Outcome measures
Primary outcome:
all-cause and aneurysm-related mortality
Secondary outcomes:
30-day operative mortality,
health related quality of life (HRQL),
costs and cost effectiveness,
durability - rates of complications and re-interventions,
renal function measured using serum creatinine.

Power calculations
The primary outcome measure was all-cause mortality and aneurysm-related mortality with both trials powered using all-cause mortality.
EVAR 1 – for 80% power at the 5% significance level, a total of 900 patients were required to detect a difference in all-cause mortality of 7.5% per year in the open repair arm versus 5% per year in the EVAR arm after an average of 3.3 years of follow-up (producing 179 deaths).
EVAR 2 - for 90% power at the 5% significance level, a total of 280 patients were required to detect a difference in all-cause mortality of 25% per year in the no intervention arm versus 15% per year in the EVAR arm after an average of 3.3 years of follow-up (producing 146 deaths).

Patients
Patients were referred to regional trial centres participating in the trials and assessed for eligibility for the trials. Patients of both sexes were eligible if they were aged at least 60 years and had an AAA measuring 5.5cm or greater in any plane according to a computed tomography (CT) scan. The CT scan was assessed for anatomically suitability for an EVAR device and if the patient was deemed suitable, then they were investigated for fitness for open repair. For pragmatic purposes, fitness for open repair was determined locally, however, recommended guidelines in terms of cardiac, respiratory and renal function were provided at each centre to ensure some level of conformity across centres. Patients considered fit for open repair were offered entry into EVAR Trial 1 and those considered unfit for open repair were offered entry into EVAR Trial 2. Patients found to be anatomically unsuitable for EVAR or who were considered too unwell for any intervention were offered standard alternative care of either open repair or medical therapy with AAA scanning surveillance. Patients were required to give informed consent before randomisation could proceed and patients who refused to enter the trial were offered standard AAA management outside of the trial.

Sample group
Inclusion criteria for EVAR Trial 1:
Males or females
Aged at least 60 years
AAA measures at least 5.5cm in any plane on a CT scan
AAA regarded as anatomically suitable for EVAR
Patient considered fit for an open repair

Exclusion criteria for EVAR Trial 1:
None

Inclusion criteria for EVAR Trial 2:
Males or females
Aged at least 60 years
AAA measures at least 5.5cm in any plane on a CT scan
AAA regarded as anatomically suitable for EVAR
Patient considered unfit for an open repair

Exclusion criteria for EVAR Trial 2:
Patients with hostile abdomen anatomically unsuitable for an open repair but anaesthetically fit enough for an open repair.

Randomisation
Randomisation was performed centrally by the trial manager using a password protected computer at the main trial office once all the baseline data and the signed consent form had been received via fax or post. Randomisation allocation was generated using a 1:1 ratio from randomly sized permuted blocks constructed by the STATA computer software package (STATA corporation, Texas, USA). Randomisation was stratified by centre and clinicians were encouraged to perform surgery within one month of randomisation.

Subvention fund for payment of EVAR devices
Randomisation to an EVAR device triggered additional funding for that hospital to cover the excess treatment costs of EVAR over standard management. In general, NHS and private company funding for EVAR devices was not available outside of the trial and it is thought that this has had a beneficial impact on the high recruitment rates into both trials.

Recruitment (Cumulative recruitment graph EVAR 1 - Cumulative recruitment graph EVAR 2)
Target recruitment was 900 patients for EVAR Trial 1 and 280 patients for EVAR Trial 2. Randomisation commenced on 1st September 1999 and planned recruitment closed on 31st December 2003 with both targets surpassed by about 20%, (1082 entered into EVAR Trial 1 and 338 entered into EVAR Trial 2). (Top recruiting centres for EVAR 1 - Top recruiting centres for EVAR 2)

Patients were required to be followed up for at least one year from randomisation before analyses of data could commence and thus a period of at least one year needed to elapse before any results could be made publicly available. Thus, it was decided that randomisation should continue into the trials until the main results were released in June 2005. By this date, 1429 patients had been entered into EVAR Trial 1 and 462 patients had been entered into EVAR Trial 2. A reduced follow-up protocol has been adopted for these additional patients randomised after 31st December 2003. Recruitment into the trials closed on June 30th 2005.

Follow-up
All patients were flagged for mortality at the Office for National Statistics (ONS) and all death certificates were reviewed by an independent Endpoint Committee, without knowledge of randomised group, to agree cause of death and assign an International Classification of Diseases (ICD) code (10th Revision). Aneurysm-related mortality was defined as all deaths occurring within 30 days of any surgery for abdominal aortic aneurysm as well as deaths where the underlying cause was attributed to ICD10 codes I713-I719 unless over-ruled by post mortem findings or if a separate procedure (unrelated to the aneurysm) took place between aneurysm repair and death and was attributed as the cause of death. Deaths occurring within 30 days of surgery were categorised as procedure-related deaths. Late complications of aneurysm repair over 30 days after operation, such as aorto-duodenal fistula or aortic rupture after endografting were also recorded as procedure-related aneurysm deaths.

For the secondary outcomes, patients were seen at 1, 3 and 12 months and annually thereafter. To monitor graft durability, CT scan data were collected annually for all patients in both treatment groups of each trial with EVAR patients having additional CT scans at 1 and 3 months post procedure. For patients randomised to EVAR or open repair in either trials, HRQL follow-up was timed at 1, 3 and 12 months from operation and included the Short Form 36 (SF36) and EuroQol 5D (EQ5D) measures. HRQL for patients randomised to no intervention in EVAR Trial 2 was monitored using the same HRQL measures but timed at 2, 4 and 13 months from randomisation to account for the 1 month delay to surgery in the EVAR arm. For costs during the main hospital admission and subsequent aneurysm-related admissions, data were collected on key resource use including length of hospital stay, duration of procedure, type of device/graft and the resource implications of complications. For cost-effectiveness analyses, a EuroQol questionnaire was completed each year from randomisation. Data on adverse events such as myocardial infarction, stroke, amputation or the need for chronic renal dialysis also were collected.

Follow-up for patients up to 31st December 2004 has been reported and these mid-term results have indicated that longer-term follow-up is required for both trials. Additional funding has been granted by the NCCHTA until June 2010 which will allow us to follow all patients for a minimum of 5 years (maximum 11 years).

Trial management and data monitoring (Flow chart of data management)
All data from the trial were stored confidentially at the main trial office and audited by an independent Data Monitoring and Ethical Committee (DMEC) who were the only body who had access to the data. One interim analysis was performed in May 2003 when the first 100 deaths had occurred in EVAR Trial 1 but stopping rules did not need to be implemented.

The day-to-day management of the trials was the responsibility of the Trial Management Committee (TMC) who had expertise in statistics and clinical trial methodology, vascular surgery, interventional radiology, HRQL and health economics.

A Trial Steering Committee (TSC) were responsible for funding matters and public dissemination of information.

The Regional Trial Participants Committee (RTPC) consisted of a nominated vascular surgeon, radiologist and coordinator from each centre. This committee met each year to feed back any problems that were obstructing the smooth running of the trials locally.