Population-based outcomes of open descending thoracic aortic aneurysm repair
Article Outline
Objective
To evaluate national outcomes after open repair of descending thoracic aortic aneurysm (DTA).
Methods
The DTA repairs were identified from the NIS database from 1988-2003 by ICD9 codes for thoracic vascular resection and replacement (38.45) and a diagnosis of intact (441.1) or ruptured (441.2) thoracic aortic aneurysm; excluding thoraco-abdominal aneurysm, abdominal aortic aneurysm repair, cardioplegia, hypothermia, cardiac surgery, or aorta to carotid or subclavian bypass. Demographics and comorbidities were noted. Outcomes included in-hospital mortality, length of stay, and complications. Annual hospital surgical volume terciles (high, medium, and low) were quantified for the series and patients assigned accordingly. Outcomes were compared between intact and ruptured aneurysm characteristics as well as annual hospital volume. Predictors of peri-operative mortality were analyzed by multivariate logistic regression.
Results
A total of 2549 DTA repairs were identified (1976 intact, 573 ruptured). Mortality was 18% overall; 10% for intact (age <65 6.2%, 65-74 11.3%, ≥75 17.6%, P < .001), 45% for ruptured (age <65 33.3%, 65-74 47.1%, ≥75 52.4%, P < .001). Mortality decreased over the 15-year time-period (P < .0001). Mortality after intact repair was lower at a high volume hospital (HVH) (8%) than a low volume hospital (LVH) (13%) or medium volume hospital (MVH) (12%). Hospital volume tercile did not predict rupture mortality. Complications after intact DTA repair were coded in 42%; including respiratory (13%), cardiac (11%), acute renal failure (8%), stroke (3%), and neurologic (non-stroke) (2%). Complications were coded in 49% after ruptured DTA repair including respiratory (13%), cardiac (13%), acute renal failure (20%), stroke (3%), and neuro (non-stroke) (2%). Predictors of mortality (for all DTA repairs) were (odd ratio [OR], 95% confidence interval [CI]): age 65-74 vs age <65 (1.8, 1.4-2.4), age ≥75 vs age <65 (2.7, 2.0-3.6), rupture (6.3, 5.1-7.9), and LVH or MVH vs HVH (1.3, 1.1-1.7).
Conclusion
Mortality after open repair of DTA is high and complications are common. Mortality is dependent upon age, rupture status, and hospital surgical volume. Results of endovascular DTA repair should be compared using similar population-based data.
The incidence of thoracic aortic aneurysms has increased significantly over the past 50 years.1, 2, 3 The rupture risk approaches 50% in large descending thoracic aortic aneurysms (DTA), therefore, early recognition and repair is crucial to patient survival.4 Operative intervention however is not without significant morbidity and mortality. Mortality rates for open surgical DTA repair have been reported at 3-12%.5, 6, 7, 8, 9, 10, 11 Most reports, however, are from high volume centers of excellence and, therefore, these results may not reflect the actual experience nationwide. Additionally, many reports in the literature include the more extensive thoraco-abdominal aortic aneurysms making it difficult to ascertain the expected outcome after isolated DTA.11, 12, 13, 14, 15 With the recent Food & Drug Administration (FDA) approval of thoracic endovascular aneurysm repair (TEVAR) for repair of DTA, it is important to have benchmarks for results of open DTA repair and be able to compare population-based outcomes. In this report, we identify patients with isolated DTA undergoing open surgical repair from a large database reflective of the national experience from the 16-year period prior to FDA approval of TEVAR. We also examine the effect of hospital surgical volume on mortality.
Methods
Database
The Nationwide In-patient Sample is a database maintained through the Healthcare Cost and Utilization Project (HCUP) and represents an approximate 20% random sample of all in-patient hospitalizations from 37 states. More than 8 million hospitalizations are represented per year and the data recorded have been used extensively in medical research to provide outcome analyses in a variety of healthcare topics.16 In using this database for evaluation of DTA repairs, we are able to study the population-based outcomes and characteristics of this relatively rare procedure.
Patients
Clinical data were extracted from the Nationwide Inpatient Sample (NIS) database from 1988-2003 using diagnosis and procedure codes from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) system codes. A procedure code for thoracic vascular resection and replacement (38.45) and a diagnosis of thoracic aortic aneurysm (441.1 or 441.2) identified patients who had undergone repair of a thoracic aortic aneurysm within the admission. To isolate descending thoracic aneurysms, exclusion criteria included concomitant diagnosis of thoraco-abdominal aneurysm (441.6, 441.7), abdominal aortic aneurysm repair (38.44), cardioplegia (39.63), hypothermia (39.62), cardiac surgery (35.00-37.99), aorta to carotid or subclavian bypass (39.22), and intrathoracic bypass (39.23). In addition, patients were required to be 18 years of age or older to minimize inclusion of congenital anomalies.
Demographics and hospital volume
Each patient was classified as having either an intact (441.2) or a ruptured (441.1) aneurysm based upon their qualifying ICD-9 diagnosis code. In addition, hospital-specific surgical volumes were tabulated for each year within the dataset and terciles of operative volume were defined as a low volume hospital (LVH), a medium volume hospital (MVH), and a high volume hospital (HVH). Each patient within the sample was then assigned into one of these categories based on their designated institution and hospitalization year.
Age, gender, race, and comorbidities were documented. Comorbid conditions were determined by ICD-9 diagnosis codes. These included coronary artery disease (CAD), prior myocardial infarction, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), chronic renal failure (CRF), cerebral vascular disease (CVD), diabetes mellitus (DM), hypertension, and peripheral vascular disease (PVD).
Outcomes
The primary outcome variable was in-hospital mortality and secondary outcomes included total length of stay (LOS) and complications. Complications were identified by ICD-9 codes for complications of surgical and medical care (996-999) and included stroke, non-stroke neurologic, cardiac, and respiratory complications. Coding for acute renal failure was also considered a complication diagnosis. Outcomes were compared between intact and ruptured groups as well as among hospital surgical volume terciles.
Statistical analysis
The NIS database querying was done using SAS 9.1 statistical software (SAS Institute Inc, Cary, NC). All statistical analysis was then performed with R statistical software (R Foundation for Statistical Computing, Vienna, Austria). Characteristics and outcomes were compared by Pearson's χ2 and Kruskal-Wallis tests for categorical and continuous variables, respectively. Mortality over time was analyzed by the χ2 test of trend as well as logistic regression. Overall predictors of mortality were then analyzed by univariate and multivariate logistic regression. Due to the inability to determine whether some coexisting diagnoses were complications versus preexisting comorbidities, separate multivariate models were derived with and without comorbidities to ensure the reliability of the results. Statistical significance was defined as P < .05.
Results
Overall demographics
A total of 2549 DTA repairs were identified from the database within the years 1988-2003. There were 1976 (77.5%) intact and 573 (22.5%) ruptured aneurysms. Overall demographic and comorbidity characteristics are shown in Table I. The median age was 68 years with a prevalence of White males. Approximately half of the sample was diagnosed with hypertension and 17% had CAD while other comorbidities were rare (<10%).
Table I. Demographic characteristics and peri-operative mortality of patients undergoing open repair of descending thoracic aneurysms
| Age (median, range) | 68 (18, 93) |
| Male (%) | 57.6 |
| White race (%)⁎ | 81.7 |
| Ruptured TAA (%) | 22.5 |
| Comorbidity (%) | |
 Hypertension | 48.2 |
| Coronary artery disease | 17.3 |
| Prior myocardial infarction | 3.0 |
| Congestive heart failure | 8.2 |
| Chronic obstructive pulmonary disease | 5.5 |
| Chronic renal failure | 1.8 |
| Cerebrovascular disease | 5.0 |
| Diabetes, without complications | 5.5 |
| Diabetes, with complications | 0.47 |
| Peripheral vascular disease | 1.6 |
| % Mortality | |
| Overall | 18.3 |
| Patient age | |
| <65 years | 10.6 |
| 65-74 years | 19.0 |
| ≥75 years | 29.6 |
| Ruptured TAA | 45.2 |
| Intact TAA | 10.4 |
⁎830 missing values. |
Outcomes
Overall, in-hospital mortality after surgery was 18.3% (Table I). Mortality decreased over the 15-year time-period with an average odds decrease of 0.94 per year (Fig 1) (P < .0001). Mortality rates varied significantly by patient age with those ≥75 years having a 20% greater mortality than those <65 years. Complications included respiratory (13.3%), cardiac (11.6%), acute renal failure (10.4%), stroke (2.9%), and non-stroke neurologic complications (1.5%). Median length of stay was 17 days (range, 1-330) for all patients.
Fig 1.
Mortality after repair of descending thoracic aortic aneurysm by year of repair.
Intact vs ruptured DTA
Patients with ruptured aneurysms were significantly older than those with intact aneurysms (Table II). Gender and race were equivalent between the two groups. There were a significantly greater proportion of patients with a diagnosis of CAD and hypertension in the intact aneurysm group while the ruptured group was more frequently diagnosed with CHF and CRF. All other comorbid conditions were similar between the groups.
Table II. Demographic data of patients undergoing open repair of descending thoracic aneurysms for intact versus ruptured aneurysms
| Intact n = 1976 | Ruptured n = 573 | P-value | |
|---|---|---|---|
| Age (median, range) | 67 (18, 92) | 72 (20, 93) | <.001 |
| <65 years | 54 (18-64) | 53 (20-64) | .90 |
| 65-74 years | 70 (65-74) | 71 (65-74) | <.001 |
| ≥75 years | 77 (75-92) | 79 (75-93) | <.001 |
| Male (%) | 57.2 | 58.8 | .53 |
| White race (%)⁎ | 81.6 | 81.9 | .96 |
| Comorbidity (%) | |||
| Hypertension | 51.1 | 38.4 | <.001 |
| Coronary artery disease | 18.7 | 12.2 | <.001 |
| Prior myocardial infarction | 3.1 | 2.8 | .88 |
| Congestive heart failure | 7.2 | 11.5 | <.01 |
| Chronic obstructive pulmonary disease | 5.6 | 4.9 | .57 |
| Chronic renal failure | 1.4 | 3.3 | <.01 |
| Cerebrovascular disease | 4.7 | 6.1 | .19 |
| Diabetes, without complications | 5.7 | 5.1 | .65 |
| Diabetes, with complications | 0.5 | 0.4 | .89 |
| Peripheral vascular disease | 1.5 | 1.9 | .57 |
⁎830 missing values. |
Mortality was 10.4% for intact and 45.2% for ruptured DTA (Table III). Mortality increased significantly with age for both intact and ruptured DTA. Median LOS was longer after intact repair. The LOS among those who died was longer than the LOS in those surviving for intact aneurysms (23 vs 17 days) but shorter for ruptured aneurysms (4 vs 21 days). Global complications and acute renal failure were more frequent after ruptured DTA repair than intact repair.
Table III. Peri-operative mortality and outcomes of patients undergoing open repair of descending thoracic aneurysms for intact versus ruptured aneurysms
| Intact n = 1976 | Ruptured n = 573 | P-value | |
|---|---|---|---|
| Mortality (%) | 10.4 | 45.2 | <.001 |
| Patient age (% mortality) | |||
| <65 years | 6.2 | 33.3 | <.001 |
| 65-74 years | 11.3 | 47.1 | <.001 |
| ≥75 years | 17.6 | 52.4 | <.001 |
| LOS (median, range) | 18 (1, 330) | 15 (1, 110) | <.001 |
| Complications (%) | |||
| Global | 42.3 | 48.5 | <.01 |
| Stroke | 2.6 | 3.3 | .46 |
| Neuro non-stroke | 1.5 | 1.7 | .78 |
| Cardiac | 11.2 | 12.7 | .36 |
| Respiratory | 13.4 | 13.1 | .90 |
| Acute renal failure | 7.7 | 20.0 | <.001 |
Hospital volume
Results of analysis by hospital volume are presented in Table IV, Table V. The lowest volume tercile was characterized by the performance of only one DTA repair annually while MVHs performed a median of 2 procedures (range, 2-3) and HVHs performed a median of 4 (range, 3-25). The total number of procedures performed within the LVH group was 685 (27%), however the LVHs accounted for 55% of the hospitals represented. Conversely, 1262 procedures were performed at high volume centers which represented 21% of the hospitals identified (Fig 2).
Table IV. Hospital volume characteristics and demographic characteristics of patients undergoing open repair of descending thoracic aneurysms by hospital volume strata
| LVH | MVH | HVH | P-value | |
|---|---|---|---|---|
| Hospital Characteristics | ||||
| Total number of procedures | 685 (26.9%) | 602 (23.6%) | 1262 (49.5%) | |
| Number of hospitals within category | 685 | 297 | 265 | |
| Median number of procedures | 1 | 2 | 4 | |
| Range | (1, 1) | (2, 3) | (3, 25) | |
| Patient Characteristics | ||||
| Age (median, range) | 68 (21, 89) | 68 (18, 93) | 68 (18, 92) | .15 |
| Male (%) | 56.9 | 57.8 | 57.8 | .94 |
| White race (%)⁎ | 83.9 | 78.9 | 81.8 | .18 |
| Ruptured TAA (%) | 25.4 | 24.9 | 19.7 | <.01 |
| Comorbidity (%) | ||||
| Hypertension | 47.2 | 47.8 | 49.0 | .73 |
| Coronary artery disease | 17.4 | 14.3 | 18.6 | .07 |
| Prior myocardial infarction | 2.2 | 2.2 | 3.9 | <.05 |
| Congestive heart failure | 7.0 | 8.1 | 8.8 | .39 |
| Chronic obstructive pulmonary disease | 6.1 | 5.8 | 4.9 | .48 |
| Chronic renal failure | 1.9 | 1.7 | 1.8 | .95 |
| Cerebrovascular disease | 5.3 | 5.3 | 4.7 | .78 |
| Diabetes, without complications | 6.9 | 6.1 | 4.5 | .07 |
| Diabetes, with complications | 0.6 | 0.5 | 0.4 | .84 |
| Peripheral vascular disease | 2.5 | 1.8 | 1.0 | <.05 |
⁎830 missing values. |
Table V. Peri-operative mortality and outcomes of patients undergoing open repair of descending thoracic aneurysms by hospital volume strata
| LVH | MVH | HVH | P-value |
|---|---|---|---|---|
| Mortality (%) | 21.7 | 20.4 | 15.5 | <.01 |
| Patient age (% mortality) | ||||
| <65 years | 12.0 | 11.1 | 9.7 | .60 |
| 65-74 years | 22.1 | 20.2 | 16.6 | .17 |
| ≥75 years | 34.7 | 36.4 | 23.6 | <.01 |
| Ruptured DTA (% mortality) | 46.5 | 45.6 | 44.2 | .89 |
| Intact DTA (% mortality) | 13.2 | 12.0 | 8.4 | <.01 |
| LOS (median, range) | 15 (1, 176) | 17 (1, 98) | 19 (1, 330) | <.01 |
| Complications (%) | ||||
| Global complications | 44.4 | 41.0 | 44.5 | .33 |
| Stroke | 2.3 | 2.5 | 3.2 | .50 |
| Neuro non-stroke | 2.2 | 1.3 | 1.3 | .26 |
| Cardiac | 11.2 | 8.6 | 13.2 | <.05 |
| Respiratory | 12.4 | 13.3 | 13.9 | .66 |
| Acute renal failure | 10.8 | 11.3 | 9.8 | .58 |
Fig 2.
Distribution of annual hospital surgical volume.
There was a similar age, gender, race, and comorbidity distribution across the hospital volume terciles. The percentage of ruptured DTA repairs was significantly higher in low and medium volume institutions versus high volume institutions.
Overall, mortality was significantly higher within low and medium volume centers versus high volume centers (Table V, Fig 3). Mortality after intact repair was lowest at HVH (8%) versus LVH (13%) or MVH (12%). Mortality after elective repair was not improved further with a higher volume cutoff (Fig 3). Hospital volume terciles did not demonstrate differences for rupture mortality.
Fig 3.
Mortality by annual hospital surgical volume.
Median LOS increased as hospital volume increased. There were no significant differences in complication rates among the volume groups.
Multivariate analysis
Independent predictors of mortality were age, rupture, and hospital volume (Table VI). When including comorbidities, CVD, CRF, and absence of hypertension were additional predictors of mortality (Table VII).
Table VI. Multivariate predictors of mortality associated with descending thoracic aortic aneurysm repair, with comorbidities included
| OR | 95% CI | P-value | |
|---|---|---|---|
| Age 65-74 vs Age <65 | 1.8 | 1.4-2.4 | <.001 |
| Age ≥75 vs Age <65 | 2.6 | 2.0-3.5 | <.001 |
| Comorbidities | |||
| Chronic renal failure | 3.0 | 1.5-6.0 | <.01 |
| Cerebrovascular disease | 2.8 | 1.9-4.3 | <.001 |
| Hypertension | 0.7 | 0.6-0.9 | <.05 |
| LVH/MVH vs HVH | 1.3 | 1.1-1.6 | <.05 |
| Ruptured vs Intact TAA | 6.1 | 4.9-7.7 | <.001 |
Table VII. Multivariate predictors of mortality associated with descending thoracic aortic aneurysm repair, with comorbidities excluded
| OR | 95% CI | P-value | |
|---|---|---|---|
| Age 65-74 vs Age <65 | 1.8 | 1.4-2.4 | <.001 |
| Age ≥75 vs Age <65 | 2.7 | 2.0-3.6 | <.001 |
| LVH/MVH vs HVH | 1.3 | 1.1-1.7 | <.01 |
| Ruptured vs Intact TAA | 6.3 | 5.1-7.9 | <.001 |
Discussion
In-hospital mortality associated with open repair of DTAs is high but varies significantly by hospital operative volume. Even in high volume hospitals, peri-operative mortality with repair of intact DTA was 8%. Ruptured aneurysms carried approximately a 45% mortality rate regardless of hospital experience. Peri-operative mortality for intact and ruptured DTA decreased over the time period of the study.
Morbidity is also very high for open DTA repair with complication rates reaching 50%. Differences in hospital volume did not significantly affect morbidity rates.
In the past, mortality rates of 3-12% have been reported for descending thoracic aneurysms.5, 6, 7, 8, 9, 10, 11 Most prior publications have been reports from academic centers of excellence, therefore, the mortality as well as morbidity outcomes are likely to be better in these studies than in an overall population-based study such as performed here. The 10% intact DTA mortality and 45% ruptured DTA mortality based on a national sample is likely representative of the true population risk over the time course reported.
A large study by Svensson et al5 reported 832 cases of open DTA repair performed from 1956-1991. The series was comprised of only 30% females and the median age was younger (65 years) than that in this series. Additionally, indications other than aneurysmal disease such as trauma and dissection were included. The 30-day mortality was 8% overall (21% for ruptured pathology and 7% for non-ruptured). Complications reported included 7% acute renal failure, 5% paraparesis, 2.3% paraplegia, 3% stroke, 10% cardiac complications, and 28% pulmonary complications.
More recently, the reported literature has focused upon comparing endovascular stent grafts for repair versus the traditional open technique. These reports include single institutional experiences as well as two multicenter comparative trials. One such institutional study by Stone et al6 reported an open cohort of 93 patients from 1996-2005. Peri-operative mortality was 15.1% overall and 9.6% for intact DTA repairs. Mean age was 71 years and length of stay 19 days (9 ICU). Complications included 8.6% paraparesis, 4.5% paraplegia, and 8% intra-operative stroke.
Multicenter trials include those published by Bavaria et al7 and Matsumura et al.11 The former had an open cohort of 94 patients and a peri-operative mortality of 11.7% while the latter had an open cohort of 70 patients and a peri-operative mortality of 5.7%. Complications reported by Bavaria et al included 20% respiratory, 13% renal, 1% myocardial infarction, 4% CVA, and paraplegia/paraparesis of 14%. Matsumura et al reported complication rates of 44% pulmonary, 14% renal, 44% cardiac, 8.5% CVA, and 6% paraplegia.
Mortality, age, and length of stay in these recent trials are similar to the current study with the exception of the mortality of 5.7% reported by Matsumura et al.11 The neurologic complication rates for the most part show a substantial difference and likely represent under-coding of post-operative complications relating to neurologic deficits, whether from spinal cord injury or stroke. Interestingly, we report cardiac and renal complications at 12% and 10% which may be more accurately coded as they are somewhat closer to rates reported in series reporting these outcomes.5, 7, 11
The second aim of our study was to evaluate differences in outcome based on ruptured pathology. Our findings of a 45% mortality associated with rupture is significantly greater than the study by Svensson et al.5 This series represented an annual surgical volume greater than any center noted in our study and, therefore, may reflect a unique expertise with the procedure as well as specifics regarding referral practices in their location. We also cannot make conclusions regarding hemodynamic status of those chosen for transfer and who survive transfer.
There are a few population studies assessing outcomes of thoraco-abdominal aneurysm repairs. Cowan et al17 reported NIS data from 1988 to 1998 on intact thoraco-abdominal aneurysm repair and found significant mortality differences between low and high volume hospitals as well as high and low volume surgeons. They also studied ruptured thoraco-abdominal aneurysms and, as we have, found no relationship of volume to outcome.18 As expected, mortality after intact DTA repair in our study (10%) was lower than mortality after intact thoraco-abdominal aortic aneurysm repair (15%) in Cowan's study. Derrow et al19 previously reported a 20% mortality after elective thoraco-abdominal aneurysm repair in the NIS from 1993 to 1997. The complication rate was high at 62% versus our finding of a 42% global complication rate. Rigberg et al20 used a California statewide database and assessed 30-day mortality in 1010 patients from 1991 to 2002. Mortality was 19% for intact and 48.4% for ruptured thoraco-abdominal aneurysm repair. They found no relationship of procedural volume to mortality outcome. Compared to our current study, this database had an older median age for elective repair (70 years vs 67 years) and a greater percentage of male patients (62% intact and 68% ruptured vs 57% intact and 59% ruptured).
In comparison to outcomes reported for open repair of thoracic aneurysms, endovascular repair for thoracic aneurysms shows promising results for substantial mortality benefits. The European Collaborators on Stent Graft Techniques for Thoracic Aortic Aneurysm and Dissection Repair (EUROSTAR) and the United Kingdom Thoracic Endograft registries reported a 30-day mortality of 5.3% for elective repair and 25% for emergent repair.21 Population-based studies as this technique gains widespread use will be important for future direct comparisons.
There are important limitations that are inherent to studies using administrative data sets. There could be considerable coding variability among institutions limiting the ability to accurately identify comorbid conditions, complications, and even confusing the primary diagnoses and procedures. The standardization of the ICD-9 coding system helps to minimize this, however, the dataset only represents what is recorded in discharge information. Because one cannot reliably determine, using this dataset, whether some comorbid conditions were in fact pre-existing or were complications, separate multivariate models were run with and without comorbidities to control for this limitation. While CRF and CVD predicted worse outcome, these may have been indicative of postoperative stroke or dialysis requirements. In the model excluding comorbidities, however, there was little impact on the other predictors of adverse outcome (age, rupture, emergency admission, and hospital volume) strengthening the independence of these findings. Additionally, due to limitations in the ICD-9 coding system captured by the NIS, we chose to exclude patients undergoing DTA repair using circulatory arrest. While some surgeons use circulatory arrest for routine DTA repair, it is more commonly used for more extensive aortic arch surgery. The exclusion was necessary to more accurately select for DTA repairs only.
This study demonstrates that DTA repair across the US is associated with a high mortality and morbidity. For intact DTA, outcomes are superior at more experienced centers while this was not seen for ruptured DTA. This study includes the period prior to the introduction of TEVAR and should serve as a benchmark for trials of TEVAR. Analysis of TEVAR outcomes should include population-based studies when these data are available to be sure that efficacy demonstrated in high volume centers of excellence can be generalized to the US population.
Author contributions
References
- Thoracic aortic aneurysms: a population-based study. Surgery. 1982;92:1103–1108
- . Improved prognosis of thoracic aortic aneurysms: a population-based study. JAMA. 1998;280:1926–1929
- . Expert consensus document on the treatment of descending thoracic aortic disease using endovascular stent-grafts. Ann Thorac Surg. 2008;85:S1–S41
- . What is the appropriate size criterion for resection of thoracic aortic aneurysms?. J Thorac Cardiovasc Surg. 1997;113:476–491
- . Variables predictive of outcome in 832 patients undergoing repairs of the descending thoracic aorta. Chest. 2003;104:1248–1253
- . Stent-graft versus open-surgical repair of the thoracic aorta: mid-term results. J Vasc Surg. 2006;44:1188–1197
- . Endovascular stent grafting versus open surgical repair of descending thoracic aortic aneurysm in low-risk patients: a multicenter comparative trial. J Thorac Cardiovasc Surg. 2007;133:369–377
- . Descending thoracic aortic aneurysm: surgical approach and treatment using the adjuncts cerebrospinal fluid drainage and distal aortic perfusion. Ann Thorac Surg. 2001;72:481–486
- Selective approach to descending thoracic aortic aneurysm repair: a ten-year experience. Ann Thorac Surg. 1996;62:1152–1157
- . Mid-term survival and costs of treatment of patients with descending thoracic aortic aneurysms; endovascular vs open repair: a case-control study. Eur J Vasc Endovasc Surg. 2005;299:28–34
- . International controlled clinical trial of thoracic endovascular aneurysm repair with the Zenith TX2 endovascular graft: 1-year results. J Vasc Surg. 2008;47:247–257
- . Contemporary management of descending thoracic and thoracoabdominal aortic aneurysms: endovascular versus open. Circulation. 2008;117:841–852
- Endovascular stenting versus open surgery for thoracic aortic disease: systematic review and meta-analysis of peri-operative results. J Vasc Surg. 2008;Jan 31: [Epub ahead of print]
- . Thoraco-abdominal aortic aneurysm: observations regarding the natural course of the disease. J Vasc Surg. 1986;3:578–582
- Prospective study of the natural history of thoracic aortic aneurysms. Ann Thor Surg. 1997;64:1533–1545
- . Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2007;April. www.hcup-us.ahrq.gov/tools_software.jsp
- . Surgical treatment of intact thoraco-abdominal aortic aneurysms in the United States: hospital and surgeon volume-related outcomes. J Vasc Surg. 2003;37:1169–1174
- . Ruptured thoraco-abdominal aortic aneurysm treatment in the United States: 1988 to 1998. J Vasc Surg. 2003;38:319–322
- The outcome in the United States after thoraco-abdominal aortic aneurysm repair, renal artery bypass, and mesenteric revascularization. J Vasc Surg. 2001;34:54–61
- J Vasc Surg. 2006;43:217–223
- . Endovascular treatment of thoracic aortic diseases: Combined experience from the EUROSTAR and United Kingdom Thoracic Endograft registries. J Vasc Surg. 2004;40:670–680
Competition of interest: none.
PII: S0741-5214(08)00779-9
doi:10.1016/j.jvs.2008.05.022
© 2008 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
