Advertisement

Mortality and complications after aortic bifurcated bypass procedures for chronic aortoiliac occlusive disease

      Objective

      Open surgery has given way to endovascular grafting in patients with aortoiliac occlusive disease. The growing use of endovascular grafts means that fewer patients with aortoiliac occlusive disease have open surgery. The declining open surgery caseload challenges the surgeon's operative skills, particularly because open surgery is increasingly used in those patients who are unsuitable for endovascular repair and hence technically more demanding. We assessed the early outcome after aortic bifurcated bypass procedures during two decades of growing endovascular activity and identified preoperative risk factors.

      Methods

      Data on patients with chronic limb ischemia were prospectively collected during a 20-year period (1993 to 2012). The data were obtained from the Danish Vascular Registry, assessed, and merged with data from The Danish Civil Registration System.

      Results

      We identified 3623 aortobifemoral and 144 aortobiiliac bypass procedures. The annual caseload fell from 323 to 106 during the study period, but the 30-day mortality at 3.6% (95% confidence interval [CI], 3.0-4.1) and the 30-day major complication rate remained constant at 20% (95% CI, 18-21). Gangrene (odds ratio [OR], 3.3; 95% CI, 1.7-6.5; P = .005) was the most significant risk factor for 30-day mortality, followed by renal insufficiency (OR, 2.5; 95% CI, 1.1-5.8; P = .035) and cardiac disease (OR, 2.1; 95% CI, 1.4-3.1; P < .001). Multiorgan failure, mesenteric ischemia, need for dialysis, and cardiac complications were the most lethal complications, with mortality rates of 94%, 44%, 38%, and 34%, respectively.

      Conclusions

      Aortic bifurcated bypass is a high-risk procedure. Although open surgery has increasingly given way to endovascular repair, 30-day outcomes have remained stable during the past decade. Thus, it is still acceptable to consider an aortic bifurcated bypass whenever endovascular management is not feasible.
      The management of aortoiliac occlusive disease has undergone substantial changes since the introduction of percutaneous transluminal techniques. The endovascular technique has improved much over time, and today, focal and complex atherosclerotic lesions are increasingly treated endovascularly.
      • Jongkind V.
      • Akkersdijk G.J.
      • Yeung K.K.
      • Wisselink W.
      A systematic review of endovascular treatment of extensive aortoiliac occlusive disease.
      This shift makes it difficult for the surgeon to keep up sufficient surgical skills because the open surgery caseload is reduced. Hypothetically, the surgeon is further challenged by the fact that with the growing sophistication of percutaneous transluminal techniques, the procedures that are left for open surgery are becoming ever more surgically demanding.
      Extant literature on 30-day mortality and complication rates in open aortic bypass surgery stems from the time before endovascular repair became an alternative, and estimated risk rates vary considerably.
      No recent randomized trials have been published on aortic bifurcated bypass procedures. A large-scale register survey, taking advantage of the unique and complete Danish civil registration system, The Danish Register of Causes of Death, offers a robust estimation of the early risk after aortic bifurcated bypass procedures during the past 20 years.
      The objective of this study was to assess the mortality and the complication rate after aortic bifurcated bypass procedures for aortic occlusive disease during a period with growing endovascular activity and to identify preoperative risk factors in a national cohort registered prospectively over 20 years.

      Methods

       The Danish Vascular Registry

      Since 1993, all vascular procedures performed in Denmark have been registered in the Danish Vascular Registry (www.karbase.dk). The registry contains data on demographics, indications for surgery, comorbidity, surgical procedures, in-hospital complications, discharge, and 30-day outcome. Other variant procedures, such as aorta-uniiliac-unifemoral or iliobifemoral, could be excluded from the data retrieval. Data are reported prospectively, continuously, and online to a central server.
      Karbase Landsregister-The Danish Vascular Registry
      The clinical data of the Registry have previously been proven to be very accurate.
      • Laustsen J.
      • Jensen L.P.
      • Hansen A.K.
      Accuracy of clinical data in a population based vascular registry.
      Every Danish citizen has a unique civil registration number that is used to link data from the Danish Vascular Registry to data from The Danish Register of Causes of Death, which gives the precise dates of death.
      The Danish National Board of Health and the Danish Data Protection Agency approved the Danish Vascular Registry and the data linkage performed in the present study. According to Danish law, a patient's consent to data capture from an officially approved registry, such as the Danish Vascular Registry, is not required where such data are used for statistical purposes.

       Patients

      Data on all patients who had an aortobifemoral (ABF) or an aortobiiliac (ABI) bypass for limb ischemia in Denmark from 1993 to 2012 were extracted from The Danish Vascular Registry. To be able to draw safe conclusions on patients planned for open aortic bifurcated bypass procedures due to chronic limb ischemia, we excluded procedures featuring a high-risk profile, defined as acute aortic or iliac thrombosis, a procedure initiated in the night shift, adjunctive mesenteric or renal procedures, and patients with previous aortoiliac open surgery. Patients with previous aortoiliac endovascular procedures were not excluded.

       End points

      The primary end point was 30-day mortality and complications after aortic bifurcated bypass procedures during the study period. Preoperative and intraoperative risk factors potentially influencing the early outcome were analyzed as secondary end points. Mortality after a major complication was assessed for each individual complication and for each institution.

       Definitions

      Mortality was defined as death ≤30 days of surgery. Major complications were defined as predischarge medical or surgical incidents necessitating treatment, extending hospitalization, or being fatal and which were registered in the Vascular Registry as having occurred ≤30 days of the primary aortic bifurcated bypass procedures. Nonfatal wound complications necessitating treatment or prolonged hospitalization were defined as minor complications. Whether the complication led to extended hospitalization was determined by the surgeon discharging the patient. Complications not necessitating intervention were not registered. The complication criteria are detailed in Table I.
      • Rutherford R.B.
      • Baker J.D.
      • Ernst C.
      • Johnston K.W.
      • Porter J.M.
      • Ahn S.
      • et al.
      Recommended standards for reports dealing with lower extremity ischemia: revised version.
      Table ICriteria for complications
      • Rutherford R.B.
      • Baker J.D.
      • Ernst C.
      • Johnston K.W.
      • Porter J.M.
      • Ahn S.
      • et al.
      Recommended standards for reports dealing with lower extremity ischemia: revised version.
      ComplicationsDefinition
      Medical complications
       CardiacMyocardial infarction, heart failure, or arrhythmia necessitating medical treatment
       PulmonaryPneumonia or respiratory distress syndrome necessitating medical treatment or mechanical ventilation for >2 days
       Renal failureIncrease in serum creatinine of >150 μmol/L
       DialysisTransient or permanent need for dialysis
       Intensive careStay in intensive care unit for >3 days
       Stroke/TIACT-confirmed incidence of stroke or clinically suspected TIA
       Deep venous thrombosisVerified by ultrasound imaging
       Pulmonary embolismRadiographically confirmed
       Compartment syndromeFasciotomy performed
       Multiorgan failureDiagnosis reported by the intensivist
      Surgical complications
       BleedingRequiring reoperation
       Rupture of the abdominal fasciaRequiring reoperation
       Bowel obstructionConfirmed by laparotomy
       Mesenteric ischemiaConfirmed by laparotomy
       EmbolismPeripheral embolization confirmed by reduction in ABI
       Graft failureTotal or partial occlusion of the graft
       Graft infectionNecessitating medical treatment
      Wound complicationsInfection, hematoma, lymphocele requiring surgical revision
      ABI, Ankle-brachial index; CT, computed tomography; TIA, transient ischemic attack.

       Statistics

      Primary end points were expressed as mean with 95% confidence intervals (CIs). For all potential explanatory preoperative and intraoperative risk factors listed in Table II suspected of being associated with an adverse 30-day outcome (except blood loss and procedure time), a multivariable model was fitted that contained all variables that were significant in the univariate logistic regression analysis at the 20% to 25% level. We then assessed whether removal of the covariate produced an important change, defined as a 20% change in the coefficients of the variables remaining in the model. If so, it was regarded as an important confounder and therefore added to the model again. We continued to do so until no variables could be deleted from the model. All previously excluded variables from the initial multivariable model were added to the model to confirm that they were neither statistically significant nor important confounders.
      • Hosmer Jr., D.W.
      • Lemeshow S.
      • May S.
      Applied survival analysis regression modeling of time-to-event data.
      P value of <.05 after multivariate regression analysis was considered to indicate statistical significance.
      Table IIThe study population
      VariableTotal (N = 3767)
      n of N (rate %)Mean (SD)
      Intraoperative details
       ABF prosthesis3623/3767 (97)
       Dacron prosthesis3273/3722 (89)
       Midline incision2152/3650 (61)
       Previous vascular surgery591/3767 (16)
       Procedure time, min183 (60)
       Blood loss, mL1126 (935)
       Combined GA and epidural3136/3636 (87)
      Demographic data
       Age, years62 (9)
       Male gender1734/3767 (46)
       BMI, kg/m224 (4)
       Claudication2377/3767 (63)
       Rest pain878/3767 (23)
       Ischemic ulcer362/3767 (10)
       Gangrene150 (4)
       Current smoker2526/3559 (76)
       Assisted living308/3730 (8)
      Comorbidity
       Renal insufficiency64/3647 (1.8)
       Pulmonary insufficiency569/3739 (15)
       CHF/IHD615/3735 (16)
       History of stroke/TIA300/3731 (8)
       Hypertension1424/3716 (38)
       Diabetes438/3749 (12)
      ABF, Aortobifemoral; BMI, body mass index; CHF, chronic heart failure; GA, general anesthesia; IHD, ischemic heart disease; n, number of procedures; N, number of data available; SD, standard deviation; TIA, transient ischemic attack.
      Pearson's correlation coefficient (r) was used to measure the degree of association between mortality, complication rates, demographics (eg, the proportion of patients with heart disease) blood loss, procedure time, and time.
      Long-term survival rates were presented with Kaplan-Meier survival curves, and we used the log-rank test for comparison between Kaplan-Meier curves. Then, the Cox proportional hazard model was used to identify independent outcome determinants for poor long-term survival. The Bonferroni correction was used to eliminate type I error by multiple testing. Data analysis was performed using SAS 9.2 software (SAS Institute Inc, Cary, NC).

      Results

       Patients

      We identified 4293 inserted ABF/ABI bypass procedures. We excluded patients with acute limb ischemia (n = 235), secondary renovascular hypertension (n = 29), secondary mesenteric ischemia (n = 14), secondary aneurysm (n = 26), and patients who had previously undergone intra-abdominal vascular surgery (n = 222). Hence, throughout the 20-year period, we identified 3623 ABF (96%) and 144 ABI (4%) bypass procedures performed in Denmark due to chronic intermittent claudication (n = 2377) or chronic critical limb ischemia (n = 1390), as detailed in Table II.

       Early mortality

      The 30-day mortality after an ABF/ABI bypass procedure was 3.6% (95% CI, 3.0%-4.1%; n = 134). Of the 134 patients who died ≤30 days, 111 (83%) died in the hospital where the operation was performed, 19 (14%) died after having being transferred to another hospital, and 4 (4%) died after discharge. The most common cause of death was cardiac (n = 52), followed by multiorgan failure (n = 30), mesenteric ischemia (n =9), stroke (n = 5), uremia (n = 2), and bleeding (n =3). The exact cause of death was not available for 32 patients.

       Early complications

      The overall incidence of major complications during the first 30 days after the operation was 20% (95% CI, 18%-21%; n = 733). Medical complications occurred in 15% (95% CI, 14%-16%) and far outnumbered surgical complications, which occurred in 8.2% (95% CI, 7.3%-9.1%).
      The 30-day mortality in patients with no complications entered was 0.5% (95% CI, 0.3%-0.8%; n = 3008). In patients with one major complications, the 30-day mortality was 10% (95% CI, 7%-13%; n = 41), and in patients with two, three, or more than three complications, the 30-day mortality was 16% (95% CI, 10%-21%; n = 30), 26% (95% CI, 16%-36%; n = 20), and 36% (95% CI, 26-47; n = 30), respectively.
      Graft infections were diagnosed in 0.7% (95% CI, 0.4%-1.0%; n = 23) and only recorded among patients undergoing the ABF bypass procedures. Multivariate regression analysis confirmed that wound complications (odds ratio [OR], 10; 95% CI, 3.1-36; P = .001) and age per 10 years (OR, 2.2; 95% CI, 1.1-4.5; P < .001) were the only independent factors associated with the development of graft infection. All complications are summarized in Table III.
      Table IIIComplications and connected mortality
      Type of complicationIncidenceMortality after complication
      No. (% rate [95% CI])No. (% rate [95% CI])
      Pulmonary complication265 (7.1 [6.2-7.9])31 (12 [8-16])
      Cardiac231 (6.2 [5.4-6.9])78 (34 [28-40])
      Dialysis32 (0.9 [0.6-1.2])12 (38 [20-55])
      Bleeding73 (2.0 [1.5-2.4])12 (16 [8-25])
      Acute graft occlusion73 (1.9 [1.5-2.4])9 (12 [5-20])
      Intensive care116 (3.1 [2.5-3.7])28 (24 [16-32])
      Mesenteric ischemia61 (1.6 [1.2-2.0])27 (44 [31-57])
      Abdominal fascial dehiscence47 (1.3 [0.9-1.6])3 (6 [0-14])
      Stroke/TIA38 (1.0 [0.7-1.3])8 (21 [7-35])
      Multiorgan failure35 (0.9 [0.6-1.2])33 (94 [86-100])
      Embolus35 (0.9 [0.6-1.2])5 (14 [2-26])
      Bowel obstruction27 (0.7 [0.5-1.0])4 (15 [0-29])
      Renal failure68 (1.8 [1.4-2.2])15 (23 [12-32])
      Graft infection24 (0.6 [0.4-0.9])3 (13 [0-26])
      Major complications in total733 (20 [18-21])118 (16 [13-19])
      Wound complications488 (13 [12-14])7 (1.4 [0.4-2.5])
       Total complications
      Represents the total number of patients with at least one complication.
      1098 (29 [28-31])120 (11 [9-13])
      CI, Confidence interval; TIA, transient ischemic attack.
      a Represents the total number of patients with at least one complication.

       Preoperative and intraoperative risk factors

      The 30-day mortality rate and the complication rates were significantly correlated with increasing age (P < .001). The highest mortality rate was observed in the group aged 75 to 79 years, of whom 13% (95% CI, 9%-17%) died. The risk-adjusted multivariable model showed that gangrene (OR, 3.3; 95% CI, 1.7-6.5; P < .001), ischemic ulcer (OR, 2.9; 95% CI, 1.8-4.8; P < .001), and renal insufficiency (OR, 2.5; 95% CI, 1.1-5.8; P = .036) were the most significant risk factors for early mortality (Table IV).
      Table IVThirty-day outcome by multivariate regression analysis
      Risk factorMortalityP value
      P < .05 indicates statistical significance.
      ComplicationsP value
      P < .05 indicates statistical significance.
      OR (95% CI)OR (95% CI)
      Renal insufficiency2.5 (1.1-5.8).0351.7 (0.9-2.9).079
      CHF/IHD2.1 (1.4-3.1)<.0011.3 (1.0-1.6).038
      Years > 101.6 (1.3-2.0)<.0011.3 (1.2-1.5)<.001
      Rest pain1.7 (1.1-2.7).0131.2 (1.0-1.4).070
      Ischemic ulcer2.9 (1.8-4.8)<.0011.4 (1.1-1.9).009
      Gangrene3.3 (1.7-6.5)<.0011.3 (0.8-1.9).287
      GA without epidural1.7 (1.1-2.7).0221.2 (0.9-1.5).125
      Pulmonary insufficiency1.8 (1.2-2.7).0051.7 (1.3-2.1)<.001
      Former stroke/TIA1.5 (0.9-1.8).9701.6 (1.2-2.1).001
      CHF, Chronic heart failure; CI, confidence interval; GA, general anesthesia; IHD, ischemic heart disease; OR, odds ratio; TIA, transient ischemic attack.
      a P < .05 indicates statistical significance.
      Current smoking, hypertension, assisted living, midline incision, type of prosthesis, diabetes, body mass index (BMI), former vascular surgery, or coronary artery bypass grafting demonstrated no effect on the 30-day outcome.
      An ABI procedure was associated with a 4.9% rate of wound complications (95% CI, 1.4-8.4), which was far less than the 13% rate (95% CI, 12%-14%) for ABF procedures (P = .001). Furthermore, logistic multivariate analysis showed that ischemic ulcer (OR, 1.6; 95% CI, 1.1-2.3; P = .022) and BMI <20 kg/m2 (OR, 1.8; 95% CI, 1.2-2.3; P = .003) were significantly associated with a higher risk of wound complications.

       Historical perspective

      Throughout the study period, the annual number of open procedures decreased from 323 to 106 (Fig 1, a). This decrease in the annual caseload did not significantly affect the 30-day mortality (r = −0.223; 95% CI, −0.606 to 0.243; P = .349) or the 30-day major complication rate (r = −0.436; 95% CI, –0.736 to 0.008; P = .0542; Fig 1, b).
      Figure thumbnail gr1
      Fig 1Historical perspective. a, Number of inserted aortic bifurcated grafts (circles) for chronic limb ischemia in Denmark. b, The development in mortality (squares) and major complications (circles) at 30 days during the study period. c, The presence of significant preoperative medical risk factors (circles, cardiac disease; squares, pulmonary insufficiency) during the study period. d, The development in blood loss (circles) and procedure time (squares).
      The 30-day surgical complication rate (r = .027; 95% CI, −0.420 to 0.464; P = .910) remained unchanged, but a decrease in the presence of preoperative heart failure and ischemic heart disease (r = −0.654; 95% CI, −0.850 to −0.297; P = .001) and pulmonary insufficiency (r = −0.568; 95% CI, −0.807 to −0.167; P = .008) was observed during the study period (Fig 1, c). The proportion of patients who had previous vascular procedures, including percutaneous transluminal angioplasty, more than doubled from 10% to 24% (P < .001). An increase in the surgical procedure time from 168 minutes to 205 minutes (r = 0.877; 95% CI, 0.071-0.951; P < .001) and an increase in blood loss from 912 mL to 1483 mL (r = 0.867; 95% CI, 0.688-0.946; P < .001) were recorded during the study period (Fig 1, d). The indication for surgery (P = .334), mean age (P = .495), age range (P = .516), and the mean BMI remained unchanged (P = .226) throughout the study period.
      We noticed a change in the preferred abdominal incision during the study period in favor of the transverse incision, as the midline incision decreased from 78% to 23% (P < .001). The retroperitoneal approach was only used in 53 patients, and we identified no laparoscopic procedures.
      The preference for the Dacron graft (DuPont, Wilmington, Del) increased from 86% to 91% (P = .002). We observed a small reduction in the mean hospital length of stay after surgery from 10 to 9 days during the study period (P = .002).

       Cumulative survival

      We found from the data from the Danish Civil Registration System a median long-term survival of 10.3 years after the aortic bifurcated bypass procedures and a mean follow-up time of 8.3 years for our cohort. Independent predictors for poor long-term survival were renal insufficiency (hazard ratio [HR], 1.9; 95% CI, 1.5-2.4; P < .001), diabetes (HR, 1.5; 95% CI, 1.3-1.6; P < .001), age per 10 years (HR, 1.35; 95% CI, 1.3-1.4; P < .001), indication for surgery (HR, 1.2; 95% CI, 1.1-1.3; P < .001), pulmonary insufficiency (HR 1.2; 95% CI, 1.1-1.3; P < .002), and former stroke/transient ischemic attack (HR 1.2; 95% CI, 1.1-1.4; P < .004; Fig 2).
      Figure thumbnail gr2
      Fig 2Kaplan-Meier survival curves. a, Patients aged <50 years, dashed line; 51-60 years, solid line; 61-70 years, dotted line; and >70 years, dashed-dotted line. b, Patients with preoperative serum creatinine >150 μmol/L (dotted line) and patients with regular renal function (solid line). c, Diabetic patients (dotted line) and nondiabetic patients (solid line). d, Claudication, dashed line; rest pain, dotted line; ischemic wounds, solid line; and gangrene, dotted-dashed line.

       Hospital volume

      Annual caseloads in the participating hospitals ranged from nine to 34 procedures, and complication and mortality rates were analyzed relative to this. Mortality after a major complication varied across the participating hospitals (range, 6%-22%, P = .089), but these differences could not be explained by hospital caseload. We found that a high annual caseload was associated with fewer major complications (P < .001) but could not show any significant effect between the annual caseload and mortality (P = .836).

      Discussion

      We hypothesized that a declining number of open aortic reconstructions could lead to an increase in 30-day mortality and in 30-day major complications due not only to a reduced caseload per hospital and surgeon but also because the more technically demanding procedures would be left to open surgery. We found that the number aortic bifurcated bypass procedures fell (Fig 1, a) by two thirds and that the number of patients with previous vascular procedures rose. The same period saw an increase in blood loss and procedure time.
      These observations support our hypothesis that the procedures left for open surgery were technically more demanding in the end of the study period than the procedures performed in the beginning of the study period. The increased blood loss and the prolonged procedures could also reflect a decline in technical skills among the surgeons due to a lower caseload. Fortunately, this seemed to have no effect on the 30-day outcome, because the mortality and major complication rates remained unchanged during the study period.
      The aortic bifurcated bypass procedure is associated with a significant periprocedural and early postprocedural risk because of the complex picture of both surgical and medical complications. Drawing on data from >3500 aortic bifurcated bypass procedures, this study shows that during the first 30 days, 3.6% of the patients died and approximately one-fifth faced a major complication. These findings are in accordance with previous reports.
      • de Vries S.O.
      • Hunink M.G.
      Results of aortic bifurcation grafts for aortoiliac occlusive disease: a meta-analysis.
      Thus, the aortic bifurcated bypass should be considered a high-risk procedure and calls for a meticulous preoperative assessment whenever an endovascular option is precluded. As reported by others, we found that renal insufficiency carried a threefold increased risk and heart disease a twofold risk of procedure-related death.
      • Brady A.R.
      • Fowkes F.G.
      • Greenhalgh R.M.
      • Powell J.T.
      • Ruckley C.V.
      • Thompson S.G.
      Risk factors for postoperative death following elective surgical repair of abdominal aortic aneurysm: results from the UK Small Aneurysm Trial. On behalf of the UK Small Aneurysm Trial participants.
      • Kantonen I.
      • Lepantalo M.
      • Luther M.
      • Salenius P.
      • Ylonen K.
      Factors affecting the results of surgery for chronic critical leg ischemia-a nationwide survey. Finnvasc Study Group.
      The correlation between age and surgical risk was very clear, especially among the oldest patients (aged 75-79 years) who presented a mortality rate of 13%. This finding represents a strong argument for conservative treatment, percutaneous transluminal angioplasty, or extra-anatomic reconstruction in this patient group.
      We were unable to demonstrate a hospital caseload-related variation in early outcome as reported by Dimick et al.
      • Dimick J.B.
      • Cowan Jr., J.A.
      • Henke P.K.
      • Wainess R.M.
      • Posner S.
      • Stanley J.C.
      • et al.
      Hospital volume-related differences in aorto-bifemoral bypass operative mortality in the United States.
      Our range of 30-day mortality was, however, smaller and we included fewer institutions than Dimick et al. We were, nevertheless, still able to demonstrate significant variation (from 5% to 17%) in mortality after major complications among hospitals. This fact and the generally high incidence of cardiopulmonary complications emphasizes that the focus in open aortic surgery should be not only on how to prevent complications but also how to discover and effectively treat these inevitable complications.
      • Ghaferi A.A.
      • Birkmeyer J.D.
      • Dimick J.B.
      Complications, failure to rescue, and mortality with major inpatient surgery in medicare patients.
      Thus, to further investigate this aspect, future studies should also include aortic aneurysm repair and perform a more thorough investigation of other hospital distinctive variables, aside from annual caseload, that may explain the variation in mortality after major complication among hospitals (eg, the number of patients accepted for surgery despite a high level of comorbidity, transfusion strategies, in-house 24/7 access to acute cardiac intervention, and surgical gastroenterologic support, etc).
      Regarding intraoperative factors, such as surgical technique and method of anesthesia, there is always an inherent risk of neglecting the need for further improvement of the existing procedures (open) when introducing a new treatment option (endovascular). Evidence-based initiatives to reduce the patient's risk must be present in every aspect of patient care. In this study, this concern was expressed as a recommendation to choose epidural anesthesia, ABI bypass, and probably, a transverse incision. One must not neglect future attempts to improve and refine the surgical approach in our endeavor to improve and expand the indication for endovascular repair.
      Our finding of unchanged 30-day mortality despite diminished surgical activity and prolonged procedures during the study period must be interpreted with caution because a closer examination of these risk factors showed that the medical risk factors decreased during the study period (Fig 1, c). Thus, we acknowledge that the unaffected 30-day outcome, despite an increased blood loss and prolonged procedure times, may first of all have been driven by improved patient selection, meaning that the most comorbid patients were preferentially treated endovascularly. One could speculate that postoperative care might have improved during the 20-year study period and had a positive effect on 30-day outcome. Unfortunately, our data set does not allow us to draw safe conclusions on these matters.
      Although, we excluded patients with a high risk profile, such as acute aortoiliac thrombosis, and previous intra-abdominal vascular surgery, the study was still favored by its large number of patients. During a study period of 20 years, some unpredictable and uncontrollable factors besides the evolvement of endovascular techniques, such as improvements in medical and critical care, might have influenced the results. However, forcing the year of procedure into our model to correct for temporal changes did not change the significance of the predictors presented in Table IV.
      The primary drawback of this study is the lack of population data on the number of patients undergoing iliac stent procedures for occlusive disease during the same period of time, as well as the missing characterization of the atherosclerotic lesions treated and the degree of aortic and peripheral calcification present. Furthermore, the data set did not allow us to discern any effect of decreased surgeon experience, whether the procedure was done in a teaching setting or by a experienced key surgeon, but we acknowledge these data would have been important.
      Although we have national recommendations that include the administration of platelet inhibitors and statins before arterial surgery, a more detailed list of medications was not available. Intraoperative details, such as clamp location, aortic cross-clamping time, and vasopressor requirements, were also not entered into the Registry. Furthermore, we did not evaluate whether adjunctive run-off procedures, extended endarterectomies, or the number of technical failures and amputations increased over time.
      Finally, because we could not retrieve any information on readmission to another hospital, these data may hide an even bigger 30-day complication rate. However, readmission ≤30 days to another hospital for complications related to the surgical procedure is very unlikely due to the organization of health care in Denmark.
      A comparison of our results with other reports shows that register-based surveys carry some inherited drawbacks, including nonaccession to the criteria used to determine complications and which parameters each study chose to include. The cardiac complication rate is irregularly reported in the available literature, where it ranges from 0.5% to 10%.
      • Nevelsteen A.
      • Wouters L.
      • Suy R.
      Aortofemoral Dacron reconstruction for aorto-iliac occlusive disease: a 25-year survey.
      • Poulias G.E.
      • Polemis L.
      • Skoutas B.
      • Doundoulakis N.
      • Papaioannou K.
      • Ershaid B.
      • et al.
      Bilateral aorto-femoral bypass in the presence of aorto-iliac occlusive disease and factors determining results. Experience and long term follow up with 500 consecutive cases.
      • Szilagyi D.E.
      • Elliott Jr., J.P.
      • Smith R.F.
      • Reddy D.J.
      • McPharlin M.
      A thirty-year survey of the reconstructive surgical treatment of aortoiliac occlusive disease.
      Compared with this, our reported cardiac complication rate of 6% is average, but most importantly, almost 30% of patients with a cardiac complication died ≤30 days. Pulmonary complications were the most common medical complications according to our Registry. Hence, nearly 10% of all patients suffered from pneumonia or respiratory distress syndrome or needed prolonged mechanical ventilation. This rate is, however, lower than the rate of major pulmonary complications of 13% to 16% reported by others.
      • Calligaro K.D.
      • Azurin D.J.
      • Dougherty M.J.
      • Dandora R.
      • Bajgier S.M.
      • Simper S.
      • et al.
      Pulmonary risk factors of elective abdominal aortic surgery.
      Finally, some known complications, such as colitis, cholecystitis, gastric bleeding events, impotence, spinal cord ischemia, and ureteral injuries, are not routinely notified in the Registry and are therefore under-reported.
      • Valentine R.J.
      • Hagino R.T.
      • Jackson M.R.
      • Kakish H.B.
      • Bengtson T.D.
      • Clagett G.P.
      Gastrointestinal complications after aortic surgery.
      • Nevelsteen A.
      • Beyens G.
      • Duchateau J.
      • Suy R.
      Aorto-femoral reconstruction and sexual function: a prospective study.
      • Rosenthal D.
      Spinal cord ischemia after abdominal aortic operation: is it preventable?.
      Thus, only severe complications necessitating laparotomy, surgical revision, or medical treatment were registered in our Registry, and this may explain why other studies report a higher incidence.

      Conclusions

      Our data show that insertion of an aortic bifurcated graft should still be considered a high-risk procedure, because every fifth patient faced a major complication. The mortality remains considerable, especially among the oldest patients and those with renal insufficiency and severe atherosclerotic manifestations. Having these numbers in mind, we must question the appropriateness of the aortic surgery performed because our data set contained a very large cohort of elderly patients who underwent open aortic reconstruction for claudication (63%). The decrease in open surgical activity, however, did not result in a worse outcome for patients during the study period, and this may reflect improved intraoperative and postoperative care or better patient selection. Thus, provided that a careful preoperative assessment is undertaken, the use of an aortic bifurcated graft and preferably ABI rather than ABF bypass procedure remains a reasonable option when less invasive possibilities are fully exhausted.

      Author contributions

      Conception and design: KB, LJ, TS, HS, JE
      Analysis and interpretation: KB, LJ, TS, HS, JE
      Data collection: KB, LJ, JE
      Writing the article: KB, LJ, TS, HS, JE
      Critical revision of the article: KB, LJ, TS, HS, HN, JE
      Final approval of the article: KB, LJ, TS, HS, HN, JE
      Statistical analysis: KB, HN
      Obtained funding: HS
      Overall responsibility: KB

      References

        • Jongkind V.
        • Akkersdijk G.J.
        • Yeung K.K.
        • Wisselink W.
        A systematic review of endovascular treatment of extensive aortoiliac occlusive disease.
        J Vasc Surg. 2010; 52: 1376-1383
        • Karbase Landsregister-The Danish Vascular Registry
        (Available at:)
        www.karbase.dk
        Date: 2013
        (Accessed April 31, 2013)
        • Laustsen J.
        • Jensen L.P.
        • Hansen A.K.
        Accuracy of clinical data in a population based vascular registry.
        Eur J Vasc Endovasc Surg. 2004; 27: 216-219
        • Rutherford R.B.
        • Baker J.D.
        • Ernst C.
        • Johnston K.W.
        • Porter J.M.
        • Ahn S.
        • et al.
        Recommended standards for reports dealing with lower extremity ischemia: revised version.
        J Vasc Surg. 1997; 26: 517-538
        • Hosmer Jr., D.W.
        • Lemeshow S.
        • May S.
        Applied survival analysis regression modeling of time-to-event data.
        2nd edition. Wiley & Sons, Hoboken, NJ2008
        • de Vries S.O.
        • Hunink M.G.
        Results of aortic bifurcation grafts for aortoiliac occlusive disease: a meta-analysis.
        J Vasc Surg. 1997; 26: 558-569
        • Brady A.R.
        • Fowkes F.G.
        • Greenhalgh R.M.
        • Powell J.T.
        • Ruckley C.V.
        • Thompson S.G.
        Risk factors for postoperative death following elective surgical repair of abdominal aortic aneurysm: results from the UK Small Aneurysm Trial. On behalf of the UK Small Aneurysm Trial participants.
        Br J Surg. 2000; 87: 742-749
        • Kantonen I.
        • Lepantalo M.
        • Luther M.
        • Salenius P.
        • Ylonen K.
        Factors affecting the results of surgery for chronic critical leg ischemia-a nationwide survey. Finnvasc Study Group.
        J Vasc Surg. 1998; 27: 940-947
        • Dimick J.B.
        • Cowan Jr., J.A.
        • Henke P.K.
        • Wainess R.M.
        • Posner S.
        • Stanley J.C.
        • et al.
        Hospital volume-related differences in aorto-bifemoral bypass operative mortality in the United States.
        J Vasc Surg. 2003; 37: 970-975
        • Ghaferi A.A.
        • Birkmeyer J.D.
        • Dimick J.B.
        Complications, failure to rescue, and mortality with major inpatient surgery in medicare patients.
        Ann Surg. 2009; 250: 1029-1034
        • Nevelsteen A.
        • Wouters L.
        • Suy R.
        Aortofemoral Dacron reconstruction for aorto-iliac occlusive disease: a 25-year survey.
        Eur J Vasc Surg. 1991; 5: 179-186
        • Poulias G.E.
        • Polemis L.
        • Skoutas B.
        • Doundoulakis N.
        • Papaioannou K.
        • Ershaid B.
        • et al.
        Bilateral aorto-femoral bypass in the presence of aorto-iliac occlusive disease and factors determining results. Experience and long term follow up with 500 consecutive cases.
        J Cardiovasc Surg (Torino). 1985; 26: 527-538
        • Szilagyi D.E.
        • Elliott Jr., J.P.
        • Smith R.F.
        • Reddy D.J.
        • McPharlin M.
        A thirty-year survey of the reconstructive surgical treatment of aortoiliac occlusive disease.
        J Vasc Surg. 1986; 3: 421-436
        • Calligaro K.D.
        • Azurin D.J.
        • Dougherty M.J.
        • Dandora R.
        • Bajgier S.M.
        • Simper S.
        • et al.
        Pulmonary risk factors of elective abdominal aortic surgery.
        J Vasc Surg. 1993; 18: 914-920
        • Valentine R.J.
        • Hagino R.T.
        • Jackson M.R.
        • Kakish H.B.
        • Bengtson T.D.
        • Clagett G.P.
        Gastrointestinal complications after aortic surgery.
        J Vasc Surg. 1998; 28: 404-411
        • Nevelsteen A.
        • Beyens G.
        • Duchateau J.
        • Suy R.
        Aorto-femoral reconstruction and sexual function: a prospective study.
        Eur J Vasc Surg. 1990; 4: 247-251
        • Rosenthal D.
        Spinal cord ischemia after abdominal aortic operation: is it preventable?.
        J Vasc Surg. 1999; 30: 391-397