Discussion

Article Outline

• Copyright

Dr Robert Rutherford (Corpus Christi, Tex). My question has to do with your stated primary end point of aneurysm-related death. It is a long-term end point, not a perioperative end point. Although it is a soft end point whose accuracy depends on whether the death was really witnessed or an autopsy was done and tends to perpetuate any perioperative advantage that EVAR [endovascular aneurysm repair] might have, my point is that you are really comparing postoperative mortality from this NIS [Nationwide Inpatient Sample] data, aren't you?

Dr Kristina Giles. The NIS is an administrative database that is primarily based upon hospital discharge information. The patients that we are able to capture in the NIS are patients that present to a hospital and then are diagnosed with either an intact AAA [abdominal aortic aneurysm] or a ruptured AAA. Further in our study, intact patients were included if they went on to have a repair, whereas ruptured AAA patients we measured whether or not they went on to repair. We do not have autopsy information and make the assumption that there should not be a significant change in the number of patients that die of aneurysm rupture before making it to a hospital.

Dr Rutherford. So you are really presenting just perioperative mortalities?

Dr Giles. Yes.

Dr Jon Matsumura (Chicago, Ill). Aneurysm-related mortality normally includes the periprocedural (inpatient and within 30 days) deaths after initial treatment and reintervention and from rupture. Since your group has done great work looking at these reinterventions after endo and open repairs, were you also counting aneurysm-related mortality after these secondary interventions in this study?

Dr Giles. This is inpatient only database; therefore, we don't have follow-up data with this particular study. We are unable to look at deaths related to secondary interventions, unless they were still in the hospital at the time.

Dr Anton Sidawy (Washington, DC). Any addition, Dr Schermerhorn?

Dr Marc Schermerhorn. We will have that for you in the near future with the deaths related to reinterventions, but using a different database.

Dr Sidawy. Am I correct that the total number of aneurysms repaired over time has remained constant?

Dr Giles. It has remained relatively stable when you account for both intact and ruptured aneurysms. The mean annual volume has increased by approximately one thousand from a pre-endovascular to postendovascular time point.

Dr Sidawy. That indicates that the appropriate operation continues to be performed for the appropriate indication and that surgeons have not changed or relaxed their indication for AAA repair just because EVAR is somewhat simpler and less time consuming. Also, some of us think that since after EVAR the aneurysms done with open repair are the more difficult ones; therefore, the results of open repair may be getting worse. However, I gather that your results have not confirmed this assumption, since mortality rate even for open repair has gone down over the years.

Dr Giles. Correct, even for open the mortality has decreased slightly. When you look at a pre-endovascular to postendovascular period, the mortality rate was 4.7% vs 4.5%. So open repair mortality is certainly not going up.

Dr Schermerhorn. And I'll just back up your comment, Dr Sidawy. We were concerned because when laparoscopic cholecystectomy came out, the procedure volume went up so much that despite the lower operative mortality there was no decrease in cholecystectomy-related death. So we were worried that perhaps we are expanding AAA repair to patients who are so old they are going to die before rupture or their AAA is so small they don't need a repair, but that does not seem to be the case. So we are happy to see these results.

Dr Wilhelm Sandmann (Düsseldorf, Germany). As I understand the message of your paper, it is to use more endovascular treatment and you will have lower mortality rates. I don't think that your paper can prove this because, since endovascular treatment arrived on the market, there are a lot more small aneurysms being repaired, which probably have lesser mortality with either open repair or EVAR. So my suspicion is that the number, which increased in the second period, has to do with better patients and better outcome. Do you know if those aneurysms which are appearing in the second period have the same morbidity and mortality criteria as in the first period?

Dr Giles. We don't have any anatomic data from this database, so I can't tell you if they are smaller or not. However, we do know from prior studies that an EVAR cohort typically has a greater number of comorbidities and is older than an open repair cohort. In this study, the age for all repairs is increasing over time. The average age in the postendovascular era, with the aging population, has actually increased by about 2 years. So just extrapolating from that, you would assume that some of these patients are more ill after EVAR became an option. So I would not conclude that it is a healthier population that is being operated on based on that.

Dr Krish Soundararajan (Philadelphia, Pa). I believe that NIS data that you have chosen for your study categorizes the hospitals based on the region and case volume. There are several reports that in general suggest better outcome of vascular interventions in the high-volume centers. Were you able to see any such difference based on the regions or volume in your analysis?

Dr Giles. We did not look at regional variations for this study; however, that is something that could be done in future work. NIS data have been used in the past to show a volume–outcome relationship, with higher volume associated with lower mortality rates for both open repair and EVAR for intact aneurysms. We did not repeat this in the current study. We did however analyze the hospital volume–outcome relationship for ruptured aneurysms in a separate study that will be presented also at this meeting. This showed that there was a mortality benefit for higher-volume centers.