Abstract
Background
This report examines the effects of regional versus general anesthesia for infrainguinal bypass procedures performed in the treatment of critical limb ischemia (CLI).
Methods
Nonemergent infrainguinal bypass procedures for CLI (defined as rest pain or tissue loss) were identified using the 2005 to 2008 American College of Surgeons National Surgical Quality Improvement Program database using International Classification of Disease, ninth edition, and Current Procedure Terminology codes. Patients were classified according to National Surgical Quality Improvement Program data as receiving either general anesthesia or regional anesthesia. The regional anesthesia group included those specified as having regional, spinal, or epidural anesthesia. Demographic, medical, risk factor, operative, and outcomes data were abstracted for the study sample. Individual outcomes were evaluated according to the following morbidity categories: wound, pulmonary, venous thromboembolic, genitourinary, cardiovascular, and operative. Length of stay, total morbidity, and mortality were also evaluated. Associations between anesthesia types and outcomes were evaluated using linear or logistic regression.
Results
A total of 5,462 inpatient hospital visits involving infrainguinal bypasses for CLI were identified. Mean patient age was 69 ± 12 years; 69% were Caucasian; and 39% were female. In all, 4,768 procedures were performed using general anesthesia and 694 with regional anesthesia. Patients receiving general anesthesia were younger and significantly more likely to have a history of smoking, previous lower-extremity bypass, previous amputation, previous stroke, and a history of a bleeding diathesis including the use of warfarin. Patients receiving regional anesthesia had a higher prevalence of chronic obstructive pulmonary disease.
Tibial-level bypasses were performed in 51% of procedures, whereas 49% of procedures were popliteal-level bypasses. Cases performed using general anesthesia demonstrated a higher rate of resident involvement, need for blood transfusion, and operative time. There was no difference in the rate of popliteal-level and infrapopliteal-level bypasses between groups. Infrapopliteal bypass procedures performed using general anesthesia were more likely to involve prosthetic grafts and composite vein. Mortality occurred in 157 patients (3%). The overall morbidity rate was 37%. Mean and median lengths of stay were 7.5 days (±8.1) and 6.0 days (Q1: 4.0, Q3: 8.0), respectively. Multivariate analyses demonstrated no significant differences by anesthesia type in the incidence of morbidity, mortality, or length of stay.
Conclusion
These results provide no evidence to support the systematic avoidance of general anesthesia for lower-extremity bypass procedures. These data suggest that anesthetic choice should be governed by local expertise and practice patterns.
INTRODUCTION
Open surgical bypass is a common procedure in vascular surgical practice, with > 70,000 procedures performed annually to treat critical lower-limb ischemia.1 Given the morbidities inherent to patients requiring such procedures, optimization of perioperative care is important. There have been multiple studies investigating the potential effects of anesthetic type on outcome of lower-extremity bypass procedures.
In 1993, the Perioperative Ischemia Randomized Anesthesia Trial study group randomized 100 patients undergoing lower-extremity revascularization to general anesthesia or epidural anesthesia.2 The resulting data demonstrated that there was no significant difference according to anesthetic type in outcomes, including cardiac events, mortality, infection, or pulmonary complications, but did demonstrate a significant increase in the need for graft revision or embolectomy among patients undergoing general anesthesia. The increased rate of reintervention was postulated by the authors to be secondary to differences in circulating catecholamine and Cortisol levels. Several single-institution studies have reported contradictory findings, demonstrating no association between anesthesia type and lower-extremity bypass outcomes.2–4 The only multi-institution study was reported by Singh et al., who echoed the findings of increased morbidity associated with the use of general anesthesia in veterans undergoing lower-extremity bypass.5 To examine the issue further, this study examined a large sample of North American vascular surgery procedures to evaluate possible associations between anesthesia type and outcomes after lower-extremity bypass for critical limb ischemia (CLI) in contemporary practice.
METHODS
Data Source
This investigation used data from the America College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP), which is a validated prospective database compiled through a systematic sampling of cases at participating U.S. hospitals. Data files are provided by year. These files included cases from 121 sites in 2005 with an increase to 211 sites by 2008. Available data elements included demographic information, preoperative data, intraoperative data, and postoperative data. A trained and quality-audited research nurse collected 141 variables for each patient at each site for sampled surgical procedures. Maximum follow-up in this program was 30 days. Definitions for the variables in the NSQIP database have been described in previous reports.6–8 Long-term graft patency data were not available.
Study Sample
Nonemergent infrainguinal bypass procedures for CLI were identified using the 2005 to 2008 ACS NSQIP Participant Use Data File. CLI was defined as rest pain or tissue loss and included patients with Rutherford classifications of 4 to 6.9 Patients were initially identified by Current Procedure Terminology codes (American Medical Association, Chicago, IL) indicating lower-extremity bypass procedure to the popliteal or tibial level (Table I). Inclusion also required the NSQIP variable for rest pain/gangrene to be “yes” or identification of at least one International Classification of Disease, ninth edition, code denoting CLI, including codes for rest pain, ulceration, or gangrene (Table I). Patients undergoing emergency procedures or undergoing concomitant surgical procedures requiring general anesthesia were excluded from further analyses.
Table I.
Procedural and diagnostic codes for inclusion
| Lower-extremity vascular bypass CPT inclusion codes
|
ICD-9 inclusion codes
|
|||
|---|---|---|---|---|
| Code | Conduit | Anatomic level | Code | Diagnosis |
| 35556 | Vein | Femoral to popliteal | 440.22 | Atherosclerosis of native arteries of the extremities with rest pain |
| 35566 | Vein | Femoral to tibial, peroneal, or other distal vessel | 440.23 | Atherosclerosis of native arteries of the extremities with ulceration |
| 35570 | Vein | Tibial to tibial, peroneal, or other distal vessel | 440.24 | Atherosclerosis of native arteries of the extremities with gangrene |
| 35571 | Vein | Popliteal to tibial, peroneal, or other distal vessel | 707.06 | Chronic ulcer of skin, ankle |
| 35583 | In situ vein | Femoral popliteal bypass | 707.07 | Chronic ulcer of skin, heel |
| 35585 | In situ vein | Femoral to tibial, peroneal, or other distal vessel | 707.1 | Ulcer of lower limb, except pressure ulcer |
| 35587 | In situ vein | Popliteal to tibial, peroneal, or other distal vessel | 707.10 | Unspecified ulcer of lower limb |
| 35656 | Grafta | Femoral popliteal bypass | 707.12 | Ulcer of calf |
| 35666 | Grafta | Femoral to tibial, peroneal, or other distal vessel | 707.13 | Ulcer of ankle |
| 35671 | Grafta | Popliteal to tibial, peroneal, or other distal vessel | 707.14 | Ulcer of heel and midfoot |
| 707.15 | Ulcer of other part of foot | |||
| 707.19 | Ulcer of other part of lower limb | |||
| 707.9 | Chronic ulcer of unspecified site | |||
| 785.4 | Gangrene | |||
CPT, Current Procedure Terminology; ICD-9, International Classification of Disease, ninth edition.
CPT codes and ICD-9 codes used for case selection.
At least one CPT code from this table was required for patient inclusion in the study.
In absence of rest pain variable in NSQIP database being yes, at least one of the above ICD-9 codes was required for study inclusion.
CPT Copyright 2010 American Medical Association. All rights reserved. CPT is a registered trademark of the American Medical Association.
“Graft” denotes use of a graft other than vein for bypass conduit.
Demographic and Medical Risk Factors
Demographic, medical, risk factor, operative, and outcomes data were abstracted for the study sample from the ACS NSQIP database. Race was considered as white or nonwhite. Age was considered as a continuous variable. The ACS NSQIP Participant Use Data File variable for age is a numerical value corresponding to the patient’s numeric age in years for all records except those with an age in excess of 90, who are coded as 90+ to minimize identifiable information within the record. Estimated glomerular filtration rate was used to assess renal function and was calculated using the abbreviated Modification of Diet in Renal Disease formula.10 Body mass index was calculated using height and weight data.
Anesthesia Type
Data regarding the type of anesthetic were extracted from the ACS NSQIP for all identified elective lower-extremity bypass cases. Anesthesia type is designated in the ACS NSQIP as general, epidural, spinal, local, monitored anesthesia care, and other. Regional anesthesia was defined as spinal, epidural, and/or regional.
Outcomes
The outcomes for analysis were morbidity, mortality, and length of stay (LOS). Morbidity included the following morbidity classes considered individually as well as in aggregate: wound (superficial or deep surgical site infection), pulmonary (pneumonia, reintubation, or failure to wean from ventilator within 48 hours), venous thromboembolic (deep vein thrombosis, or pulmonary embolus), genitourinary (acute renal insufficiency, acute renal failure, or urinary tract infection), cardiovascular (myocardial infarction, cardiac arrest, or stroke), or operative (return to operating room, postoperative bleeding, or graft failure). Postoperative mortality was defined as death within 30 days or during the same acute care hospital stay. LOS was denned as the time from surgery to hospital discharge or death.
Preoperative characteristics, medical risk factors, and procedural data were described using mean ± standard deviation, count (%), or median (Q1, Q3). Characteristics were compared using univariate techniques, including χ2 or Fisher exact test for categorical variables and t tests for continuous variables.
Morbidity and mortality associations were examined using logistic regression. LOS was log-transformed before analysis with linear regression to satisfy normality assumptions. The LOS comparison was back-transformed for presentation as percent difference. All multivariate analyses were adjusted for age, race, gender, operative time, and total work relative value units (RVUs) of the component Current Procedure Terminology codes defining the surgical procedure. Work RVUs were used as a surrogate marker for case complexity to address the concerns from earlier work that patients undergoing general anesthesia had more complex procedures. Additional covariates were included in the analyses of each of the grouped morbidity classes as well as mortality. Covariates for the multivariate analyses were chosen according to previous full-sample analyses of the ACS NSQIP by central ACS statistical faculty (which are available to each participating site) examining predictors for each of the morbidity classes detailed previously and mortality in vascular surgery patients. The covariates for each analysis are detailed in the tabulated results that follow this report. All analyses were performed using SAS software, version 9.2 (SAS Institute, Cary, NC).
RESULTS
Patient Characteristics
A total of 5,462 lower-extremity bypass procedures for critical ischemia meeting the inclusion and exclusion criteria were identified. Demographic data are summarized in Table II. In all, 4,768 cases (87%) were done with general anesthesia, whereas 694 cases (13%) were done under regional anesthesia. The regional anesthesia group consisted of 273 (39%) cases using epidural anesthesia, 373 (54%) cases using spinal anesthesia, and 48 (7%) classified as “regional”. Thirty-nine percent of identified patients were female, and 69% were white.
Table II.
Selected demographic characteristics among entire sample and according to anesthetic group
| Variable | Overall n = 5,462 |
General anesthesia n = 4,768 |
Regional anesthesia n = 694 |
P value* |
|---|---|---|---|---|
| Age (yr) | 68.8 ± 12.0 | 68.3 ± 12.1 | 71.8 ± 11.5 | <0.0001 |
| Age ≥80 yr (%) | 1,205 (22.1) | 997 (20.9) | 208 (30.0) | <0.0001 |
| Nonwhite race | 1,719 (31.5) | 1,485 (31.2) | 234 (33.7) | 0.1728 |
| Female gender | 2,147 (39.3) | 1,878 (39.4) | 269 (38.8) | 0.7521 |
| BMI | 27.0 ± 6.2 | 27.0 ± 6.0 | 26.8 ± 7.0 | 0.3606 |
| Diabetes | 2,657 (48.7) | 2,308 (48.4) | 349 (50.3) | 0.3540 |
| Current smoker | 2,194 (40.2) | 1,955 (41.0) | 239 (34.4) | 0.0010 |
| Functional status | ||||
| Independent | 4,189 (76.7) | 3,682 (77.2) | 507 (73.1) | 0.0152 |
| Partially/totally dependent | 1,273 (23.3) | 1,086 (22.8) | 187 (27.0) | |
| History of | ||||
| COPD | 728 (13.3) | 617 (12.9) | 111 (16.0) | 0.0270 |
| CHF | 201 (3.7) | 170 (3.6) | 31 (4.5) | 0.2386 |
| MI | 138 (2.5) | 123 (2.6) | 15 (2.2) | 0.5118 |
| Angina | 131 (2.4) | 118 (2.5) | 13 (1.9) | 0.3331 |
| Previous CABG | 1,407 (25.8) | 1,237 (25.9) | 170 (24.5) | 0.4150 |
| Previous PTCI | 1,028 (18.8) | 908 (19.0) | 120 (17.3) | 0.2698 |
| HTN | 4,633 (84.8) | 4,044 (84.8) | 589 (84.9) | 0.9700 |
| Revascularization or amputation | 3,156 (57.8) | 2,784 (58.4) | 372 (53.6) | 0.0171 |
| Dialysis dependence | 540 (9.9) | 477 (10.0) | 63 (9.1) | 0.4449 |
| TIA | 375 (6.9) | 323 (6.8) | 52 (7.5) | 0.4843 |
| Stroke without residual disability | 409 (7.5) | 370 (7.8) | 39 (5.6) | 0.0453 |
| Stroke with residual disability | 491 (9.0) | 428 (9.0) | 63 (6.1) | 0.9305 |
| Bleeding disordera | 1,315 (24.1) | 1,220 (25.6) | 95 (13.7) | <0.0001 |
| Preoperative sepsis (SIRS/sepsis/septic shock) | 384 (7.0) | 340 (7.1) | 44 (6.3) | 0.4464 |
| 10% or more weight loss in past 6 months | 122 (2.2) | 102 (2.1) | 20 (2.9) | 0.2161 |
| Acute renal failure | 109 (2.0) | 93 (2.0) | 16 (2.3) | 0.5321 |
| Transfer status | ||||
| Other hospital or facility | 426 (7.8) | 370 (7.8) | 56 (8.1) | 0.7766 |
| Admitted directly from home | 5,036 (92.2) | 4,398 (82.2) | 638 (91.9) | |
| ASA class | ||||
| No disturb, mild, severe (1–3) | 4,297 (78.7) | 3,742 (78.5) | 555 (78.0) | 0.3812 |
| Life-threatening or moribund (4–5) | 1,163 (21.3) | 1,024 (21.5) | 139 (20.0) | |
| eGFR (mL/min) | 64.5 ± 34.2 | 64.9 ± 34.6 | 61.6 ± 31.0 | 0.0106 |
| Chronic dyspnea | 1,080 (19.8) | 929 (19.5) | 151 (21.8) | 0.1600 |
| Wound class >1b | 834 (15.3) | 711 (14.9) | 123 (17.7) | 0.0544 |
| Chronic steroid use | 260 (4.8) | 234 (4.9) | 26 (3.8) | 0.1794 |
| Rest pain/gangrene | 4,244 (77.7) | 3,713 (77.9) | 531 (76.5) | 0.4212 |
| WBC | 8.6 ± 3.2 | 8.6 ± 3.2 | 8.4 ± 3.0 | 0.1130 |
| HCT | 36.2 ± 5.8 | 36.2 ± 5.8 | 36.0 ± 5.6 | 0.3879 |
| Platelet count | 275.0 ± 110.4 | 275.4 ± 110.4 | 272.0 ± 110.6 | 0.4520 |
ASA, American Society of Anesthesiologists; eGER, estimated glomerular filtration rate; MAC, monitored anesthesia care; PTCI, percutaneous transluminal coronary intervention; COPD, chronic obstructive pulmonary disorder; MI, myocardial infarction; CHF, congestive heart failure; CABG, coronary artery bypass graft; HTN, hypertension; TIA, transient ischemic attack; SIRS, systemic inflammatory response syndrome; WBC, white blood cell count; HCT, hematocrit; BMI, body mass index.
Bleeding disorder denotes disorders that increase risk for bleeding, such as presence of hemophilia, thrombocytopenia, vitamin K deficiency, or long-term anticoagulation therapy that was not discontinued before surgery.
Wound class ranges from 1 to 4, where 1 = clean, 2 = clean–contaminated, 3 = contaminated, and 4 = dirty.
P values for tests of differences between general and regional anesthesia from χ2 test for categorical variables or t test for continuous variables.
Patients receiving general anesthesia had a higher baseline prevalence of current smoking, history of revascularization or amputation, history of stroke without disability, preoperative anticoagulation, and a higher glomerular filtration rate. Patients receiving regional anesthesia were older with a higher prevalence of chronic obstructive pulmonary disease and dependent preoperative functional status. The remaining demographic characteristics were similar among the two groups (see Table II).
Procedural Characteristics, by Anesthesia Type
Procedural details are summarized in Table III. Mean operative time was 237 minutes. Twenty percent of patients required transfusion, with a mean transfusion amount of 2 units. Tibial-level bypasses were performed in 2,787 (51%) of procedures, whereas 2,675 (49%) procedures were popliteal-level bypasses. Cases performed using general anesthesia demonstrated a higher rate of resident involvement, need for blood transfusion, and operative time. There was no difference in the rate of popliteal versus infrapopliteal bypasses between groups. Infrapopliteal bypass procedures performed using general anesthesia were more likely to involve prosthetic grafts and composite vein.
Table III.
Selected procedural characteristics among entire sample and according to anesthetic group expressed as count (%) or mean ± SD
| Variable | Overall n = 5,462 |
General anesthesia n = 4,768 |
Regional anesthesia n = 694 |
P value* |
|---|---|---|---|---|
| Surgeon specialty | ||||
| Vascular surgeon | 5,268 (96.5) | 4,598 (96.4) | 670 (96.5) | 0.8866 |
| Other | 194 (3.6) | 170 (3.6) | 24 (3.5) | |
| Resident involved | 3,651 (66.9) | 3,243 (68.1) | 408 (58.9) | <0.0001 |
| Patients requiring transfusion (%) | 1,092 (20.0) | 1,001 (21.0) | 91 (13.1) | <0.0001 |
| Units transfused (among those >0) | 2.0 ± 1.2 | 2.0 ± 1.2 | 2.0 ± 1.1 | 0.7365 |
| Patients requiring >4 units transfused (%) | 46 (0.8) | 45 (0.9) | 1 (0.1) | 0.0312 |
| Operative time (min) | 237.0 ± 101.9 | 240.3 ± 103.2 | 214.6 ± 89.6 | <0.0001 |
| Total work RVUs | 32.4 ± 12.5 | 32.4 ± 12.5 | 32.0 ± 11.9 | 0.4622 |
| Infrapopliteal bypass | 2,787 (51.0) | 2,448 (51.3) | 339 (48.9) | 0.2193 |
| Vein bypass | 2,249 (41.2) | 1,963 (41.2) | 286 (41.2) | 0.9840 |
| Single vein bypass | 2,146 (39.3) | 1,862 (39.1) | 284 (40.9) | 0.3459 |
| Composite vein | 103 (1.9) | 101 (2.1) | 2 (0.3) | 0.0009 |
| Any prosthetic | 538 (9.9) | 485 (10.2) | 53 (7.6) | 0.0363 |
| Prosthetic only | 489 (9.0) | 440 (9.2) | 49 (7.1) | 0.0617 |
| Prosthetic and vein | 49 (0.9) | 45 (0.9) | 4 (0.6) | 0.3375 |
| Popliteal bypass | 2,675 (49.0) | 2,320 (48.7) | 355 (51.2) | 0.2193 |
| Vein bypass | 1,563 (28.6) | 1,341 (28.1) | 222 (32.0) | 0.0354 |
| Single vein bypass | 1,544 (28.3) | 1,322 (27.7) | 222 (32.0) | 0.0198 |
| Composite vein | 19 (0.4) | 19 (0.4) | 0 (0) | 0.1586a |
| Any prosthetic | 1,112 (20.4) | 979 (20.5) | 133 (19.2) | 0.4029 |
| Prosthetic only | 1,080 (19.8) | 953 (20.0) | 127 (18.3) | 0.2970 |
| Prosthetic and vein | 32 (0.6) | 26 (0.6) | 6 (0.9) | 0.3032a |
| Use of prosthetic graft | 1,645 (30.1) | 1,464 (30.7) | 181 (26.1) | 0.0131 |
RVU, relative value unit.
Fisher exact test used in place of χ2 test owing to low expected cell counts.
P values for tests of differences between general and regional anesthesia from χ2 test or Fisher exact test for categorical variables or from t test for continuous variables.
Outcomes, by Procedure Type
Morbidity by class, overall morbidity, mortality, and LOS results are displayed in Table IV. Mortality occurred in 157 patients (3%). The overall morbidity rate was 37%. Mean and median lengths of stay were 7.5 days (± 8.1) and 6.0 days (Q1: 4.0, Q3: 8.0), respectively. The most common morbidity observed was return to the operating room, which occurred in 22% of patients. In particular, graft thrombosis was found in 7.3% of patients, with an equal rate in both groups. Wound complications occurred in 12% of patients, infectious complications (including any septic morbidity) in 6% of patients, and pulmonary morbidity occurred in 4% of patients. The rate of cardiovascular complications was similar between groups, affecting 2.8% of general anesthesia patients and 2.2% of regional anesthesia patients. Genitourinary complications were the same in the two groups, with a rate of 1.2%. Venous thromboembolism rates were similar, affecting 0.9% of general anesthesia patients and 0.4% of regional anesthesia patients.
Table IV.
Morbidity, mortality, and LOS results expressed as count (%) or mean ± SD
| Variable | Overall n = 5,462 |
General anesthesia n = 4,768 |
Regional anesthesia n = 694 |
|---|---|---|---|
| Morbidity (any type) | 2,012 (36.8) | 1,776 (37.3) | 236 (34.0) |
| Wound problems | |||
| Superficial wound infection | 465 (8.5) | 409 (8.6) | 56 (8.1) |
| Deep wound infection | 190 (3.5) | 167 (3.5) | 23 (3.3) |
| Organ space wound infection | 30 (0.6) | 27 (0.6) | 3 (0.4) |
| Wound dehiscence | 94 (1.7) | 82 (1.7) | 12 (1.7) |
| Any superficial or deep wound infection | 645 (11.8) | 566 (11.9) | 79 (11.4) |
| Pulmonary | |||
| Pneumonia | 100 (1.8) | 88 (1.9) | 12 (1.7) |
| Unplanned reintubation | 125 (2.3) | 111 (2.3) | 14 (2.0) |
| Failure to wean from ventilator | 95 (1.7) | 88 (1.9) | 7 (1.0) |
| Any pulmonary morbidity | 224 (4.1) | 201 (4.2) | 23 (3.3) |
| Venous thromboembolic | |||
| Deep venous thrombosis | 41 (0.8) | 38 (0.8) | 3 (0.4) |
| Pulmonary embolism | 7 (0.1) | 7 (0.2) | 0 (0) |
| Any venous thromboembolic morbidity | 48 (0.9) | 45 (0.9) | 3 (0.4) |
| Genitourinary | |||
| Acute renal insufficiency | 38 (0.7) | 34 (0.7) | 4 (0.6) |
| Acute renal failure | 32 (0.6) | 27 (0.6) | 5 (0.7) |
| Urinary tract infection | 146 (2.7) | 130 (2.7) | 16 (2.3) |
| Any renal insufficiency or renal failure | 63 (1.2) | 55 (1.2) | 8 (1.2) |
| Cardiovascular | |||
| Stroke | 38 (0.7) | 34 (0.7) | 4 (0.6) |
| Cardiac arrest | 67 (1.2) | 59 (1.2) | 8 (1.2) |
| Myocardial infarction | 48 (0.9) | 45 (0.9) | 3 (0.4) |
| Any cardiovascular morbidity | 146 (2.7) | 131 (2.8) | 15 (2.2) |
| Operative | |||
| Postoperative hemorrhage | 66 (1.2) | 59 (1.2) | 7 (1.0) |
| Graft failure | 396 (7.3) | 346 (7.3) | 50 (7.2) |
| Return to operating room | 1,214 (22.2) | 1,078 (22.6) | 136 (19.6) |
| Any operative morbidity | 1,306 (23.9) | 1,157 (24.3) | 149 (21.5) |
| Septic | |||
| Sepsis | 223 (4.1) | 200 (4.2) | 23 (3.3) |
| Septic shock | 114 (2.1) | 100 (2.1) | 14 (2.0) |
| Any septic morbidity | 331 (6.1) | 294 (6.2) | 37 (5.3) |
| Mortality | 157 (2.9) | 137 (2.9) | 20 (2.9) |
| Length of postoperative surgical stay (days) | 7.5 ± 8.1 | 7.4 ± 7.7 | 7.6 ± 10.6 |
| Length of postoperative surgical stay (days)a | 6.0 (4.0, 8.0) | 6.0 (4.0, 8.0) | 5.0 (4.0, 8.0) |
LOS, length of stay.
LOS expressed as median (Q1, Q3).
No significant differences were observed according to anesthetic type in the occurrence of individual morbidity classes, overall morbidity, mortality, or LOS. These observations were consistent across unadjusted and multivariate analyses (Table V).
Table V.
Multivariate analysis of select morbidity classes, mortality, and LOS between the general and regional anesthetic groups
| Outcomes | General versus regional
|
||
|---|---|---|---|
| OR | 95% CI | P value | |
| Any morbidity | |||
| Unadjusted | 1.15 | 0.97, 1.36 | 0.0982 |
| Multivariate modela | 1.12 | 0.93, 1.34 | 0.2382 |
| Wound morbidity | |||
| Unadjusted | 1.05 | 0.82, 1.35 | 0.7100 |
| Multivariate modelb | 0.96 | 0.73, 1.26 | 0.7812 |
| Pulmonary morbidity | |||
| Unadjusted | 1.28 | 0.83, 1.99 | 0.2647 |
| Multivariate modelc | 1.41 | 0.88, 2.26 | 0.1551 |
| Cardiovascular morbidity | |||
| Unadjusted | 1.28 | 0.75, 2.20 | 0.3726 |
| Multivariate modeld | 1.38 | 0.78, 2.43 | 0.2695 |
| Operative morbidity | |||
| Unadjusted | 1.17 | 0.97, 1.42 | 0.1069 |
| Multivariate modele | 1.18 | 0.95, 1.45 | 0.1306 |
| Mortality | |||
| Unadjusted | 1.00 | 0.62, 1.61 | 0.9900 |
| Multivariate modelf | 1.21 | 0.72, 2.02 | 0.4649 |
| Ratio days | 95% CI | P value | |
|
|
|||
| LOS | |||
| Unadjusted | 1.01 | 0.97, 1.06 | 0.6244 |
| Multivariate modela | 1.01 | 0.97, 1.06 | 0.6203 |
Model covariates for any morbidity and LOS: age, gender, race, ASA class, work RVU, preoperative sepsis, functional status, revascularization/amputation, WBC, history of COPD, platelets, rest pain/gangrene, HCT, BMI, steroid use, dyspnea, wound class, current smoker, weight loss >10%, history of CHF, preoperative renal failure, eGFR, cerebrovascular accident, transfer status, operative time.
Model covariates for wound: age, gender race, work RVU, revascularization/amputation, BMI, current smoker, history of PTCI, transfer status, history of COPD, platelets, WBC, return to operating room, graft failure, diabetes, operative time.
Model covariates for pulmonary: age, gender, race, work RVU, functional status, ASA class, preoperative sepsis, history of COPD, wound class, WBC, preoperative cerebrovascular accident, dyspnea, red blood cell count >4, HCT, diabetes, return to operating room, operative time.
Model covariates for cardiovascular: age, gender, race, ASA class, work RVU, functional status, WBC, HCT, platelets, revascularization/amputation, dyspnea, history of angina, history of PTCI, history of CABG, history of TIA, HTN, history of CHF, current smoker, operative time.
Model covariates for operative: age, gender, race, infrapopliteal bypass, graft, work RVU, history of diabetes mellitus, bleeding disorder, platelets, sepsis, revascularization/amputation, history of PTCI, wound class, BMI, current smoker, functional status, ASA class, operative time.
Model covariates for mortality: age, gender, race, functional status, ASA class, work RVU, WBC, dyspnea, platelets, BMI, history of CHF, dialysis, steroid use, preoperative sepsis, history of COPD, graft failure, return to operating room, operative time.
DISCUSSION
This report represents a large observational analysis of patients with CLI undergoing lower-extremity bypass intended to investigate the effects of anesthesia technique on outcome. We found that regional anesthesia was used more frequently in patients with pulmonary disease. However, we found little difference in outcomes between patients undergoing bypass procedures using general anesthesia or regional anesthesia (Table IV). Specifically, there was no difference in rates of cardiac events, graft failure, postoperative pneumonia, or return to the operative room that were seen in previous studies. In contrast to previous reports, anesthetic type was not associated with postoperative complications or mortality in either unadjusted or multivariate analyses (Table V).
There has been considerable controversy regarding the optimal anesthetic method for performing lower-extremity bypass for the past 20 years. Proponents of regional anesthesia cite several potential advantages. The sympathetic blockade associated with subarachnoid and epidural anesthesia decreases circulating catecholamine levels and improves lower-extremity blood flow, with the potential to decrease the incidence of deep vein thrombosis and prevent lower-extremity arterial bypass thrombosis.11–13 In addition, improved intraoperative and postoperative hemodynamic stability may be achieved with regional anesthesia,14 potentially leading to decreased cardiac complications.5 Finally, the lack of airway instrumentation may lead to improved postoperative pulmonary function.
There have been a number of studies that have attempted to examine the clinical impact of these potential biologic advantages, which have yielded conflicting results. The Perioperative Ischemia Randomized Anesthesia Trial demonstrated an increased graft patency rate with regional anesthesia, which was supported in a subsequent subgroup analysis performed by Perler et al.15 The hypothesis that general anesthesia had a negative outcome on lower-extremity revascularization was further supported with an analysis of the Veteran’s Administration (VA) NSQIP data. The group found an increased risk of graft failure and higher rates of myocardial infarction and pneumonia in patients undergoing general anesthesia compared with spinal anesthesia.5
However, other studies of the effects of anesthetic type and outcome have not supported these results.3,4 A prospective report by Pierce et al. found no differences among anesthetic groups for 30-day graft patency, death, amputation, or LOS. That study has been criticized for having a high participant exclusion rate, which could have confounded the results.3 Another study by Bode et al. demonstrated no difference in cardiac morbidity based on anesthetic type.16 The current study also suggests that there is no difference in complication rates associated with type of anesthesia for lower-extremity bypass.
Our study design and use of the NSQIP database is similar to the previous VA study performed by Singh et al.,5 but we came to opposite conclusions. One potential explanation for the different conclusion is that gender is a confounding variable. Our study included approximately 40% women, whereas the VA study contained 99% men, although gender had no effect on outcome in our analysis. An alternative explanation is that the difference in group selection may account for the contradictory findings. In the previous study, patients undergoing regional anesthesia were subdivided into epidural and spinal groups, with spinal anesthesia used as the reference group. One criticism of the previous article was that the findings of decreased morbidity did not extend to the patients who had epidural anesthesia, despite having a similar causal pathway. Therefore, we combined patients with epidural anesthesia and those with spinal anesthesia under the heading of regional anesthesia. This group did not show a significant difference in morbidity when compared with patients undergoing general anesthesia. The results from our study failed to demonstrate a significant association between anesthesia type and graft failure. We believe that this lack of association reflects a real trend, rather than lack of statistical power, because of the large number of patients involved. In addition, it appears that more complex cases were performed preferentially under general anesthesia, given the increased work-related RVUs. The assumption of increased complexity is also supported by the observation of increased resident involvement, transfusion need, and operative time observed in those procedures performed using general anesthesia in this study. The lack of difference in outcome among patients undergoing regional anesthesia compared with general anesthesia may also reflect advances in anesthetic technique and improved medication safety profile.
It is important to note that this retrospective study has several inherent limitations. First, we attempted to account for case complexity by controlling for work RVUs and operative time. However, it is possible that this may provide incomplete adjustment, and confounding by case complexity may still persist. Second, this analysis was not randomized and is unable to protect from surgeon and/or anesthesiologist biases in selection of anesthesia. Long-term graft patency and survival were also not evaluated secondary to the absence of such data in the NSQIP database. Participation in the NSQIP is voluntary and is not a randomly collected sampling of institutions, so an institutional bias may persist as well. The ACS NSQIP database is a quality improvement database, which was not designed for critical assessment of specific issues related to vascular surgical procedure performance or anesthesia, such as those addressed herein. As is often the case, only a well-developed multicenter randomized trial or carefully constructed registry can appropriately address the questions at hand. Third, our study noted a high complication rate, which approached 37% in both treatment arms. This is in contrast to a previous study performed by LaMuraglia et al., who found a 19% rate of complications when they analyzed NSQIP data from 2005 to 2006.8 The difference in complication rate may be accounted for by our more inclusive definition of morbidity and our selection of patients with CLI, who have a higher reoperation rate than patients with claudication.17
CONCLUSION
In summary, choice of anesthetic type did not appear to affect outcomes after lower-extremity bypass for CLI. The results suggest that there is no evidence to support the avoidance of general anesthesia over regional anesthetic modalities for this procedure. Anesthetic choice should be governed by local expertise and practice patterns.
References
- 1.Nowygrod R, Egorova N, Greco G, et al. Trends, complications, and mortality in peripheral vascular surgery. J Vasc Surg. 2006;43:205–16. doi: 10.1016/j.jvs.2005.11.002. [DOI] [PubMed] [Google Scholar]
- 2.Christopherson R, Beattie C, Frank SM, et al. Perioperative morbidity in patients randomized to epidural or general anesthesia for lower extremity vascular surgery. Perioperative Ischemia Randomized Anesthesia Trial Study Group. Anesthesiology. 1993;79:422–34. doi: 10.1097/00000542-199309000-00004. [DOI] [PubMed] [Google Scholar]
- 3.Pierce ET, Pomposelli FB, Jr, Stanley GD, et al. Anesthesia type does not influence early graft patency or limb salvage rates of lower extremity arterial bypass. J Vasc Surg. 1997;25:226–32. doi: 10.1016/s0741-5214(97)70345-8. [DOI] [PubMed] [Google Scholar]
- 4.Schunn CD, Hertzer NR, O’Hara PJ, et al. Epidural versus general anesthesia: does anesthetic management influence early infrainguinal graft thrombosis? Ann Vasc Surg. 1998;12:65–9. doi: 10.1007/s100169900117. [DOI] [PubMed] [Google Scholar]
- 5.Singh N, Sidawy AN, Dezee K, et al. The effects of the type of anesthesia on outcomes of lower extremity infrainguinal bypass. J Vasc Surg. 2006;44:964–8. doi: 10.1016/j.jvs.2006.06.035. [DOI] [PubMed] [Google Scholar]
- 6.Crawford RS, Cambria RP, Abularrage CJ, et al. Preoperative functional status predicts perioperative outcomes after infrainguinal bypass surgery. J Vasc Surg. 2010;51:351–8. doi: 10.1016/j.jvs.2009.08.065. [DOI] [PubMed] [Google Scholar]
- 7.Hua HT, Cambria RP, Chuang SK, et al. Early outcomes of endovascular versus open abdominal aortic aneurysm repair in the National Surgical Quality Improvement Program-Private Sector (NSQIP-PS) J Vasc Surg. 2005;41:382–9. doi: 10.1016/j.jvs.2004.12.048. [DOI] [PubMed] [Google Scholar]
- 8.LaMuraglia GM, Conrad MF, Chung T, et al. Significant perioperative morbidity accompanies contemporary infrainguinal bypass surgery: an NSQIP report. J Vasc Surg. 2009;50:299–304. doi: 10.1016/j.jvs.2009.01.043. [DOI] [PubMed] [Google Scholar]
- 9.Rutherford RB, Baker JD, Ernst C, et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg. 1997;26:517–38. doi: 10.1016/s0741-5214(97)70045-4. [DOI] [PubMed] [Google Scholar]
- 10.Levey AS, Coresh J, Greene T, et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–54. doi: 10.7326/0003-4819-145-4-200608150-00004. [DOI] [PubMed] [Google Scholar]
- 11.Parker SD, Breslow MJ, Frank SM, et al. Catecholamine and cortisol responses to lower extremity revascularization: correlation with outcome variables. Crit Care Med. 1995;23:1954–61. doi: 10.1097/00003246-199512000-00003. [DOI] [PubMed] [Google Scholar]
- 12.Li Y, Zhu S, Yan M. Combined general/epidural anesthesia versus general surgery for upper abdominal surgery. Anesth Analg. 2008;106:1562–5. doi: 10.1213/ane.0b013e31816d1976. [DOI] [PubMed] [Google Scholar]
- 13.Breslow MJ, Parker SD, Frank SM, et al. Determinants of catecholamine and cortisol responses to lower extremity revascularization. The PIRAT Study Group. Anesthesiology. 1993;79:1202–9. doi: 10.1097/00000542-199312000-00010. [DOI] [PubMed] [Google Scholar]
- 14.Gold MS, DeCrosta D, Rizzuto C, et al. The effect of lumbar epidural and general anesthesia on plasma catecholamines and hemodynamics during abdominal aortic aneurysm repair. Anesth Analg. 1994;78:225–30. doi: 10.1213/00000539-199402000-00006. [DOI] [PubMed] [Google Scholar]
- 15.Perler BA, Christopherson R, Rosenfeld BA, et al. The influence of anesthetic method on infrainguinal bypass graft patency: a closer look. Am Surg. 1995;61:784–9. [PubMed] [Google Scholar]
- 16.Bode RHJ, Lewis KP, Zarich SW, et al. Cardiac outcome after peripheral vascular surgery: comparison of general and regional anesthesia. Anesthesiology. 1996;84:3–13. doi: 10.1097/00000542-199601000-00002. [DOI] [PubMed] [Google Scholar]
- 17.DeRubertis BG, Faries PL, McKinsey JF, et al. Shifting paradigms in the treatment of lower extremity vascular disease: a report of 1000 percutaneous interventions. Ann Surg. 2007;246:415–24. doi: 10.1097/SLA.0b013e31814699a2. [DOI] [PMC free article] [PubMed] [Google Scholar]