If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany.
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
c E. Sebastian Debus and A. Sedrakyan share last authorship.
Eike S. Debus
Footnotes
c E. Sebastian Debus and A. Sedrakyan share last authorship.
Affiliations
Department of Vascular Medicine, Working Group German Vasc, German Aortic Center Hamburg, University Heart Center Hamburg, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany
Colonic ischaemia (CI) is a severe complication following abdominal aortic aneurysm (AAA) repair, leading to high morbidity and mortality. The aim of the study was to determine the incidence, predictors, and outcomes of CI following AAA repair.
Methods
National claims from Germany's third largest insurance provider, DAK-Gesundheit, were used to investigate CI after intact (iAAA) and ruptured (rAAA) AAA repairs. Patients undergoing endovascular (EVAR) or open surgical (OSR) repairs between January 2008 and December 2017 were included in the study.
Results
There were 9145 patients (8248 iAAA and 897 rAAA) undergoing EVAR or OSR procedures and the median follow up was 2.28 years. Most patients were male (79.2% iAAA, 79.3% rAAA); the median age was 73.0 years (iAAA group) and 76.0 years (rAAA group). Overall, CI occurred 97 (1.2%) times after iAAA and 95 (10.6%) after rAAA. In univariable analyses CI occurred less often after EVAR than after OSR (0.6% vs. 3.7%; p < .001). Acute post-operative renal and respiratory insufficiencies were also related to the occurrence of CI (p < .001). CI was associated with greater in hospital mortality (42.2% vs. 2.7% for iAAA, 64.2% vs. 36.3% for rAAA; p < .001) and lower long-term survival for iAAA (Kaplan–Meier analysis). In multivariable analyses, rAAA (odds ratio [OR] 5.59), and higher van Walraven comorbidity score (OR 1.09) were independently associated with greater risk of CI occurrence. EVAR use (OR 0.30) was protective. EVAR use remained protective in stratified analyses within iAAA (OR 0.32) and rAAA (OR 0.26).
Conclusion
Post-operative CI after AAA repair is not common but is associated with worse in hospital outcomes and lower long-term survival. EVAR was protective after both rAAA and iAAA repairs. When discussing the treatment of AAA with patients the protective effect of EVAR should be considered. Future studies should validate predictive scores and advance preventive strategies.
Available real world data evidence is limited regarding the incidence of colonic ischaemia (CI), its predictors, and the possible benefits of endovascular techniques in abdominal aortic aneurysm repair. Although practice patterns have changed significantly since endovascular repair was introduced in the early 1990s, very few studies include patients treated within the last decade. Moreover, available studies are limited to short-term outcomes. This large scale study using contemporary health insurance claims data identified independent predictors for CI, and its link with several major complications and worse long-term survival. Endovascular repair is found to have a protective effect and preventive strategies need evidence based implementation.
Introduction
Colonic ischaemia (CI) is a serious and life threatening complication after both endovascular (EVAR) and open surgical (OSR) repair of abdominal aortic aneurysms (AAA).
The reported incidence of this major complications varies from 0.5% to 3.6% for elective AAA repairs and from 3.7% to 23.0% for urgent AAA repairs (Supplemental Table 1).
However, practical guidelines for the management of the diseases of mesenteric arteries and veins do not cover CI after AAA repairs. D-dimer remains the only sensitive and highly non-specific biomarker for intestinal ischaemia but is increased in most cases following surgery or intervention, which lowers its diagnostic value.
Editor's choice—management of the diseases of mesenteric arteries and veins: clinical practice guidelines of the European Society of Vascular Surgery (ESVS).
and computed tomography angiography can be used to diagnose acute occlusion of the mesenteric arteries. Under specific circumstances, laparotomy is indicated to diagnose CI or to decompress abdominal compartment syndrome as another severe complication associated with the occurrence of CI.
There is little evidence concerning the impact of the adoption of new technology in vascular surgery and it is unclear what the other predictors of CI are. Furthermore, the impact of CI on short and long-term mortality is also not well understood. In a systematic review and meta-analysis of 13 studies reporting specific outcomes of CI after elective AAA repair, EVAR was associated with a lower incidence of CI than OSR.
Better knowledge of the effect of EVAR on CI in real world settings is much needed and may help make policy recommendations. In addition, understanding other predictors of CI occurrence will help improve care guidelines.
Accordingly, the aim of this study was to determine the occurrence, predictors, and consequences of this devastating complication following AAA repair using real world evidence.
Methods
The health insurance claims data of Germany's third largest insurance provider, DAK-Gesundheit (DAK-G), includes the outpatient and in hospital medical care provided to approximately 6.5 million German citizens (8% of Germany's population). In contrast to most registry based data on AAA from Germany, the DAK-G database is not restricted to vascular surgeons but includes all medical specialties treating the insured cohort for AAA (e.g., cardiac surgeons, cardiologists, angiologists, interventional radiologists, and general surgeons). The DAK-G cohort includes nationally generalisable data with comparable sex and age distribution (40.4% female and 29.1% over 65 years of age) and has been validated before.
Perioperative mortality following repair for abdominal aortic aneurysm in Germany: comparison of administrative data of the DAK health insurance and clinical registry data of the German Vascular Society.
For this study, the DAK-G database was used to determine in hospital treatments for AAA using the International Classification of Diseases (ICD-10) coding I71.4 (intact [iAAA]) or I71.3 (ruptured [rAAA]) and Operations and Procedures Codes (OPS) coding for OSR (OPS code 5–384*) or EVAR (OPS code 5-38a*, 8–842*) repair of infrarenal AAA. The cohort included procedures conducted between 1 January 2008 and 31 December 2017. The German OPS code is adapted to the International Classification of Procedures in Medicine. For the identified cases that matched the abovementioned basic search criteria, data were collected on the occurrence of colonic ischaemia (OPS K55.0, K55.9), coding of a colonoscopy (OPS 1–65*), occurrence of abdominal compartment syndrome (ICD R19.*), demographics, procedures done while in hospital (OPS codes), coded comorbidities (World Health Organisation [WHO] ICD-10 codes at the time of discharge), and reason for discharge. For the long-term survival analyses, patients whose insurance contract expired within the follow up period were censored. Baseline differences in demographics, comorbidities, and primary and secondary end points were assessed. The first AAA repair procedure was included as the primary case (no re-intervention cases during the study period were included).
Regarding comorbidities, the Elixhauser coding was used to summarise all comorbidities present in 30 categories via WHO ICD-10 code.
Mean and SD or median and interquartile range are reported for continuous variables. Proportions and 95% confidence interval are reported for categorical variables. Tests of normality were conducted using Kolmogorov–Smirnov test. Student t-test was used for normally distributed data and Mann–Whitney U test and Kruskal–Wallis H test were used for non-normally distributed data. Rates and univariate differences were compared using Pearson's chi-square test and Fisher's exact test. The multivariable regression models were used for the entire cohort and the model included age, sex, operative procedure (endovascular repair, open repair), rupture, and van Walraven comorbidity score. Predictive models were developed for the entire cohort. Two additional models were developed for subgroups stratified for rupture status (iAAA vs. rAAA).
Kaplan–Meier survival curves were used to study survival by occurrence of CI over time. Patients with unknown mortality were censored. Sensitivity analysis using the landmark approach for all patients who survived at least 30 days after procedure was performed. To address multiple testing in univariate analyses (Table 1, Table 2), the Holm-Bonferroni method was used and a p value of < .01 was considered statistically significant. All statistical analyses were performed with software R version 3.3.2 (The R Foundation for Statistical Computing, Vienna, Austria).
Table 1Baseline demographics, comorbidities, and clinical characteristics by occurrence of colonic ischaemia.
iAAA
rAAA
CI: no
CI: yes
p
CI: no
CI: yes
p
Patients (n)
8151
97
802
95
EVAR
4614 (56.6)
26 (26.8)
<.001
209 (26.1)
8 (8.4)
.013
Median (IQR) patient age (y)
73.0 (66–78)
73.0 (70–77)
1.0
76.0 (69–84)
78 (70.5–82.5)
1.0
Male patients (n)
6468 (79.4)
67 (69.1)
.592
630 (78.6)
81 (85.3)
1.0
van Walraven comorbidity score
6.30 (5.93)
12.60 (8.25)
<.001
9.48 (6.59)
12.81 (7.88)
.007
CHF
1005 (12.3)
21 (21.6)
.314
147 (18.3)
15 (15.8)
1.0
Cardiac arrhythmias
1587 (19.5)
24 (24.7)
1.0
233 (29.1)
28 (29.5)
1.0
AF
1157 (14.2)
22 (22.7)
.803
182 (22.7)
25 (26.3)
1.0
Chronic pulmonary disease
1423 (17.5)
25 (25.8)
1.0
112 (14.0)
16 (16.8)
1.0
Chronic renal failure (all stages)
1681 (20.6)
17 (17.5)
1.0
208 (25.9)
14 (14.7)
.939
High grade renal failure (GFR < 30 mL/min)
212 (2.6)
2 (2.1)
1.0
44 (5.5)
5 (5.3)
1.0
Hypertension
5762 (70.7)
64 (66.0)
1.0
437 (54.5)
37 (38.9)
.248
Diabetes uncomplicated
1066 (13.1)
12 (12.4)
1.0
94 (11.7)
9 (9.5)
1.0
Diabetes complicated
404 (5.0)
4 (4.1)
1.0
39 (4.9)
5 (5.3)
1.0
Note. Categorical values are reported as n (%) and continuous variables as median (interquartile range [IQR]). Significant p values (Holm–Bonferroni method) < .01 are marked bold. iAAA = intact abdominal aortic aneurysm; rAAA = ruptured abdominal aortic aneurysm; CI = colonic ischaemia; EVAR = endovascular aortic aneurysm repair; CHF = congestive heart failure; AF = atrial fibrillation; GFR = glomerular filtration rate.
For a retrospective analysis of anonymised health insurance claims data, no local ethics committee approval was required (exempt), and no patient informed consent was obtained for the study.
Results
A total of 9145 patients underwent invasive treatment for AAA during the study period. Of these patients, 8248 were diagnosed with an iAAA and 897 patients with a rAAA. Of the iAAA repair patients, 97 (1.2%) developed CI (0.56% after EVAR, 1.97% after OSR; p < .001). Among rAAA repair patients, 95 (10.6%) developed CI (3.7% after EVAR, 12.8% after OSR; p = .001). Median follow up was 2.28 years (range 0–9.78 years) and loss of follow up owing to the end of insurance contract was similar in both groups following both iAAA and rAAA repairs. Baseline demographics, comorbidities, and clinical characteristics of patients with or without CI are listed in Table 1. Among iAAA repair patients, patients developing CI were more likely to be female, to undergo OSR, and to have a higher van Walraven comorbidity score. Among rAAA repair patients, patients with post-operative CI were more likely to undergo OSR and to have a higher van Walraven comorbidity score.
Unadjusted in hospital outcomes are listed in Table 2. A considerably higher rate of in hospital mortality was found among patients with acute post-operative CI in both iAAA (41.2% vs. 2.7%; p < .001) and rAAA (64.2% vs. 36.3%; p < .001). This difference remained stable after 90 days (iAAA: 44.3% vs. 4.9% [p < .001]; rAAA: 67.4% vs. 40.8% [p < .001]). CI was also associated with other major complications such as acute renal insufficiency, acute respiratory insufficiency, and pneumonia. Among iAAA and rAAA patients, post-operative length of stay was higher in the CI group than in the non-CI group (iAAA: 20 vs. 8 days [p < .001]; rAAA: 17 vs. 12 days [p = .004]). A colonoscopy was performed in 51.5% of iAAA patients and in 43.2% of rAAA patients if CI occurred. Diagnostic laparotomy was performed in 32.0% of iAAA patients and in 51.6% of rAAA patients if CI occurred.
Long-term survival following CI
Kaplan–Meier analyses of survival following iAAA and rAAA repair are reported in Fig. 1. Examination of the survival curves show that occurrence of CI was associated with worse overall survival after both iAAA (p < .001) and rAAA repair (p < .001). In additional analyses of survival conditional upon surviving the procedure (patients who survived at least the first 30 days after the procedure), this difference remained significant after iAAA repair (p = .0014).
Figure 1Kaplan–Meier survival analysis for intact (top left) and ruptured (top right) abdominal aortic aneurysm (AAA) repair. (Bottom) Survival analysis conditional upon surviving the first 30 days following intact (bottom left) and ruptured (bottom right) AAA repair.
Independent predictors of CI following AAA repair in the entire cohort are listed in Table 3. In adjusted analysis, rAAA and higher van Walraven comorbidity score were the most important determinants of CI. EVAR was associated with a lower risk of post-operative CI.
Table 3Independent predictors for the occurrence of colonic ischaemia in the entire cohort.
Entire cohort
OR
95% CI
p
Ruptured aneurysm (vs. intact AAA)
5.587
4.053–7.697
<.001
van Walraven comorbidity score (increase by one point)
1.088
1.069–1.108
<.001
EVAR (vs. OSR)
0.297
0.197–0.436
<.001
Age (increase by 1 year)
1.011
0.994–1.028
.203
Male sex (vs. female sex)
1.008
0.699–1.427
.964
Note. Significant p values < .05 are marked bold. OR = odds ratio; CI = confidence interval; AAA = abdominal aortic aneurysm; EVAR = endovascular aneurysm repair; OSR = open surgical repair.
Subgroup analyses (stratified by iAAA vs. rAAA) are listed in Table 4. In adjusted analysis EVAR remained protective within each subgroup, whereas higher van Walraven comorbidity score was associated with a higher risk of CI after AAA repair.
Table 4Independent predictors for the occurrence of colonic ischaemia stratified by intact abdominal aortic aneurysm (iAAA) versus ruptured abdominal aortic aneurysm (rAAA).
Stratified model
iAAA (n = 8248)
rAAA (n = 897)
OR
95% CI
p
OR
95% CI
p
van Walraven comorbidity score (increase by one point)
1.100
1.075–1.125
< .001
1.068
1.036–1.100
<.001
EVAR (vs. OSR)
0.317
0.193–0.507
< .001
0.260
0.114–0.519
<.001
Age (increase by 1 year)
1.023
1.000–1.048
.058
1.002
0.978–1.027
.882
Male sex (vs. female sex)
1.432
0.907–2.209
.112
0.626
0.326–1.126
.136
Note. Significant p values < .05 are marked bold. OR = odds ratio; CI = confidence interval; EVAR = endovascular aneurysm repair; OSR = open surgical repair.
In this large scale investigation of German health insurance claims data, it was found that post-operative CI was rare after EVAR for iAAA (0.6%) and relatively common following OSR for rAAA (12.8%). The occurrence of this major complication was associated with various major complications, higher in hospital mortality, and worse long-term survival. Independent predictors of CI were ruptured AAA and higher van Walraven comorbidity score. EVAR was protective for CI for both iAAA and rAAA.
The evidence from large cohorts related to CI occurrence, impact, and predictors is limited. In an early analysis including 2930 open surgical aorto-iliac or aorto-femoral procedures conducted between 1987 and 1993 in Sweden, Björck et al. reported an overall CI incidence of 2.8% (7.3% for rAAA), most frequently affecting the left colon.
More recently, Senekowitsch et al. reported that the incidence of histologically proven CI was three times higher than the incidence measured by clinical symptoms (15% vs. 5.7%), emphasizing the challenges related to screening for inconsistent symptoms and asymptomatic ischaemia.
Interestingly, in the present study, only 51.5% of patients diagnosed with CI following iAAA repair and 43.2% following rAAA repair underwent a colonoscopy. A diagnostic laparotomy was performed in 51.6% and 32.0% of affected patients, respectively. Another large scale analysis of the 2003 and 2004 Nationwide Inpatient Sample in the USA reported ruptured aneurysm presentation, increasing disease severity, female sex, and open surgery (for iAAA) to be predictors of CI.
Bowel ischaemia was less frequent following iAAA repair (0.6% after EVAR, 3.6% after OSR) than in rAAA repair (6.4% after EVAR, 19.3% after OSR). The authors found older age, female sex, OSR, ruptured aneurysm, and various other covariates to be predictors of ischaemic bowel complications.
The current study confirms and extends prior knowledge related to the incidence of CI for iAAA and rAAA repair using health insurance claims data in Europe. Consistent with prior studies, rAAA, higher van Walraven comorbidity score, and OSR were found to be independent predictors of CI. It was found that CI is associated with serious in hospital morbidity and mortality. Specifically, acute post-operative renal and acute respiratory insufficiency were both associated with CI. The severity of clinical presentation and deterioration may be attributed to the fact that CI may lead to multi-organ syndrome and eventually to death. This hypothesis is supported by the novel finding that CI following iAAA repair has an effect on long-term survival.
Interestingly, CI has been the focus of research in the past 30 years, but its incidence does not seem to decrease. It is time to implement preventive measures to reduce CI after AAA repair and address these in the guidelines that have no recommendations regarding CI prevention.
This study supports this recommendation and suggests use of EVAR whenever possible to reduce the chance of CI in higher risk groups. Future studies should assess and evaluate the value of predictive scores and the effectiveness of preventive strategies.
Limitations
Health insurance claims data are not collected for scientific evaluation, but rather for reimbursement purposes.
Although coding errors are possible, it is believed that these errors would affect predictors equally because data collection occurs independently from scientific enquiries such as in the present investigation: there is no self selection of cases by researchers. It has to be highlighted that without external and internal validation of data, results need to be interpreted with caution with consideration of the chance of some misleading results.
German health insurance companies randomly conduct internal validation of claims data by cross checking coded information with the original patient files. A cohort validation study was also conducted by comparing data from the DAK-G database and the AAA registry hosted by the German Vascular Society.
It was found that mortality, length of hospital stay, and use of EVAR had good internal validity for DAK-G data. Additional advantages of DAK-G data over AAA registry data include the availability of post-discharge follow up data and the lack of reporting bias due to self selection of submitted cases by physicians. Unfortunately, health insurance claims data do not provide valid, detailed information regarding anatomical differences such as neck anatomy, clinical symptoms, or pre-existing mesenteric vessel patency. There might be a significant variation between clinical symptoms and histopathological evidence of CI. Variation in definitions and diagnostic algorithms might also limit the comparability of the results.
Conclusions
Post-operative CI following AAA repair is not common but is associated with worse in hospital outcomes and lower long-term survival. EVAR was protective after both rAAA and iAAA repair. When discussing the treatment of AAA with patients the protective effect of EVAR for CI should be considered. Future studies should address the application of predictive scores and advance preventive strategies.
Acknowledgements
The authors thank the German Stifterverband and the CORONA foundation grant number (S199/10061/2015) for financial support in the organisation and publication of this study. The authors thank the DAK-Gesundheit and Mark Dankhoff for supporting this quality improvement and research project.
Conflicts of interest
None.
Funding
None.
Appendix A. Supplementary data
The following is the supplementary data related to this article:
Editor's choice—management of the diseases of mesenteric arteries and veins: clinical practice guidelines of the European Society of Vascular Surgery (ESVS).
Perioperative mortality following repair for abdominal aortic aneurysm in Germany: comparison of administrative data of the DAK health insurance and clinical registry data of the German Vascular Society.
Colonic ischaemia (CI) is a well known complication after repair of abdominal aortic aneurysms (AAAs), with very serious consequences. When studying such infrequent events, single centre series are meaningless and prone to publication bias. Previous publications based on vascular registry data have described the risk factors for CI,1 the most important being rupture, major bleeding, and abdominal compartment syndrome (ACS).2
To submit a comment for a journal article, please use the space above and note the following:
We will review submitted comments as soon as possible, striving for within two business days.
This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
We require that commenters identify themselves with names and affiliations.
30369-1&id=gr1.jpg){kind=link}