall perioperative mortality rate was 3.2% with a range from 5.2% for low-volume hospitals to 2.6% for high-volume hospitals. Operations in low volume hospitals were longer with a higher rate of suprarenal clamping, a longer length of stay and a higher frequency of blood transfusions (p<0.001, Cochran Armitage Trend Test, unadjusted data). In contrast, all other preoperative, intraoperative, and postoperative variables, such as patient age, ASA classification and surgical techniques (excluding suprarenal clamping) did not differ between low volume and high volume hospitals. General and specific complications (excluding perioperative fatality) occurred equally across all volume groups (Table 2).
Table 2. Distribution of all perioperative variables for all volume groups for 10163 conventional AAA-repairs for non-ruptured AAAs in 131 hospitals, 1999–2004. (unadjusted data, Cochran Armitage Trend)
| all (n=10163) | 0–9 (n=367) | 10–19 (n=1881) | 20–29 (n=1634) | 30–39 (n=1551) | 40–49 (n=739) | >/= 50 (n=3991) | P | |
|---|---|---|---|---|---|---|---|---|
| Preoperative variables | ||||||||
| Age (Years, Median) | 68 | 69 | 68 | 68 | 68 | 68 | 68 | – |
| ASA-Score I/II (%) | 37.1% | 42.8% | 39.3% | 36.7% | 33.3% | 41.5% | 39% | – |
| ASA-Score III/IV (%) | 62.9% | 56.9% | 60.6% | 63.1% | 66.6% | 58.5% | 62.3% | – |
| Maximal AAA diameter (mm, median) | 55 | 52 | 55 | 55 | 56 | 57 | 55 | – |
| Additional aneurysm of the A. iliaca (%) | 35.1% | 35.7% | 35.3% | 39.6% | 35.7% | 30.5% | 32.9% | – |
| Inflammatory AAA (%) | 4.9% | 5.2% | 3.2% | 4.5% | 4.1% | 4.3% | 6.1% | – |
| Intraoperative Variables | ||||||||
| Duration of procedure (min, median) | 155 | 170 | 171 | 163 | 155 | 155 | 143 | <0.001 |
| Suprarenal clamping (%) | 6.5% | 2.7% | 4.8% | 7.4% | 6.7% | 7.3% | 6.8% | 0.0019 |
| Inferior mesenteric artery reimplantation (%) | 7.3% | 4.4% | 9.6% | 5.5% | 4.3% | 3.4% | 9.0% | – |
| Surgical technique | ||||||||
| Tube graft | 54.2% | 51.2% | 52.8% | 48.3% | 46.7% | 65.6% | 58.3% | – |
| Aorto-bi-iliac(-femoral) graft | 45.7% | 49.1% | 47.2% | 51.7% | 52.9% | 34.0% | 41.7% | – |
| Prosthetic material | ||||||||
| Dacron | 91.0% | 87.5% | 96.4% | 86.5% | 74.7% | 97.7% | 95.7% | – |
| PTFE | 8.2% | 12.3% | 2.9% | 13.2% | 24.1% | 1.1% | 3.4% | – |
| Intraoperative cardiovascular stability (%) | 87.4% | 78.2% | 85.4% | 85.4% | 85.8% | 79.2% | 92.2% | – |
| Need for transfusion yes/no (%) | 71.6% | 81.2% | 76.3% | 77.0% | 74.9% | 68.5% | 65.6% | <0.001 |
| Postoperative Variables | ||||||||
| Inpatient Intensive/ICU (median, days) | 2 | 4 | 3 | 2 | 2 | 1 | 2 | <0.001 |
| Duration of hospitalization (median, days) | 16 | 19 | 17 | 17 | 17 | 16 | 15 | <0.001 |
| General complications | 26.3% | 33.8% | 26.1% | 29.3% | 27.9% | 19.5% | 25.2% | |
| Pulmonary complications | 8.7% | 10.4% | 9.9% | 9.9% | 8.1% | 4.9% | 8.5% | – |
| Cardiovascular complications | 7.5% | 11.4% | 8.5% | 8.1% | 7.0% | 3.3% | 7.5% | – |
| Dialysis | 1.5% | 1.6% | 1.7% | 1.8% | 1.5% | 1.4% | 1.3% | – |
| Specific complications | ||||||||
| Secondary bleeding | 2.0% | 1.1% | 2.5% | 2.1% | 2.3% | 2.6% | 1.7% | – |
| Intestinal ischemia | 1.0% | 1.1% | 1.1% | 1.0% | 0.5% | 0.7% | 1.2% | – |
| Peripheral arterial thrombosis/embolism | 2.4% | 2.5% | 2.1% | 3.4% | 2.1% | 2.0% | 2.3% | – |
| Perioperative fatality | 3.2% | 5.2% | 3.8% | 4.0% | 3.2% | 3.1% | 2.6% | <0.001 |
| Discharged home | 75.7% | 69.8% | 78.0% | 74.3% | 78.5% | 60.9% | 77.5% | – |
| Transferred to another hospital or rehabilitation clinic | 21.0% | 25.1% | 18.3% | 21.7% | 18.4% | 36.0% | 19.9% | – |
The step-wise logistic regression model identified the following parameters as highly significant predictors for an increased perioperative fatality: increasing age, increasing maximal AAA-diameter, higher ASA status, length of the surgical procedure, need for suprarenal clamping, need for blood transfusion. The existence of an inflammatory AAA, a co-existent aneurysm of the iliac artery, the surgical technique (including inferior mesenteric artery reimplantation and the type of prosthesis material used) did not significantly influence the results (Table 3).
Table 3. Multivariate analysis to determine statistically independent predictors of an elevated perioperative risk during conventional surgery on an unruptured AAA (131 hospitals, n=10163, 1999–2004). Annual volume in different volume groups
| Variable | Odds-Ratio (95% CI) | P∘ |
|---|---|---|
| Age (continuous) | 1.087 (1.069–1.105) | <0.001 |
| Maximum AAA-diameter | 1.009 (1.002–1.017) | 0.017 |
| Duration of the procedure (min) | 1.008 (1.006–1.009) | <0.001 |
| ASA-Score (I-IV) | 2.652 (2.142–3.282) | <0.001 |
| Suprarenal clamping | 1.533 (1.065–2.206) | 0.021 |
| Blood substitution | 1.786 (1.268–2.514) | <0.001 |
| Annual volume | ||
| 1–9 | 1.903 (1.124–3.222) | 0.064 |
| 10–19 | 1.237 (0.897–1.705) | 0.778 |
| 20–29 | 1.375 (0.990–1.908) | 0.578 |
| 30–39 | 1.092 (0.765–1.559) | 0.225 |
| 40–49 | 1.249 (0.779–2.003) | 0.896 |
| >/=50* | – | – |
Annual volume as a continous variable had a moderate influence of perioperative mortality with an OR of 1.003 (1–1.006) but failed to reach statistical significance in the step-wise regression model (p=0.07). The analysis of different volume groups together with the predictive parameters from the step-wise regression model confirmed that patient's age, ASA status, AAA-diameter, length of procedure, blood transfusion and suprarenal clamping were predictive factors. In addition a low annual volume of 1-9 open AAA-repairs was associated with a significantly higher risk than a high annual volume of more than 50 procedures (OR 1.903, 95% CI 1.124–3.222). The relative risk thus increased by 90%. There were no significant differences with regard to the other volume groups (Table 4, Fig. 1, Fig. 2).
Table 4. Statistically independent predictors of an elevated perioperative mortality during conventional surgery on an unruptured AAA (131 hospitals, n=10163, 1999–2004). Annual volume as continuous variable
| Variable | Odds-Ratio (95% CI) | P∘ |
|---|---|---|
| Age (continuous) | 1.084 (1.066–1.102) | <0.001 |
| AAA-diameter | 1.008 (1.001–1.016) | 0.017 |
| Length of the procedure | 1.008 (1.006–1.009) | <0.001 |
| ASA-Score (I-IV) | 2.636 (2.129–3.264) | <0.001 |
| Suprarenal clamping | 1.447 (1.008–2.078) | 0.045 |
| Blood substitution | 1.786 (1.268–2.514) | <0.001 |
| Annual volume | 1.003 (1–1.006) | 0.075 |
Fig. 1
Forest plot illustrating statistically independent predictors of an increase of perioperative mortality (stepwise logistic regression model, annual volume as continuous variable).
Fig. 2
Forest plot illustrating statistically independent predictors of higher perioperative risk including different volume groups (multivariate analysis). An annual volume of 50 AAA-repairs was used as reference category.
Discussion
This study evaluated the influence of several preoperative and intraoperative variables on the perioperative death rate following elective open repair of unruptured AAAs. This is the first report with a special emphasis on the relationship between annual volume and perioperative mortality in German hospitals. We identified age, ASA status, AAA diameter and a need for blood transfusion as the strongest predictors for perioperative mortality, followed by suprarenal clamping and length of the surgical procedure. Annual volume as a continous variable had a moderate influence on outcome (OR 1.003, 95% CI 1–1.006), but failed to reach statistical significance in the stepwise logistic regression model (p=0.076). Further multivariate analysis for different volume groups showed that an annual volume <10 is a statistically independent predictor of an increased perioperative mortality with a relative risk increase of 90% compared to hospitals with more than 50 open AAA- repairs per year.
In the literature several several studies and a recent review were able to demonstrate a positive volume effect for elective open repair for unruptured AAA.5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 In a meta-analysis from Holt et al. looking at 421.299 elective and 45.796 ruptured AAAs a weighted odds ratio of 0·66 (0·65 to 0·67) for elective repair at a threshold of 43 AAAs per annum and 0·78 (0·73 to 0·82) for ruptured aneurysm repair at a threshold of 15 AAAs per annum were found, both in favour of high-volume institutions.15 These data are corroborated by a retrospective review of surgery for AAA in the UK between 2000 and 2005 where an overall mortality of 7.4% and a threshold of 32 elective AAA repairs per annum was found. Interestingly Hennebiens et al. found only some evidence in their literature review including papers between 1966 and 2006. A significantly lower mortality rate for high-volume hospitals was found in 14/24 studies with cut-off values in the range from 8 to 50 procedures annually. In addition a wide range of peri-operative mortality was found both in low volume hospitals (3.0 to 13.8%, median 6.2%) and in high volume hospitals (1.8%–7.4%, median 4.3%). Hennebiens concluded that quality of care may also be dependent on other factors such as hospital related factors (infrastructure), surgeon's experience (surgeon volume and certification), co-morbidity and population based factors such as age and sex.13 Our data confirm this careful assessment.
In another systematic review about the impact of the volume of surgery and specialization on patient outcome, Chowdhury addressed several methodological issues that have to be considered when drawing conclusions from volume-outcome studies.7 It is of major concern that in almost every study a different definition of “high volume” was used. The different definitions are often chosen arbitrarily, or the study population is divided in halfs by calculating the median value, or in trintiles, quartiles or quintiles. Depending on the data source and the number of procedures this will inevitably lead to different volume groups. Use of volume as continuous variable is reasonable to prove a significant effect of volume but not helpful in defining a threshold. Therefore Chowdhury recommended using graded volumes to identify a true cut-off and to allow a proper definition of high volume. We followed these suggestions using the annual volume of open AAA-repair as continuous variable in the logistic regression model followed by another regression analysis with volume groups for low volume hospitals (1-9 AAA repairs per annum) and high volume hospitals (>50 AAA repairs per annum).
Another matter of debate is that some studies have found an even stronger association between surgeon volume and specialisation than hospital volume. This may also be true for AAA repairs. Due to privacy legislation, no data regarding complication rates for individual surgeons are available in Germany. For the same reason no information is available about qualifications and specialisation. Previous studies have shown that surgical experience and a professional qualification as a vascular surgeon are associated with a lower surgical risk.7, 23
The source of data and selection of patients and hospitals is also important when interpreting data from volume-outcome studies, since significant differences between routine hospital administrative data and the Vascular Database may occur. Under-reporting cannot be excluded in society-based registries.26, 27 This concern may be applicable for this study which was based on data from a voluntary AAA registry organised by the German Society of Vascular Surgery. Data from the Federal Statistical Agency and the German Hospital Institute indicate that about 30% of hospitals and 40% of all surgically treated AAAs are represented in this registry.24, 25 Unfortunately no further pooled data on AAAs exists in Germany and therefore our results cannot be verified. It could be assumed that hospitals and (vascular) surgeons with a special interest in AAAs are overrepresented in the registry and this would also explain the low overall mortality rate of 3.2% in this study. On the other hand, low-volume hospitals may be underrepresented in our registry. This assessment is supported by a survey of the German Hospital Institute which showed that 415 of 494 hospitals (84%) only conduct 1-29 AAA-repairs per year. This indicates that the volume-outcome relationship found in our study would be even more significant if all low-volume hospitals had been analysed. The definition of a mandatory minimum of AAA repairs in Germany should also reflect that hospitals conducting very few operations actually may also have low complication rates and that the availability of emergency care for ruptured AAAs should not be endangered. Therefore, documentation of AAA repairs should become mandatory for all hospitals treating AAAs. This would enable smaller hospitals in areas with poor infrastructure to continue conducting AAA repairs, as long as they were able to document good outcomes.
In conclusion, this registry-based study showed that age, ASA status, AAA diameter, length of surgical procedure, suprarenal clamping and need for blood transfusion are significant predictors for perioperative mortality following elective open AAA repair in German hospitals that participated in the Registry. Annual volume as a continous variable is also moderately predictive and a low annual volume of 1-9 AAA repairs was associated with a significantly higher mortality compared to a high annual volume of 50 AAA repairs or more.
References
- . Guidelines for the treatmeno of abdominal aortic aneuryms. Report of a subcommitte of the Joint Counsil of the American Association for Vascular Surgery and Society for Vascular Surgery. J Vasc Surg. 2003;37:1106–1117
- Two-year outcomes after conventional or endovascular repair of abdominal aortic aneurysms. N Engl J Med. 2005;352:2398–2405
- . Endovascular aneurysm repair versus open repair in patients with abdominal aortic aneurysm (EVAR trial 1): randomised controlled trial. Lancet. 2005;365:2179–2186
- . Comparison of endovascular aneurysm repair with open repair in patients with abdominal aortic aneurysm (EVAR trial 1), 30-day operative mortality results: randomised controlled trial. Lancet. 2004;364:843–848
- Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346:1128–1137
- . Surgeon volume and operative mortality in the United States. N Engl J Med. 2003;349:2117–2127
- . A systematic review of the impact of volume of surgery and specialization on patient outcome. Br J Surg. 2007;94:145–161
- . The Leapfrog volume criteria may fall short in identifying high-quality surgical centers. Ann Surg. 2003;238(4):447–455[discussion 455–7]
- . Results of elective abdominal aortic aneurysm repair in the 1990s: a population-based analysis of 2335 cases. J Vasc Surg. 1999;30:985–995
- . The volume-outcome effect for abdominal aortic surgery: differences in case-mix or complications?. Arch Surg. 2002;137:828–832
- . Survival after ruptured abdominal aortic aneurysm: effect of patient, surgeon and hospital factors. J Vasc Surg. 2004;39:1253–1260
- . 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
- . Relation between hospital volume and outcome of elective surgery for abdominal aortic aneurysm: a systematic review. Eur J Vasc Endovasc Surg. 2007;33:285–292
- . Epidemiological study of the relationship between volume and outcome after abdominal aortic aneurysm surgery in the UK from 2000 to 2005. Br J Surg. 2007;94:441–448
- . Meta-analysis and systematic review of the relationship between volume and outcome in abdominal aortic aneurysm surgery. Br J Surg. 2007;94:395–403
- . Is there a minimum caseload that achieves acceptable operative mortality in abdominal aortic aneurysm operations?. Eur J Vasc Endovasc Surg. 2006;32:273–276
- . Mortality in abdominal aortic aneurysm surgery – the effect of hospital volume, patient mix and surgeon's case load. Eur J Vasc Endovasc Surg. 1997;14:375–379
- . Relation of surgical volume to outcome in eight common operations: results from the VA National Surgical Quality Improvement Program. Ann Surg. 1999;230(3):414–429[discussion 429–32]
- . The importance of surgeon volume and training in outcomes for vascular surgical procedures. J Vasc Surg. 1999;29(5):768–776[discussion 777–8]
- . Is there a positive volume-outcome relationship in peripheral vascular surgery? Results of a systematic review. Eur J Vasc Endovasc Surg. 2000;20(4):326–335
- . Differences in operative mortality between high- and low-volume hospitals in Ontario for 5 major surgical procedures: estimating the number of lives potentially saved through regionalization. CMAJ. 2003;68:1409–1414
- . Is volume related to outcome in health care? A systematic review and methodologic critique of the literature. Ann Intern Med. 2002;137(6):511–520
- . The influence of surgical specialty training on the outcomes of elective abdominal aortic aneurysm surgery. J Vasc Surg. 2001;33:447–452
- www.destatis.de
- . Umverteilungswirkungen der Mindestmengenregelung, Repräsentativerhebung des Deutschen Krankenhaus-Instituts. Das Krankenhaus. 2004;10:787–790
- . Descriptive study comparing routine hospital administrative data with the vascular society of Great Britain and Ireland's National Vascular Database. Eur J Vasc Endovasc Surg. 2007;33:461–465
- . Accuracy of clinical data in a population based vascular registry. Eur J Vasc Endovasc Surg. 2004;27:216–219
☆ Results from the AAA-Registry by the German Society for Vascular Surgery (Deutsche Gesellschaft für Gefäßchirurgie) 1999–2004.
PII: S1078-5884(07)00345-0
doi:10.1016/j.ejvs.2007.05.006
© 2007 European Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
Volume 34, Issue 3 , Pages 260-266, September 2007
