This is a worthy topic and the accurate answer would be of interest. The method used by Baas, Grobbee and Blankensteijn required the 1000 vascular surgeons and 400 trainees to think back how they would have managed patients before the trials were published and afterwards. It emerges that 72% preferred open repair (OR) against 28% for EVAR before publication, which moved to 74% and 26% respectively after trial publication. The authors noted that about a third of the respondents changed their opinion after trial publication but there was no significant overall change.
Another approach to the subject is that used in the European Vascular and Endovascular Monitor. This consists of 250 vascular and endovascular clinicians across Europe. The membership is consistently updated. The clinicians involved receive a quarterly report on procedure numbers of vascular and endovascular in various domains including abdominal aortic aneurysm.
Vascular and endovascular surgeons, like all enthusiastic clinicians, tend to overstate their numbers and so the total activity numbers are suspect. However, the methodology is more reliable for the proportions and techniques used at one time or change against time. More recently, Vascular News North America launched the USVEM similar to the one in Europe. In figure 1, the EVEM stent graft usage is given from 1st July 2005 for a year, by country. In UK and France, there is a continuing rise. In Belgium and Luxembourg, the usage is flat and almost flat in The Netherlands. This is against a background of previous relentless rise in the proportion of patients managed by EVAR in the whole of Europe, according to the EVEM panel estimate (Figure 2).
The trials seem to have had no impact in the US (Figure 3) where the penetration of EVAR is more than 50% of the whole.
The EVEM panel data for the whole of Europe appeared to demonstrate stentgraft increase after the trials were published (Figure 4) but there is an apparent flattening out at about the time that cost effectiveness data emerged. This goes to show that the method chosen gives a different answer. I have little doubt that the methodologies quoted in the Dutch paper and also used in the EVEM and USVEM panels is less than robust, but it is what we have at this stage.
I am indebted to surgeons of the Swedish Vascular Society for sending their vascular registry data (Figure 5). This shows a stepwise increase in aortic aneurysm repair annually from 2003, entirely by virtue of increase in stentgraft performance, which shows no signs of slowing. This is a further angle to look at the problem. There is simply no constant message across the methods.