Noninfectious Fever Following Aortic Surgery: Incidence, Risk Factors, and Outcomes

Noninfectious Fever Following Aortic Surgery: Incidence, Risk Factors, and Outcomes

Chin Med Sci J December 2009 Vol. 24, No. 4 P. 213-219 CHINESE MEDICAL SCIENCES JOURNAL ORIGINAL ARTICLE Noninfectious Fever Following Aortic Surge...

192KB Sizes 0 Downloads 62 Views

Chin Med Sci J December 2009

Vol. 24, No. 4 P. 213-219

CHINESE MEDICAL SCIENCES JOURNAL ORIGINAL ARTICLE

Noninfectious Fever Following Aortic Surgery: Incidence, Risk Factors, and Outcomes Yun-tai Yao, Li-huan Li*, Qian Lei, Lei Chen, Wei-peng Wang, and Wei-ping Chen Department of Anesthesiology, Fuwai Cardiovascular Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100037, China

Key words: fever; noninfectious; aortic surgery Objective To determine the incidence, course, potential risk factors, and outcomes of noninfectious fever developed in patients after aortic surgery. Methods Patients who received operation for aortic aneurysm or dissection in our center from January 2006 to January 2008 were reviewed. Patients who met one of the following criteria were excluded: having a known source of infection during hospitalization; having a preoperative oral temperature greater than or equal to 38.0°C; undertaking emergency surgery; having incomplete data. Univariate analysis was performed in patients with noninfectious postoperative fever and those without, with respect to demographics, intraoperative data, etc. Risk factors for postoperative fever were considered for the multivariate logistic regression model if they had a P value less than 0.10 in the univariate analysis. Results Totally 463 patients undergoing aortic surgery were enrolled for full review. Among them, 345 (74.5%) patients had noninfectious postoperative fever, the other 118 (25.5%) patients didn’t develop postoperative fever. Univariate analysis demonstrated that several risk factors were associated with the development of noninfectious postoperative fever, including weight, surgical procedure, minimum intraoperative bladder temperature, temperature upon intensive care unit (ICU) admission, discharge, and during ICU stay, as well as blood transfusion. In a further multivariate analysis, surgical site of thoracic and thoracoabdominal aorta (odds ratio: 4.861; 95% confidence interval: 3.029-5.801; P=0.004), lower minimum intraoperative bladder temperature (odds ratio: 1.117; 95% confidence interval: 1.01-1.24; P=0.04), and higher temperature on admission to the ICU (odds ratio: 2.57; 95% confidence interval: 1.28-5.18; P=0.008) were found to be significant predictors for noninfectious postoperative fever. No difference was found between the febrile and afebrile patients with regard to postoperative hospitalization duration (P=0.558) or total medical costs (P=0.896). Conclusion Noninfectious postoperative fever following aortic surgery is very common and closely related with perioperative interventions. Received for publication September 2, 2009. *Corresponding author Tel: 86-10-88372598, Fax: 86-10-88372598, E-mail: [email protected]

P

OSTOPERATIVE fever is a common occurrence in cardiac1-5 and noncardiac6-8 surgical patients and has been shown to be related with many adverse consequences.9-12 Postoperative fever can be

214

CHINESE MEDICAL SCIENCES JOURNAL

roughly classified into infectious or noninfectious catego13

December 2009

larly 30 minutes before surgery. General anesthesia was

Previous studies have shown that the vast majority

induced with fentanyl (5-15 Njg/kg), midazolam (0.05-0.1

of patients who develop postoperative fever have no in-

mg/kg), etomidate (0.2-0.6 mg/kg), rocuronium (0.6

ries.

1-8,13

However, the etiology and pathophysiology of

mg/kg), and maintained by isoflurane inhalation (end tidal

noninfectious postoperative fever remain poorly under-

concentration of 0.5%-1.5%) and propofol infusion (2-4

stood. Inflammatory response caused by surgical stress

mg.kg-1.h-1). Prophylactic antibiotic (cetriaxone 1 g or ce-

and cardiopulmonary bypass has been regarded as one

foperazone 1 g) was administered intravenously before

major culprit leading to noninfectious postoperative fever

skin incision. Inotropic drugs (dopamine or dobutamine)

in cardiac surgical patients.4,5 Aortic surgery shares many

and/or vasoactivators (ephedrine, phenylephrine, sodium

similarities with cardiac surgery, but the phenomenon of

nitroprusside, or nitroglycerine) were administered as

noninfectious postoperative fever in patients undergoing

needed to avoid excessive hypertension or hypotension.

aortic surgery has been seldom described. Therefore, we

Both nasopharyngeal and bladder temperatures were

performed this study to determine the incidence of non-

monitored intraoperatively using Agilent temperature

infectious postoperative fever following aortic surgery,

probes, components of the Agilent Multifunctional Monitors

understand its course, and identify some potential factors

(Agilent Technologies, USA). After operation, all patients

contributing to noninfectious postoperative fever in this

were transferred to the intensive care unit (ICU), sedated

subgroup of surgical patients.

with propofol and ventilated till endotracheal extubation.

fection.

Chest X-ray was taken after admission to the ICU on the

PATIENTS AND METHODS

operation day, the first postoperative day (POD 1), the second postoperative day (POD 2), and at any time

Subject enrollment

deemed necessary by attending physicians. Complete

After being approved by the Institutional Review Board of

blood count (CBC) test was performed at least once per day

Fuwai Cardiovascular Hospital and obtaining patients’

within the first postoperative week.

written informed consents, we reviewed all patients who had open surgery performed for aortic aneurysm or dis-

Data collection

section in our center during the period from January 2006

Every patient’s demographics (age, gender, weight, height,

to January 2008. Patients undertaking emergency surgery

etc.), intraoperative data (surgical site, surgery time,

for aneurysm rupture were not enrolled. Patients with a

anesthesia time, volume of blood transfusion, etc.), and

known source of infection during hospitalization, either

postoperative treatment (volume of postoperative blood

preoperatively or postoperatively, and patients who had

transfusion, antipyretic treatment, etc.) were recorded.

preoperative oral temperature greater than or equal to

Fever in this study was defined as oral temperature equal

38.0°C were excluded. Diagnostic laboratory testing for

to or above 38.0ć according to our institutional protocols.

infection was in accordance with routine criteria for the

Oral temperatures, measured at four-hour interval every

adult surgical unit of our institution. Briefly, cultures of

day before and after operation in general ward and every

urine, sputum, blood, and wound secretion as well as chest

two hours during ICU stay, were transcribed from patient’s

X-ray examination, were obtained when attending physi-

temperature chart and then reviewed by an investigator to

cians determined that there were clinical signs or symp-

detect and eliminate erroneous artifact. The radiologist’s

toms of infection. Cultures positive for pathogenic organ-

report of every chest X-ray examination was recorded.

isms were considered positive evidence for infection.

White blood cell (WBC) count and neutrophil percentage of every CBC test were also recorded; if there was more than

Perioperative management

one CBC test performed on a single day, the mean value

Perioperative management of all patients was in accor-

was calculated for analysis. Patient’s hospitalization cost

dance with our institutional protocols. Preoperatively,

was transcribed from every case chart.

hemodynamic stability was achieved by combined administration of vasoactivators, and morphine 5-10 mg was

Statistical analysis

given intramuscularly for pain relief. Patients took oral

To compare patients who did and did not develop post-

benzodiazepines (diazepam 10 mg or estazolam 2 mg)

operative fever, respectively classified as the febrile group

before sleep on the night prior to the operation. Anticho-

and afebril group, we used Student’s t-test or analysis of

linergics (scopolamine 0.3 mg or penehyclidine 0.5 mg)

variance for continuous variables and chi-square or Fisher’s

with or without morphine 10 mg were given intramuscu-

exact test for categorical variables. The variables were

Vol. 24, No.4

CHINESE MEDICAL SCIENCES JOURNAL

215

treated as continuous or categorical according to their data distribution, the goodness of fit of the overall model, and the presence or absence of a clinically meaningful threshold value. To determine independent risk factors, we created a stepwise multivariate logistic regression model with postoperative fever as the dependent variable. Risk factors for postoperative fever were considered for the multivariate logistic regression model if they had a P value less than 0.10 in the univariate analysis. All statistical tests were two-sided. Values were presented as mean±SD or number (proportion). A value of P<0.05 was considered statistically significant. All analyses were conducted with SPSS 13.0 statistical software (Chicago, IL, USA).

RESULTS Incidence We reviewed 549 patients undergoing aortic surgery during the period from 2006 to 2008 in our center. Eighty-six patients who met the exclusion criteria were not enrolled for further analysis. Of the remaining 463 patients who constituted the study population, 345 patients had noninfectious postoperative fever. The operation day, POD 1, and POD 2 saw most occurrence of fever, with the daily percentage of 41.2%, 42.9%, and 23.7%, respectively (Fig.

Figure 2. Maximum (A) and minimum (B) temperatures in the

1). The other 118 patients didn’t develop postoperative

afebrile and febrile groups.

fever. The temperature profiles of the febrile and afebrile

*P<0.05, **P<0.001 compared with afebrile group.

patients are illustrated in Figure 2.

19.3%, aminophenazone 14.7%) and ethanol sponge bath (11.3%). No afebrile patient received antipyretic drug

Risk factors

treatment.

The ICU admission temperature was lower in the afebrile

The included 463 patients’ demographic and pe-

patients than in the febrile ones (Table 1). During the ICU

rioperative data are shown in Table 1. There was no dif-

stay, peak temperature in the afebrile and febrile patients

ference between the two groups with respect to age,

(37.6±0.3°C vs. 38.4±0.5°C, P<0.001) was reached after

gender, height, anesthesia time, surgery time, extracor-

19.2±7.6 hours and 19.6±5.4 hours after admission, re-

poreal circulation duration, and rewarming time.

spectively. Two hundred and eighty-four patients received

Univariate analysis demostrated that several risk fac-

postoperative antipyretic treatment. Icepack cooling was

tors were associated with postoperative fever. Patients

the most frequently used method (67.3%), followed by

who became febrile weighed more than those did not

antipyretic medicines (bupleurum 30.1%, indomethacin

(73.7±12.9 kg vs. 66.3±9.7 kg, P<0.001). More febrile patients received thoracic and thoracoabdominal procedures than those in the afebrile group (95.7% vs. 56.8%, P<0.001). Febrile patients had lower minimum intraoperative

bladder

temperature

than

afebrile

patients

(23.2±5.1°C vs. 26.4±5.7°C, P<0.001). Febrile patients had higher temperatures than afebrile patients upon ICU admission and discharge, and during ICU stay as well (P<0.001). Febrile patients received more intraoperative Figure 1. Accumulative and new-onset fever cases following aortic surgery within the first 7 postoperative days. PreOP: preoperative period; OP: operation day; POD: postoperative day.

and total (intraoperative plus postoperative) transfusion of packed red blood cells (PRBC), fresh frozen plasma (FFP), and platelet concentration (PC) (P<0.001) (Table 1).

216

CHINESE MEDICAL SCIENCES JOURNAL

December 2009

Table 1. Demographic and perioperative data of patients with aortic surgery§ Demographic and perioperative data Age (yr) Male, n (%) Weight (kg) Height (cm) ASA classification, n (%) ĉ,Ċ ċ,Č Hypertension, n (%) Diabetes mellitus, n (%) Smoking, n (%) Alcoholism, n (%) Anesthesia duration (min) Surgical site, n (%) Thoracic aorta Abdominal aorta Thoracoabdominal aorta Surgery duration (min) ECC duration (min) Rewarming time (min) Temperature (°C) Nasopharyngeal temperature after anesthesia induction Bladder temperature after anesthesia induction Minimum nasopharyngeal temperature Minimum bladder temperature Nasopharyngeal temperature upon ECC weaning Bladder temperature upon ECC weaning Temperature upon ICU admission Minimum temperature during ICU stay Maximum temperature during ICU stay Temperature when discharging from ICU Blood transfusion† Intraoperative PRBC (U) Intraoperative FFP (mL) Intraoperative PC (U) Total PRBC (U) Total FFP (mL) Total PC (U) Postoperative mechanical ventilation duration (h) Duration of ICU stay (h) Surgical drainage duration (d) Urinary catheterization duration (d) Central venous catheterization duration (d) Duration of postoperative antibiotics (d)

Afebrile (n=118)

Febrile (n=345)

P value

45.8±14.1 94 (79.7%) 66.3±9.7 173.7±7.0

45.8±13.7 280 (81.2%) 73.7±12.9 172.5±8.9

0.986 0.787 <0.001 0.184 0.669

51 (43.2%) 67 (56.8%) 40 (34.7%) 27 (22.9%) 55 (46.6%) 45 (38.1%) 383.2±120.9

158 (45.8%) 187 (54.2%) 110 (31.9%) 85 (24.6%) 179 (51.9%) 118 (34.2%) 345.6±141.9

57 (48.3%) 51 (43.2%) 10 ( 8.5%) 331.0±119.9 165.9±53.1 86.9±35.5

289 (83.8%) 15 ( 4.3%) 41 (11.9%) 288.8±134.9 140.9±60.1 70.5±33.9

0.174 0.067 0.064

36.3±0.6 36.8±0.7 24.3±6.4 26.4±5.7 36.8±0.6 36.3±0.9 35.9±0.9 35.7±0.8 37.6±0.3 36.8±0.5

36.4±0.7 36.9±0.6 20.2±4.5 23.2±5.1 36.8±2.3 36.4±1.2 36.3±0.7 36.1±0.6 38.4±0.5 37.2±0.6

0.166 0.166 0.410 <0.001 0.915 0.528 <0.001 <0.001 <0.001 <0.001

6.9±4.4 1 143.5±731.2 1.1±0.4 7.6±3.0 1 258.8±948.2 1.3±0.7 15.2±10.5 67.9±52.1 4.6±2.1 3.6±1.9 6.5±1.4 10.6±3.0

7.3±4.0 1 429.6±758.4 1.6±0.8 9.4±3.6 1 870.4±1008.2 2.9±1.1 15.0±10.3 51.5±38.9 4.5±3.4 3.7±1.7 6.2±2.2 11.2±5.4

<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.589 0.077 0.863 0.593 0.166 0.558

0.733 0.803 0.339 0.437 0.238 <0.001

§: Plus-minus values are means±SD. †

Intraoperative blood transfusion included those added to the priming solution of extracorporeal circulation and those transfused intra-

venously. Total blood components included those used intraoperatively and postoperatively. ASA: American Society of Anesthesiologists; ECC: extracorporeal circulation; ICU: intensive care unit; PRBC: packed red blood cell; FFP: fresh frozen plasma; PC: platelet concentration.

Based on the univariate analysis results, the following

of thoracic and thoracoabdominal aorta (odds ratio: 4.861;

characteristics were included in a further multivariate

95% confidence interval: 3.029-5.801; P=0.004), lower

analysis: body weight, surgical site, minimum intraopera-

minimum intraoperative bladder temperature (odds ratio:

tive bladder temperature, temperature upon ICU admis-

1.117; 95% confidence interval: 1.01-1.24; P=0.04), and

sion and discharge, maximum and minimum temperature

higher temperature on admission to the ICU (odds ratio:

during ICU stay, intraoperative and total volume of blood

2.57; 95% confidence interval: 1.28-5.18; P=0.008) were

components (PRBC, FFP, and PC) transfusion. Surgical site

found to be significant predictors for noninfectious post-

Vol. 24, No.4

CHINESE MEDICAL SCIENCES JOURNAL

217

study also suggested surgical site was the most significant

operative fever. No difference was found in the proportion of patients

predictor of noninfectious postoperative fever: 83.5% and

with radiological atelactasis between febrile and afebrile

80.4% patients who had surgery performed in thoracic and

groups on the operation day (39.8% vs. 39.3%, P=0.35),

thoracoabdominal aorta became febrile postoperatively,

POD 1 (32.5% vs. 34.5%, P=0.846), and POD 2 (39.4% vs.

and only 22.7% of those undertaking abdominal aortic

37.3%, P=0.558). The prevalence of radiological pleural

surgery developed postoperative fever. Interleukin (IL)-6

effusion in febrile patients on the first 3 days were all lower

is one of the major cytokines produced after tissue injury

than those in afebrile patients, but only significantly dif-

after aortic surgery.14 Increased level of plasma IL-6 has

ferent on the operation day (16.4% vs. 21.4%, P=0.006).

been shown to be one major culprit causing postoperative

When compared with preoperative values, WBC count and

fever in cardiac5 and noncardiac15 surgical patients, and

neutrophil percentage remarkably increased after opera-

also in experimental animals.16 IL-6 acts via several

tion in both febrile and afebrile patients, peaking between

mechanisms to promote inflammation and fever. It pro-

POD 2 [WBC: (15.1±3.9)×109 cell/L vs. (16.3±5.5)×109

motes neutrophil activation and prolongs its functional

cell/L; neutrophil: (88.2±8.5)% vs. (90.2±7.1)%] , and

lifespan,5 and it can also mediate the pyrogenic effect of

POD 3 [WBC: (18.3±5.6)×109 cell/L vs. (18.6±3.8)×109

both IL-1 and tumor necrosis factor (TNF)-Į.17 In patients

cell/L; neutrophil: (90.6±4.5)% vs. (89.1±7.3)%], and did

undertaking thoracic and abdominal aneurysmectomy, the

not restore to normal range [WBC: (4-10)×109 cell/L;

concentration of IL-6 increases immediately after surgery

neutrophil: 50%-70%] even after one week postopera-

and reaches peak value within 3 days of surgery,14 the

tively. Intra-group comparison of the WBC count and

period in which we observed most fever cases in our study.

neutrophil percentage before operation and on the first 7

Intraoperative hypothermia might be another possible

days after the operation didn’t show any significant dif-

cause of postoperative fever.4,18 Roe18 noted that patients

ference (all P>0.05).

who became hypothermic intraoperatively developed a moderate hyperthermia postoperatively, whereas patients

Outcomes

who maintained their intraoperative temperature did not

No patient in either of the two groups died or developed

display a postoperative fever. Anesthetic-induced ther-

major complications during hospitalization. After endo-

moregulatory inhibition can significantly decrease core

tracheal extubation, 16 patients in the febrile group and 4

temperature during anesthesia and surgery. Additionally,

in the afebrile group developing confusion or agitation were

cardiovascular surgery performed with hypothermic car-

recorded (Fisher’s exact test, P=0.793), which didn’t sig-

diopulmonary bypass involves great thermal perturbation.

nificantly influence their postoperative recovery. There was

Our study suggested that minimum intraoperative bladder

no difference between the two groups with regard to

temperature was a predictor of postoperative fever, which

postoperative ICU stay length (67.9±52.1 hours vs. 51.5±

could also be a consequence of more extensive and com-

38.9 hours, P=0.077) , hospitalization duration (12.2±5.7

plicated surgeries that required greater surgical interven-

days vs. 13.2±7.1 days, P=0.176), and total medical costs

tion. Therefore, it might be expected that those patients

(86 570.6±20 425.3 yuan vs. 89 565.1±10 623.5 yuan, P=

having larger operations would have a greater inflamma-

0.896).

tory response and more postoperative fever. Our study found that the initial ICU temperatures were higher in patients who subsequently developed fever,

DISCUSSION

which was similar with the findings previously reported in

In this present study, we found that more than 70%

coronary artery bypass surgery patients.3 Targeted re-

(345/463) patients undertaking aortic surgery developed

warming at the end of cardiopulmonary bypass is a

postoperative fever, which can’t be attributed to infection.

common practice among many institutions.3,4 Overly ag-

The exact etiology of noninfectious postoperative fever

gressive intraoperative rewarming to prevent postopera-

remains unknown. One possible reason for it may be in-

tive hypothermia, and too rapid establishment of nor-

flammatory response triggered by surgical trauma. The

mothermia might result in unintentional postoperative

postoperative cytokine levels in the early period directly

hyperthermia.3,4 For those patients with higher initial ICU

reflect the magnitude of operative trauma. More complex

temperatures, whether was it due to aggressive rewarming

and traumatic than abdominal aorta surgeries, thoracic and

or were they already developing fever upon their admission

thoracoabdominal aorta surgeries triggered more intensive

to the ICU couldn’t be elucidated in this study.

14

inflammatory response and higher level of cytokines.

Our

We also found that transfusion of blood products to be

218

CHINESE MEDICAL SCIENCES JOURNAL

December 2009

another possible pyrogenic cause. Passenger leukocytes

still possible that some infection-related febrile patients

are the cause of alloimmunization to leukocyte-specific

without clinical or laboratory evidence might be included as

antigens in transfusion recipients.19 It is also well known

noninfectious febrile patients. Finally, treatment of fever,

that erythrocyte and platelet transfusion can result in

either by physical or pharmaceutical means, might affect

febrile nonhemolytic transfusion reactions due to high

the natural history of temperature pattern after surgery.

concentration of leukocyte-derived and platelet-derived 20

cytokines found in stored blood components.

To summarize, our findings indicate that noninfectious

WBCs and

postoperative fever might be a manifestation of perioper-

neutrophils are integral components of the inflammatory

ative stress. Considering the high incidence of postopera-

response network. But leukocytosis and elevated neutro-

tive fever following aortic surgery and its potential adverse

phil percentage are not reliable predictors of infection, nor

effects, it is essential to understand exactly the underlying

are they sensitive or specific indicators for fever.14,21

mechanism to aid in the development of proper preventive

Whether atelectasis can cause fever remains controver-

or therapeutic intervention strategy. To do so, well-de-

sial.2,22 We found no association between atelectasis and

signed prospective randomized controlled clinical as well as

noninfectious fever in our study, which was in accordance

experimental studies are urgently warranted.

2

with the results of the study conducted by Engoren. We did not find any association between pleural effusion and

ACKNOWLEDGEMENT

noninfectious fever, either.

The authors are grateful to Professor Li-zhong Sun and Dr.

Higher maximum postoperative temperature has been

Bin Li of Vascular Surgery Center, Fuwai Cardiovascular

reported to be correlated with a greater degree of cognitive

Hospital for their kind help during the preparation and

12

impairment 6 weeks after cardiac surgery.

Fortunately,

conduction of this study.

no difference was found in the proportion of confusion and agitation between febrile and afebrile patients during their

REFERENCES

postoperative recovery. Admittedly, exploration of the relationship between postoperative fever and neurocogni-

1.

Livelli FD Jr, Johnson RA, McEnany MT, et al. Unexplained

tive dysfunction is not the aim of our study, and given the

in-hospital fever following cardiac surgery. Natural history,

retrospective nature of the study and lack of comprehen-

relationship to postpericardiotomy syndrome, and a

sive neurocognitive tests, it is inappropriate to conclude

prospective study of therapy with indomethacin versus

that postoperative fever dose not influence neuropsychological function in aortic surgical patients based on our findings. Several limitations of the current study deserve men-

placebo. Circulation 1978; 57:968-75. 2.

fever. Chest 1995; 107:81-4. 3.

temperature monitoring. Its accuracy might also be un-

Analg 2002; 95:1489-95. 4.

nary artery bypass surgery. J Cardiothorac Vasc Anesth

intubation. Additionally, there is no universally-accepted

2005; 19:426-9. 5.

postoperative fever. Cytokine 2007; 38:37-42.

it is a treatment trigger in our center, and that according to 6.

probe within urinary catheter. The second limitation is that,

Infect Control 1985; 6:273-7. 7.

fever which were not analyzed in this study and haven’t been studied by others (e.g. vasoactivators, anesthetics,

Fanning J, Neuhoff RA, Brewer JE, et al. Frequency and yield of postoperative fever evaluation. Infect Dis Obstet

given the multifactorial nature of fever phenomenon, there might be other covariates accounting for noninfectious

Garibaldi RA, Brodine S, Matsumiya S, et al. Evidence for the non-infectious etiology of early postoperative fever.

operatively when patient is in general ward and bladder temperature is only monitored intraoperatively by thermal

Mitchell JD, Grocott HP, Phillips-Bute B, et al. Cytokine secretion after cardiac surgery and its relationship to

oral temperature > 38°C as the diagnostic criterion is that our local protocol, oral temperature is measured peri-

Clark JA, Bar-Yosef S, Anderson A, et al. Postoperative hyperthermia following off-pump versus on-pump coro-

dermined by some factors, such as transoral endotracheal criterion for fever diagnosis yet. The reason why we chose

Thong WY, Strickler AG, Li S, et al. Hyperthermia in the forty-eight hours after cardiopulmonary bypass. Anesth

tion. One is that the oral temperature which is routinely recorded in our institution, is not the gold standard for

Engoren M. Lack of association between atelectasis and

Gynecol 1998; 6:252-5. 8.

Shaw JA, Chung R. Febrile response after knee and hip arthroplasty. Clin Orthop Relat Res 1999; 367:181-9.

9.

Reith J, Jørgensen HS, Pedersen PM, et al. Body tem-

and circadian rhythm of temperature). The third limitation

perature in acute stroke: relation to stroke severity, in-

is that, although routine and frequent surveillance tests

farct size, mortality, and outcome. Lancet 1996; 347:

were performed for suspected infection in our center, it is

422-5.

Vol. 24, No.4

CHINESE MEDICAL SCIENCES JOURNAL

10. Ginsberg MD, Busto R. Combating hyperthermia in acute stroke: a significant clinical concern. Stroke 1998; 29: 529-34. 11. Cook DJ. Cerebral hyperthermia and cardiac surgery: consequences and prevention. Semin Thorac Cardiovasc Surg 2001; 13:176-83. 12. Grocott HP, Mackensen GB, Grigore AM, et al. Postop-

219

after total knee arthroplasty: the role of cytokines. Clin Orthop Relat Res 2003; 415:221-31. 17. Zetterström M, Sundgren-Andersson AK, Östlund P, et al. Delineation of the proinflammatory cytokine cascade in fever induction. Ann N Y Acad Sci 1998; 856:48-52. 18. Roe CF. Fever and energy metabolism in surgical disease. Monogr Surg Sci 1996; 3:85-132.

erative hyperthermia is associated with cognitive dys-

19. Ezidiegwu CN, Lauenstein KJ, Rosales LG, et al. Febrile

function after coronary artery bypass graft surgery.

nonhemolytic transfusion reactions. Management by

Stroke 2002; 33:537-41.

premedication and cost implications in adult patients. Arch

13. Pile JC. Evaluating postoperative fever: a focused approach. Cleve Clin J Med 2006; 73:S62-6.

Pathol Lab Med 2004; 128:991-5. 20. Heddle NM. Pathophysiology of febrile nonhemolytic

14. Shindo S, Ogata K, Kubota K, et al. Vascular prosthetic

transfusion reactions. Curr Opin Hematol 1999; 6:420-6.

implantation is associated with prolonged inflammation

21. Pullicino EA, Carli F, Poole S, et al. The relationship be-

following aortic aneurysm surgery. J Artif Organs 2003; 6:

tween the circulating concentrations of interleukin 6 (IL-6),

173-8.

tumor necrosis factor (TNF) and the acute phase response

15. Lahat N, Zlotnick AY, Shtiller R, et al. Serum levels of IL-1, IL-6 and tumour necrosis factors in patients undergoing coronary artery bypass grafts or cholecystectomy. Clin Exp Immunol 1992; 89:255-60. 16. Andres BM, Taub DD, Gurkan I, et al. Postoperative fever

to elective surgery and accidental injury. Lymphokine Res 1990; 9:231-8. 22. Welsh RG, Michael D, Juras K, et al. Reasons for extending hospital stay beyond three days post coronary artery bypass surgery. Cardiac Chronicle 1993; 7:1-5.