• 
• 

Three Part Question

In [patients presenting with a ruptured abdominal aortic aneurysm] does [the establishment of a multidisciplinary EVAR protocol] improve [mortality and morbidity outcomes]?

Clinical Scenario

A patient presents with a known abdominal aortic aneurysm presents to your hospital with symptoms indicating rupture. The patient is haemodynamically stable enough to have a CT scan, which confirms this diagnosis. The patient is treated by traditional open surgery, though it is noted that the anatomy of the aneurysm is suitable for endovascular repair (EVAR) using a stent graft. Subsequently the case prompts discussion about establishing a service for the treatment of ruptured abdominal aortic aneurysms (rAAA) preferentially by EVAR. You decide to search for the evidence for this in order to investigate whether such a service would be beneficial to patients.

Search Strategy

Medline 2005-August 2009. Aortic Rupture"[Mesh] AND endovascular[All Fields]) OR EVAR[All Fields]

In view of successive generational advances in aortic endograft design and expertise, series published prior to 2005 were not used for this analysis. Reports preceding this year were felt to not be translatable to contemporary endovascular practice. Additionally, studies that included symptomatic but unruptured aneurysms were discarded.

Search Outcome

1327 papers were found of which 26 were deemed to be relevant. These papers are presented in Table 1.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
F J Veith et al 2009 USA	1037 patients treated by EVAR for rAAA across 49 centres between 2002-08. Questionairre sent out to centres whom perform EVAR-first approach to treatment of rAAA between 2002-07 Updated questionnaire was sent December 2008 yielding results from 13 centres whom were known to treat hameodynamically unstable patients by EVAR-first protocol. From these, 680 patients were treated by EVAR (included in overall 1037 analysed) and 763 treated by OR.	Co-operative multi-centre cohort study across a total of 49 institutions. (level 2b)	30-day mortality	Overall 30-day mortality amongst EVAR patients = 21.2% Amongst 680 patients from the 13 centres sent second questionnaire, 30-day mortality for EVAR was 19.7% with a range of 0% to 32. 30-day mortality amongst open-repair patients = 35.8%	Largest study of its type – pooling all consecutive patients treated by EVAR for rAAA across a large number of centres. Authors claim this pooled data shows EVAR to be unequivocally superior to OR for the treatment of anatomically suitable rAAA. Several aspects of care in the 46 original centers varied greatly – such as exclusion of haemodynamically unstable patients, method of insertion, and type of graft used. Analysis of the 680 patients from 13 centres known to treat all patients with anatomically-suitable neck yielded favourable survival figures amongst a known high-risk population. Statistically superior to OR patients treated over this period at same centres: 30-day mortality 19.7% for EVAR vs. 36.3% for OR (P <0.0001). The validity of this comparison is limited as the cases of RAAA were still not comparable in the 2 groups. More anatomically difficult patients were subjected to OR, and they may have also had more high risk factors.
C D Karkos et al 2009 Greece	Systematic review of literature published Jan 1996 - Dec2006. Only patients with true ruptures were included. Additionally, information on mortality after concurrent open repair was sought. 29 studies with total of 897 patients who underwent endovascular repair of rAAA	Systematic review of non-randomised observational studies. (level 2a)	In-hospital and/or 30-day mortality	In-hospital and/or 30-day mortality ranged between 0% and 54% in different series. Pooled mortality after endovascular repair was 24.5% (95% confidence interval [CI], 19.8%-29.4%). In 19 studies reporting results of both endovascular and concurrent open repair from the same unit, the pooled mortality after open repair was 44.4% (95% CI, 40.0%-48.8%), and the pooled overall mortality for RAAA undergoing endovascular or open repair was 35% (95% CI, 30%-41%).	All included studies non-randomised so selection bias present throughout. Procedure protocols varied across included studies, and lack of standardised reporting of outcomes in included studies. Some studies considered haemodynamic instability a contraindication to endovascular repair, selecting a cohort of patients for this intervention with a likely better prognosis than the corresponding patients treated by open repair in the same unit during the period of study. 19 of the included studies provided mortality data for contemporary open repair during the same period with a mortality of 44.4%. When pooling data from these 19 studies from surgical units that offered their patients both open and endovascular repair, the overall RAAA mortality of 35% is much lower than that which is generally accepted.
Vogel TR, Dombrovskiy VY, Haser PB, Graham AM. 2009 USA	State inpatient data used to identify all 700 patients who underwent repair of rAAA during period 2001-2005. 618 patients (88%) were treated by open surgery, and 82 (12%) with EVAR.	Retrospective large, national cohort study (level 2b)	In-hospital mortality, hospital length of stay, cost	Mortality was not significantly different between EVAR and open repair groups (45.1% vs. 52.4%, p=0.21). Length of stay was not significantly different between the EVAR and open-repair groups.	No significant difference in outcomes between EVAR and open repair groups observed in this large national dataset.
K A Giles et al 2009 USA	567 patients who underwent repair of rAAA between 2005-07 were identified from the American College of Surgeons NSQIP database. Of these, 121 patients (21%) underwent EVAR, and 446 were repaired by open surgery.	Observational (level 2b)	30-day mortality, pre-operative status, and co-morbidities	Mortality of successful EVAR was 24% vs. 36% for open repair (p<0.05). Preoperative hemodynamic status was similar based on need for .4 units of blood (3% vs. 6%, p=0.31), intubation (12% vs. 17%, p=0.18), impaired sensorium (7% vs. 11%, p=0.25), coma (4% vs. 5%, p=0.65), acute renal failure (2% vs. 2%, p=0.60), and ASA class 5 (29% vs. 34%, p=0.29). EVAR patients had greater incidences of recent myocardial infarction (7% vs. 2%, p<0.05), revascularization or amputation for peripheral vascular disease (8% vs. 3%, p<0.05), and cerebrovascular disease (22% vs. 11%, p<0.01).	The NSQIP database is contributed prospectively to by 221 institutions, both academic and community hospitals, across the USA. In this large, multi-centre cohort of patients, the rate of mortality was found to be significantly lower in patients treated by EVAR despite similar preoperative status, and greater co-morbidities in the EVAR cohort. Complication rates were similar in the two groups. Pre-operative status was described by the markers: preoperative intubation, acute renal failure, coma, impaired sensorium, and preoperative blood transfusion. More specific information on blood pressure parameters, tachycardia, and pressor dependence was, however, not available.
J Visser et al 2009 Netherlands	201 consecutive patients with rAAA treated at seven institutions in the Netherlands between 22nd December 2004 and 31st October 2006. All of the hospitals preferentially treated patients with rAAA by EVAR. 58 patients underwent EVAR, and 143 patients were treated by open surgery.	Multi-centre prospectively recruited cohort study (level 2b)	30-day mortality	Thirty-day mortality was 15/58 (26%) for patients treated with endovascular repair and 57/143 (40%) for patients treated with open surgery (P =0.06).	A survival benefit in 30-day mortality was observed in the EVAR group, however this did not quite reach statistical significance. The treatment protocols between the participating hospitals were slightly different. The criteria for patients being hemodynamically unstable differed across the hospitals, and anatomical criteria differed as different models of endograft were employed.
J McPhee et al 2009 USA	Nationwide inpatient sample used to identify all patients treated with open or EVAR for rAAA in the period 2001-06 at a national level in USA = 27,750. Used ICD-9 diagnostic & procedure codes	Retrospective large, national cohort study (level 2b)	In-hospital mortality	11.5% of RAAA treated with EVAR –proportion increased over time (5.9% in 2001 to 18.9% in 2006, P < .0001) EVAR had a lower overall in-hospital mortality than open repair (31.7% vs. 40.7%, P < .0001), an effect which amplified when stratified by institutional volume EVAR had a shorter length of stay (11.1 vs. 13.8 days, P < .0001), higher discharges to home (65.1% vs. 53.9%, p < .0001), and lower charges ($108,672 vs. $114,784, p < .0001).	No longer-term data on survival, complication, or cost reported. Incidence of rAAA remained relatively constant during study period NIS, unlike Medicare data, incorporates all-payer information, hence including patients <65yrs and uninsured patients
D Mayer et al 2009 Switzerland	102 consecutive patients treated by EVAR for rAAA at single tertiary centre over 10 year period from Jan 1998 to April 2008. An EVAR-first protocol was in place from January 2000.	Single-centre retrospective cohort study (level 2b)	30-day mortality, major morbidity.	30 day mortality for this cohort was 13%. Major 30-day morbidity rate was 35%	Hetrogeneous population, with patients in the first 2 years being treated by EVAR on selection of surgeon and team, whereas after Jan 2000 an EVAR-first protocol was introduced. The only exclusion criteria was unsuitable anatomy.
U Sadat et al 2009 UK	50 consecutive patients presenting between Jan 2006 –Dec 2007 (once EVAR protocol for rAAA established), comparison with 71 consecutive patients presenting between Jan 2003 – Dec 2005 (before eEVAR protocol established). Patients with ruptured AAA only – did not include acute, non-ruptured AAA. No significant differences between 2 groups in terms of age and co-morbidity	Cohort study – data collected for cases collected prospectively & compared to historical group. (level 2b)	In-hospital mortality before & after introduction of eEVAR protocol	Group 1 = 17 eEVAR, 29 open repairs, 4 palliated (after the introduction of eEVAR), Group 2 = 54 underwent open repair and 17 were palliated. Total in-hospital mortality was significantly lower in Group 1 20% (eEVAR (n=1), 6%: Open (n=5), 17%: palliated (n=4), 100%) when compared to Group 2 54% (Open (n=21), 39%: palliated (n=17), 100%) (p=0.000001). Significant differences in 30-day operative mortalities between the two groups 13% in Group 1 versus 39% in Group 2 (p=0.0003).	Comparison over 2 different time periods may be biased by changes in staff and improvements in peri-operative care. Proportion of patients who were palliated significantly decreased following introduction eEVAR protocol (8% Group 1 versus 24% Group 2, p=0.01).
A Azizzadeh et al 2008 USA	Systematic review of literature – 1200 pooled patients treated for rAAA in 34 publications deemed appropriate for analysis. 531 patients (44.3%) underwent endovascular repair.	Systematic review of non-randomised observational studies. (level 2b)	Mortality rate.	Mortality rate of 30.2%. A 3.8% reduction in mortality observed for each 10% increase in percentage of ruptures treated endovascularly at each centre.	All included studies non-randomised so selection bias present throughout. Procedure protocols varied across included studies, and lack of standardised reporting of outcomes in included studies. The observation of increased mortality in centres where lower proportion of ruptures treated by endovascular means suggests that a team’s familiarity with the procedure improves the outcomes, and would support the proposal that high-volume centres should be treating this condition. The higher mortality rates in centres performing a greater proportion of repairs of rAAA by endovascular methods refutes claims that selection of more haemodynamically stable patients for endovascular repair is responsible for the improved mortality seen in many studies.
jonathan ghosh 23 Oct 2009 Netherlands	135 consecutive patients admitted at single centre over 5 year period 2002-2006 with confirmed rAAA following introduction of strict protocol for selective use of EVAR for rAAA. 89 open repair, 36 EVAR, 10 palliative	Single-centre Cohort study, prospective collection of data. (level 2b)	30-day and in-hospital mortality	Overall mortality by intention-to-treat was 27%. Mortality as per on-treatment, with 93% of all RAAA treated, was 24%, with an open mortality of 28.1% and EVAR mortality of 13.9% (p=0.092) No significant difference in mortality over 25 month follow-up between EVAR and OR	Single centre trial with small numbers. 93% patients in study suitable for treatment – suggests not representative of general rAAA population. Haemodynamically unstable patients treated immediately by OR - likely influencing lower survival in OR cohort. Significant difference in selection of EVAR by surgeon suggests protocol not rigid enough to exclude selection bias. Evaluation for EVAR increased from 62.5% in the first year to 80% and more in the last 3 years of the study. The percentage of patients treated by EVAR increased gradually, from 18% to 37.5%, during the 5-year study period
N Egorova et al 2008 USA	ll 43,033 patients treated for rAAA identified from Medicare national USA dataset using ICD-9 codes, 41,969 treated by open repair; 1064 treated by EVAR during the period 1995-2004.	Case control study – 1044 patients treated with EVAR were paired to controls from the open repair group using a propensity score to match them for patient, surgeon and hospital characteristics. (level 2b)	Mortality evaluated over 4 year follow up period	Evaluation of EVAR vs. OR, prior to propensity matching, showed an initial mortality benefit of EVAR however this advantage disappeared at 90 days. Survival analysis of patients matched by propensity score showed a benefit of EVAR over OAR that persisted throughout the 4 years of follow-up (p=0.0042). Perioperative and long-term survival after rAAA repair correlated with increasing annual surgeon and hospital volume in OAR and EVAR and also with rAAA experience. rEVAR repair had a protective effect (HR=0.857, p=0.0061) on long-term survival controlling for co-morbidities, demographics, and hospital and surgeon volume	J Vasc Surg. 2008;48:1092–1100.
H S Rayt 2008 UK	Literature search of Medline & Embase databases revealed data from 31 studies concerning 982 patients treated with EVAR for rAAA. Excluded studies that assessed acute AAA rather than rAAA.	Separate meta-analyses of pooled mortality from 31 studies (taken as either in-hospital or 30-day). (level 2a)	Pooled morbidity from 21 studies (all morbidity grouped together) using random-effects models.	Overall pooled mortality for rAAA treated by EVAR = 24% with moderate degree of heterogeneity between studies. Overall pooled morbidity 44% with high degree of heterogeneity between studies.	All studies non-randomised observational studies. Demonstrates a large degree of publication bias - funnel plot for mortality showed a large degree of asymmetry with a suggestion that studies corresponding to high rates of mortality are missing. Data on morbidity variable and no description of severity of morbidity
A Wibmer et al 2008 Austria	89 consecutive patients treated for rAAA - 47 patients treated (16 EVAR, 31 open) over 41month period after introduction of protocol, compared with 41 patients treated over preceding 41 month period by open repair alone	Single centre retrospective cohort study Patients’ survival rates were calculated at 3-monthly intervals and compared using Kaplan-Meier method (2b)			Single centre, small numbers of patients. Comparison over 2 different time periods may be biased by changes in staff and improvements in peri-operative care.
			The 90-day mortality rate for patients undergoing rAAA repair in the first period was 54.8%, compared with 27.7% during the second period (p<0.01). Haemodynamically stable patients benefited from the introduction of EVAR for rAAA (40% vs. 10.7%; p<0.02), but the risk of death did not change for haemodynamically unstable patients (68.2% vs. 52.6%; p<0.35). Patients in the second period had a statistically significantly higher chance of surviving the early postoperative period and the first year after rupture (p<.04).
K Lesperance et al 2008 USA	The Nationwide Inpatient Sample (NIS) was used to identify 28,123 admissions for rAAA in USA from 2001 through 2004. Detailed analysis performed on 9931 patients from 2003-4 presenting with rAAA, of whom 8982 underwent open repair, and 949 underwent EVAR.	Multicentre, national, retrospective cohort study. (level 2b)	In-hospital mortality, length of hospital stay	EVAR use increased significantly from 6% of all rAAA in 2001 to 11% in 2004 (p< .01). Mortality for EVAR declined significantly from 43% to 29% (p< .01), but mortality with open repair showed no change (40% to 43%). From the 2003-4 data set, the EVAR patients had lower mortality (31% vs. 42%), shorter hospital stay (6 vs. 9 days), and were more likely to be discharged to home (59% vs. 37%) - all p < .01, than open repairs.	Analysis excluded patients transferred from other facilities to treating centre, and patients with rAAA who were palliated. Observed decrease in ruptured AAA presentation may reflect increased screening for AAA in USA during this period, which may confound mortality results. Retrospective review of a large prospectively collected administrative database, which can be affected by missing data fields and inaccurate or miscoded entries.
K Kubin et al 2008 Austria	22 patients who underwent EVAR for rAAA between 2000 and 2006 at single centre where EVAR used preferentially for rAAA– 17 male, 5 female; median age 74	Retrospective cohort study (level 2b)	30-day mortality, long-term mortality, early and late complication rates	30-day mortality rate =30%, early complication rate = 54%, With a mean clinical follow-up 744 ±480 days, the long-term mortality rate was 14% in the study cohort.	Patients only excluded from treatment with EVAR based on anatomical suitability – more haemodynamically unstable patients treated with EVAR than in other studies, likely resulting in higher early mortality rates.
R W Lee et al 2008 UK	52 consecutive patients treated for rAAA at single centre between July 2002 and May 2006 following the implementation of defined protocol with emphasis on EVAR for rAAA. 15 patients had para-renal AAA and were treated by open surgery. Of the 37 infra-renal rAAA, 20 were treated by open repair and 17 by EVAR.	Single-centre retrospective observational cohort (level 2b)	Mortality rates	Overall mortality was 53%. Mortality was significantly lower in the EVAR group compared to open – 35% EVAR vs. 75% in open infrarenal rAAA group (p<0.02). Mortality in the para-renal group was 47%.	Single centre non-randomised study with small numbers of patients involved. Despite significantly lower mortality in EVAR group, overall mortality remained unchanged from historical values. Review of case notes showed that many of the infra-renal rAAA treated by open surgery would have been anatomically suitable for EVAR, and that this was not performed due to haemodynamic instability. These findings suggest that the lower mortality rates observed for the patients treated by EVAR is due to selection bias.
S F Najjar et al 2007 USA	37 consecutive patients treated for ruptured AAA at single centre between January 1, 2000 and December 31, 2005. 15 Patients treated by EVAR, 22 patients treated by open repair.	Cohort study, single centre, non-randomised (level 2b)	30-day mortality, 30-day morbidity	EVAR 30-day mortality was 6.7%, compared with 13.6% for open repair group. 30-day morbidity was 6.7% for EVAR group (single case of abdominal compartment syndrome) vs. 50% for open repair group	Does not describe clear inclusion criteria for selecting EVAR approach, other than anatomic suitability. Only includes patients with confirmed ruptured AAA – symptomatic non-ruptured aneurysms not included. Aortic occlusion not required for any of the EVAR cases. The final 11 EVAR cases performed via entirely percutaneous system.
P Anain et al 2007 USA	All 40 consecutive patients presenting with rAAA to a single community hospital following introduction of an EVAR-first approach to patients presenting with CT/operation-confirmed rupture between Feb 2001 – Oct 2006. 30 underwent attempted EVAR. EVAR completed in 29 cases.	Retrospective cohort study of consecutive patients. (level 2b)	30 day mortality	Overall 30 day mortality =22.5 %, EVAR group 30 day mortality =16.6 %, open repair 30 day mortality = 40% (p=0.19)	Since EVAR-first approach, open rAAA repair does not represent a representative control group for comparison as selection bias of surgeon in selecting which patients to undergo EVAR. Overall 30 day mortality rate of 22.5% for all patients in study. Balloon excluder used if patients became too haemodynamically unstable during procedure. Mortality in patients who required balloon occlusion was 44% vs. 4% in those who did not (p=.019) in the EVAR group.
J Visser et al 2007 Netherlands	Systematic review of literature from Jan 1994 – Mar 2006 for studies with comparison of EVAR and open surgery for rAAA. Ten studies, in which the results of 478 procedures (148 EVAR, 330 open surgery) were reported.	Systematic review of non-randomised observational studies (level 2a)	30-day mortality and morbidity adjusted for inclusion criteria	Pooled 30-day mortality 22% (95% CI: 16%, 29%) for endovascular repair and 38% (95% CI: 32%, 45%) for open surgery. The crude odds ratio for 30-day mortality of endovascular repair versus open surgery was 0.45 (95% CI: 0.28, 0.72). Heterogeneity was demonstrated for the patients’ hemodynamic condition at presentation to the hospital across the studies for both treatment groups (p <0.01). After adjustment for patients’ hemodynamic condition at presentation to the hospital, the OR was 0.67 (95% CI: 0.31, 1.44; p=0.37), and indicates that the difference in 30-day mortality was, in part, explained by this variable. After adjustment for haemodynamic condition the difference in 30-day mortality no longer statistically significant. Pooled percentage of total systemic complications (i.e., the sum of cardiac, pulmonary, cerebrovascular, and renal complications; multiorgan failure; and sepsis) was 28% (95% CI: 17%, 48%) for endovascular repair and 56% (95% CI: 37%, 85%) for open surgery.	Adjustments made to account for selection of patients more haemodynamically unstable for open repair, however other selection bias not accounted for as all studies non-randomised. Procedure protocols varied across included studies, and lack of standardised reporting of outcomes in included studies. Funnel plot constructed by authors of study was not symmetrically shaped; small studies with higher mortality rates for endovascular repair than for open surgery seemed to be underrepresented. No long-term follow-up data reported by included studies.
R J Hinchliffe et al 2006 UK	Patients presenting with AAA who were considered fit for open repair randomised to EVAR/open repair after consent obtained. 32 patients recruited to study between 1st September 2002 and 31st December 2004. 17 patients were randomized to the open repair group. 13 patients had attempted EVAR, 2 required conversion to open repair.	Single-centre, prospective, randomised controlled trial (level 1b)	30-day mortality, moderate or severe complications	30-day intention-to-treat mortality 53% in EVAR group, 53% in open repair group. Moderate or severe operative complications occurred in 77% in the EVAR group and in 80% in the OAR group. More patients in the EVAR group suffered severe renal complications (6 (55%) versus 1 (8%) in open repair group (p=0.02). 4 patients died prior to surgery.	This represents the only randomised trial between EVAR and open repair for ruptured AAA. Demonstrates equal mortality between EVAR and open repair groups, in contrast to the many non-randomised studies described here. Patients were excluded if considered too haemodynamically unstable to undergo CT scan. Patients who were so unstable that the surgeon deemed CT scanning unethical did badly with open repair. Logistical difficulties experiences in availability of 24hr endovascular service at study centre. Power calculation required 100 patients to be randomised; however this was based on published data from non-randomised trials on selected patients. A much larger number may need to be randomised to provide statistically significant data given results of this study.
M A Sharif et al 2007 UK	126 patients treated for rAAA in a single unit over a 5-year period. 52 patients underwent EVAR and 74 patients underwent open repair.	Single centre, retrospective cohort study. (level 2b)	Five Hardman factors (age >76, History of loss of consciousness, ECG evidence ischaemia, Hb<9.0g/dl, and Serum creatanine >0.19mmol/l) were assessed and their association with in-hospital or 30-day mortality was assessed.	Mortality for open repair was 51.4% in comparison to 32.7% for the endovascular group (p=0.05). Mortality rates for open repair patients with Hardman scores <2 were 43.5% vs. 22.9% for the endovascular group (p=0.06). Mortality for patients with scores >/= 2 were 64.3% vs. 52.9% for the respective groups (p=0.54).	Difference observed between the open and endovascular repair groups bordered upon statistical significance. Those patients with Hardman score <2 showed a trend toward better survival following endovascular repair compared with open repair.
R Hassan-Khodja et al 2007 France	10 consecutive patients treated by EVAR for rAAA between April 2004 and December 2005 following introduction of protocol to consider all rAAA for EVAR repair.	Single-centre retrospective cohort study (level 2b)	30-day mortality	The 30-day mortality rate was 20%. Following the introduction of the endovascular repair protocol 59% of rAAA were treated by endovascular means.	Single-centre study with low volume of patients treated in this 21 month period. Patients whose haemodynamic status was too unstable to permit a preoperative CT scan and patients with unfavourable anatomy for endovascular repair were treated with open surgery. Endovascular repair was with aorto-uni-iliac system with femoro-femoral cross-over bypass and deployment of an occlude in the contralateral common iliac artery. Does not report data on outcome for those patients treated by open surgery in this period. The overall mortality for all patients treated for rAAA following introduction of the protocol may have remained unchanged.
S Ockert et al 2007 Germany	58 consecutive patients treated for rAAA with open or endovascular repair between January 2000 and December 2005. 29 patients were treated by endovascular techniques, and 29 patients with open repair.	Single-centre, retrospective cohort study. (level 2b)	30-day mortality, major morbidity, midterm mortality (after mean 40.25 month follow-up).	30-day mortality was 31% in both groups (p=1.0). Morbidity rates were 55.2% in the EVAR group vs. 62% in open repair group (p=0.9)	Only included patients with definitive rupture not acute symptomatic AAA. No significant difference observed in the 30-day or midterm mortality rate, nor major morbidity rate, between those patients treated by endovascular means and by open repair.
Hinchliffe RJ, Braithwaite BD. 2007 UK	54 patients who underwent endovascular repair of rAAA at single centre over a 10-year period (1994-2004).	Single-centre, retrospective cohort study. (level 2b)	Perioperative mortality, 3-year and 5-year survival rates.	Perioperative mortality rate was 37%. 3-year survival rate was 36%, and 5-year survival rate 26%.	Single centre retrospective study. Cohort includes only those patients who were treated by EVAR and hence selection bias as to which patients received this treatment. Suggests that EVAR does not appear to confer any overall survival advantage in the mid- to long-term compared with the published results for open repair.
J J Visser et al 2006 Netherlands	All 55 consecutive haemodynamically stable, conscious patients with rAAAs presenting to single tertiary referral centre between January 1, 2001, and December 31, 2005. Patients excluded if too haemodynamically unstable to undergo CT scan. Patients underwent EVAR if met anatomical criteria: 26 patients underwent endovascular repair, and 29 patients underwent open surgery.	Cohort study - part retrospective part prospective data collection, observational. (level 2b)	Intraoperative mortality, 30-day mortality, systemic complications, complications necessitating surgical intervention, and mortality and complications during 1-year follow-up.	Intra-operative mortality = 4% for endovascular repair group, 14% for open repair group (p=0.18). 30-day mortality = 31% for endovascular repair group, and 31% for open repair group (p=0.98). Complications requiring surgical intervention 19% EVAR group vs. 28% of open repair group (p=0.47). One-year follow-up was completed for 45 (82%) of 55 patients. 2 patients in the endovascular repair group died during this follow-up period. Total deaths at 1year follow-up were 40% of endovascular repair group vs. 38% of open repair group (p=0.78).	Exclusion of haemodynamically unstable patients makes comparison of endovascular and open repair groups more adequate than some other studies, however still a degree of selection bias and heterogeneity between the study groups. “Logistical reasons” cited as reason for open as opposed to EVAR repair in 3 patients, representing unexplained factors likely introducing bias. Small numbers of patients in non-randomised, single centre trial.
G Coppi et al 2006 Italy	124 consecutive patients presenting to a single centre with confirmed rAAA between Dec 1999 and April 2006 were considered according to an intention-to-treat model with EVAR. 33 patients underwent EVAR, 91 patients underwent open repair.	Cohort study, retrospective, single centre, non-randomised. (level 2b)	30-day mortality, overall reintervention rate, and severe complication rate.	Overall 30-day mortality for EVAR was 30% (unstable, 53%; stable, 11%), vs. 46% for open repair (unstable, 61%; stable, 21%). The EVAR postoperative re-intervention rate (within 30 days) was 15% (unstable, 20%; stable, 11%), and for open repair it was 10% (unstable, 9%; stable, 15. Severe complication rate for EVAR patients was 27% (unstable, 40%; stable, 17%), and for patients treated with open repair, it was 33% (unstable, 35%; stable, 29%).	31 of the 91 open repair patients would have been anatomically suitable for EVAR but were treated with open repair at the beginning of this centre’s experience: 6 patients due to young age (initially considered a criterion of exclusion) and 25 patients due to unavailability of adequately trained staff and endovascular supplies. Seventy-two patients (EVAR, 45%; open, 63%) were classified as hemodynamically unstable at arrival, and 52 were classified as stable (EVAR, 55%; open, 37%). 9% of EVAR group required immediate conversion to open repair

Comment(s)

Direct comparison between EVAR and open surgery for ruptured abdominal aortic aneurysm is made challenging as not all patients are anatomically suitable for the former approach. In addition, haemodynamically decompensating patients may not survive the time interval for radiological assessment prior to definitive treatment. This best evidence topic search does not aim to determine the optimal form of endovascular repair or explore economic issue, but rather determine whether an ‘EVAR-first’ approach to ruptured AAA is associated with favourable mortality compared to the traditional open approach. To address this clinical question, various sources of papers are to be considered. The most prevalent studies published addressing this issue are observational studies. There are a number of single-centre cohort studies, as well as larger observational studies of multiple centres or national datasets. There are also papers that systematically review data from multiple non-randomised, observational studies. There is just a single paper published which reports upon a randomised control trial, Hinchliffe et al. This single-centre study showed a higher EVAR related mortality compared to observational studies and did not find any overall survival advantage in the mid- to long-term with EVAR compared to open repair. Numerous observational studies have been performed whereby the outcomes of ruptured AAA are reported following the introduction of an EVAR-first protocol for the treatment of rAAA. The specifics of the protocol varied from study-to-study, and the proportion of rAAA treated by EVAR also differs. This heterogeneity reflects the inherent selection bias present in these observational studies. Notably there is considerable variance in the definition of haemodynamic stability between series. A multi-centre prospectively recruited cohort study showed near statistical significance for a 30-day survival benefit with EVAR over open surgery, although this study was constrained by variation in treatment protocols between the participating hospitals. Most recently, a study describing pooled results obtained from 49 separate centres showed No observational study concluded any superiority of open surgery over EVAR with reference to mortality or morbidity, two found equivalent mortality and the remainder found improved 30-day survival with EVAR. Nevertheless, limited patient numbers, selection bias and heterogeneity in treatment protocols limit the generality of such findings. Specifically it is difficult to assess the impact of patient stability to patient selection and thus outcome. Indeed, it is pointed out by several authors that patients stable enough for CT scanning are already a self-selecting group compared to those that are in extremis and cannot wait for imaging. A study by Veith et al attempts to address this issue of haemodynamic instablility by pooling data from 13 centres whom perform EVAR on all anatomically suitable rAAA, and found siginificantly better 30-day survival amongst those patients treated by EVAR in comparison with open repair. Finally, synthesis of published data has been performed in systematic reviews and meta-analyses. Karkos et al. found that, in institutions performing both procedures, overall mortality following undergoing endovascular repair was lower than open (24.5% v 44.4%). Similarly, Rayt et al found the post- EVAR mortality and morbidity rates to be 24% and 44% respectively, comparing favorably with published outcomes of open surgery. Funnel plots of mortality data however did suggest possibility of publication bias in favour for EVAR and both reviews comment on the potential for bias in the source data. These biases will be reduced following synthesis of future level 1b evidence, such as the forthcoming IMPROVE trial (www.improvetrial.org), though it should be noted that such data is still several years away.

Editor Comment

ICVTS

Clinical Bottom Line

Endovascular repair as the primary treatment for ruptured AAA is achievable and appears to be associated with favourable mortality over open repair with appropriate case selection. Appraisal is however constrained by heterogeneity between individual series, bias and dependence on predominantly level 2b data.

References

1. Veith FJ, Lachat M et al. Collected World and single center experience with endovascular treatment of ruptured abdominal aortic aneurysms. Ann. Surg. 1009; 818-24.
2. Karkos CD, Harkin DW, Giannakou A, Gerassimidis TS. Mortality after endovascular repair of ruptured abdominal aortic aneurysms – A systematic review and meta-analysis. Arch Surg. 770-8.
3. Vogel TR, Dombrovskiy VY, Haser PB, Graham AM. Has the implementation of EVAR for ruptured AAA improved outcome? Vasc Endovascular Surg. 2009;43:252-7.
4. Giles KA, Pomposelli FB, Hamdan AD, Wyers MC, Schermerhorn ML. Comparison of open and endovascular repair of ruptured abdominal aortic aneurysms from the ACS-NSQIP 2005–07. J Endovasc Ther. 2009;16:365-72.
5. Visser JJ, Williams M, Kievit J, Bosch JL. Prediction of 30-day mortality after endovascular repair or open surgery in patients with ruptured abdominal aortic aneurysms. j Vasc Surg 2009;49:1093-9.
6. McPhee J, Eslami MH, Arous EJ, Messina LM, Schanzer A. Endovascular treatment of ruptured AAA in the United States [2001-2006]: A significant survival benefit over open repair is independently associated with increased institutional volume. J Vasc Surg 2009;49:817-26
7. Mayer D, Pfammatter T, Rancic Z, Hechelhammer L, Wilhelm M, Veith FJ, Lachat M. 10 years of emergency endovascular aneurysm repair for ruptured abdominal aortoiliac aneurysms: lessons learned Ann Surg 2009; 249:510 –515.
8. Sadat U, Hayes PD, Kullar PJ, Cousins C, Varty K, Boyle JR. An emergency EVAR service reduces mortality in ruptured abdominal aortic aneurysms. EJVES 2009;37:189-93
9. Azizzadeh A, Villa MA, Miller CC 3rd, Estrera AL, Coogan SM, Safi HJ. Endovascular repair of ruptured abdominal aortic aneurysms: systematic literature review. Vascular 2008;16:219-24.
10. Verhoeven EL, Kapma MR, Groen H, et al. Mortality of ruptured abdominal aortic aneurysm treated with open or endovascular repair. J Vasc Surg. 2008;48:1396–1400. 2008;48:1396–1400.
11. Egorova N, Giacovelli J, Greco G, et al. National outcomes for the treatment of ruptured abdominal aortic aneurysm: comparison of open versus endovascular repairs. J Vasc Surg 2008;48:1092–1100.
12. Rayt HS, Sutton AJ, London NJ, Sayers RD, Bown MJ. A systematic review and meta-analysis of endovascular repair (EVAR) for ruptured abdominal aortic aneurysm. EJVES 2008;143:544-549.
13. Wibmer A, Schoder M, Wolff KS, et al. Improved survival after abdominal aortic aneurysm rupture by offering both open and endovascular repair. Arch Surg 2008;143:544-549.
14. Lesperance K, Andersen C, Singh N, Starnes B, Martin MJ. Improved survival after abdominal aortic aneurysm rupture by offering both open and endovascular repair. J Vasc Surg 2008;47:1165-70.
15. Kubin K, Sodeck GH, Teufelsbauer H, Nowatschka B, Kretschmer G, Lammer J, Schoder M Endovascular therapy of ruptured abdominal aortic aneurysm: mid- and long-term results. Cardiovasc Intervent Radiol 2008;31:496-503.
16. Lee RW, Rhodes JM, Singh MJ, Davies MG, Wolford HY, Diachun C, Norton R, Illig KA, Is there a selection bias in applying endovascular aneurysm repair for rupture? Ann Vasc Surg 2008;22:215-20
17. Najjar SF, Mueller KH, Ujiki MB, Morasch MD, Matsumura JS, Eskandari MK. Percutaneous endovascular repair of ruptured abdominal aortic aneurysms. Arch Surg 2007;142:1049-52.
18. Anain PM, Anain JM Sr, Tiso M, Nader ND, Dosluoglu HH. Early and mid-term results of ruptured abdominal aortic aneurysms in the endovascular era in a community hospital. J Vasc Surg. 2007;46:898-905
19. Visser JJ, van Sambeek MR, Hamza TH, Hunink MG, Bosch JL. Ruptured abdominal aortic aneurysms: endovascular repair versus open surgery–systematic review. Radiology 2007;245:122-9.
20. Hinchliffe RJ, Bruijstens L, MacSweeney STR, Braithwaite BD. A randomised trial of endovascular and open surgery for ruptured abdominal aortic aneurysm: results of a pilot study and lessons learned for future studies. EJVES 2006;32(5):506-51
21. Sharif MA, Lee B, Makar RR, Loan W, Soong CV. Role of the Hardman index in predicting mortality for open and endovascular repair of ruptured abdominal aortic aneurysm. J Endovasc Ther. 2007;14:528-35
22. Hassan-Khodja R, Jean-Baptiste E, Haudebourg P, Declemy S, Batt M, Bouillanne PJ. Endovascular repair of ruptured aneurysms of the infrarenal abdominal aorta: feasibility and results J Cardiovasc Surg (Torino). 2007;48:491-5.
23. Ockert S, Schumacher H, Böckler D, Megges I, Allenberg JR. Early and midterm results after open and endovascular repair of ruptured abdominal aortic aneurysms in a comparative analysis. J Endovasc Ther 2007;14:324-32.
24. Hinchliffe RJ, Braithwaite BD. Ruptured abdominal aortic aneurysm: endovascular repair does not confer any long-term survival advantage over open repair. Vascular 2007; 15:191-6.
25. Visser JJ, Bosch JL, Hunink MG, van Dijk LC, Hendriks JM, Poldermans D, van Sambeek MR. Endovascular repair versus open surgery in patients with ruptured abdominal aortic aneurysms: clinical outcomes J Vasc Surg 2006;44:1148-55.
26. Coppi G, Silingardi R, Gennai S, Saitta G, Ciardullo AV. A single-center experience in open and endovascular treatment of hemodynamically unstable and stable patients with ruptured abdominal aortic aneurysms J Vasc Surg. 2006;44:1140-1147.