For quite some time, Vascular Surgeons have had the privilege and obligation to practice and merge the élite data available about Abdominal Aortic Aneurysms (AAA) in their mutual decision-making with their patients.
Unfortunately, in well-studied conditions such as AAA, multiple therapeutic interventions have been tested in randomized controlled clinical trials demonstrating heterogeneity in clinical outcomes. Based largely on the results of these trials, professional societies develop practice guidelines to provide some weighted consideration for evidence-based therapeutic options. The heavy reliance of treatment guidelines in Vascular Surgery on the results of randomized controlled treatment trials has fueled a reductionist view of clinical trial evidence in practice. Trials are commonly viewed as ‘positive’, ‘neutral’, or, even worse, ‘negative’, with the implication that neutral results are largely uninformative, while the positive trials may be practice changing. However, in my professional opinion, the intended 2018 AAA NICE guidelines have to be labelled negative or at least neutral.
The NICE guideline is a data synthesis, which is substantially based on a published Cochrane review from 2014 and then updated by the NICE committee in consultation with the Cochrane Vascular group in 2017. This meta-analysis collates data from four randomized controlled trials which recruited patients between 1999 and 2004. The trials included are EVAR-1 (U.K.), DREAM (The Netherlands), ACE (France) and OVER (USA), number with a combined study population of 2,783, 94% of which are male. This constitutes level one evidence, however the recurring theme in Vascular Surgery practice is as to whether or not level one evidence reflects real life practice, and this is a pertinent question when considering whether NICE guidelines will impact international standards of care for patients with AAA. Randomized controlled trials and meta-analysis thereof, are regarded as the pinnacle and most robust forms of evidence. However the data is restricted by the timing of the studies and the inherent limitations of RCTs that make them difficult to apply to real-world practice.
The timing of the major EVAR trials meant that the operators, devices and necessary infrastructures were immature. The trials were undertaken from 1999 to 2004 when EVAR was in its relative infancy and learning curves had not been overcome. While the EVAR trials only required the hospital (not operator) to have experience with 20 EVARs before being permitted to join the trial, the OVER trial was even more liberal in its recruitment of investigators, with operators only required to have experience of five EVAR procedures prior to enrolling patients. This lack of experience is exemplified by the fact that so many EVAR procedures were abandoned or converted to open repair in the trials, for such basic reasons as poor planning and access vessel issues. It is also reflected in the fact that 54% of those randomised to EVAR and 43% of those allotted open repair in the DREAM trial had “unfavorable features of the intrarenal aortic neck” rendering their inclusion in the trial questionable.
The devices were also under developed and many of those used in the trials, such as Talent (Medtronic, Minneapolis, MN, USA), Quantum or Teramed (Cordis, Waterloo, Belgium), and AneuRx (Medtronic), have subsequently been withdrawn from the market. The latter device, in fact, was associated with 60% of the aneurysm related deaths in the EVAR group of the OVER trial.
The crux of aortic treatment is its provision within high volume centers, by operators who are equally proficient with open and endovascular repair and who understand and are experienced in deciphering which treatment applies best to each patient. This is about a personalized, patient-centered approach to optimal therapy. From our own experience in Galway we know that following twenty years of experience with both open and endovascular aortic repair in almost a thousand patients that both methods of repair are cost-effective. In a paper from our group which is due to be published shortly in the Journal of Vascular Surgery, we reported a long-term parallel group comparison of 494 patients, followed over 15 years. We found no difference in cost (EVAR €15,647 vs. OSR €18,281) or cost-per-QALY (EVAR €6386 vs. OSR €7649). These costs are most certainly influenced by the fact that surgeons alone provide both open and EVAR treatments without the need for a separate interventional radiology team, as would be required in a lot of ceners in the UK. The favorable economic outcomes are, furthermore, due to the judicious use of surveillance imaging, using CTA post-operatively only when Duplex Ultrasound demonstrates an increased in aortic sac diameter. The favorable costs were also as a result of low long-term rates of re-intervention in both EVAR and open surgical groups. The similar re-intervention rates between groups reflects the outcomes reported by the OVER trial and contrasts that of the EVAR-1 trial, most likely because the latter did not collate data for abdominal hernia repairs or other complications related to open repair, thereby rendering the comparison of long-term re-interventions invalid. Our results reflect high deliberate practice volume with contemporary endo grafts in elective management of AAA, with no learning curve issue. No post-operative ruptures occurred despite what some may consider less fastidious follow up with duplex ultrasound rather than CTA.
Our results, along with those from other high volume centers, confirm that high deliberative practice volume allows for patient specific treatment allocation and efficient use of contemporary resources. This renders the management of abdominal aortic aneurysm consistently cost-effective. Experienced interventionalists, equally proficient in both open and endovascular techniques can use precision medicine with judicious adaptation of graft type, configuration and post-operative surveillance to maximize individual patient care.
EVAR and open repair are not competitive therapies, each have their own merits and the secret to aortic therapy, is the ability to decide which one to use and when.
The NICE system and the guidelines that emanate from it are meant to guide treatments within a specific UK-based healthcare system. NICE committees are formed to reflect UK society are a made up of a diverse range of members including people who use health and social care services, caretakers and experts in health and social care. The NICE committee tasked with the EVAR guidelines in particular was chaired by a vascular surgeon and comprised two other vascular surgeons, a radiologist, radiographer, theater nurse, anesthetist, emergency department doctor, vascular scientist, vascular nurse and two lay members.
NICE guidelines are an important factor in healthcare delivery in England but their impact on other countries and healthcare systems is not necessarily transferable. The committee’s recommendations are largely based on economic parameters, all of which have been analyzed and presented only in the context of the UK National Health Service. The economic models of healthcare delivery vary greatly though the world and so this report is not ubiquitous or universally applicable.
The NICE document represents a biased analysis of existing evidence. The only new evidence in the NICE document is the use of a “new model” for economics analysis which is not validated outside of this report and which pertains only to the UK. The delivery of vascular services within the UK, particularly the combined use of Vascular Surgery and radiology staff in aortic repair cases, is not a model which is used in the majority of countries.
The interpretation of the evidence was undertaken by a committee of three vascular surgeons and one radiologist and although one has to respect the professional integrity of the vascular surgeons involved, there is no doubt that unconscious bias may have been influential in this committee’s output. The aortic practices of those involved, i.e., whether mostly open or endovascular, should be made transparent, so that we can get a better understanding of how their conclusions were drawn.
The evidence is not concrete and in fact the terms “low”, “very low” and “moderate” are used a lot in the NICE guideline to describe the quality of the evidence, while the term “high” is used less frequently. When the evidence is not black and white then the interpretation of that evidence is more difficult and more likely to be influenced by the underlying bias of the committee. This is a huge factor in the recommendations put forward in this document.
Recommendations to reassess the costs around service provision, including medical teams, devices costs and surveillance program efficiencies were not addressed by the NICE committee, yet surely these are all areas in which savings can be made, and this could easily tip the balance for EVAR in terms of economic viability, once the anatomical and physiological suitability of the patient are correctly adjudicated and the operating team sufficiently experienced. These are my views and comments are welcome.
Dr. Sultan has provided commentary with Medical Device News Magazine on this subject in June of 2018. View it here.