Professional Documents
Culture Documents
To examine the interaction of epidural anesthesia, eral arterial graft, coronary artery or deep vein throm-
coagulation status, and outcome after lower extrem- boses). The rates of cardiovascular, infectious, and
ity revascularization, 80 patients with atherosclerotic overall postoperative complications, as well as dura-
vascular disease were prospectively randomized to tion of intensive care unit stay, were significantly
receive general anesthesia combined with postopera- reduced in the GEN-EPI group. Stepwise logistic
tive epidural analgesia (GEN-EPI) or general anesthe- regression demonstrated that the only significant
sia with on-demand narcotic analgesia (GEN). Demo- predictors of postoperative cardiovascular complica-
graphics did not differ between groups except that tions were preoperative congestive heart failure and
the GEN-EPI group had a higher incidence of diabe- general anesthesia without epidural analgesia. We
tes mellitus and of previous myocardial infarction.
Coagulation status was monitored using throm- conclude that in patients with atherosclerotic vascular
boelastography. An additional 40 randomly selected disease undergoing arterial reconstructive surgery (a)
patients without atherosclerotic vascular disease un- thromboelastographic evidence of increased platelet-
dergoing noncardiovascular procedures served as fibrinogen interaction is associated with early postop-
controls for coagulation status. Vascular surgical pa- erative thrombotic events, and (b) epidural anesthe-
tients were hypercoagulable compared with control sia and analgesia is associated with beneficial effects
patients before operation and on the first postopera- on coagulation status and postoperative outcome
tive day. Postoperatively, this hypercoagulability was compared with intermittent on-demand opioid anal-
attenuated in the GEN-EPI group and was associated gesia.
with a lower incidence of thrombotic events (periph- (Anesth Analg 1991;73:69&704)
These disease-related changes in coagulation may be (L-3-T-10) epidural catheters inserted using a stan-
exacerbated by surgical and postoperative stress (10). dard loss of resistance technique before induction of
As most cardiovascular morbidity occurs in the general anesthesia. Epidural catheters were evalu-
hours and days after completion of vascular surgery, ated for accidental intravascular or intrathecal place-
postoperative stress-induced hypercoagulability ment and for evidence of dermatomal spread (tl
could play a causal role. Epidural anesthesia and spinal segment) of analgesia using a 3-mL test dose of
analgesia continuing into the postoperative period lidocaine (1.5%)with epinephrine (1:200,000). Intra-
can attenuate the postoperative stress response in operatively, the anesthesia for the GEN-EPI patients
specific patient populations (4,ll-13). Epidural anes- was maintained with inhaled nitrous oxide, low con-
thesia and analgesia is also associated with direct and centrations of isoflurane (10.5% inspired), 5 5 pg/kg
indirect effects on the hemostatic system and may of intravenous fentanyl, and incremental injections of
result in diminished postoperative hypercoagulability 1.5% lidocaine administered through the epidural
(14-16). If postoperative anesthetic management, catheter to achieve surgical anesthesia. After initial
such as the use of continuous epidural analgesia, dosing with local anesthetic, the epidural catheter
positively influences outcome in high-risk patients in was redosed at the discretion of the anesthesiologist
the hours and days after major peripheral vascular using approximately one-half of the original dose of
surgery, it may do so by altering the coagulation local anesthetic if hemodynamic changes and the
system. requirement for general anesthetic drugs suggested
With these considerations in mind, this study was that the level of analgesia provided by the local
designed to reevaluate the association of epidural- anesthetic had decreased. The total volume of local
general anesthesia and postoperative epidural anal- anesthetic used intraoperatively (12-20 mL) was de-
gesia with improved cardiovascular and overall out- termined by analgesic and anesthetic requirements of
come after major vascular surgery. We also sought to the individual patient. Nondepolarizing muscle relax-
study the previously untested hypothesis of an asso-
ants were used in both groups to facilitate control of
ciation between coagulation state, postoperative com- ventilation, primarily during surgery of the abdomi-
plications, and choice of anesthetic in high-risk vas-
nal aorta.
cular patients. An additional goal of this study was to
In the GEN-EPI group, a continuous epidural
determine the relative importance of anesthetic man-
infusion of fentanyl (0.001%) and bupivacaine (0.1%)
agement compared with preoperative risk factors for
adverse cardiovascular outcome in patients undergo- at 5-8 mL/h was initiated at least 30 min before
ing major peripheral vascular surgery. anticipated completion of the surgical procedure.
When necessary, parenteral narcotics were used to
supplement pain relief, and the duration of postop-
Methods erative epidural analgesia was a clinical decision not
dictated by study protocol. Patients receiving epidu-
Perioperative Management ral analgesia were evaluated at least daily by a mem-
With institutional review board approval, 80 consent- ber of the Department of Anesthesiology acute post-
ing adults undergoing major vascular surgery of the operative pain service, and care was given according
abdominal aorta and lower extremities were ran- to established pain service protocols.
domly assigned to receive general anesthesia com- The trachea was intubated, and ventilation was
bined with epidural anesthesia and analgesia (GEN- controlled, for at least the duration of the operation,
EPI), or general anesthesia with postoperative in both anesthetic groups. In both groups, choice of
parenteral and/or oral narcotic administration (GEN). premedication and specific doses of anesthetic and
Patients with preoperative coagulopathies or liver nonanesthetic agents were at the discretion of the
disease, receiving anticoagulant or antiplatelet medi- anesthesiologist and depended on the patients con-
cations, or with any contraindication to the use of an dition and the requirements for providing amnesia,
epidural catheter, or any of the standard anesthetic/ muscle relaxation, and hemodynamic stability.
analgesic agents employed, were excluded from en- Chronically administered cardiac medications were
try into this study. given on the morning of operation. All patients had
Intraoperatively, patients in the GEN group re- inonitoring of electrocardiographic leads I1 and V,
ceived an anesthetic consisting of inhaled nitrous and of intraarterial catheters for arterial blood pres-
oxide, oxygen, isoflurane (51.5% inspired), and sure measurements and for analysis of blood gas
515 pglkg of intravenous fentanyl. Postoperatively, tensions and hematocrit. All patients undergoing
GEN patients received parenteral and/or oral narcotic procedures involving the abdominal aorta had central
analgesics as requested for pain relief. Patients in the venous catheters inserted in the operating room.
GEN-EPI group had high lumbar or low thoracic During the perioperative period, hemodynamic aber-
698 CARDIOVASCULAR ANESTHESIA TUMAN ET AL. ANESTH ANALG
EPIDURAL ANESTHESIA AND ANALGESIA FOR VASCULAR SURGERY 1991;73:696-704
rations exceeding 20% of baseline, serious dysrhyth- chronic hypertensive therapy were noted. Left ven-
mias, and ischemia were treated. tricular dysfunction was defined by an ejection frac-
Intraoperatively, all patients undergoing vascular tion of less than 0.4. Renal dysfunction (defined as a
clamping received 50-75 mg of heparin, reversed preoperative creatinine level greater than 1.4 mg/dL),
with protamine. No blood products except packed diabetes mellitus, age, preoperative antianginal ther-
red blood cells were administered. Postoperatively, apy (padrenergic or calcium channel blocking drugs
no patients received anticoagulants except for one and topical or oral nitrates), duration of operation,
patient in the GEN group who received a heparin and estimated blood loss were also noted. Intraoper-
infusion in the immediate postoperative period be- ative hypertension was considered present if the
cause of an acute vascular graft thrombosis. systolic pressure exceeded 120% of the mean pre-
Coagulation monitoring was performed using a operative systolic blood pressure for more than 5 min;
commercially available thromboelastograph (BICLOT hypotension, if the systolic pressure was less than
ELVI 816 Thromboelastograph, Logos Scientific, 80% of the mean preoperative systolic pressure for
Henderson, Nev.) (17,18). Forty patients without more than 5 min; and tachycardia, if the heart rate
clinical evidence of atherosclerotic vascular disease exceeded 120% of the mean preoperative rate or 100
undergoing noncardiovascular procedures involving beats/min for more than 5 min.
the abdomen or lower extremities were randomly Postoperative observations of clinical outcome
selected to serve as controls for coagulation status. were made using prospectively defined definitions of
One-half of these control patients received general morbidity (Table 1). Electrocardiographic ischemia
anesthesia, and one-half received general-epidural was defined as 20.1 mV positive or negative horizon-
anesthesia and analgesia as described. Thromboelas- tal ST segment shift, which extended at least 80 ms
tography was performed on whole blood after induc- beyond the J point of the QRS complex in any lead.
tion of anesthesia and on the first postoperative day Hard copies of lead I1 or CS, were obtained every 4 h
for the control patients and for the 80 vascular surgi- in the ICU and more frequently if changes suggestive
cal patients. of ischemia (ST segment or T wave abnormalities) or
Postoperatively, all patients were transferred to an of arrhythmias appeared on continuous oscillo-
intensive care unit (ICU) and received postoperative graphic displays. In addition, standard 12-lead elec-
intensive care following standardized protocols. Post- trocardiograms were recorded twice daily in the ICU.
operative hemodynamic variables, arterial blood gas All hard-copy electrocardiographic recordings were
tensions, electrolytes, hemoglobin, and urine output calibrated using a 1-mV standard.
were monitored closely under the direction of a staff Postoperative vascular graft patency was deter-
surgeon and an anesthesiologist who were not one of mined by clinical examination and by Doppler mea-
the authors and who were unaware of the goals of the surements. Early vascular graft failure was defined as
study. Discharge of patients from the ICU was deter- a return of preoperative clinical status, lack of signif-
mined by the surgical staff and was dependent on the icant improvement in Doppler measurements, or
assessment of overall patient course and cardiopul- need of reoperation for thrombectomy or unplanned
monary stability. Patients were transferred to a reg- amputation of any portion of the affected limb. All
ular hospital bed only when they did not require patients in whom graft failure developed had intra-
vasoactive infusions, invasive monitoring, fluid ad- operative arteriograms performed at the completion
ministration above maintenance levels, or aggressive of the initial procedure. These completion arterio-
pulmonary therapy for at least 24 h. grams were independently reviewed by two cardio-
vascular surgeons not involved in the primary oper-
Data Collection ation; none of the patients with graft failure had any
evidence of technical error such as intimal flap or
The physicians and nurses involved in the care of the anastomotic problems, and all had adequate outflow
patients were unaware of the outcome variables that on completion of the operation. Deep venous throm-
were being monitored. Data were collected by a bosis of the lower extremities was diagnosed by
full-time data manager. Preoperative patient demo- clinical examination and was documented with
graphics that might be associated with greater mor- venography .
tality and morbidity after vascular surgery (1-3) were
selected for examination. Evidence of congestive Data Analysis
heart failure (defined as classic chest radiographic
changes in conjunction with either rales on lung Patient demographics, perioperative risk factors, and
auscultation or an S 3 gallop on cardiac auscultation), the incidence of postoperative morbidity were com-
history of angina, prior myocardial infarction, pre- pared between the GEN and GEN-EPI group. Patient
operative arrhythmias requiring treatment, and age, duration of operation, estimated blood loss,
ANESTH ANALG CARDIOVASCULAR ANESTHESIA TUMAN ET AL. 699
1991;73:696-704 EPIDURAL ANESTHESIA AND ANALGESIA FOR VASCULAR SURGERY
Table 5. Thromboelastographic Coagulation Data for Patients With Vascular Graft, Coronary Artery, or Deep
Vein Thromboses
~~ ~
Baseline Postoperative
Thrombosis No thrombosis Thrombosis No thrombosis
Variables (n = 12) (n = 68) ( n = 12) (ti = 68)
GEN group was also associated with a larger inci- graft thrombosis (12 patients experiencing 13 events)
dence of reoperation on the affected limb (7/40 vs were compared with that for patients without these
1/40, P = 0.025). The incidence of postoperative complications (Table 5). There were no baseline dif-
electrocardiographic evidence of ischemia did not ferences in coagulation state detectable by throm-
differ between the GEN and GEN-EPI groups (15/40 boelastography between patients with or without
vs 13/40, respectively; P = 0.639). In addition to the postoperative thrombotic complications. However,
larger overall incidence of postoperative infections, accelerated coagulation was observed postoperatively
there was a larger incidence of pulmonary infections in those patients experiencing thrombotic events
in the GEN group (4/40 vs 0/40). Five of the 40 compared with those without thrombotic events ( n =
patients in the GEN group required tracheal intuba- 68). Of those patients in whom postoperative throm-
tion and mechanical ventilation for longer than 24 h bosis of a coronary artery, deep vein, or vascular graft
postoperatively, and one of these patients required developed, 11 received general anesthesia alone and
tracheal reintubation and mechanical ventilation. one received GEN-EPI anesthesia (P = 0.002). Multi-
None of the patients in the GEN-EPI group required variate analysis using stepwise logistic regression
tracheal reintubation, and only one of 40 patients in considered all of the perioperative factors listed in
the GEN-EPI group required mechanical ventilation Table 2 as well as the type of anesthesia. The pres-
for more than 24 h. ence of preoperative congestive heart failure (P coef-
Postoperatively, three patients in the GEN group ficient = 3.02, P = 0.009) and the use of general
experienced neurologic complications: one patient anesthesia without epidural analgesia ( p coefficient =
suffered a hemispheric cerebral infarct, another had 1.85, P = 0.022) were significant predictors of post-
severe exacerbation of multiinfarct dementia requir- operative cardiovascular complications with odds ra-
ing prolonged rehabilitation, and one had mental tios of 20.51 and 6.36, respectively (Table 6). None of
obtundation and symptoms consistent with a tran- the other perioperative variables had predictive
sient ischemic event that resolved without sequelae. value.
None of the patients receiving GEN-EPI anesthesia
experienced postoperative neurologic complications. Discussion
Significant gastrointestinal complications developed
in three patients after operation: one patient required The choice of anesthesia for patients undergoing
reoperation for removal of a portion of ischemic small major peripheral vascular surgery remains controver-
bowel and gangrenous cholecystitis; ischemic colitis sial. Yeager et al. (4) included patients undergoing
with lower gastrointestinal bleeding developed in major peripheral vascular surgery as a subset of
one, requiring a diverting colostomy; and one expe- high-risk surgical patients exhibiting a significantly
rienced a severe form of Olgilvie's syndrome postop- larger cardiovascular complication rate after general
eratively. There was no difference in the occurrence anesthesia compared with combined regional-general
of these events between anesthesia groups, nor were anesthesia. Another randomized trial (20) and a ret-
there differences in any outcome variable between rospective review (5) were not able to demonstrate
type of surgical procedure. differences in mortality or cardiac morbidity in pa-
To determine if the occurrence of thrombotic tients receiving central neuraxial anesthesia com-
events in the postoperative period was associated pared with general anesthesia for peripheral vascular
with hypercoagulability as measured by thromboelas- surgery. These conflicting results may be attributed
tography, coagulation data for patients with clinical to differences in population characteristics, intraop-
evidence of coronary artery, deep vein, or vascular erative management, and, perhaps most importantly,
702 CARDIOVASCULAR ANESTHESIA TUMAN ET AL. ANESTH ANALG
EPIDURAL ANESTHESIA AND ANALGESIA FOR VASCULAR SURGERY 1991;73:h9&704
Table 6. Multivariate Predictors of Cardiovascular ated protein breakdown, increased hepatic synthesis
Complications of selected proteins (acute-phase reactants including
~ ~~~ ~
complications (manifested as a lower incidence of high-risk patients. Although pain scores were inten-
prolonged tracheal intubation and respiratory infec- tionally not obtained in the GEN group to avoid
tions) were observed in the patients who received altering the intrinsic nature of intermittent, on-
epidural anesthesia and analgesia. Previous findings demand pain control and other aspects of these
of shorter duration of ventilatory support and ICU patients care, it is very likely that the analgesia was
stay in patients receiving postoperative epidural an- less than that observed in the epidural group. The
algesia have been questioned because of the inclusion current study design does not allow confirmation or
of high-dose opioid anesthesia in the nonepidural rebuttal of the hypothesis that improved outcome in
group (4).A strength of the current study is the the setting described may have been related primarily
confirmation of these findings in the absence of this to different degrees of postoperative analgesia.
potentially confounding factor. Despite the use of Rather, it confirms earlier findings that epidural an-
general anesthesia techniques with differing effects esthesia and analgesia, used in the manner de-
on the neuroendocrine response to surgical stress, scribed, is associated with better outcomes compared
these data and those of Yeager et al. (4) both demon- with on-demand, intermittent opioid analgesia. Al-
strate that epidural anesthesia and analgesia is asso- though plausible, it seems unlikely that visitation by
ciated with a beneficial effect on postoperative mor- acute pain service personnel of those patients not
bidity. These findings suggest that any potential receiving epidural analgesia after major vascular sur-
benefits of epidural anesthesia and analgesia may be gery could result in similar improvements in out-
related more to postoperative rather than intraoper- come, as analgesia would still remain intermittent.
ative influences. Several mechanisms may explain the Further studies will be required to clarify whether
latter effects of epidural anesthesia and analgesia: other forms of analgesic therapy, such as patient-
better analgesia, decreased duration of endotracheal controlled analgesia with or without continuous opi-
intubation and mechanical ventilation which dimin- oid infusions, will also be associated with similar
ish pulmonary defense mechanisms against infec- outcome differences. Defining the relative contribu-
tion, longer duration of postoperative ICU stay which tion of effective analgesia per se, neuraxial sym-
is associated with a higher risk of nosocomial infec- patholysis, improvements in lower extremity blood
tions, and better maintenance of postoperative im- flow, and reduction in hypercoagulability to clinical
munocompetence (31). Our study design does not outcomes will also require further studies to deter-
allow us to confirm or refute these hypotheses but mine the precise mechanism(s) by which epidural
does reconfirm the beneficial effects of a combination anesthesia and analgesia may be related to these
of general with epidural anesthesia and analgesia findings. Because our study was designed as an
compared with general anesthesia with routine post- observational critique of the clinical outcome of two
operative pain management on several aspects of methods of perioperative anesthetic management,
outcome (4). interpretation of the findings is limited in that specific
The major goal of this study was to compare the mechanisms were not sought to explain the inter-
use of combined general-epidural anesthesia and group differences.
continuous epidural analgesia with a specific general Although previous data have suggested an effect
anesthetic technique followed by the on-demand, of epidural anesthesia and analgesia on postoperative
intermittent administration of opioids for postopera- mortality in high-risk patients (4),we are not able to
tive analgesia. No attempt was made to provide equal confirm or refute these findings as no in-hospital
degrees of postoperative analgesia in the study mortality was observed in this study. Differences in
groups, but rather, these modalities were compared surgical populations, sample sizes, and patient man-
as to how they would commonly be used in clinical agement may be responsible for differences in the
practice. As a consequence, to avoid confounding incidence of mortality in different studies. Our data
influences on the administration of and response to do, however, confirm that preoperative congestive
intermittent, on-demand analgesia as it is usually heart failure is a significant predictor of adverse
administered, these patients were not visited by a cardiovascular outcome in high-risk patients under-
member of the acute pain service. In contrast, an going noncardiac surgery ( 3 ) .
essential standard of postoperative epidural analgesia Despite the levels of significance of our major
is the involvement of an acute analgesia service that findings, an important limitation of this investigation
frequently evaluates patients to assess the efficacy of is the small number of patients assigned to each
analgesia, to ensure the use of supplemental analge- anesthetic group. Although univariate and multivari-
sic medications, and to evaluate and treat side effects ate statistical analysis failed to detect any significant
such as pruritus and nausea. The excellent pain bias against those patients assigned to the GEN
scores in the epidural analgesia group attest to the group, there may have been unquantifiable differ-
efficacy of the epidural analgesia service in these ences between the randomized anesthesia groups. In
704 CARDIOVASCULAR ANESTHESIA TUMAN ET AL. ANESTH ANALG
EPIDURAL ANESTHESIA AND ANALGESIA FOR VASCULAR SURGERY 1991;73.69&704
addition, a very select patient population was exam- extradural morphine on the adrenocortical and hyperglycae-
ined, so that the results of this study may not be mic response to surgery. Br J Anaesth 1982;43:23-7.
13. Kehlet H, Brandt MR, Prange-Hansen A, Alberti KFMM.
generalizable to patient subsets undergoing different Effects of epidural analgesia on metabolic profiles during and
surgical procedures. after surge&. Br J Surg-1979;66:5434.
In summary, we have demonstrated that patients 14. Modig J, Borg T. Bagge L, Saldeen T. Role of extradural and of
with atherosclerotic occlusive vascular disease are general anesthesia in fibrinolysis and coagulation after total
hip replacement. Br 1 Anaesth 1983;55:625-8.
hypercoagulable to patients*and that 15. Borg T, Modig J. Potential antithrombotic effects of local
the use of epidural anesthesia and analgesia attenu- anaesthetics due to their inhibition of platelet aggregation.
ates this hypercoagulability after lower extremity Acta Anaesthesiol Scand 1985;29:73942.
revascularization. The use of general anesthesia 16. Nielsen TH, Nielsen HK, Husted SE, Hansen SL, Olsen KH,
Fjeldborg N. Stress response and platelet function in minor
bined with epidural analgesia using dilute solutions surgery during epidural buyivacaine and general anesthesia:
of local anesthetic and narcotic is associated with effect of epidural morphine addition. Eur J Anaesthesiol 1989;
improved outcome compared with general anesthesia 6:409-17.
and routine, on-demand narcotic analgesia for major 17. Spiess BD, Tuman KJ, McCarthy RJ, DeLaria GA, Schillo R,
Ivankovich AD. Thromboelastography as an indicator of post-
peripheral vascular surgery. cardiopulmonary bypass coagulopathy. ] Clin Monit 1987;3:25
30.
18. Tuman KJ, Spiess BD, McCarthy RJ, Ivankovich AD. Effects of
The authors thank Elizabeth R. Corey, RN, and Sharon Grande, RN,
progressive blood loss on coagulation as measured by throm-
for their assistance in data collection. belastography. Anesth Analg 1987;66:85&63.
19. Haviland MG. Yate's correction for continuity and the analysis
of 2 x 2 contingency tables. Stat Med 1990;9:363-83.
20. Cook IT,Davies MJ, Cronin KD, Moran P. A prospective
References randomized trial comparing spinal anesthesia using hyperbaric
cinchocaine with general anesthesia for lower limb vascular
1. Hertzer NR. Fatal myocardial infarction following abdominal surgery. Anaesth Intensive Care 1986;14:37340.
aortic aneurysm resection: 343 patients followed 6-11 years 21. Brandt MR, Olguard K, Kehlet H. Epidural analgesia inhibits
postoperatively. Ann Surg 1980;192:667-73. renin and aldosterone response to surgery. Acta Anaesthesiol
2. Hertzer NR. Fatal myocardial infarction following lower ex- Scand 1979;23:267-72.
tremity revascularization: 273 patients followed 6-1 1 postop- 22. Engquist A, Brandt MR, Fernandes A, Kehlet H. The blocking
erative years. Ann Surg 1981;193:492-8. effects of epidural analgesia on the adrenocortisol and hyper-
3. Goldman L, Caldera DI., Nussbaum SR, et al. Multifactorial glycemic responses to surgery. Acta Anaesthesiol Scand 1977;
index of cardiac risk in noncardiac surgcal procedures. N Engl 21 :330-5.
J Med 1977;29784550. 23. Engquist A, Fog-Moller F, Christiansen C, Thode J, Anderson
4. Yeager MP, Glass DD, Neff RK, Brinck-Johnsen T. Epidural T, Nistrop-Madsen S. Influence of epidural analgesia on the
anesthesia and analgesia in high-risk surgical patients. Anes- catecholamine and cyclic AMP response to surgery. Acta
thesiology 1987;66:729-36. Anaesthesiol Scand 1980;24:17-21.
5. Manolio TA, Beattie C, Christopherson R, Pearson TA. Re- 24. Ardlie NG, Cameron HA, Garrett J . Platelet activation by
gional versus general anesthesia in high-risk surgical patients: circulating levels of hormones: a possible link in coronary heart
the need for a clinical trial. J Clin Anesth 1989;1:414-21. disease. Thromb Res 1984;36:315-22.
6. Meade TW, Mellows S, Brozovic M, et al. Haemostatic func- 25. Uza G, Crisnic I. Effects of angiotensin I1 upon platelet adhe-
tion and ischemic heart disease: principal results of the North- siveness and the thrombelastogram in patients with essential
wick Park Heart Study. Lancet 1986;ii:53>7. arterial hypertension. Pathol Eur 1975;10:327-32.
7. Trip MD, Cats VM, vancapelle FJL, Vreeken J. Platelet hyper- 26. Weissman C. The metabolic response to stress: an overview
eactivity and prognosis in survivors of myocardial infarction. and update. Anesthesiology 1990;73:30&27.
N Engl J Med 1990;322:1549-54. 27. Simpson PJ, Radford SG, Lockyer ]A. The influence of anaes-
8. Flinn WR, McDaniel MD, Yao JST, Fahey VA, Green D. thesia on the acute phase protein response to surgery. Anaes-
Antithrombin 111 deficiency as a reflection of dynamic protein thesia 1987;42:69M.
metabolism in patients undergoing vascular resconstruction. 28. Brandt MR, Fernandes A, Mordhorst R, Kehlet H. Epidural
J Vasc Surg 1984;1:88%95. analgesia improves postoperative nitrogen balance. Br Med J
9. McDaniel MD, Pearce WH, Yao JST, et al. Sequential changes 1978;l :I 106-8.
in coagulation and platelet function following femorotibial 29. Christensen T, Waaben J, Lindeburg T, Vesterberg K, Vinnars
bypass. J Vasc Surg 1984;1:261-8. E, Kehlet H. Effect of epidural analgesia on muscle amino acid
10. Naesh 0, Friis JT, Hindberg I, Winther K. Platelet function in pattern after surgery. Acta Chir Scand 1986;152:407-11.
surgcal stress. Thromb Haemostasis 1985;54:849-52. 30. Haljamae H, Frid I, Holm J, Akerstrom G. Epidural versus
11. Rutberg H, Hakanson E, Anderberg B, Jorfeldt L, Martensson general anesthesia and leg blood flow in patients with occlu-
J, Schildt 8 . Effects of the extradural administration of rnor- sive atherosclerotic disease. Eur I Vasc Surg 1988;2:395.400.
phine, or bupivacaine, on the endocrine response to upper 31. Hole A, Unsgaard G. The effect of epidural and general
abdominal surgery. Br J Anaesth 1984;56:2334. anaesthesia on lymphocyte functions during and after major
12. Christensen P, Brandt M, Rem J, Kehlet H. Influence of orthopaedic surgery. Acta Anaesthesiol Scand 1983;27:13541.