Factors Affecting Perioperative Mortality and Wound-Related Complications Following Major Lower Extremity Amputations

Article Outline

• Abstract
• INTRODUCTION
• PATIENTS AND METHODS
• RESULTS
  • Factors Affecting Perioperative Mortality
  • Factors Affecting Wound Complications (90 Days)
• DISCUSSION
  • Perioperative Mortality
  • Wound Complications (90 Days)
• REFERENCES
• Copyright

Major lower extremity amputations continue to be associated with significant morbidity and mortality, yet few recent large series have evaluated factors associated with perioperative mortality and wound complications. The purpose of this study was to examine factors affecting perioperative mortality and wound-related complications following major lower extremity amputation. A retrospective review was conducted of all adult patients who underwent non-traumatic major lower extremity amputations over a 5-year period at a single tertiary-care center in southern West Virginia. Demographic and clinical data, perioperative data, and outcomes were collected and analyzed to identify any relationship with perioperative mortality, as well as wound complications and early revisions (within 90 days) to a more proximal level. Variables were examined using chi-squared, two-tailed t-tests, and logistic regression. Three hundred eighty patients (61% male) underwent 412 major lower extremity amputations during 1999–2003. The initial level of amputation included 230 below-knee (BKA), 149 above-knee (AKA), and one hip disarticulation. Perioperative mortality was 15.5% (n = 59). From a regression model, age, albumin level, AKA, and lack of a previous coronary artery bypass graft (CABG) were independently related to mortality. Patients who did not have a previous CABG were nearly three times more likely to die than those who did (p = 0.038). Overall early wound complications were noted in 13.4% (n = 51). Four factors were independently related to experiencing a 90-day wound complication: BKA, community (rather than care facility) living, type of anesthesia, and preoperative hematocrit >30%. Major lower extremity amputation in patients with peripheral vascular disease continues to be associated with considerable perioperative morbidity and mortality. Even though the surgical procedure itself may not be challenging from a technical standpoint, underlying medical conditions put this group at high risk for perioperative death. Wound-healing problems are frequently encountered and must be minimized to facilitate early mobilization and hospital discharge.

Back to Article Outline

INTRODUCTION 

Despite employing aggressive endovascular and surgical bypass procedures for limb-threatening ischemia, over 100,000 amputations per year are required for end-stage peripheral vascular disease in the United States.¹ Unfortunately, perioperative survival has not particularly improved despite advances in perioperative care and developments in anesthesia management. In a recent series of 959 major lower extremity amputations at an academic tertiary-care center, Aulivola et al.² reported perioperative mortality of 5.7% for below-knee amputation (BKA) and 16.5% for above-knee amputation (AKA). In a 13-year review, Sandnes et al.³ showed some improvement in long-term survival after amputation but no improvement in 30-day mortality. Similarly, in a 5-year review of functional outcomes after amputation, Nehler et al.⁴ reported an overall 30-day mortality of 10%.

In addition to high perioperative mortality, wound-related complications are frequently seen following amputations, with some reports citing up to 30% developing wound complications.⁵ Factors such as glycemic control, appropriate perioperative antibiotic use, maintaining perioperative body temperature, and elevated hemoglobin levels have been implicated in affecting wound healing in major surgical procedures.

As there have been no significant improvements in perioperative mortality following lower extremity amputations, we sought to identify factors that may be associated with perioperative mortality and major wound complications following amputation of the lower extremity secondary to peripheral arterial disease.

Back to Article Outline

PATIENTS AND METHODS 

After securing institutional review board approval, a retrospective review was conducted of 380 patients who underwent 412 nontraumatic major lower extremity amputations from January 1999 through December 2003 at Charleston Area Medical Center, a tertiary-care, teaching center in Charleston, West Virginia. Data collected from hospital charts included demographics, preexisting medical conditions, indications for surgery, laboratory results, perioperative data, and outcomes.

Patient demographics included age, gender, comorbid conditions, body surface area, level of independence (community living, nursing home), mobility status (ambulatory outdoor, ambulatory indoor, nonambulatory), preoperative evaluations [ankle-brachial index (ABI, vascular bypass procedures], and type of presenting symptoms (primary infection, chronic ischemia, or acute ischemia). Acute ischemia was defined as stage III acute limb ischemia, meaning irreversible ischemia resulting in major tissue loss or inevitable permanent nerve damage, with profound sensory loss, muscle paralysis, and inaudible arterial and venous Doppler signals.⁶

Perioperative data included laboratory values (serum albumin, hematocrit and hemoglobin levels, glucose, ejection fraction), anesthesia type (general or spinal/epidural), amputation type and side, peak operative glucose, lowest operative body temperature, and any complications. The attending surgeon determined level of amputation, with preoperative and intraoperative physical exam findings dictating the level of amputation. Transcutaneous oximetry (TcO₂) was not performed in any patient to determine level of amputation.

Main, outcome measures included perioperative mortality, deep surgical site incision complications as defined by the Centers for Disease Control (i.e., the infection appears to be related to the operation and to infection of the deep soft tissues and at least one of the following is present: purulent drainage, wound dehiscence, abscess, fever, or other indication necessitating further intervention), or amputation revision to a more proximal level (within 90 days). Amputations that were “revised” less than 5 days from the original surgery date were considered single, staged amputations rather than an amputation and subsequent failure. Given that some patients may have experienced multiple amputations on the same limb (i.e., BKA progressing to AKA), analyses were completed based on the level of initial amputation, which indicates that the BKA patient pool includes patients who survive to eventually undergo an AKA. All revisions made to AKAs were considered local revisions as none progressed to hip disarticulation.

Data were analyzed using SAS software (version 8.02; SAS Institute, Gary, NC) and broken down into demographics, perioperative data, and outcomes. Data were examined using descriptive statistics, chi-squared test, Fisher's exact test, and two-tailed t-tests. A subset of variables that were found to be significantly related to perioperative mortality in univariate analysis were examined using a stepwise logistic regression model to determine their independent effects upon mortality. A multivariate analysis was also completed to determine predictors of wound complication, p ≤ 0.05 was considered significant for all analyses.

Back to Article Outline

RESULTS 

From January 1, 1999, through December 31, 2003, 380 patients (61% men, median age 67 ± 13 years) underwent a major lower extremity amputation at our institution. Demographic and preoperative clinical data are presented in Table I. Indications for amputation included acute ischemia in 78 patients (20.5%), chronic ischemia in 173 (45.5%), and primary infection in 129 (34%). The mean ABI of patients with acute ischemia was 0.23 ± 0.08, whereas for chronic ischemia it was 0.60 ± 0.3. It should be noted that ABIs wre available only for 142 patients. As a great percentage of the patients were diabetic with noncompressible vessels, more often than not they did not have ABI values listed in their records. This is a well-established failing of the ABI.

Table I. Demographic and preoperative variables

COPD, chronic obstructive pulmonary disease; MI, myocardial infarction.

Sixty-one percent (n = 230) of patients underwent BKA initially, where as 39% underwent AKA (n = 149). One patient underwent hip disarticulation primarily. An equivalent number of procedures were performed on left and right limbs, while 11 patients (2.9%) underwent simultaneous bilateral procedures (five below-knee, six above-knee). Overall, 76 patients (20%) progressed to bilateral amputations, including 23 patients who had one major amputation prior to the inception of the study period. The mean ABIs were 0.6 ± 0.3 and 0.33 ± 0.21 in BKA and AKA patients, respectively. Procedures were performed using general anesthesia in 208 patients (54.7%) as opposed to either epidural or spinal anesthesia. Additional clinical and laboratory values are presented in Table II.

Table II. Clinical and laboratory values

Factors Affecting Perioperative Mortality 

Mean age was significantly related to perioperative mortality (72 ± 10 years for deaths vs. 66 ± 14 years for survivors; p = 0.0002, two-tailed t-test). Body surface area was not significantly related to mortality (p = 0.7), regardless of gender. Likewise, ABI was not significantly related to mortality (p = 0.1).

Table III illustrates the impact of additional patient medical and functional status factors, amputation indication and type, and clinical and laboratory factors on perioperative mortality from univariate analysis. It should be noted that functional status did have an impact on amputation type; nonambulatory patients more frequently had AKA (56%) versus BKA. Similarly, 42% of care facility residents underwent AKA. Preoperative nonambulatory patients were also much more likely to be care facility residents (p < 0.0001).

Table III. Risk of perioperative mortality (n, %, or mean ± standard deviation)

COPD, chronic obstructive pulmonary disease; MI, myocardial infarction. Significance testing by chi-squared or Fisher's exact test.

The perioperative mortality rate in the whole series was 15.5%. Consistent with other series, patients undergoing AKA were at a significantly higher risk of perioperative mortality (16% AKA, 5% BKA; p < 0.001) but a significantly lower risk of developing 90-day wound complications (p = 0.006). However, undergoing a 90-day revision of the same limb did not put one at a higher risk of perioperative mortality in this population (p = 0.22). Simultaneous bilateral amputation, regardless of level, was significantly related to perioperative mortality (p = 0.003). However, progression of disease leading to an amputation on the previously intact side did not make one more likely to suffer perioperative death (p = 0.32).

One hundred eighteen patients had more than one surgery (primary amputation with local revision or secondary amputation at higher level) during the study period, with a mean of 1.36 ± 0.59 (range 1–3). Ten of these 118 died perioperatively, a significantly lower perioperative mortality rate than the population as a whole (p = 0.01). This would seemingly indicate a protective effect of multiple surgeries (p = 0.01). However, this is most likely an artifact of amputation level in these patients as those who underwent multiple surgeries and died were much more likely to have undergone AKA rather than BKA (p = 0.03).

Factors Affecting Wound Complications (90 Days) 

The overall (90-day) wound complication rate in the series was 13.4%, with 62.7% of the complications leading to a more proximal amputation as opposed to a local revision. There were no significant differences between those undergoing local revision and those undergoing revision to a more proximal level. Age, body surface area (regardless of gender), ABI, and progression of disease leading to a contralateral amputation were not significantly related to wound complications. The one exception to this was that patients initially undergoing BKA were more likely to undergo revision to a more proximal level, whereas those undergoing AKA were more likely to undergo local revision (p = 0.0006). Also at greater risk of wound complication were patients who had undergone a previous infrainguinal bypass (p = 0.003). The vast majority of bypasses (approximately 80%) were done with prosthetic grafts. It is suspected that the increased rate of wound complications seen in this subset of patients may be related to a greater risk of infections associated with partially amputated prosthetic grafts, as originally reported by Rubin et al.⁷ The impacts of additional patient medical and functional status factors, amputation indication and type, and clinical and laboratory factors on the development of 90-day wound complications from univariate analysis are reported in Table IV.

Table IV. Risk of 90-day wound complications (n, %, or mean ± standard deviation)

COPD, chronic obstructive pulmonary disease: MI. myocardial infarction. Significance testing by chi-squared. Fisher's exact, or two-tailed r-test.

Variables that were found to be significantly related to perioperative mortality or wound complications in univariate analysis were also examined with a multiple logistic regression model, the results of which are presented in Table V.

Table V. Probability of perioperative mortality or 90-day wound complications

Multiple logistic regression (stepwise).

Back to Article Outline

DISCUSSION 

Patients undergoing lower extremity amputation for peripheral vascular disease suffer from numerous comorbidities, which place them at a high risk for even minor surgical procedures. The severity of diffuse atherosclerosis seen in these patients is reflected by perioperative mortality rates higher than those seen in elective aortic surgery. In an attempt to better understand factors associated with increased perioperative mortality and surgical site-related complications, a retrospective review was conducted of 380 patients undergoing lower extremity amputations, to identify demographic and clinical predictors associated with increased early mortality. These predictors may help to identify patients at highest risk for complications and perhaps allow preoperative and postoperative modifications of patient care.

Perioperative Mortality 

Perioperative mortality following lower extremity amputation is significant, with most series, including the current one, reporting mortality rates higher than those seen with elective aortic surgery or emergent coronary artery bypass grafting (CABG). Identifying those at high risk for early death is an important step in improving patient management. Multivariate analysis found several factors that were significantly associated with increased early mortality. Three of the four - increasing age, a more proximal amputation level, and decreasing serum albumin - are intuitive and previously demonstrated in various populations.⁸, ⁹, ¹⁰, ¹¹, ¹², ¹³, ¹⁴, ¹⁵

As expected, patients of increased age often have more comorbid conditions, more progressive disease, and generally a lower physiological capacity to heal and overcome complications secondary to the natural limitations imposed by aging. Similarly, those patients requiring a more proximal initial amputation level have more diffuse atherosclerosis, which makes their increased perioperative mortality predictable. Along these same lines, a low serum albumin level is an excellent clinical marker for chronic medical illness and may infer poor physical reserve, also making patients more susceptible to higher perioperative mortality.

Patients who had undergone previous CABG comprised the only subset that had an independent survival advantage within 30 days of surgery. A small number of patients had a history of recent percutaneous transluminal coronary angioplasty (PTCA, within 2 years of the amputation), which did not bestow a survival advantage. However, when combined with CABG, a history of any previous revascularization (CABG or PTCA) did confer improved survival in univariate analysis. This raises the question of the potential benefits from preoperative coronary evaluation in those patients requiring amputation, except for those in whom acute limb ischemia precludes such consideration.

Many patients with peripheral arterial disease have concomitant coronary artery disease (CAD), and most early deaths are attributed to cardiac complications. Recent studies have shown that a CABG within 5 years of peripheral vascular surgery may provide modest protection against adverse perioperative cardiac events resulting from the subsequent vascular surgery.¹⁶ Results from the Coronary Artery Surgery Study (CASS) likewise demonstrated a protective effect of CABG for higher-risk surgery and vascular surgery.¹⁷ However, recent results from the Coronary-Artery Revascularization Prophylaxis (CARP) study seem to suggest otherwise.¹⁸ The CARP study randomized 510 male patients to either coronary artery revascularization before surgery for abdominal aortic aneurysm/peripheral arterial occlusive disease or no revascularization before surgery. Fifty-nine percent of the patients underwent percutaneous coronary intervention, whereas 41% underwent CABG. The perioperative mortality rate was 3% for both the medically and surgically treated groups. The perioperative myocardial infarction rate was 12% in the surgical group vs. 14% in the medical group. Also, there was no difference in mortality at 2 years between the groups.

However, several factors must be considered in interpreting these conflicting findings. Although limited by the retrospective study design, it can be noted from the demographic data that this patient population, as a whole, may be medically worse off than the average patient included in the CARP study (71% diabetic as opposed to 38%, 29% nonambulatory, 33% with renal insufficiency of varying severity, 32% with previous CABG). Additionally, the mean left ventricular ejection fraction in the CARP patients was >50% in both groups, whereas in our patient pool the mean was only 42%. Notably, the CASS investigators concluded that the greatest benefits of CABG were seen in the highest-risk patients with multivessel disease and were inversely proportional to ejection fraction. It may be that the patients in the current cohort had a high incidence of multivessel disease that was undetected and may have benefited from CABG as seen in the CASS.

It is also intriguing that survival improvement was seen in the current study in those undergoing CABG but not necessarily PTCA. The CARP study did not offer postoperative outcomes analysis by type of revascularization. Given that a substantially larger number of patients underwent PTCA as opposed to CABG, it would be interesting to know if the type of revascularization conferred any survival advantage. It may be that, compared to PTCA, CABG may offer substantially superior perfusion to those in need of revascularization.

An additional important consideration is the exceptional medical management that patients received in the nonsurgical arm of the CARP study. The prevalence of β-blockers, aspirin, statins, and angiotensin-converting enzyme inhibitors was decidedly superior to that available to the CASS patients and most likely the current cohort. Unfortunately, we did not complete a careful examination of the medical treatment of our patients during data collection. Historically, given the setting of our institution, it is not utilized by the majority of this cohort for primary and preventive care – perhaps unlike the CARP patients, who receive the vast majority of their health services through the Veterans Administration system.

Wound Complications (90 Days) 

Efforts have been made to limit the frequency of failed amputations, while preserving limb length. The increased energy expenditure with more proximal amputations as well as the dramatic drop in ambulation rates/prosthetic use associated with higher amputations have been well documented in the literature. Numerous articles have been written on various noninvasive tests and techniques used to predict the most distal level of healing tissue in lower extremity amputations. Yet, these techniques have not been widely used, and clinical judgment continues to be paramount in the effort to minimize amputation revisions while maximizing ambulatory capacity. Hence, we reviewed clinical factors that may affect major wound revisions and further proximal amputations.

Four factors were found to be independently related to experiencing a 90-day wound complication: BKA, community (rather than care facility) living, type of anesthesia, and preoperative hematocrit >30%. Interestingly, care facility residents were found to have significantly fewer wound complications in multivariate analysis. However, as discussed with mortality, this appears to be a direct result of these patients having significantly more AKAs. Numerous studies have shown that proximal amputations have fewer wound complications and revisions.², ¹⁹, ²⁰ We found similar results in that 17.4% of patients initially undergoing BKA had a wound complication as opposed to only 7.4% with AKA (p = 0.005).

Bailey et al.²¹ previously reported preoperative hemoglobin as a predictor of failed amputations in 59 diabetic patients. In that study, all amputations performed on patients with hemoglobin levels <12.0 g/dL healed successfully, whereas all amputations in those with levels >13 g/dL did not heal. Also, Eneroth et al.²² found that preoperative hemoglobin >12 g/dL had a higher risk of wound failure. In our current series, this was examined using the closely related preoperative hematocrit level, which did appear to have an independent effect on future major wound complications, as other series have shown. Those with a hematocrit >30 mg/dL had a significantly increased risk of revision (p = 0.03). This has been thought to be secondary to increased blood viscosity seen with increasing hematocrit levels.

In some studies, anesthesia type has been shown to have effects on collagen synthesis, vasoconstriction, and immunosuppression, all of which influence wound healing.²³ General anesthesia and use of postoperative oxygen have been linked with preserving collagen deposition by maintaining tissue oxygen partial pressure; however, most general anesthetics also cause transient immunosuppression.²⁴, ²⁵ Regional anesthesia has been associated with decreased numbers of wound complications, felt to be secondary to less vasoconstriction related to perioperative and postoperative pain.²⁶ We found that general anesthesia may have fewer associated wound complications than an epidural or a spinal anesthetic (p = 0.0007) regardless of amputation level. This result deserves further study.

In summary, the cohort of vascular patients requiring major lower extremity amputations is often saddled with end-stage comorbidities. Even though the surgical procedure itself may not be challenging from a technical standpoint, these underlying medical conditions put this group at high risk for perioperative death. Wound-healing problems are frequently encountered and must be minimized to facilitate early mobilization and hospital discharge.

Back to Article Outline

REFERENCES 

1. Feinglass J , Brown JL , LoSasso A , et al.   Rates of lower extremity amputation and arterial reconstruction in the United States, 1979 to 1996 . Am J Public Health . 1999;89:1222–1227
   • View In Article
   • MEDLINE
   • CrossRef
2. Aulivola B , Hile CN , Hamdan AD , et al.   Major lower extremity amputation: outcome of a modern series . Arch Surg . 2004;139:395–399
   • View In Article
   • MEDLINE
   • CrossRef
3. Sandnes DK , Sobel M , Flum DR . Survival after lower-extremity amputation . J Am Coll Surg . 2004;199:394–402
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (254 KB)
   • CrossRef
4. Nehler MR , Coll JR , Hiatt WR , et al.   Functional outcome in a contemporary series of major lower extremity amputations . J Vasc Surg . 2000;38:7–14
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (101 KB)
   • CrossRef
5. Cruz CP , Eidt JF , Capps C , Kirtley L , Moursi MM . Major lower extremity amputations at a Veterans Affairs hospital . Am J Surg . 2003;186:449–454
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (152 KB)
   • CrossRef
6. Rutherford RB , Baker JD , Ernst C , et al.   Recommended standards for reports dealing with lower extremity ischemia: revised version (SVS/ISCVS Clinical Categories of Acute Limb Ischemia) . J Vasc Surg . 1997;26:517–538
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (1996 KB)
   • CrossRef
7. Rubin JR , Marmen C , Rhodes RS . Management of failed prosthetic grafts at the time of major lower extremity amputation . J Vasc Surg . 1988;7:673–676
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (378 KB)
   • CrossRef
8. Collins TC , Johnson M , Daley J , Henderson WG , Khuri SF , Gordon HS . Preoperative risk factors for 30-day mortality after elective surgery for vascular disease in Department of Veterans Affairs hospitals: is race important? . J Vasc Surg . 2001;34:634–640
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (62 KB)
   • CrossRef
9. Felix WR , Sigel B , Gunther L . The significance for morbidity and mortality of Doppler-absent pedal pulses . J Vasc Surg . 1987;5:849–855
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (690 KB)
   • CrossRef
10. Yeager RA , Moneta GL , Taylor LM , Hamre DW , McConnell DB , Porter JM . Surgical management of severe acute lower extremity ischemia . J Vasc Surg . 1992;15:385–393
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (885 KB)
    • CrossRef
11. Campbell WB , Marriott S , Eve R , Mapson E , Sexton S , Thompson JF . Amputation for acute ischemia is associated with increased comorbidity and higher amputation level . Cardiovasc Surg . 2003;11:121–123
    • View In Article
    • MEDLINE
    • CrossRef
12. Harpole DH , DeCamp MM , Daley J , et al.   Prognostic models of thirty-day mortality and morbidity after major pulmonary resection . J Thorac Cardiovasc Surg . 1999;117:969–979
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (63 KB)
    • CrossRef
13. Rady MY , Ryan T , Starr NJ . Perioperative determinants of morbidity and mortality in elderly patients undergoing cardiac surgery . Crit Care Med . 1998;26:225–235
    • View In Article
    • MEDLINE
    • CrossRef
14. Longo WE , Virgo KS , Johnson FE , et al.   Risk factors for morbidity and mortality after colectomy for colon cancer . Dis Colon Rectum . 2000;43:83–91
    • View In Article
    • MEDLINE
    • CrossRef
15. Braga M , Gianotti L , Nespoli L , et al.   Nutritional approach in malnourished surgical patients: a prospective randomized study . Arch Surg . 2002;137:174–180
    • View In Article
    • MEDLINE
    • CrossRef
16. Back MR , Stordahl N , Cuthbertson D , Johnson BL , Bandyk DF . Limitations in the cardiac risk reduction provided by coronary revascularization prior to elective vascular surgery . J Vasc Surg . 2002;36:526–533
    • View In Article
    • Abstract
    • Full-Text PDF (168 KB)
    • CrossRef
17. Eagle KA , Rihal CS , Mickel MC , Holmes DR , Foster ED , Gersh BJ . Cardiac risk of noncardiac surgery: influence of coronary artery disease and type or surgery in 3368 operations. CASS Investigators and the University of Michigan Heart Care Program. Coronary Artery Surgery Study . Circulation . 1997;96:1882–1887
    • View In Article
    • MEDLINE
18. McFalls EO , Ward HB , Moritz TE , et al.   Coronary-artery revascularization before elective major vascular surgery . N Engl J Med . 2004;351:2795–2804
    • View In Article
    • CrossRef
19. Keagy BA , Schwartz JA , Kotb M , Burnham SJ , Johnson G . Lower extremity amputations: the control series . J Vasc Surg . 1986;4:321–326
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (647 KB)
    • CrossRef
20. Miller N , Dardik H , Wolodiger F , et al.   Transmetatarsal amputation: the role of adjunctive revasularization . J Vasc Surg . 1991;13:705–711
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (575 KB)
    • CrossRef
21. Bailey MJ , Johnston CL , Yates CJ , Somerville PG , Dormandy JA . Preoperative haemoglobin as predictor of outcome of diabetic amputations . Lancet . 1979;2:168–170
    • View In Article
    • MEDLINE
22. Eneroth M , Persson BM . Risk factors for failed healing in amputation for vascular disease. A prospective, consecutive study of 177 cases . Acta Orthop Scand . 1993;64:369–372
    • View In Article
    • MEDLINE
23. Buggy D . Can anesthetic management influence surgical wound healing? . Lancet . 2000;356:355–357
    • View In Article
    • Full Text
    • Full-Text PDF (57 KB)
    • CrossRef
24. Allen DB , Maguire JJ , Mahdavian M , et al.   Wound hypoxia and acidosis limit neutrophil bacterial killing mechanisms . Arch Surg . 1997;132:991–996
    • View In Article
    • MEDLINE
25. McBride WT , Armstrong MA , McBride SJ . Immunomodulation: an important concept in modern anaesthesia . Anaesthesia . 1996;51:465–473
    • View In Article
    • MEDLINE
    • CrossRef
26. Ballantyne JC , Carr DB , deFerranti S , et al.   The comparative effects of postoperative analgesic therapy on pulmonary outcome: cumulative analysis of randomized controlled trials . Anesth Analg . 1998;86:598–612
    • View In Article
    • MEDLINE
    • CrossRef