Bongard Frederic , Darryl Sue. Diagnosis and Treatment Critical Care

Подождите немного. Документ загружается.



INFECTIONS IN THE CRITICALLY ILL

377

group off the β-lactam ring—found in the penicillins and

cephalosporins—imipenem has a short hydroxyethyl side

chain that protects the ring from degradation by bacteria-

produced β-lactamases. The carbapenems are active against

most aerobic and anaerobic gram-positive and gram-

negative organisms. Imipenem and meropenem have activity

against P. aeruginosa and Acinetobacter species, whereas

ertapenem does not. Imipenem is available only in combina-

tion with cilastatin sodium; the latter is added to prevent

renal metabolism of the imipenem, which decreases nephro-

toxicity. Imipenem–cilastatin sodium has been associated

with adverse CNS reactions, including seizures. Slight struc-

tural modifications have rendered meropenem resistant to

degradation by the dehydropeptidases of the renal brush

border, conferred slightly greater activity against aerobic

gram-negative bacilli but slightly less activity against aerobic

gram-positive cocci, and decreased the epileptogenic poten-

tial of the drug.

Quinolones

The quinolones are a class of antimicrobial agents that has

expanded significantly in recent years. Quinolones act by

binding to bacterial DNA gyrase and topoisomerase intra-

venous, thus preventing DNA replication. Resistance occurs

via accrual of mutations at the enzyme target or by active

efflux of the drug out of bacterial cells. Ciprofloxaxin, an

early quinolone, has activity primarily against gram-negative

organisms and staphylococci, with limited activity against

streptococci and anaerobes. It is the only fluoroquinolone

with reliable activity against P. aeruginosa. Some newer

agents (eg, levofloxacin, gemifloxacin, and moxifloxacin)

have greater potency against streptococci, particularly S.

pneumoniae, including penicillin-resistant S. pneumoniae.

They also have excellent activity against aerobic gram-

negative bacilli but less activity against P. aeruginosa. All

quinolones have activity against atypical pneumonia

pathogens, including Legionella, C. pneumoniae, M. pneumo-

niae, C. trachomatis, and some mycobacteria. In general, the

quinolones have excellent bioavailability and may be given

either orally or intravenously.

Tigecycline

Tigecycline, a new semisynthetic glycylcycline related to

tetracycline, possesses in vitro antibacterial activity against

many clinically important gram-positive and gram-

negative aerobic and anaerobic bacteria, including some

with resistance to other classes of agents (eg, vancomycin-

resistant enterococci [VRE], MRSA, and multidrug-

resistant Acinetobacter). Tigecycline may be useful in the

treatment of complicated skin and skin-structure infections

and complicated intraabdominal infections. Because of its

limited activity against P. aeruginosa and Proteus species,

tigecycline is not recommended for the treatment of serious

infections caused by these organisms. Tigecycline should be

used carefully to avoid selective pressure leading to the

development of resistance. It is available for parenteral use

only, has a large volume of distribution, and is cleared by

hepatic metabolism and excretion. There is no dose adjust-

ment for renal insufficiency. Nausea and vomiting are the

most frequently reported side effects.

Triazoles

The triazoles are fungistatic drugs that inhibit the fungal

cytochrome P450 enzyme to block the transformation of

lanosterol to ergosterol, the major sterol in the fungal cell

membrane. The result is osmotic instability and loss of

integrity of the fungal cell wall. Fluconazole is available in

both oral and parenteral forms, and essentially equivalent

blood levels are achieved by either route. Fluconazole has a

wide volume of distribution, achieving high levels in urine,

sputum, saliva, peritoneal fluid, vaginal secretions, bile, cere-

brospinal fluid, skin, liver, and prostate. Achievable cere-

brospinal fluid levels are about 60–80% of serum levels.

Fluconazole is 80% excreted unchanged in the urine, and

dose adjustment is required for renal insufficiency.

Fluconazole is active against C. albicans, C. tropicalis, C. para-

psilosis, C. immitis, and C. neoformans. It has been shown to

be effective in the treatment of hematogenous, mucosal, and

end-organ candidal infection. Caution should be used when

treating infection with certain non-albicans species such as

C. krusei and C. glabrata because these organisms are less

sensitive to fluconazole.

Itraconazole is another triazole that is available for

clinical use. A parenteral form is available. Oral formula-

tions have inconsistent absorptions and so should be

avoided in the critically ill patient. Compared with flu-

conazole, itraconazole has greater activity against

Aspergillus, B. dermatitidis, H. capsulatum, and Sporothrix

schenkii.

Voriconazole is available in both parenteral and oral

forms. Voriconazole is active against all Candida species,

including C. krusei and C. glabrata,many Aspergillus species,

B. dermatitidis, C. immitis, H. capsulatum, Scedosporim/

Pseudoallescheria, and Fusarium species. Like fluconazole,

voriconazole has a large volume of distribution and excellent

oral bioavailability. Voriconazole is indicated for the treat-

ment of invasive aspergillosis, Candida esophagitis, and inva-

sive fungal infections caused by Pseudallescheria boydii and

Fusarium species.

Amphotericin B Lipid Formulations

Amphotericin B desoxycholate has long been the standard

for treatment of severe mycoses. It is a fungicidal drug that

binds to ergosterol in the cell wall and causes osmotic insta-

bility and cell death. The major weakness of amphotericin B

desoxycholate is its numerous adverse effects, especially

nephrotoxicity. The new lipid formulations of amphotericin

B are costly but have the advantage of inducing less



CHAPTER 15

378

nephrotoxicity. There are currently three formulations avail-

able: amphotericin B lipid complex, amphotericin B colloidal

dispersion, and liposomal amphotericin B. They are

approved for use in patients with invasive fungal infections

who are intolerant of amphotericin B desoxycholate.

Liposomal amphotericin B has an additional indication for

empirical therapy in febrile patients with neutropenia. There

are some data, mostly from retrospective comparisons and a

few prospective trials, suggesting improved response rates

with lipid formulations and a lower incidence of nephrotox-

icity. Owing to the significant cost differential between the

lipid formulations of amphotericin B and amphotericin B

desoxycholate, use of the lipid formulations generally is

reserved for specific subgroups of patients. For example, in

high-risk patients such as bone marrow transplant recipients

with suspected invasive aspergillosis, earlier use of the lipid

formulations may be warranted in an attempt to improve

response rates and to decrease renal toxicity.

Echinocandins

Caspofungin, micafungin, and anidulafungin belong to the

class of antifungal drugs called the echinocandins, which

inhibit β-(1,3) glucan synthesis with resulting damage to fun-

gal cell walls. Caspofungin is available for parenteral use only.

It has activity against Candida and Aspergillus species.

Caspofungin has been demonstrated to be equivalent to

amphotericin B in the treatment of invasive candidiasis and

candidemia. Caspofungin has an indication for empirical

therapy in the febrile neutropenic patient. Anidulafungin is

approved for parenteral treatment of candidemia and candi-

dal peritonitis and intraabdominal abscesses, as well as

esophageal candidiasis. Micafungin is approved for can-

didemia, acute disseminated candidiasis, candida peritonitis,

candidal esophagitis, and antifungal prophylaxis for

hematopoietic stem cell transplant recipients.

Macrolides

The three macrolides available for clinical use in the United

States are erythromycin, clarithromycin, and azithromycin.

In comparison with erythromycin, the latter two have greater

activity against H. influenzae, C. trachomatis, M. avium com-

plex, and other nontuberculous mycobacteria. The role of

these agents in the critically ill patient is limited to treatment of

patients with suspected atypical pneumonias. All three agents

possess good activity against M. pneumoniae, C. pneumoniae,

and Legionella species. Erythromycin and azithromycin are

available in a parenteral formulation, whereas clarithromycin

is available only in an oral form.

Quinupristin-Dalfopristin

Quinupristin-dalfopristin is a formulation of two bacterio-

static antimicrobial agents that, when combined, are bacteri-

cidal. Its antimicrobial activity is the result of inhibition of

protein synthesis at the 50S ribosome. Quinupristin-

dalfopristin is available for intravenous use only. Its spectrum of

activity is similar to that of vancomycin and includes S.

pneumoniae, other streptococci, S. aureus, and coagulase-

negative staphylococci. It is bacteriostatic for Enterococcus

faecium and has no activity against E. faecalis. The niche for

quinupristin-dalfopristin is in the treatment of vancomycin-

resistant E. faecium and possibly glycopeptide-insensitive

S. aureus.

Linezolid

Linezolid belongs to the class of drugs known as oxazolidi-

nones. These drugs block bacterial protein synthesis at the

ribosome at a very early stage. Because of its novel mech-

anism of action, linezolid does not share cross-resistance

with other antimicrobial agents. Linezolid’s spectrum of

activity is essentially identical to that of vancomycin.

Much like quinupristin-dalfopristin, the major indication

for use of this agent is the treatment of vancomycin-

resistant enterococcus and glycopeptide-insensitive S.

aureus infections. Linezolid is approved for treatment of

pneumonia, bacteremia, and skin and soft tissue infec-

tions owing to gram-positive organisms. Linezolid is

available in both oral and parenteral formulations, with

excellent oral bioavailability.

Daptomycin

Daptomycin belongs to a new class of antimicrobial agents; it

is a cyclic lipopeptide antibiotic that is rapidly bactericidal

against S. aureus, streptococci, and enterococci, including

multidrug-resistant strains. It is available for parenteral use

only and is approved for use in complicated skin and soft

tissue infections caused by staphylococci, streptococci, and

E. faecalis.

Denning DW: Echinocandin antifungal drugs. Lancet 2003;

362:1142–51. [PMID: 14550704]

Dix SP, Andriole VT: Lipid formulations of amphotericin B. Curr

Clin Top Infect Dis 2000;20:1–23. [PMID: 10943516]

Johnson LB, Kauffman CA: Voriconazole: A new triazole antifun-

gal agent. Clin Infect Dis 2003:36:630–7. [PMID: 12594645]

Lundstrom TS, Sobel JD: Antibiotics for gram-positive bacterial

infections: Vancomycin, quinopristin-dalfopristin, linezolid,

and daptomycin. Infect Dis Clin North Am 2004;18:651–68.

[PMID: 15308280]

Martin SI, Kaye KM: Beta-lactam antibiotics: Newer formulations

and newer agents. Infect Dis Clin North Am 2004;18:603–20.

[PMID: 15308278]

O’Donnell JA, Gelone SP: The newer fluoroquinolones. Infect Dis

Clin North Am 2004;18:691–716. [PMID: 15308282]

Walsh TJ et al: Caspofungin versus liposomal amphotericin B for

empirical antifungal therapy in patients with persistent fever

and neutropenia. N Engl J Med 2004;351:1391–402. [PMID:

15459300]

Zuckerman JM: Macrolides and ketolides: Azithromycin, clar-

ithromycin, telithromycin. Infect Dis Clin North Am 2004;18:

621–50. [PMID: 15308279]



INFECTIONS IN THE CRITICALLY ILL

379



Evaluation of the ICU Patient with New

Fever

The evaluation of new fever in the ICU setting involves con-

sideration of diagnoses much different from those encoun-

tered in patients in a community setting. Many noninfectious

causes of fever must be considered and excluded in order to

avoid missing important diagnoses and perhaps administer-

ing antimicrobial agents unnecessarily.

Clinical Features

A. History and Examination—When eliciting a history, the

physician should pay special attention to changes in quality

and quantity of respiratory secretions, changes in the patient’s

tolerance to oral intake, the condition of the patient’s skin, and

the presence or absence of diarrhea. Because many patients in

an ICU will be unable to provide a history, information should

be elicited from family members and caregivers, and a careful

review of medical records and laboratory data and a thorough

physical examination should be performed.

It is recommended that temperatures be taken using an

electronic probe in the mouth, rectum, or external auditory

canal. Axillary temperature measurements may be spuriously

low and so should not be used. Sinusitis should be consid-

ered in patients with nasotracheal or nasogastric tubes—

which block the sinus ostia, thus predisposing to infection.

Careful auscultatory examination should be performed to

look for new pulmonary findings that may indicate a noso-

comial pneumonia or a new or changing murmur that may

indicate endocarditis. Abdominal examination should be

performed to assess the presence or absence of bowel sounds

and any focal tenderness or masses. All intravascular catheter

sites should be scrutinized for signs of inflammation at the

catheter exit site or along the subcutaneous tunnel. The

patient’s skin requires careful inspection for any lesions that

may indicate hematogenous seeding (either bacterial or fun-

gal), skin eruptions that may indicate a drug reaction, and

decubitus ulcers or surgical wounds that may be infected.

B. Laboratory and Imaging Studies—Laboratory evalua-

tion should include a complete blood count, routine

chemistries with liver enzymes, and urinalysis. Cultures of

blood and urine should be obtained routinely. Other evalua-

tions should be directed by the results of the history and

physical examination. A chest radiograph and sputum sam-

ple for culture should be obtained when nosocomial pneu-

monia is suspected. Stool specimens for C. difficile toxin

should be collected from patients who develop diarrhea

while hospitalized. Routine stool cultures and stool for ova

and parasites are not useful in evaluation of new fever in a

patient in the ICU. Any wounds that appear infected should

be debrided, with cultures performed on any purulent mate-

rial. If the abdominal examination or the liver enzymes sug-

gest an intraabdominal process, a right upper quadrant

ultrasound or CT scan of the abdomen and pelvis may be

requested. If sinusitis is suspected, CT scan or MRI of the

sinuses may be helpful, although impractical in the unstable

or ventilated patient.

Treatment

Empirical therapy will depend on the results of the physical

examination and initial laboratory data.

Avecillas JF, Mazzone P, Arroliga AC: A rational approach to the

evaluation and treatment of the infected patient in the intensive

care unit. Clin Chest Med 2003;24:645–69. [PMID: 14710696]

O’Grady NP et al: Practice guidelines for evaluating new fever in

critically ill adult patients. Task Force of the Society of Critical

Care Medicine and the Infectious Diseases Society of America.

Clin Infect Dis 1998;26:1042–59. [PMID: 9597223]

Peres Bota D et al: Body temperature alterations in the critically ill.

Intensive Care Med 2004;30:811–6. [PMID: 15127194]



Nosocomial Pneumonia

ESSENTIALS OF DIAGNOSIS



Fever, tachypnea, tachycardia, and abnormal breath

sounds.



Risk factors include altered mental status, lung disease,

endotracheal tube, or tracheostomy.



Change in volume or appearance of sputum.



Infiltrates or other changes on chest x-ray.



Hypoxemia or worsening arterial blood gases.

General Considerations

It is estimated that at least 5% of hospitalized patients will

develop a nosocomial infection. Pneumonia is the second

most common nosocomial infection after urinary tract

infection and is the leading cause of death owing to hospital-

acquired infection. Although often considered separately,

ventilator-associate pneumonia (VAP) is a very important

subset of nosocomial pneumonia.

Pathophysiology

Pneumonia develops from one of three mechanisms:

(1) inhalation of an aerosol containing infectious microor-

ganisms, (2) hematogenous seeding of microorganisms in

the lung, or (3) aspiration of oropharyngeal flora.

A. Inhalational Pneumonia—Infectious particles less than

3–5 μm in diameter are capable of reaching the terminal

bronchi and alveoli during inhalation. M. tuberculosis,

Legionella, and influenza virus—as well as many fungi—are

known to cause nosocomial pneumonia by this route of

spread. Contaminated respiratory equipment has been rec-

ognized as a source of infected aerosols and a cause of gram-

negative bacillary necrotizing pneumonia. Hospital policies



CHAPTER 15

380

and procedures requiring rigorous, frequent changing and

disposal of equipment have markedly decreased the inci-

dence of inhalation pneumonia in the hospital setting.

B. Hematogenous Pneumonia—Pneumonia caused by the

hematogenous seeding of infection is uncommon. Patients

predisposed to hematogenous pneumonia are those with

S. aureus or Candida infections associated with infected

intravenous catheters, septic thrombophlebitis, or endo-

carditis of the right side of the heart. Pneumonia develops as

a consequence of infected thrombi forming within the

intravascular space and traveling through the bloodstream to

reach the pulmonary circulation (septic pulmonary emboli).

C. Aspiration Pneumonia—Aspiration of oropharyngeal

bacterial flora is the most common mechanism responsible

for nosocomial pneumonia. Although all individuals aspirate

microscopic quantities of oropharyngeal secretions, certain

factors increase the risk of pneumonia. Such factors include

(1) high frequency and large volume of secretions aspirated,

(2) aspiration of particularly virulent bacterial species, (3)

the presence of particulate matter in the aspirate, and (4)

abnormalities of the lower airway and alveolar defense mecha-

nisms. Aerobic enteric gram-negative bacilli are not usually

part of the flora colonizing the oropharynx in normal, healthy,

nonhospitalized individuals. However, oropharyngeal colo-

nization by gram-negative bacilli occurs within a days of

admission to the ICU. Patients receiving antibiotics develop

gram-negative colonization more rapidly. Cross-infection and

cross-colonization from patient to patient contribute to the

pathogenesis when infection control measures are substandard.

Prophylaxis of stress ulcers with the use of antacids or

-receptor antagonists is considered a risk factor for gram-

negative aspiration pneumonia via gastric acid neutralization,

allowing subsequent overgrowth of enteric gram-negative

bacilli, although this theory remains controversial.

Microbiologic Etiology

The microbiology of pneumonia in hospitalized patients is

significantly different from that of pneumonia acquired

outside the hospital. Early-onset nosocomial pneumonia

(<4–7 days) in patients who have not received prior antibi-

otic therapy is typically caused by Enterobacteriaceae,

Haemophilus species, S. aureus, and pneumococci. Patients

who develop late-onset pneumonia (>4–7 days) and who have

received prior antibiotic therapy are at risk for infection with

P. aeruginosa, A. baumanii, Stenotropomonas maltophilia, and

MRSA. Approximately 20–40% of nosocomial pneumonias

are polymicrobial in etiology. Anaerobic bacteria are not

considered an important pathogen in nosocomial pneumo-

nia. Legionella species have been implicated as the cause of

nosocomial pneumonia in some hospitals.

Clinical Features

The diagnosis of pneumonia depends on physical examina-

tion findings, microbiologic analysis of lower respiratory tract

secretions, and review of the chest x-ray. Although risk factors

for development of pneumonia should be considered in assess-

ing the likelihood of pneumonia, any critically ill patient—

especially one with underlying heart or lung disease—should

be considered at risk for nosocomial pneumonia. Studies have

identified a number of risk factors in the development of

nosocomial pneumonia. Among the most important of these

factors are neurologic impairment, witnessed aspiration,

mechanical ventilation (with increasing risk associated with

prolonged need for mechanical ventilation), underlying

chronic lung disease, the presence of ARDS, increasing age,

use of a nasogastric tube, enteral feeding, endotracheal cuff

pressure, severity of underlying illness, need for tra-

cheostomy, and supine position of the patient.

A. Symptoms and Signs—Patients with suspected nosoco-

mial pneumonia should undergo prompt assessment prior to

initiation of therapy. The clinical features of pneumonia in hos-

pitalized patients are subtle and variable, and in some cases,

there may be no symptoms or examination findings. Use of

clinical and radiographic features alone to diagnose nosoco-

mial pneumonia will lead to the misclassification of many dis-

orders as nosocomial pneumonia. Some diagnostic clues

include altered mental status, fever, tachypnea, tachycardia, and

abnormal breath sounds. Physical findings commonly include

crackles or rales rather than evidence of lung consolidation.

Wheezing may be present, and localized wheezing may suggest

aspiration of a foreign body. The presence of preexisting abnor-

mal lung findings may make the diagnosis difficult. Patients with

a tracheostomy or endotracheal tube may first manifest a lower

respiratory tract infection by a change in gross appearance of the

respiratory secretions or an increase in oxygen requirement. The

physician should query the respiratory therapist or nurse regard-

ing changes in quantity and appearance of these secretions. The

development of purulent sputum or an increase in the quantity

of secretions suggests heavy bacterial colonization or infection,

often prior to the development of a radiographic infiltrate.

B. Laboratory Findings—Most patients in the ICU are too

debilitated to produce an adequate sputum sample; samples

of respiratory secretions may be obtained in such cases by

endotracheal suctioning or fiberoptic bronchoscopy. It is

clear that endotracheal aspirates alone are often inaccurate in

diagnosing the etiologic agent of a nosocomial pneumonia.

Quantitative endotracheal aspirates, a protected specimen

brush, or bronchoalveolar lavage often will provide more

useful information. However, most diagnostic modalities for

nosocomial pneumonia are neither sensitive nor specific.

Respiratory cultures obtained by any means must be inter-

preted with caution because the airways of hospitalized

patients are almost always colonized with various potentially

pathogenic bacteria. A Gram stain of the sputum may help to

identify the etiologic agent and suggest appropriate empiri-

cal therapy. A sputum Gram stain that reveals 10 or fewer

squamous epithelial cells and more than 25 polymorphonu-

clear cells per high-power field can be considered an adequate

specimen. Typically, the Gram stain is more helpful than cul-

ture results in patients already receiving antimicrobial agents

that may inhibit bacterial growth in culture.



INFECTIONS IN THE CRITICALLY ILL

381

Invasive methods of diagnosis such as bronchoscopy with

or without protected specimen brush should not be used to

determine whether or not to initiate therapy for hospital-

acquired pneumonia, although results of such studies may be

useful in modifying subsequent therapy. Good outcome

remains most closely correlated with correct early antimicro-

bial therapy.

A complete blood count, blood cultures, and a chest x-ray

are essential. The chest x-ray may reveal an infiltrate, but non-

specific findings such as atelectasis may be the only finding.

Patients with prior pulmonary disease often have preexisting

abnormalities on chest x-ray that make it difficult to identify

new infiltrates. Blood cultures, when positive, should be con-

sidered definitive for the etiologic agent; 10–20% of hospital-

acquired pneumonias are associated with a positive blood

culture. Pleural fluid, if present, should be sampled by thora-

centesis for Gram stain, culture for aerobic and anaerobic

organisms, and determination of cell count, pH, total protein,

and LDH concentration. Serologic studies are seldom useful

in determining the cause of nosocomial pneumonia.

Differential Diagnosis

The differential diagnosis of nosocomial pneumonia includes

virtually all processes associated with pulmonary infiltrates.

ARDS, pulmonary emboli with infarction, cardiogenic pul-

monary edema, lung cancer, and atelectasis often are difficult

to differentiate from an infectious lung process. Less common

diseases to be considered include collagen-vascular disease,

pulmonary hemorrhage, radiation pneumonitis, hypersensi-

tivity pneumonitis, sarcoidosis, occupational lung diseases,

and pulmonary alveolar proteinosis.

Treatment

A. Antibiotics—Empirical antimicrobial therapy for nosoco-

mial pneumonia should be based on clinical features, epidemi-

ologic and host factors, and results of an initial sputum Gram

stain. Surveillance data relating to local nosocomial flora

leading to pneumonia in the ICU should be reviewed.

Antibiotic therapy should be administered intravenously and

should target a broad spectrum of bacteria. The possibility of

resistant microorganisms should be considered in the hospi-

talized patient with nosocomial pneumonia. In patients with

early-onset pneumonia, a third-generation non-

antipseudomonal cephalosporin or a β-lactam/β-lactamase

inhibitor (such as piperacillin-tazobactam), or in some cases

a fluoroquinolone should be adequate. In patients with

late-onset pneumonia, cefipime, imipenem, meropenem,

ceftazidime, or a β-lactam/β-lactamase inhibitor in combina-

tion with an aminoglycoside or fluoroquinolone provides the

broadest coverage for infections caused by most

Enterobacteriaceae, P. aeruginosa, and S. aureus. If the

prevalence of MRSA strains is significant, and staphylococ-

cal pneumonia is a consideration, vancomycin should be

included in the initial drug regimen. Quinolones have

excellent activity against most Enterobacteriaceae as well as

H. influenzae but possess less activity against streptococci and

staphylococci (except for levofloxacin and the newer

quinolones) and little to no activity against most anaerobes.

New data suggest that shorter courses of antibiotics (5-8

days) than traditionally used are effective, safe, and result in

less antibiotic resistance.

B. Supportive Care—Patients with pneumonia require suc-

tioning of respiratory secretions, postural drainage, and occa-

sionally, fiberoptic bronchoscopy. Coughing is the most effective

way to clear the large airways of respiratory secretions. In

patients with endotracheal tubes or tracheostomies, use of

appropriate suctioning must substitute for cough. Postural

drainage and chest percussion may be useful in selected

patients, but only if it can be demonstrated that removal of res-

piratory secretions is improved. In certain patients, fiberoptic

bronchoscopy can be helpful in identifying endobronchial

obstruction, and this technique may aid in suctioning secretions

from particular airways. Most patients with pneumonia receive

bronchodilator therapy, but the effectiveness of these agents in

patients without obstructive lung disease is not known.

Prevention

Prevention of nosocomial pneumonia is of utmost impor-

tance in decreasing morbidity and mortality rates and con-

trolling the costs of hospital care. Recognition of the

aspiration-prone patient is essential. Patients with nasogas-

tric tubes for enteral feeding should have the head of the bed

elevated 30–45 degrees during feeding and be monitored for

excess gastric residuals that lead to aspiration. All patients

should be turned frequently whenever possible. Appropriate

disposal, disinfection, or sterilization of respiratory equip-

ment is critical for prevention of contamination and subse-

quent inhalation pneumonias. Nurses, physicians, and

respiratory therapists must use sterile technique for endotra-

cheal suctioning. Meticulous hand washing before and after

patient examination and the wearing of gloves when appro-

priate will help to decrease the overall incidence of nosocomial

infections in the ICU. In intubated patients, subglottic suc-

tioning has been shown to be effective in reducing ventilator-

associated pneumonia. Other interventions that may help in

preventing nosocomial pneumonia include placing patients in

the semirecumbent position, avoidance of prolonged nasal

intubation (which may lead to nosocomial sinusitis), and use

of noninvasive positive-pressure ventilation rather than intu-

bation whenever possible.

American Thoracic Society and the Infectious Diseases Society of

America: Guidelines for the management of adults with hospital-

acquired, ventilator-associated, and healthcare-associated

pneumonia. Am J Respir Crit Care Med 2005;171:388–416.

[PMID: 15699079]

Canadian Critical Care Trials Group: A randomized trial of diag-

nostic techniques for ventilator-associated pneumonia. N Engl J

Med 2006;355:2619–30. [PMID: 17182987]

Chastre J: Antimicrobial treatment of hospital-acquired pneumo-

nia. Infect Dis Clin North Am 2003;17:727–38. [PMID:

15008595]



CHAPTER 15

382

Chastre J et al: Comparison of 8 vs 15 days of antibiotic therapy for

ventilator-associated pneumonia in adults: A randomized trial.

JAMA 2003;290:2588–98. [PMID: 14625336]

Diaz O, Diaz E, Rello J: Risk factors for pneumonia in the intu-

bated patient. Infect Dis Clin North Am 2003;17:697–705.

[PMID: 15008592]

Klompas M: Does this patient have ventilator-associated pneumo-

nia? JAMA 2007;297:1583–93. [PMID: 17426278]

Leroy O, Soubrier S: Hospital-acquired pneumonia: Risk factors,

clinical features, management, and antibiotic resistance. Curr

Opin Pulm Med 2004;10:171–5. [PMID: 15071367]

Soto GJ: Diagnostic strategies for nosocomial pneumonia. Curr

Opin Pulm Med 2007;13:186–91. [PMID: 17414125]



Urinary Catheter–Associated Infections

ESSENTIALS OF DIAGNOSIS



Although the patient may be asymptomatic, suprapubic

tenderness suggests lower tract infection; fever and

flank pain suggest upper tract infection.



Pyuria and white blood cell casts.



Positive urine culture.

General Considerations

A urinary (bladder) catheter provides a portal of entry into the

urinary tract for microorganisms. Urinary catheter–associated

infections account for up to 40% of all nosocomial infec-

tions. Fewer than 5% of patients who develop bacteriuria will

become bacteremic; however, the tremendous frequency

of nosocomial bacteriurias accounts for the observation

that urinary catheter–associated infections account for

15% of all nosocomial bacteremias. Early recognition and

appropriate treatment of urinary catheter-associated infec-

tions can reduce morbidity and length of stay in the ICU

significantly.

The urinary catheter allows transit of microorganisms

colonizing the perineum and the urethral meatus to pass

through the urethra and enter the bladder. Most catheter-

associated bacteriuria occurs by extraluminal ascent of bac-

teria via the mucous film coating the outer surface of the

catheter; these organisms colonize the patient’s perineum.

Bacterial also ascend intraluminally through the catheter

because of failure of the closed drainage system or contami-

nation of the collection bag. Bacteriuria occurs in catheter-

ized patients at a rate of 3–10% per catheter-day, with

10–25% of these patient developing symptoms of local infec-

tion and bacteremia and sepsis occurring in 2–4%.

The most common microorganisms causing infection in

short-term catheterized patients in the ICU include

Enterobacteriaciae (such as E. coli, Klebsiella species),

P. aeruginosa, Enterobacter species, enterococci, and Candida

species. Polymicrobial infections occur in up to 15% of

catheterized patients; anaerobic infections are extremely

rare. Candiduria is common in patients receiving corticos-

teroids or broad-spectrum antibiotics and in those with dia-

betes mellitus. Urinary tract infections with Candida species

or S. aureus should raise the question of hematogenous

spread because a urinary tract infection with one of these

organisms can be a marker of disseminated disease.

Risk factors for catheter-associated bacteriuria include

prolonged duration of catheterization, older age, severe

underlying illness, diabetes mellitus, absence of systemic

antibiotic use, female sex, abnormal serum creatinine, errors

in catheter care, and periurethral colonization with potential

uropathogens. Once bacteriuria has occurred, it is difficult to

prevent subsequent infection; thus prevention of the initial

bacteriuria is critical. The only two factors shown to decrease

the risk of nosocomial catheter-associated bacteriuria are

maintenance of a closed catheter system with aseptic inser-

tion of catheter and early removal of the urinary catheter.

Clinical Features

A. Symptoms and Signs—Diagnosis of catheter-related

urinary tract infection is straightforward, although many

patients with catheter-associated urinary tract infections are

asymptomatic. Up to 30% of patients will have fever or other

symptoms of urinary tract infection. Fever or other systemic

signs of infection, as well as pain localized to the flank, sug-

gest upper tract infection.

B. Laboratory Findings—The white blood cell count may

be elevated with serious infection. Definitive diagnosis is

based on urinalysis and results of urine culture. Direct exam-

ination of the urine continues to be valuable for early diag-

nosis of bacterial infections of the genitourinary tract. Pyuria

indicates the presence of urinary infection rather than sim-

ple bacterial colonization of the urine; the presence of white

blood cell casts suggests upper tract disease. The presence

of one or two leukocytes per high-power field (400×) or

bacteria seen under oil immersion (1000×) in unspun urine

has a 95% correlation with the presence of more than

100,000 cfu/mL of urine. Thus microscopy is useful for

identification of urinary tract involvement with a high urine

bacterial count.

The presence of bacteria in a concentration greater than

100,000 cfu/mL with accompanying pyuria is consistent with

infection. In catheterized patients, bacterial counts of more

than 1000 cfu/mL, if untreated, will increase to more than

100,000 cfu/mL within 24–48 hours. Urine culture for bacte-

ria and fungi must be submitted to the laboratory with

prompt processing. Alternatively, specimens may be refriger-

ated at 4°C, where bacterial counts will remain stable for up

to 24 hours. Urine left at room temperature for over 2 hours

after collection will have significantly higher urine bacterial

colony counts, thereby confounding the diagnosis of urinary

tract infection.

Most microbiology laboratories routinely identify and

perform antimicrobial susceptibility tests on microorganisms



INFECTIONS IN THE CRITICALLY ILL

383

present in numbers of 10

CFU/mL or more. If polymicro-

bial bacteriuria is anticipated (eg, chronic indwelling

catheter or neurogenic bladder), the laboratory should be

alerted to this possibility.

C. Imaging Studies—Seriously ill patients presenting with

a urinary tract infection, whether catheter-related or not,

who have high fever, flank pain, or urosepsis require further

evaluation for upper tract disease. Ultrasonography is use-

ful to assess the anatomy of the genitourinary tract and to

rule out an obstructed ureter. Further studies, including an

intravenous pyelogram, retrograde pyelogram, or CT scan

may be necessary to diagnose renal abscess, perinephric

abscess, or nephrolithiasis.

D. Complications—Complications of bladder and kidney

infections are common and potentially serious. The urinary

tract is the most common site of origin of gram-negative

bacteremia and sepsis. Acute pyelonephritis, chronic

pyelonephritis, emphysematous pyelonephritis, renal

abscess, and urosepsis may complicate an untreated urinary

tract infection.

Differential Diagnosis

Urinary tract infections should be differentiated from

asymptomatic bacteriuria, urethritis, prostatitis, sexually

transmitted diseases, pelvic inflammatory disease, divertic-

ulitis, intraabdominal abscess, and peritonitis.

Treatment

A. Asymptomatic Patient with Bacteriuria—Antibiotic

therapy in the asymptomatic bacteriuric patient with an

indwelling catheter and an unremarkable urinalysis is dis-

couraged because bacteriuria will recur when treatment is

discontinued and selection of antibiotic-resistant organisms

is likely.

B. Patient with Abnormal Urinalysis and Bacteriuria—In

patients with abnormal urinalysis (typically pyuria) and bac-

teriuria, the urinary catheter should be removed and antibi-

otics initiated. If the patient requires an indwelling catheter

for bladder drainage, a new catheter can be reinserted. In gen-

eral, empirical therapy should be directed against nosocomial

gram-negative pathogens; a third-generation cephalosporin,

a quinolone, or an aminoglycoside is usually appropriate ini-

tial antibiotic treatment. Therapy may be discontinued in

uncomplicated cases at 7 days. However, more seriously ill

patients with prolonged fever, suspected upper tract involve-

ment, positive blood cultures, renal insufficiency, or sepsis

require a longer course of intravenous antibiotics and further

evaluation for upper tract disease.

C. Candiduria—Asymptomatic candiduria in immunocom-

promised patients or symptomatic candidal urinary tract

infections in all patients require antifungal treatment. Oral

fluconazole and a short course of intravenous amphotericin B

are both effective for candidal infections confined to the blad-

der. Upper tract infection requires systemic treatment with

oral or intravenous fluconazole or intravenous amphotericin

B; complications such as fungal ball, renal or perirenal

abscess, or hematogenous dissemination should be consid-

ered. In all cases, the catheter should be removed or changed.

Asymptomatic candiduria in most patients without immuno-

compromise usually will clear once the indwelling catheter

has been removed and antimicrobial therapy discontinued.

Current Controversies and Unresolved Issues

The appropriate course of action to take with a hospitalized

patient who develops asymptomatic nosocomial bacteriuria

associated with a urinary catheter remains controversial. The

literature suggests that the treatment of bacteriuria in an

attempt to avoid complications is not useful. One approach

is to remove the catheter as soon as possible after discovery

of bacteriuria, with a repeat urinalysis and urine culture

3–5 days later. If the urinalysis is normal and the urine culture

is sterile, it is likely that the infection has cleared. If bacteri-

uria persists, the patient must be monitored for resolution or

be treated with a short course of antibiotic therapy.

Exceptions to this rule include the presence of bacteriuria

with organisms that have a high predilection for causing sub-

sequent bacteremia (eg, Serratia marcescens) and bacteriuria

in patients at risk for serious complications (eg, neutropenic

or pregnant patients).

Questions about the value of screening tests for bac-

teriuria in patients who have urinary catheters are unre-

solved. Several tests correlate reasonably well with

bacteriuria of 100,000 cfu/mL or more. The leukocyte

esterase–nitrate test has a sensitivity of about 85% in

detecting bacteria at 10

cfu/mL or more. However, the

predictive value of a positive test is much lower (25–50%)

than that of a negative test.

Attempts to decrease the incidence of bacterial coloniza-

tion by use of nitrofurazone- or silver alloy–coated urinary

catheters have been investigated. It appears that such

catheters may decrease the incidence of bacteriuria.

Economic evaluations seem to indicate that these catheters

are cost-effective when used in patients requiring in-dwelling

catheters for short duration.

Johnson JR, Kuskowski MA, Wilt TJ: Systematic review: antimicro-

bial urinary catheters to prevent catheter-associated urinary

tract infection in hospitalized patients. Ann Intern Med

2006;144:116–26. [PMID: 16418411]

Kauffman CA: Candiduria. Clin Infect Dis 2005;41:S371–6.

[PMID: 16108001]

Leone M et al: Risk factors of nosocomial catheter-associated uri-

nary tract infection in a polyvalent intensive care unit. Intensive

Care Med 2003;29:1077–80. [PMID: 12743682]

Saint S, Chenoweth CE: Biofilms and catheter-associated urinary

tract infections. Infect Dis Clin North Am 2003;17:411–32.

[PMID: 12848477]



CHAPTER 15

384



Intravenous Catheter–Associated Infections

ESSENTIALS OF DIAGNOSIS



Exit-site infection: erythema, tenderness, induration,

and exudate at cutaneous exit site.



Tunnel infection: induration, erythema, tenderness at

least 1 cm deep to the skin exit site and without puru-

lent drainage through the exit site.



Catheter sepsis: fever, tachypnea, and tachycardia with

positive blood cultures and no other site of infection

identified.

General Considerations

Obtaining short-term central intravenous and arterial access

by means of indwelling catheters is an integral part of the

monitoring and management of patients in the ICU. These

catheters are required for administration of antimicrobial

therapy, blood products, fluid and electrolyte replacement,

and monitoring of hemodynamic status. Short-term catheters

may be metal needles or Teflon or other synthetic catheters

inserted into vessels with only a short distance between the

skin and intravascular space. These include radial, dorsalis

pedis, and femoral artery catheters, intravenous catheters

inserted into superficial veins on the extremities, and large-

bore central venous catheters inserted into subclavian or inter-

nal jugular veins in the neck. The wide-lumen catheters

through which pulmonary artery catheters and temporary

transvenous cardiac pacemakers are placed are also considered

short-term catheters.

Patients requiring long-term central venous catheters,

used primarily for administration of cancer chemotherapy

or for total parenteral nutrition, are often admitted to the

ICU. These long-term catheters have in common a subcuta-

neous tunnel through which the catheter passes after enter-

ing the skin and before entering the vein. The presence of a

Dacron cuff just inside the exit site of the catheter stimulates

growth of the surrounding tissue, thus preventing microor-

ganisms from entering the catheter tract. This permits long-

term use of the catheter with a greatly decreased incidence of

catheter-related infection.

Two relatively recent additions to the armamentarium of

long-term venous access catheters are the midline catheters

and the peripherally inserted central venous catheters

(PICCs). The midline catheter is inserted into the proximal

basilic or cephalic vein or into the distal subclavian vein via

the antecubital fossa. PICCs are inserted into the superior

vena cava via the antecubital fossa as well. These methods of

venous access are associated with fewer mechanical complica-

tions and lower rates of phlebitis and bloodstream infection

and are easier to maintain. It appears that the PICC can be left

in place for a long duration, but the specifics are not yet known.

Pathophysiology

Infection of intravascular catheters typically results from one of

four events. The most common mechanism of infection is

migration of cutaneous colonizing organisms from the inser-

tion site to the catheter tip, frequently seen with short-term

catheters. Inadvertent contamination of the catheter hub by the

physician or nurse at the time of insertion or by frequent device

manipulation also can lead to infection and is an important factor

in long-term catheter infection. Contamination of the infusate

and hematogenous seeding of the intravascular device from

another primary site of infection are less common problems.

Microbiologic Etiology

Gram-positive organisms—most commonly coagulase-

negative staphylococci, followed by S. aureus—and entero-

cocci are implicated in most bloodstream infections owing to

intravenous devices. The incidence of nosocomial infections

owing to Candida species, especially C. albicans, has

increased dramatically in recent years. Approximately

15–20% of bloodstream infection are due to gram-negative

bacilli (eg, E. coli, Klebsiella species, Enterobacter species, and

P. aeruginosa). It is important to remember that virtually any

organism can cause intravenous catheter infection in an

immunocompromised host.

Clinical Features

A. Symptoms and Signs—When evaluating a patient with a

possible catheter-related infection, several syndromes need to

be considered. If an intravascular catheter is removed, growth

of 15 colony-forming units (semiquantitative) or more or 10

organisms (quantitative culture) or more from a distal or prox-

imal segment of the intravascular catheter in the absence of

accompanying clinical symptoms defines bacterial colonization

of the catheter. An exit-site infection is characterized by the

presence of erythema, tenderness, induration, and purulence

within 2 cm of the exit site of the catheter. A tunnel infection is

manifested by erythema, tenderness, and induration in the tis-

sues over the catheter and more than 2 cm from the exit site.

Finally, a catheter-related bloodstream infection is defined as

isolation of the same organism from both the catheter tip and

peripherally drawn blood cultures in the appropriate clinical

setting with no other source of infection identified.

It is important to distinguish patients with severe neu-

tropenia and catheter-related infections from their immuno-

competent cohorts. These patients may not manifest typical

inflammatory changes at the site of infection—especially

notable is the lack of purulent drainage. However, erythema

is still present with infection, and patients may complain of

local tenderness; either finding should alert the physician to

the probability of infection.

Complications that may occur as a result of catheter-

related infections include septic thrombophlebitis and

endocarditis. Associated risk factors for catheter-related



INFECTIONS IN THE CRITICALLY ILL

385

infections include extremes of age, altered host defenses

(especially skin diseases), severity of the underlying disease,

use of a multilumen catheter, and the presence of infections

remote from the catheter site. However, the most important

hospital-related risk factor is the duration of use of a partic-

ular catheter site. Total parenteral nutrition also has been

identified as a risk factor for catheter-related infection, per-

haps because of the length of time the catheter remains in

place, as well as the fact that total parenteral nutrition solu-

tions may act as a culture medium to promote the growth of

various organisms. Location of the central venous catheter

influences the risk of infection. Femoral catheters are most

prone to infection, followed by internal jugular catheters; the

latter is probably the result of contamination with oropha-

ryngeal secretions and difficulty with immobilization.

Subclavian catheters are least prone to infection.

B. Laboratory Findings—Local catheter-related infections

(eg, exit-site or tunnel infections) may be associated with few

or no laboratory abnormalities. Culture of drainage fluid or

exudate associated with the exit site may be helpful in deter-

mining a causative organism.

Bacteremia (positive blood cultures) may be related to

intravascular devices and may occur with or without sepsis.

Device-related bacteremia has been defined using quantitative

blood culture techniques.A device-related bacteremia is defined

as (1) more than a 5–10-fold increase in colony-forming units

of bacteria per milliliter of blood obtained through the device

compared with blood drawn from a distant peripheral vein,

(2) in the absence of positive peripheral blood cultures, more

than 1000 cfu/mL in blood obtained through the device, or

(3) organisms isolated from a catheter tip on removal in the

appropriate clinical setting. When no other clinical source of

infection is evident, resolution of fever on removal of the device

is also suggestive of catheter-related infection.

Differential Diagnosis

Exit-site, catheter-tunnel, or catheter infection must be differ-

entiated from chemical phlebitis, which can cause erythema,

tenderness, and induration of the vessel wall and overlying

skin. Fever also may be present in chemical phlebitis, but sys-

temic symptoms and signs consistent with sepsis are absent,

and exudate cannot be expressed from the exit site.

Thrombophlebitis may complicate catheter-related infection.

Pain, tenderness, and erythema at the infected site, as well as

edema distal to the site, are common presenting signs.

Thrombosis of a central venous catheter can cause edema

proximal to the site of thrombosis, but tenderness, erythema,

and fever are typically absent. Catheter tip sepsis usually pres-

ents with no symptoms or signs of local infection and must be

differentiated from infection from any other source.

Treatment

A. Exit-Site Infections—Short-term catheters suspected of

being infected should be removed. However, there is increasing

interest in maintaining long-term catheters in the presence of

certain infections. Few data are available to assist in developing

recommendations regarding removal or maintenance of an

infected long-term catheter. From the published data that do

exist, it appears that treatment of an infected catheter with

antibiotics alone is successful in about 50% of cases. Thus, in

some settings, it may be reasonable to attempt to eradicate a

long-term catheter exit-site infection with local care plus intra-

venous antibiotics. Empirical therapy should include van-

comycin to cover coagulase-negative staphylococci and S. aureus,

as well as an aminoglycoside or a cephalosporin with activity

against P. aeruginosa and other gram-negative bacilli. Definitive

therapy can be selected using culture results. If there is no

response within a few days, or if the exit-site infection appears to

be progressing to involve the catheter tunnel—or if fungi or

resistant bacteria are isolated—the catheter should be removed.

B. Tunnel Infections—Patients with tunnel infections

require removal of the catheter for definitive therapy. The

tunnel itself may require surgical incision and debridement.

Empirical antibiotic therapy should consist of vancomycin to

treat coagulase-negative staphylococci and S. aureus and a

third-generation cephalosporin or an aminoglycoside to

treat gram-negative bacilli.

C. Catheter-Tip Sepsis—Short-term catheters should be

removed and antibiotics administered. For a long-term

catheter, catheter-tip sepsis may be treated initially with intra-

venous antibiotics while leaving the catheter in place. However,

if the patient is in septic shock, the long-term catheter should

be removed immediately. Vancomycin in conjunction with a

third-generation cephalosporin or aminoglycoside is appropri-

ate initial therapy. Definitive therapy should be guided by the

results of the blood culture. Patients should be treated for a

total of 2–3 weeks. Catheter-tip sepsis in certain patient popu-

lations mandates immediate removal of a long-term catheter.

These include patients with neutropenia, severe immunocom-

promise, persistent bacteremia or sepsis despite appropriate

antibiotics, septic thrombosis, and those whose catheters are

infected with S. aureus, Candida or other fungi, or antibiotic-

resistant bacteria.

Management decisions must be based on the clinical cir-

cumstance. The intravascular device should be removed

immediately in any patient in whom rapid resolution of the

signs of infection does not occur. It should be emphasized

that neutropenic patients may not develop the classic local

signs of infection; therefore, the diagnosis of catheter-

associated infection in these patients may be difficult.

Finally, the risk of catheter-related infection may be signifi-

cantly reduced by appropriate care of the catheter and by

observing strict hand washing, dressing, and glove precautions.

Current Controversies and Unresolved Issues

A. Prevention of Catheter-Related Infection—The

challenge of preventing intravenous catheter infections in

the ICU while providing reliable access for intravenous



CHAPTER 15

386

antimicrobial therapy, total parenteral nutrition, hemody-

namic monitoring, or temporary hemodialysis is daunting.

Many recommendations have been put forward for preven-

tion of catheter-related infections, but only a few are sup-

ported by data. The most important are (1) use of full

sterile barrier precautions during insertion; (2) skin decon-

tamination with chlorhexidine; and (3) routine inspection

and care of the catheter site.

Recently, the use of antibiotic-impregnated catheters has

been proposed in an attempt to decrease the incidence of

catheter-site colonization and subsequent bloodstream

infection. A meta-analysis suggested that central venous

catheters impregnated with chlorhexidine–silver sulfadiazine

were effective in decreasing catheter-related bacteremia in

patients at high risk for infection (ie, patients who require

short-term catheters and multilumen catheters). Minocycline-

rifamipin-impregnanted central venous catheters (CVCs)

appear to have a lower rate of catheter-related bacteremia.

Second-generation antiseptic catheters reduced colonization

of bacteria on the catheters at the time of removal. The

impact of these catheters on the development of subsequent

infection is not as clear.

The cost associated with catheter-tip sepsis has been esti-

mated to be about $6000 per bacteremic episode. Adopting

approaches to decrease catheter infection will reduce mor-

bidity and the overall cost of care.

B. Diagnosis of Catheter-Related Bacteremia—The lab-

oratory diagnosis of catheter-related bacteremia continues to

be difficult. Semiquantitative culture of the external surface

of the catheter tip is the method used most often to detect

catheter colonization and device-related bacteremia.

However, this method may fail to detect significant coloniza-

tion of the internal lumens of catheters. Several quantitative

methods have been described, including time to positivity

and quantitative blood cultures drawn through the catheter

(compared with peripheral blood cultures).

C. Duration of Intravascular Catheter Use—Recent studies

have attempted to clarify the utility of routine change of the

catheter site and the safety of exchanging a central venous

catheter over a guidewire. Pulmonary artery catheters—despite

the need for manipulation of the catheter during measure-

ments—can be left in place unless there is unexplained fever,

positive blood culture, or evidence of skin-site infection.

However, maintenance of a catheter at a given site for a maxi-

mum of 7 days would appear to be a reasonable limit. Central

venous catheters may be left in place without a predetermined

duration of placement; they should be removed in the presence

of a positive blood culture, an infected insertion site, suspicion

of catheter-related infection, or when there is no further need

for the catheter. It does not appear that routine replacement of

central venous catheters is indicated.

Microorganisms most commonly cause catheter-related

infection by migrating along the outer site of the catheter

from the skin surface to the tip. Nevertheless the, exchange of

a central venous or pulmonary artery catheter at the same

site using a guidewire has become accepted practice in many

ICUs if no evidence of catheter infection is present.

Catton JA et al: In situ diagnosis of intravascular catheter-related

bloodstream infection: A comparison of quantitative culture,

differential time to positivity, and endoluminal brushing. Crit

Care Med 2005;33:787–91. [PMID: 15818106]

Eggimann P et al: Long-term reduction of vascular access-

associated bloodstream infection. Ann Intern Med 2005;142:

875–6. [PMID: 15897546]

O’Grady NP et al: Guidelines for the prevention of intravascular

catheter-related infections. Centers for Disease Control and

Prevention. MMWR Recomm Rep 2002;51:1–29. [PMID:

12233868]

Pronovost P et al: An intervention to decrease catheter-related

bloodstream infections in the ICU. N Engl J Med 2006;355:

2725–32. [PMID: 17192537]

Rupp ME et al: Effect of a second-generation venous catheter

impregnated with chlorhexidine and silver sulfadiazine on cen-

tral catheter-related infections: A randomized, controlled trial.

Ann Intern Med 2005;143:570–80. [PMID: 16230723]



Clostridium Difficile

–Associated Diarrhea

ESSENTIALS OF DIAGNOSIS



Watery diarrhea and low-grade fever.



Previous or current treatment with antibiotics—but may

occur as long as 6 weeks after antibiotic has been

stopped.



Presence of C. difficile toxin in stool.



Sigmoidoscopy may reveal white or yellow plaques of

pseudomembranous colitis.

General Considerations

Diarrhea developing in an ICU patient can be due to infec-

tious or noninfectious causes. The most important infectious

cause of nosocomial diarrhea is infection with C. difficile.

Nosocomial gastroenteritis caused by Salmonella, Shigella, E.

coli, and Campylobacter is exceedingly rare and should be

considered only in outbreak situations.

At least two things must occur for a patient to develop C.

difficile–associated diarrhea: acquisition of or colonization

with C. difficile and administration of an inducing agent. More

than 20% of patients hospitalized for over a week become col-

onized with C. difficile. The risk of acquisition of C. difficile

increases with length of hospital stay, reaching 50% in patients

hospitalized over 1 month. It appears that the administration

of certain antibiotics or chemotherapeutic agents induce the

organisms to elaborate toxins: toxin A is an enterotoxin, and

toxin B is a cytotoxin. The spectrum of disease encountered