DISCLAIMER: These guidelines were prepared by the Department of Surgical Education, Orlando Regional Medical Center.

They
are intended to serve as a general statement regarding appropriate patient care practices based upon the available medical
literature and clinical expertise at the time of development. They should not be considered to be accepted protocol or policy, nor are
intended to replace clinical judgment or dictate care of individual patients.

FEVER ASSESSMENT
SUMMARY
The onset of fever in the intensive care unit patient must be approached systematically and guided by
clinical findings. Current literature emphasizes utilizing a cost-conscious approach, minimizing the use of
low yield tests that have little impact on clinical outcome and may be detrimental to the patient.
RECOMMENDATIONS
Level 1
If catheter-related sepsis is suspected, two peripheral blood cultures should be obtained with
an additional culture from each indwelling catheter.
If a lower respiratory tract infection is suspected, obtain a portable AP chest radiograph.
Gram stains of centrifuged urine should be used to select antimicrobial therapy.
Level 2
Core body temperature measurements from an intravascular or urinary catheter or
esophageal thermistor should be used when available.
If clinical evaluation does not strongly suggest a non-infectious cause, blood cultures should
be obtained within the first 24 hours of fever.
Expressed purulence from an intravascular catheter insertion site should be cultured.
Use quantitative catheter cultures to determine the source of bacteremia/fungemia.
Do not routinely culture removed intravascular catheters. Culture only those suspected of
being the source of infection.
Quantitative cultures obtained by either bronchoscopy or catheter lavage should be obtained
if pneumonia is suspected.
Pleural fluid should be cultured if an adjacent infiltrate is noted or infection is suspected.
Evaluation for C. difficile infection should begin with a C. difficile toxin EIA.
Send stool cultures for enteric pathogens or ova and parasite only if diarrhea was present
prior to ICU admission, the patient is immunocompromised or it is epidemiologically indicated.
Obtain urine for microscopic exam, Gram stain and culture in all high risk patients showing
signs or symptoms of UTI.
Surgical wounds with suspected infection should be opened to obtain samples for Gram stain
and culture. Cultures of the skin overlying a wound should not be performed.
If there is sufficient clinical suspicion, a CT scan of the sinuses should be obtained.
If CNS infection is suspected, send CSF for Gram stain and culture, glucose, protein, and cell
count with differential.
Level 3
Chest radiographs, urinalysis, or urine cultures are not indicated in the first 72 hours postoperatively unless history and clinical findings suggest a high probability of infection.
Noninfectious causes of fever should be investigated, including new medications and
administration of blood products.
If fever is accompanied by altered consciousness or focal neurologic deficits, lumbar puncture
or evaluation of CSF from an indwelling ventriculostomy should be considered.
EVIDENCE DEFINITIONS
Class I: Prospective randomized controlled trial.
Class II: Prospective clinical study or retrospective analysis of reliable data. Includes observational, cohort, prevalence, or case
control studies.
Class III: Retrospective study. Includes database or registry reviews, large series of case reports, expert opinion.
Technology assessment: A technology study which does not lend itself to classification in the above-mentioned format.
Devices are evaluated in terms of their accuracy, reliability, therapeutic potential, or cost effectiveness.
LEVEL OF RECOMMENDATION DEFINITIONS
Level 1: Convincingly justifiable based on available scientific information alone. Usually based on Class I data or strong Class II
evidence if randomized testing is inappropriate. Conversely, low quality or contradictory Class I data may be insufficient to
support a Level I recommendation.
Level 2: Reasonably justifiable based on available scientific evidence and strongly supported by expert opinion. Usually
supported by Class II data or a preponderance of Class III evidence.
Level 3: Supported by available data, but scientific evidence is lacking. Generally supported by Class III data. Useful for
educational purposes and in guiding future clinical research.
1
Approved 04/30/01
Revised 10/08/07, 01/30/2013

INTRODUCTION
Fever is defined as an elevation in core body temperature greater than 38.3C (101F) and is one of the
most frequently detected abnormal signs in the intensive care unit (ICU) patient population. This
physiological response is known to have direct antimicrobial effects, in addition to its role in augmenting
humoral and cellular defense mechanisms. Although in some instances fever is indicative of an adequate
systemic response, acute onset of fever has been associated with ICU mortality in 12% of cases [1].
Appropriate evaluation of fever, and institution of early, goal directed therapy when indicated, is
associated with a clear survival benefit for patients who are septic, experiencing endocrine emergencies,
and those with other causes of temperature dysregulation. Fever evaluations in the ICU setting should be
guided by clinical assessment instead of automatic laboratory and radiologic tests.

LITERATURE REVIEW
In 2008, the Society of Critical Care Medicine (SCCM) and the Infectious Diseases Society of America
convened a task force to update the 1998 practice parameters for the assessment of fever in ICU
patients. The 2008 guidelines presented numerous recommendations based on discussion of the
published literature and panel members expertise [2]. The Cochrane Library and the National Clearing
House databases were used to identify trials, meta-analyses, literature reviews and most recent clinical
recommendations pertaining to the evaluation of the acute onset of fever in the ICU population. Finally, a
review of literature published in critical care medicine journals within the last 4 years was conducted and
considered in this review.
Measuring Temperature
The first step in evaluating fever is to accurately assess temperature. Although the gold standard for
temperature assessment is core temperature obtained in the central circulation, its use is not always
indicated in ICU patients. Bladder catheters with thermistors, although costly, have been shown to
provide essentially identical readings to thermistors in intravascular sites, are less invasive and provide
stable measurements, regardless of urine flow rate [2]. In spite of the fact that esophageal probes have
demonstrated temperature monitoring accuracy comparable to that of central venous and bladder
catheter thermistors, they are not commonly utilized in the ICU.
Less invasive modes of monitoring core body temperature incorporate the use of infrared thermometry to
record temporal artery and tympanic membrane temperatures. Although initially the temporal artery
thermometer was considered to be as accurate and precise as invasive core temperature measurements,
observational studies have demonstrated a mean difference of -0.44C when compared to bladder
catheter temperatures. Due to the lack of consensus with other established modes of core body
temperature monitoring, the use of temporal artery thermometry is not recommended in situations where
accurate and precise body temperature monitoring is indicated (Level 2) [3]. Infrared ear thermometry has
been shown to yield inaccurate readings in the setting of inflammation of the auditory canal or tympanic
membrane as well as poor correlation with core temperature readings.
A study among 110 ICU patients comparing central venous, axillary, tympanic membrane, and urinary
bladder thermistors found that tympanic membrane measurements showed only modest correlation with
central venous (r=0.77), urinary (r=0.69), and axillary (r=0.76) temperatures. The average difference
between central venous and urinary temperature was small at -0.05, with a statistically significant
correlation (p=0.92). This paper adds to a growing body of evidence that questions tympanic membrane
measurements and suggests that urinary bladder temperature monitoring is the preferred alternative
when central venous monitoring is not indicated and Foley catheter placement is indicated (Level 2) [4].
Evaluation of the Patient
With new onset of fever, a thorough search for possible etiologies should be performed. It is important,
however, to maintain a lower temperature threshold among the immunocompromised population. It
should also be considered that infections may be concealed by euthermia or hypothermia in certain
patient populations, such as the elderly, those with open abdominal wounds, extensive burns, and those
receiving extracorporeal membrane oxygenation or continuous renal replacement therapy [2]. Regardless

Approved 04/30/01
Revised 10/08/07, 01/30/2013

of the patients status it is important to conduct both a careful physical examination and thorough review
of the patients medical record. This should include:
Review of:
Patients injuries
Previous cultures
Prior antibiotic therapies
Medications
Recent chest radiograph (if available)

Examination of:
Intravascular catheter insertion sites
Drainage tubes and their output
Wounds
Extremities for swelling

Infectious vs. Non-infectious Causes of Fever

The initial goal in fever assessment should be to determine if it is infectious or non-infectious in etiology.
Some common non-infectious causes for fever seen in the ICU are listed in Table 1. Many of these are
diagnoses of exclusion and many patients will have risk factors for both infectious and non-infectious
etiologies. Unless there is strong evidence that the fever is non-infectious in nature, diagnostic evaluation
to identify the potential source of infection should be promptly initiated. This approach, however, will vary
when assessing the febrile postoperative ICU patient. In patients with fever within 72 hour of surgery, no
additional laboratory, culture, or radiographic evaluation is necessary unless directed by clinical findings,
suspicion for aspiration or a compromise of sterile technique [2,5]. Serum procalcitonin levels and
endotoxin activity assays can be employed as adjunctive diagnostic tools for discriminating infection as
the cause for fever or sepsis presentations (Level 2) [2].
Table 1: Non-Infectious Causes of Fever
Hyperthermia Syndrome
Intra-abdominal
Environmental (heatstroke)
Acalculous cholecystitis
Drug-induced
Pancreatitis
Neuroleptic malignant
Ischemic bowel
syndrome
Cirrhosis (without primary
Malignant hyperthermia
peritonitis)
Serotonin syndrome
Hematoma
Solid organ injury
Neurological
GI bleed
Subarachnoid hemorrhage
Immunologic
Cerebral infarction
Alcohol withdrawal
Postoperative fever
Transfusion reactions
Soft tissue
Organ transplant rejection
Decubitus ulcers
Systemic lupus erythematosus
Adult Stills disease
Vascular
Rheumatoid Arthritis
Stroke
Endocrine
Myocardial infarction
Thrombophlebitis
Adrenal Crisis
Thyrotoxicosis

Hematologic/Immunologic
Hematomas
Deep vein thrombosis
Fat emboli
Pulmonary embolus
Pulmonary infarct
Acute hemorrhage
Neoplastic
Hodgkins, non-Hodgkins
lymphoma
Leukemia
Multiple myeloma
Sarcoma
Tumors of the liver, brain,
kidney, colon, gallbladder and
pancreas
Renal
IV contrast reaction

Adapted from Potential causes of abnormally elevated body temperatures in the ICU patient Critical Care Med 2009 Vol. 37, No. 7

Blood Cultures
If clinical evaluation of the patient suggests that the fever is likely infectious in etiology, then blood
cultures should be obtained (Level 2) [2]. The sensitivity of blood cultures depends on several factors, the
most important being collection prior to antibiotic initiation and volume of blood drawn. It is recommended
that 3 to 4 blood cultures be collected within 24 hours of the onset of fever. Two blood cultures should be
drawn from peripheral sites, but not via a peripheral line at the time of insertion because of the potential
for contamination with the skins flora. If a central venous catheter is present and considered to be a
potential infectious source, at least one blood culture should be drawn from it. Following this, further blood
cultures should be obtained based upon clinical judgment rather than routinely with each temperature
3

Approved 04/30/01
Revised 10/08/07, 01/30/2013

elevation (Level 2) [2]. Obtaining blood cultures more than every 24 hours is rarely useful. Additional
blood cultures should be drawn thereafter only when there is clinical suspicion of continuing or recurrent
infection or to asses antimicrobial response 4896 hours after initiation of appropriate therapy [2].
Intravascular Catheters
Intravascular catheters are a major source of nosocomial blood stream infections. The highest risk of
infection is associated with short term, non-cuffed hemodialysis catheters. In spite of this, intravascular
catheters should not be routinely cultured unless a central line-associated bloodstream infection
(CLABSI) is suspected. When new onset fever is seen in patients who lack significant risk factors for
sepsis (young, immunocompetent, etc.) and who have no other obvious source of infection, it is
reasonable to suspect CLABSI. Although absent in most cases, the presence of inflammation (with or
without purulence) at the catheter insertion site in patients with signs and symptoms of sepsis, is highly
predictive of catheter-related bacteremia. Other findings suggesting that the intravascular catheter is the
source of infection include difficulty drawing or infusing through it. Furthermore, catheters that have been
in place for less than 3 days are less likely to be infected [2].
The diagnosis of CLABSI is made when a colonized catheter is associated with a contaminant blood
stream infection with no other plausible cause. While negative blood cultures from a central line can help
to exclude CLABSI, positive blood cultures drawn from the catheter may indicate either infection or
colonization. Ideally, quantitative cultures should be used in order to identify the catheter as the source of
the infection. If the quantitative culture is drawn through an infected catheter, a ten-fold or greater
increase in concentration of organisms will be noted when compared to the blood culture drawn
simultaneously from a peripheral vein. Unfortunately, due to their expense, quantitative cultures are not
often used. These provide useful information, however, when surgically implanted catheters that cannot
be readily removed are the suspected source of infection. An alternative to quantitative cultures with
equivalent specificity but lower sensitivity is the method of assessing the differential time to positivity for
peripheral vs. catheter blood cultures. In this method, if both sets of cultures are positive for the same
organism and the set drawn through the catheter becomes positive at least 120 minutes earlier than the
peripheral culture, the diagnosis of CLABSI is made [6]. If there is evidence of a tunnel infection, embolic
phenomenon, vascular compromise, or septic shock, the catheter should be removed and a new catheter
inserted at a different site (Level 2) [2].
Pulmonary Infections
The presence of respiratory secretions and abnormal chest radiographs are customary in the surgical and
trauma ICU setting, but not always representative of respiratory infection. It has been estimated that up to
50% of trauma patients with clinical evidence of pneumonia are undergoing a SIRS response [7]. Clinical
parameters for the diagnosis of nosocomial pulmonary infection, such as the Clinical Pulmonary Infection
Score (CPIS) and National Nosocomial Infection Surveillance [8] developed by the CDC, provide a
scoring system for the diagnosis of ventilatory associated pneumonia (VAP) [9]. Although neither has
been shown to consistently predict the incidence of VAP, the CPIS appears to be a superior tool in
diagnosing VAP. There is, however, considerable inter-observer variability [10]. A meta-analysis
conducted on the diagnostic performance of the CPIS for VAP has concluded that its diagnostic accuracy
is moderate, with a sensitivity between 72 and 77% and specificity between 42 and 85% [11]. A
prospective observational study on the role of CPIS in the diagnosis of trauma-associated pneumonias
4
5
(TAP) concluded that CPIS does not reliably predict a positive bronchoalveolar lavage (BAL) at 10 or 10
CFU/ml and therefore should not be used to determine the indication for a BAL when VAP is suspected
[12].

Approved 04/30/01
Revised 10/08/07, 01/30/2013

Table 2: Clinical Criteria for Pulmonary Infection Diagnosis

Clinical Pulmonary Infection Score
Temperature
0 point: 36.5 and 38.4 C
1 point: 38.5 and 38.9 C
2 point s: 39.0 or 36.5 C
Oxygenation (PaO2/FiO2)
0 point: PaO2/FiO2 > 240, ARDS, or pulmonary contusion
2 point s: PaO2/FiO2 < 240 and no evidence of ARDS
White Blood Cell Count
0 point: 4,000 and 11,000
1 point: < 4,000 or > 11,000
2 points: < 4,000 or > 11,000 AND > 500 band forms

Total score of > 6 points suggests VAP

Chest Radiograph
0 point: No infiltrate
1 point: Diffuse or patchy infiltrates
2 points: Localized infiltrate
Tracheal Secretions
0 point: None or scant
1 point: > Non-purulent
2 points: Purulent sputum
Culture of tracheal aspirate
0 pt: Minimal or no growth
1 pt: Moderate or more growth
2 pt: Moderate or greater growth
Adapted from Critical Care 2008; 12(2): R56.

NNIS Pulmonary Infection Criteria

Radiographic Signs
Clinical Signs
Two or more serial chest radiographs with at
At least 1 of the following:
least 1 of the following:
Fever (temperature > 38 C)
New or progressive and persistent infiltrate
WBC < 4000 or > 12000 cells/L
Consolidation
Altered mental status, for adults 70 years, with no
Cavitation
other recognized cause
Plus at least 2 of the following:
Microbiological Criteria
New onset of purulent sputum, or change in
At least one of the following:
character of sputum
Positive growth in blood culture without
Increased respiratory secretions, or increased
another infectious source
suctioning requirements
Positive growth in culture or pleural field
New-onset or worsening cough, or dyspnea, or
Positive quantitative culture from BAL
tachypnea
4
(> 10 ) or protected specimen brushing (>
Wales or bronchial sounds
3
10 )
Worsening gas exchange
5% of cells with intracellular bacteria on
Increased oxygen requirements
Gram-stained BAL fluid
Histopathological evidence of pneumonia
Portable chest radiography (CXR), with its high sensitivity but poor specificity, is adequate in the initial
evaluation of fever. Preferably, these should be obtained in an erect sitting position. Among all
radiographic findings, unilateral air bronchograms have the best predictive value for the diagnosis of
pneumonia. Although routinely used, data has shown that even in controlled exposure conditions, more
than 30% of the CXR films are considered suboptimal and in many instances correlate poorly with CT
5

Approved 04/30/01
Revised 10/08/07, 01/30/2013

scan findings. Bedside lung ultrasound (US) has been shown to carry a diagnostic accuracy for ARDS
comparable to that of bedside radiography and is being increasingly used in ICU patients [13]. According
to a prospective study of 42 general ICU patients conducted by Xirouchaki et al, chest US was found to
have better diagnostic performance than CXR in the identification of consolidation, interstitial syndrome
as well as pleural effusions. This study concluded that bedside US could be used as an alternative to
chest CT scans in the ICU setting [9]. CT scans, however, are reasonable in immunocompromised
patients in order rule out opportunistic diseases [2].
Pulmonary secretions should be sent for Gram stain and culture only when a pulmonary infectious source
is strongly suspected. However, even in the presence of clinical signs and laboratory values indicative of
infection, sputum evaluations might not be high yield. A retrospective study of the relationship between
fever, leukocytosis and a positive respiratory culture concluded that no level of fever or range of
leukocytosis demonstrated a correlation with positive respiratory cultures. Furthermore, the authors did
not recommend obtaining respiratory culture during the initial 14 days of hospitalization based on fever
and/or leukocytosis alone in traumatically injured patients [7].
Current literature suggests that expectorated or suctioned specimens are adequate for initial evaluation of
pulmonary infections in the non-intubated ICU population (Level 2) [2]. Specimens that demonstrate a
preponderance of epithelial cells on initial Gram stain are not suitable and should be discarded. If a
diagnosis of pneumonia is not definite based on clinical and radiographic findings, respiratory secretions
can be obtained from expectorated sputum, nasotracheal or endotracheal aspirate. Due to concerns of
sample contamination and dilution, the use of saline infusion should be avoided unless adequate
specimens cannot be otherwise obtained [2].
Intubated patients suspected of pulmonary infection based upon radiographic findings should undergo
either bronchoscopically directed aspiration or catheter-directed lavage of pulmonary secretions.
Quantitative cultures of such aspirates should be performed to confirm the presence of a clinically
significant infection. Quantitative cultures obtained via bronchoalveolar lavage (BAL), protected specimen
brush, and tracheobronchial aspirates are equivalent in diagnosing VAP and should be performed in the
presence of a discrete lobar infiltrate [1,10]
Due to the need for expeditious identification of the source of infection in an ICU patient, awareness of
the reliability of preliminary culture results is of great importance. Two prospective and retrospective
studies conducted at Presley Regional Trauma Center assessed the utility of preliminary BAL results in
suspected VAP and showed that preliminary BAL results are highly predictive of final culture results.
According to this study, preliminary BALs had a positive predictive value of 100%, whereas the negative
predictive values of no growth to date and insignificant BALs were 99% and 95%, respectively. Although
preliminary BALs with significant growth are also highly predictive of VAP, the risk of additional significant
isolates appearing on the final BAL culture results precludes the use of these results in the selection of
antimicrobial agents [14].
Pleural fluid, identified in up to 62% of ICU patients, is transudative in 50% of cases. Therefore, its culture
is not indicated in every febrile patient. Due to low sensitivity and specificity, CXRs are considered
unreliable in the identification of pleural effusions in the ICU population. Bedside ultrasonography is an
alternative that is highly sensitive and capable of identifying even small effusions (< 5ml) [15].
Furthermore, bedside US has been shown to be superior to CXR and comparable to CT scanning, for the
diagnosis of pleural effusions in mechanically ventilated patients [16]. Indications for pleural fluid sampling
under US guidance are listed in Table 4. (Level 2) [15]. Pleural fluid obtained via thoracentesis should be
sent for Gram stain, culture and cytology. In order to determine if the effusion is exudative or transudative,
the pleural fluids pH, protein, glucose, and lactate dehydrogenase concentration should be obtained as
well (Level 2) [15]. The patients risk for exposure should be used to determine the need for fungal and
mycobacterial cultures [2].

Approved 04/30/01
Revised 10/08/07, 01/30/2013

Table 4. Indications for Pleural Fluid Sampling (Level 2) [15]

Evidence suggestive of infection (otherwise
Suspicion of malignancy
unexplained):
Persistent or enlarging effusion
o Ipsilateral pneumonic process
Clinical deterioration
o Fever
Potential for contamination of pleural space:
o Leukocytosis
o Recent thoracic surgery
o Elevated serum procalcitonin level
o Trauma
Suspicion of tuberculosis
o Fistula
Urinary Tract Infections
A catheter associated urinary tract infection (CAUTI) is defined as signs or symptoms indicative of a
urinary tract infection as well as a single catheter specimen or midstream voided urine containing at least
3
10 CFU/mL bacterial species. Although bacteriuria is the most common source of gram-negative
bacteremia in hospitalized patients, it accounts for a minority of bacteremias in the ICU setting. In
general, bacteriuria and candiduria represent colonization and are rarely the cause of fever in
catheterized patients. However, the presence of urinary tract obstruction, recent urologic manipulation or
surgery, and granulocytopenia all increase the likelihood that fever is secondary to a CAUTI [2]. The most
common signs and symptoms of CAUTI are listed below. In the catheterized patient, the presence or
absence of odorous or cloudy urine should not be used to distinguish colonization from infection or to
guide diagnostic work up (Level 1) [17].
Table 5. UTI Signs and Symptoms
General Patient
New onset or worsening fever
Costovertebral angle tenderness
Altered mental status
Flank pain
Rigors
Acute hematuria
Malaise
Pelvic discomfort
Lethargy

Spinal Cord Injury Patient

Increased spasticity
Autonomic dysreflexia
Sense of unease

When an ICU patient develops fever and clinical assessment is suggestive of a UTI, a urine sample
should be obtained for culture. A retrospective study looking at 510 surgical ICU patients concluded that
fever and/or leukocytosis are not associated with the development of UTI within the first 14 days of
admission [18]. It has been suggested that urinalysis and culture are not mandatory during the initial 72
hours postoperatively (unless urological procedure was performed) if fever is the only clinical finding
(Level 3) [2].
Urine specimens should be collected by clean catch or sampling port if the patient is catheterized and
these should be processed in the laboratory within an hour. The sample should never be obtained from
3
the collecting bag (Level 2) [2]. The presence of pyuria and bacterial counts greater than 10 CFU/mL
does not necessarily prove that the patients fever is secondary to a CAUTI. Studies have consistently
shown that in most cases it is not the cause of fever (Level 2) [2] [19]. Obtaining Gram stains of the
centrifuged sample reliably shows the presence of infecting organisms and can help determine
appropriate treatment of catheter-associated urosepsis (Level 2) [2]. The absence of pyuria in a
symptomatic patient suggests a diagnosis other than CAUTI (Level 1) [17].
Wound Evaluation
Surgical and superficial wounds should be inspected at least once daily for signs of infection as part of
the fever evaluation (Level 2) [2]. Drainage emanating from a superficial wound does not require cultures
or Gram stain because these can be treated without anti-microbial agents. However, if purulence is
expressed from within the tissue it should be opened in order to obtain Gram stain and cultures from deep
within the wound site (Level 2) [2]. Tissue biopsies are preferable to superficial cultures obtained with
swabs. Superficial cultures of skin overlying a wound should not be performed [2].

Approved 04/30/01
Revised 10/08/07, 01/30/2013

Sinusitis
Ventilator associated sinusitis (VAS) is a common cause of fever of unknown origin in critically ill patients.
Infectious sinusitis affects 27% of mechanically ventilated patients and is the cause in 25% of fever of
unknown origin in the ICU. Furthermore, this review reported that ventilator associated pneumonia was
present in 41% of critically ill adults diagnosed with VAS [20]. Sinusitis is most commonly caused by
obstruction of the ostia draining the sinuses due to a nasotracheal or nasogastric tube. Clinical findings
are not always present in the ICU population, where as little as 25% of those affected will have purulent
nasal discharge [2]. Gram-negative bacilli represent 60% of the bacterial isolates in nosocomial sinusitis,
with Pseudomonas aeruginosa being most common. Staphylococcus aureus is the most common grampositive bacteria, responsible for about one third of cases. If clinical evaluation suggests that sinusitis may
be the source of fever, a CT scan of the facial sinuses should be obtained (Level 2). Sinus fluid sampling
should be pursued via puncture and aspiration of the involved sinuses if there is no response to empiric
therapy. The aspirated fluid should be sent for Gram stain and culture (Level 1) [2].
Clostridium difficile and Enteric Pathogens
Clostridium difficile colitis is the most common cause of diarrhea-related fever and should be suspected in
any patient with fever, diarrhea, and a history of receiving antibacterial or chemotherapeutic agents within
the previous 60 days. Although most ICU patients will present with diarrhea if C. difficile is the culprit,
postoperative patients may instead present with ileus, toxic megacolon or leukocytosis. Current
recommendations suggest that evaluation should be initiated with one stool sample for C. difficile
common antigen, EIA for toxin A and B, or tissue culture assay. If the first specimen for C. difficile is
negative and testing is performed by an EIA method, send an additional sample for C. difficile EIA
evaluation. A second specimen is not necessary if the common antigen test was negative (Level 2) [2].
If severe illness is present and rapid tests for C. difficile are negative or unavailable, consider flexible
sigmoidoscopy, which is a highly specific test (Level 3). Due to the associated risk of bowel perforation
with this invasive procedure, direct visualization of pseudomembranes via endoscopy should be reserved
for cases in which laboratory results are not readily available or if there is a high likelihood of having toxin
assays with false negative results. Although the use of empiric therapy is not generally recommended if
two stool evaluations (using a reliable assay) are negative, consider empiric therapy with vancomycin
while awaiting diagnostic studies in the setting of severe illness (Level 2) [2].
Other organisms that can cause fever and diarrhea include Salmonella, Shigella, Campylobacter jejuni,
Aeromonas, Yersinia, Escherichia coli, Entamoeba histolytica, and multiple viruses. These are typically
community-acquired organisms, however, and are unlikely to be the source of fever in an ICU patient
unless the patient was immunocompromised or experiencing diarrhea prior to ICU admission. Therefore,
sending stools for bacterial cultures or ova and parasite examination should generally be avoided as part
of a fever evaluation. These tests should only be done if the host is immunocompromised or meets any of
the previously mentioned reasons for increased suspicion (Level 2) [2]. Testing for norovirus is usually
performed in outbreak settings and should involve infection control and public health authorities (Level 3)
[2].
Evaluation of the CNS
If fever is accompanied by altered consciousness or focal neurologic deficits, lumbar puncture should be
obtained, unless contraindicated (Level 3). However if new focal neurologic findings raise suspicion for
intracranial disease, an imaging study is usually required before performing a lumbar puncture (Level 2).
Cerebrospinal fluid (CSF) should be sent for Gram stain and culture. Further testing for tuberculosis,
fungal disease or neoplasm is dependent on the clinical situation (Level 2). In febrile patients with an
intracranial device, CSF should be obtained for analysis from the CSF reservoir. If CSF flow to the
subarachnoid space is obstructed, it may also be prudent to obtain CSF from the lumbar space (Level 3).
In patients with ventriculostomies who develop stupor or signs of meningitis, the catheter should be
removed and the tip cultured (Level 3) [2].

Approved 04/30/01
Revised 10/08/07, 01/30/2013

References
1. Laupland KB et al. Occurrence and outcome of fever in critically ill adults. Crit Care Med 2008, 36(5):
1531-1535.
2. O'Grady NP et al. Guidelines for evaluation of new fever in critically ill adult patients: 2008 update
from the American College of Critical Care Medicine and the Infectious Diseases Society of America.
Crit Care Med 2008, 36(4): 1330-1349.
3. Stelfox HT et al. Temporal artery versus bladder thermometry during adult medical-surgical intensive
care monitoring: An observational study. BMC Anesthesiol, 2010, 10: 13.
4. Moran JL et al. Tympanic temperature measurements: are they reliable in the critically ill? A clinical
study of measures of agreement. Crit Care Med 2007, 35(1):155-164.
5. Lesperance R et al. Early postoperative fever and the "routine" fever work-up: results of a prospective
study. J Surg Res 2011, 171(1): 245-250.
6. Mermel LA et al. Clinical practice guidelines for the diagnosis and management of intravascular
catheter-related infection: 2009 Update by the Infectious Diseases Society of America. Clin Infect Dis
2009, 49(1): 1-45.
7. Claridge JA et al. The "fever workup" and respiratory culture practice in critically ill trauma patients. J
Crit Care 2010, 25(3): 493-500.
8. Bedoyan JK et al. A complex 6p25 rearrangement in a child with multiple epiphyseal dysplasia. Am J
Med Genet A 2011, 155A(1): 154-163.
9. Xirouchaki N et al. Lung ultrasound in critically ill patients: comparison with bedside chest
radiography. Intensive Care Med 2011, 37(9): 1488-1493.
10. Rea-Neto A et al. Diagnosis of ventilator-associated pneumonia: a systematic review of the literature.
Crit Care 2008, 12(2): R56.
11. Shan J, Chen HL, Zhu JH. Diagnostic accuracy of clinical pulmonary infection score for ventilatorassociated pneumonia: a meta-analysis. Respir Care 2011, 56(8): 1087-1094.
12. Patel CB et al. Trauma-associated pneumonia in adult ventilated patients. Am J Surg 2011, 202(1):
66-70.
13. Lichtenstein D et al. Comparative diagnostic performances of auscultation, chest radiography, and
lung ultrasonography in acute respiratory distress syndrome. Anesthesiology 2004, 100(1): 9-15.
14. Swanson JM et al. Utility of preliminary bronchoalveolar lavage results in suspected ventilatorassociated pneumonia. J Trauma 2008, 65(6): 1271-1277.
15. Maslove DM et al. The diagnosis and management of pleural effusions in the ICU. J Intensive Care
Med, 2011.
16. Rocco, M., et al., Diagnostic accuracy of bedside ultrasonography in the ICU: feasibility of detecting
pulmonary effusion and lung contusion in patients on respiratory support after severe blunt thoracic
trauma. Acta Anaesthesiol Scand, 2008. 52(6): p. 776-84.
17. Hooton, T.M., et al., Diagnosis, prevention, and treatment of catheter-associated urinary tract
infection in adults: 2009 International Clinical Practice Guidelines from the Infectious Diseases
Society of America. Clin Infect Dis, 2010. 50(5): p. 625-63.
18. Golob, J.F., Jr., et al., Fever and leukocytosis in critically ill trauma patients: it's not the urine. Surg
Infect (Larchmt), 2008. 9(1): p. 49-56.
19. Shuman, E.K. and C.E. Chenoweth, Recognition and prevention of healthcare-associated urinary
tract infections in the intensive care unit. Crit Care Med, 2010. 38(8 Suppl): p. S373-9.
20. Agrafiotis, M., et al., Ventilator-associated sinusitis in adults: systematic review and meta-analysis.
Respir Med, 2012. 106(8): p. 1082-95.

Surgical Critical Care Evidence-Based Medicine Guidelines Committee

Primary Reviewer: Aura Fuentes
Editor: Michael L. Cheatham, MD, Matthew W. Lube, MD
Last revision date: January 30, 2013
Please direct any questions or concerns to: webmaster@surgicalcriticalcare.net

Approved 04/30/01
Revised 10/08/07, 01/30/2013