A surgical infection is an infection that (1) is unlikely to respond to nonsurgical treatment (it usually must be excised or
drained) and occupies an unvascularized space in tissue or (2) occurs
in an operated site. Common examples of the first group are appendicitis,
empyema, gas gangrene, and most abscesses.
Surgeons are regrettably familiar with the vicious circle of operation or injury, infection, malnutrition, immunosuppression, organ failure,
reoperation, further malnutrition, and further infection. One of the fine arts of surgery is to know when to intervene with excision, drainage, physiologic support, antibiotic therapy, and nutritional
therapy. For infections arising in a space or in dead tissue, by far the most important aspect of treatment is to establish surgical drainage.
Three elements are common to surgical infections: (1) an infectious agent, (2) a susceptible host, and (3) a closed, unperfused space.
Although a few pathogens cause most surgical infections, many
organisms are capable of doing so. Among the aerobic organisms,
streptococci may invade even minor breaks in the skin and spread through
connective tissue planes and lymphatics. Staphylococcus
aureus is the most common pathogen in wound infections
and around foreign bodies. Klebsiella often invades the inner ear and
enteric tissues as well as the lung. Enteric organisms, especially
the Enterobacteriaceae and enterococci, are often found together
with anaerobes. Among the anaerobes, bacteroides species and Peptostreptococci are often present in surgical infections, and clostridium species are
major pathogens in ischemic tissue.
Pseudomonas and serratia are usually nonpathogenic surface contaminants
but may be opportunistic and even lethal invaders in critically
ill or immunosuppressed patients. Some fungi (histoplasma, coccidioides)
and yeasts (candida), along with nocardia and actinomyces, cause
abscesses and sinus tracts, and even animal parasites (amebas and
echinococcus) may cause abscesses, especially in the liver. Destructive
granulomas, such as tuberculosis, once required excision, but antibiotic
therapy has now superseded operation for this purpose in most cases. Other
rare diseases such as cat-scratch fever, psittacosis, and tularemia
may cause suppurative lymphadenitis and require drainage or excision.
Identification of the pathogen by smear and culture remains a
cardinal step in therapeutic decision making. The surgeon must inform
the microbiologist of peculiar circumstances associated with any
given specimen, so that appropriate smears and cultures can be done;
serious errors may otherwise result.
Surgical infections such as appendicitis and furuncles occur
in patients whose only defect in immunity is a closed space in tissue.
However, patients with suppressed immune systems are being seen with
increasing frequency, and their problems have become a major surgical challenge. Immunosuppression seems
a simple concept but in fact usually represents a combination of
defects of the multifaceted immune mechanism.
The immune process that depends upon prior exposure to an antigen
involves detection and processing of antigen by macrophages, mobilization
of T and B lymphocytes, synthesis of specific antibody, and other
functions. Its importance is illustrated in AIDS, transplant immunosuppression,
and agammaglobulinemia, each of which is associated with only a slight
increase in the frequency and severity of some surgical infections.
In general, isolated defects contribute little to the severity of
ordinary surgical infections. Major defects contribute substantially
to morbidity, mortality, and resource consumption.
Innate or nonspecific immunity serves to limit damage during
the first few hours after infection. Despite the emphasis in the
literature on specific immune mechanisms, nonspecific immunity,
which depends on phagocytic leukocyte migration, ingestion, and
cidal activity for microorganisms, is the principal means by which
the host defends against abscess-forming and necrotizing infections.
Chemoattraction and Phagocytosis
Invading microbes display molecular patterns that are shared
among groups of pathogens. Examples include lipopolysaccharides (LPS)
of gram-negative bacteria, lipoteichoic acid of gram-positive bacteria,
mannans of yeast, and double-stranded RNA of certain viruses. To
control infection, the host uses an array of pattern recognition
receptors (complement, adhesins, collectins, bactericidal permeability-increasing
protein [BPI], LPS-binding protein [LBP])
that bind to these molecular moieties, acting together with effector
cells to eliminate them. Typically, granulocytes internalize these
pattern-receptor complexes by engulfment into a phagocytic vacuole. The
subsequent release of chemoattractants causes the movement (diapedesis)
of leukocytes from the bloodstream to the tissue, increasing leukocyte
numbers locally and the likelihood that the invading microbe will
be destroyed. These steps require little or no oxygen, but chemotaxis
is vulnerable to a number of disorders, particularly anti-inflammatory
steroid hormones and malnutrition, which reduce the number of granulocytes
that arrive at a contaminated site in a given time.
Once the phagosome is formed, other cytoplasmic granules (lysosomes)
fuse with it and release into it preformed and increasingly acidic
proteolytic solutions that kill most bacteria and fungi.
A second process, “oxidative killing,” is particularly
important to the killing of organisms such as staphylococci, which are
commonly responsible for surgical infections. This mechanism consumes and
requires molecular oxygen, which it converts to superoxide anion.
In this process, a membrane-bound NADPH oxidase is activated, and
a burst of respiration (oxygen consumption) follows. Part of the
consumed oxygen is converted to a series of oxygen radicals (including
superoxide, hydroxyl radical, and hypochlorite), which are released into
phagosomes and assist in bacterial killing. This process is progressively inhibited
when extracellular oxygen tension falls below about 30 mm Hg. When oxygen
tension is 0 mm Hg, the antibacterial capacity of normal granulocytes
for S aureus and E coli, for instance,
falls by half—to the same capacity observed in granulocytes
taken from victims of chronic granulomatous disease, which results
from the genetic absence of membrane-bound oxidase and which without
aggressive antibiotic therapy is lethal in early childhood.
Whether a given inoculum will establish an infection and become
invasive depends to a great extent on how well tissue perfusion—and
therefore oxygenation—can meet the increased metabolic demands
of the granulocytes. Inflammatory signals from complement factors and
histamine, for instance, dilate vessels and help direct blood flow
to infected areas, but if blood volume or regional vascular supply
is so poor that tissue perfusion cannot increase, invasive infection
ensues. Tissue oxygen supplies can often be raised by increasing
blood volume and arterial Po2 and are lowered
by hypovolemia and pulmonary insufficiency.
Patients with pulmonary disease, severe trauma, congestive heart
failure, hypovolemia, or excessive levels of vasopressin, angiotensin,
or catecholamines have hypoxic peripheral tissues and are unusually
susceptible to infection. They are truly immunosuppressed. Support
of the circulation is just as important to immune defense as is
nutrition or antibiotic therapy.
Anergy is defined as the lack of inflammatory response to skin
test antigens. It characterizes a population of immunosuppressed
patients who tend to develop infections and die from them. The skin
tests used to diagnose anergy are those often used to test recall
antigens and delayed hypersensitivity—but in fact they
test much of the spectrum of antibacterial immunologic events, including
antigen detection and processing by macrophages, release of lymphokines,
antibody synthesis, and the inflammatory response, including leukocyte chemotaxis.
One event they do not detect is, conspicuously, the final crucial
step of actually killing bacteria. Anergy has many causes, including
defective T and B lymphocytes, the presence of excess anti-inflammatory
corticosteroids, defective antigen processing, and increased numbers
of suppressor T cells. Among surgical patients, severe malnutrition,
trauma, shock, and sepsis suppress skin test responses, which become active
again after resolution of the acute process.
Immunity in Diabetes Mellitus
Diabetes mellitus impairs immunity. Well-controlled diabetics
resist infection normally except in tissues made ischemic by arterial
disease, while uncontrolled diabetics do not. The mechanism is unknown,
except that leukocytes from poorly controlled diabetics adhere,
migrate, and kill bacteria poorly. They improve their performance
when glucose control is regained. Leukocytes also function poorly
without insulin, and insulin is consumed in wounds and other poorly
perfused spaces, resulting in low ambient insulin levels.
Most surgical infections start in a susceptible, usually poorly
vascularized place in tissue such as a wound or a natural space.
The common denominators are poor perfusion, local hypoxia, hypercapnia,
and acidosis. Some natural spaces with narrow outlets, such as those
of the appendix, gallbladder, ureters, and intestines, are especially
prone to becoming obstructed and then infected.
The peritoneal and pleural cavities are potential spaces, and
their surfaces slide over one another, thereby dispersing contaminating
bacteria. Foreign bodies, dead tissue, and injuries interfere with
this mechanism and predispose to infection. Fibrin inhibits the
clearing of bacteria. It polymerizes around bacteria, trapping them;
this encourages abscess formation but at the same time prevents
dangerous spread of infection.
Foreign bodies may have spaces in which bacteria can reside.
Infarcted tissue is markedly susceptible to infection. Thrombosed
veins, for example, rarely become infected unless intravenous catheters
enter them and act as entry points for bacteria.
Spread of Surgical Infections
Surgical infections usually originate as a single focus and become
life threatening by spreading and releasing toxins. Spreading occurs
by several mechanisms.
Necrotizing infections tend to spread along anatomically defined
paths. Necrotizing fasciitis spreads along poorly perfused fascial
and subcutaneous planes, its toxins causing thrombosis even of large
vessels ahead of the necrotic area, thus creating more ischemic
and vulnerable tissue.
If not promptly drained, abscesses enlarge, killing more tissue
in the process. Leukocytes contribute to necrosis by releasing lysosomal
enzymes during phagocytosis. Natural boundaries can be breached;
eg, intestinal cutaneous fistulas may form, or blood vessel walls
may be penetrated.
Phlegmons and Superficial Infections
Phlegmons contain little pus but much edema. They spread along
fat planes and by contiguous necrosis, combining features of both
of the above kinds of spread. Retroperitoneal peripancreatic inflammation
or infection is typical. Superficial infections may spread along skin
not only by contiguous necrosis but also by metastasis.
Spread of Infection Via the Lymphatic System
Lymphangiitis produces red streaks in the skin and travels proximally
along major lymph vessels. However, it may also occur in hidden
places such as the retroperitoneum in puerperal sepsis.
Spread of Infection Via the Bloodstream
Empyema and endocarditis caused by intravenously injected contaminated
recreational drugs are now common. Brain abscesses resulting from
infections elsewhere in the body (especially the face) occur in
infants and diabetics. Liver abscesses may complicate appendicitis and
inflammatory bowel disease, sometimes as a result of suppurative
phlebitis of the portal vein (pylephlebitis).
Fistulas and Sinus Tracts
Fistulas and sinus tracts often result when abdominal abscesses
contiguous to bowel open to the skin. When tissue necrosis compounds
the development of sinus tracts and erodes major blood vessels,
severe bleeding may occur. This is most troublesome in irradiated
tissue of nonhealing neck wounds and in infected groin wounds after
Some intestinal fistulas originate in poorly fashioned or necrotic
suture lines, and some result from contiguous abscesses that eventually
penetrate both bowel and skin, often helped along by the surgeon
who must drain the abscess.
Suppressed wound healing is a consequence of infection. The mechanism
is probably stimulation by bacteria of cytokines, which in turn
stimulates proteolysis, especially collagenase production.
Immunosuppression and Superinfection
Immunosuppression is a common consequence of injury, which includes
surgery, trauma, shock, or infection or sepsis. Superinfection occurs
when immunosuppression provides an opportunity for invasion by opportunistic,
often antibiotic-resistant organisms.
Bacteremia is the presence of bacteria in blood. The significance
of bacteremia is variable. Bacteremia that follows dental work is usually rapidly cleared and harmless, except in patients with damaged heart valves; cardiac, vascular, or orthopedic prostheses; or impaired immunity. It occurs predictably during instrumentation of the gastrointestinal
tract or infected urinary tract. Patients in these groups are at increased risk and should receive an appropriate prophylactic antibiotic regimen.
Organ Dysfunction, Sepsis, and the Systemic Inflammatory Response Syndrome
Infection and tissue damage initiate the inflammatory response,
a very tightly controlled, adaptive response to eliminate dead or
infected tissue. At the site of injury, endothelial cells and leukocytes
coordinate the local release of mediators of the inflammatory response,
including cytokines (tumor necrosis factor-α),
interleukins, interferons, leukotrienes, prostaglandins, nitric
oxide, reactive oxygen species, and products of the classic inflammatory
pathway (complement, histamine, and bradykinin) (Table 8–1). When localized to diseased tissue, these mediators are highly effective
at recruiting and arming cells of the innate and adaptive immune
systems to destroy invading organisms and elicit reparative mechanisms
in wounded tissue. However, if the degree of the infectious or traumatic
insult exceeds the ability of the host to contain it, the inflammatory response
becomes systemic. The result is whole-body activation of the inflammatory
response, with resultant disruption of normal cellular metabolism
and microcirculatory perfusion. This leads to clinical deterioration,
manifested as dysfunction of the brain (delirium), lungs (hypoxia), heart
and blood vessels (shock and edema), kidneys (oliguria), intestines (ileus),
liver (hyperbilirubinemia), and the hematologic (coagulopathy, anemia)
and immunologic systems (immunosuppression). This syndrome is referred
multiple organ dysfunction syndrome
). The risks of organ failure in general are directly proportionate
to the duration and severity of shock and inversely proportionate to the age and underlying health of the patient. It is frequently difficult or impossible to determine whether the cause of organ dysfunction
in critically ill patients is severe infection or inflammation.
The term sepsis is used when the systemic response
results from infection. In contrast, when the systemic response occurs
in the absence of infection, as it does in severe burns, trauma,
and pancreatitis, it is called systemic inflammatory response
syndrome (SIRS). The interrelationships among
infection, bacteremia, sepsis, and SIRS are depicted in Figure 8–1.
Table 8–1. Cytokines and Growth Factors. |Favorite Table|Download (.pdf)
Table 8–1. Cytokines and Growth Factors.
|Peptide||Site of Synthesis||Regulation||Target Cells||Effects|
|G-CSF||Fibroblasts, monocytes||Induced by IL-1, LPS, IFN-α||Committed neutrophil progenitors (CFU-G, Gran)||Supports the proliferation of neutrophil-forming colonies.
Stimulates respiratory burst.|
|GM-CSF (IL-3 has almost identical effects)||Endothelial cells, fibroblasts, macrophages, T lymphocytes,
bone marrow||Induced by IL-1, TNF||
cells (CFU-GEMM, CFU-MEG, CFU-Eo, CFU-GM)||Supports the proliferation of macrophage-, eosinophil-,
neutrophil-, and monocyte-containing colonies.|
|IFN-α, IFN-β, IFN-γ||
Epithelial cells, fibroblasts, lymphocytes, macrophages, neutrophils
||Induced by viruses (foreign nucleic acids), microbes, microbial
foreign antigens, cancer cells||Lymphocytes, macrophages, infected cells, cancer cells||Inhibits viral multiplication. Activates defective phagocytes,
direct inhibition of cancer cell multiplication, activation of killer
leukocytes, inhibition of collagen synthesis.|
|IL-1||Endothelial cells, keratinocytes, lymphocytes, macrophages||Induced by TNF-α, IL-1, IL-2, C5a; suppressed
by IL-4, TGF-β||Monocytes, macrophages, T cells, B cells, NK cells, LAK cells||Stimulates T cells, B cells, NK cells, LAK cells. Induces
tumoricidal activity and production to other cytokines, endogenous
pyrogen (via PGE2
release). Induces steroidogenesis, acute
phase proteins, hypotension; chemotactic neutrophils. Stimulates
|IL-1ra||Monocytes||Induced by GM-CSF, LPS, IgG||Blocks type 1 IL-1 receptors on T cells, fibroblasts, chondrocytes,
endothelial cells||Blocks type 1 IL-1 receptors on T cells, chondrocytes, endothelial
cells. Ameliorates animal models of arthritis, septic shock, and
inflammatory bowel disease.|
|IL-2||Lymphocytes||Induced by IL-1, IL-6||T cells, NK cells, B cells, activated monocytes||Stimulates growth of T cells, NK cells, and B cells|
|IL-4||T cells, NK cells, mast cells||Induced by cell activation, IL-1||All hematopoietic cells and many others express receptors||Stimulates B cell and T cell growth. Induces HLA class II
|IL-6||Endothelial cells, fibroblasts, lymphocytes, some tumors||Induced by IL-1, TNF-α||T cells, B cells, plasma cells, keratinocytes, hepatocytes,
stem cells||B cell differentiation. Induction of acute phase proteins,
growth of keratinocytes. Stimulates growth of T cells and hematopoietic
|IL-8||Endothelial cells, fibroblasts, lymphocytes, monocytes||Induced by TNF, IL-1, LPS, cell adherence (monocytes)||Basophils, neutrophils, T cells||Induces expression of endothelial cell LECAM-1 receptors, β2 integrins,
and neutrophil transmigration. Stimulates respiratory burst.|
|M-CSF||Endothelial cells, fibroblasts, monocytes||Induced by IL-1, LPS, IFN-α ||Committed monocyte progenitors (CFU-M, mono)||Supports the proliferation of monocyte-forming colonies.
|MCP-1, MCAF||Monocytes; some tumors secrete a similar peptide||Induced by IL-1, LPS, PHA||Unstimulated monocytes||Chemoattractant specific for monocytes|
|TNF-α (LT has almost identical effects)||Macrophages, NK cells, T cells, transformed cell lines, B
cells (LT)||Suppressed by PGE2
, TGF-β, IL-4;
induced by LPS||Endothelial cells, monocytes, neutrophils||Stimulates T cell growth. Direct cytotoxin to some tumor
cells. Profound proinflammatory effect via induction of IL-1 and
. systemic administration produces many symptoms of
sepsis. Stimulates respiratory burst and phagocytosis.|
Interrelationships among systemic inflammatory response syndrome (SIRS), sepsis, and infection. (Modified, with permission, from Crit Care Med 1992;20:864.)
The aim of management is to detect and treat sepsis before it evolves into more advanced stages.
Physical examination is the easiest way to localize a surgical
infection. When infection is suspected but cannot be identified
initially, repeated examination will often reveal subtle warmth, erythema,
induration, tenderness, or splinting due to a developing abscess. Failure
to repeat the physical examination is the most common reason for delayed
diagnosis and therapy.
Laboratory data are of limited value. Leukocytosis may give way
to leukopenia when the infection is severe. Acidosis is helpful
in diagnosis, and signs of disseminated intravascular coagulation
are useful as well. Otherwise unexplained respiratory, hepatic,
renal, and gastric (ie, stress ulcers) failure is strong evidence
Positive cultures help to differentiate SIRS from sepsis even
though 50% of cases of sepsis are culture-negative. If infection
is suspected, cultures of blood, sputum, and urine are collected
routinely initially, especially in hospitalized patients given the high frequency of nosocomial
pneumonia and urinary tract infections (see below). This is particularly
important because data from the Centers for Disease Control and
Prevention (CDC) suggest that 70% of the bacteria causing
hospital-associated infections are resistant to at least one of the
drugs most commonly used to treat them. Other fluids, such as cerebrospinal
fluid, pleural and joint effusions, and ascites, can be aspirated
and cultured on the basis of signs or symptoms that specifically
indicate these sites as potential sources of infection. In general, pus
from abscesses should be cultured unless the causative organism
is known. In rapidly advancing cases, two separate blood cultures
should be taken within 15 minutes. In less urgent situations, cultures
should be taken over a 24-hour period, and up to six cultures should
be taken if the patient has enigmatic fevers and either a cardiac
or joint prosthesis or vascular shunt. False-negative blood culture
results occur in about 20% of cases. False-positive results
are difficult to define, since skin commensals (even some diphtheroids
and Staphylococcus epidermidis), regarded as contaminants
in the past, have proved occasionally to be true pathogens. Arterial
blood cultures may be necessary to detect fungal endocarditis.
Radiologic examination is frequently helpful, particularly for
the diagnosis of pulmonary infections. Whenever infection is close
to bone, radiologic examination is indicated to detect early signs
of osteomyelitis, which might require more aggressive surgical or
antibiotic therapy. MRI imaging is most useful in detecting bone
edema, an early sign of osteomyelitis. For detecting abscesses in
solid organs, CT scanning is useful. CT scanning and ultrasonography
are particularly useful in localizing occult infection.
Numerous radionuclide scans have been tested, all with fair results.
The best radionuclides for labeling leukocytes are gallium (67Ga)
and indium (111In). Nuclear imaging modalities are rarely used
today for localization of infection.
An early diagnosis of sepsis is usually based on a combination
of suspicion and inconclusive evidence, since the results of blood
cultures are often unavailable during this stage. An important initial
step is to identify the source. Surgical or traumatic wounds, surgical
infections in the abdomen or thorax, and clostridial infections
are all common, but so are urinary tract infections, pneumonia,
and even sinus infections. Once identified, any septic focus amenable
to surgical therapy should be excised or drained.
Abscesses must be opened and bacteria, necrotic tissue, and toxins
drained to the outside. The pressure and the number of bacteria
in the infected space are lowered; this decreases the spread of toxins
and bacteria. An abscess with systemic manifestations is a surgical
Fluctuation is a reliable but late sign of a subcutaneous abscess.
Abscesses in the parotid or perianal area may never become fluctuant,
and if the surgeon waits for this sign, serious sepsis may result.
Drainage creates an open wound, but the tissue will heal by second
intention with remarkably little scarring. Deep abscesses difficult
to drain surgically may be drained by a catheter placed percutaneously
under guidance by CT scanning or ultrasonography.
It may appear that a patient with sepsis cannot withstand operation.
In fact, operation to drain an abscess may be the most important
of all therapeutic measures. One can hardly imagine delaying removal
of infarcted bowel because the patient is in shock. There is no
substitute for obliteration of the focus of infection when it is
Some surgical infections may be excised (eg, an infected appendix
or gallbladder). In these cases, drainage may not be necessary,
and the patient is cured on the operating table. Clostridial myositis may
require amputation of the infected limb. The success of such operations
is greatly facilitated by intensive specific adjuvant antimicrobial
Just as infections due to vascular ischemia are cured by restoring
arterial patency, chronic infections in poorly vascularized areas,
as in osteoradionecrosis, may be cured by transplanting a functioning
vascular bed (eg, a musculocutaneous flap or omental transposition) into
the affected area.
Antimicrobial agents are not necessary for simple surgical infections
that respond to incision and drainage alone—furuncles and
uncomplicated wound infections. Infections likely to spread or persist
require antimicrobial therapy, best chosen on the basis of therapy
targeted to the evidence of pathogen(s) via cultures and sensitivity
tests. In “toxic” infections, including septic
shock, antimicrobial therapy must be started promptly; empiric regimens
can be modified later based on procured specimen results. The preemptive
or empiric choice of drugs must take into account the organisms
most often cultured from similar infections in previous patients,
the results of body fluid Gram stains, and specific characteristics
of the patient.
In malnourished, septic, or severely traumatized patients, the
ability to ward off or recover from infection is often enhanced by
aggressive nutritional therapy. Specific measurable effects include improved
immunocompetency and blunting or reversal of catabolism. Protection or
restoration of visceral and skeletal muscle allows the patient to
cough better and be more mobile.
The mortality rate ranges from 10% in septic patients
with manifestations limited to fever, chills, and toxicity, to almost 100% in
those who manifest shock and multiple organ failure. Factors that
have independent influences on outcome include the causative microorganism,
blood pressure, body temperature (inverse relationship), primary
site of infection, age, predisposing factors, and place of acquisition
of infection (hospital or home). Of patients with low-grade fever
and an elevated leukocyte count after antibiotics have been discontinued,
60% will have a relapse. Nevertheless, continuation of antibiotics
in questionable cases is often contraindicated because it only delays
recognition of infection and may enhance morbidity as well as increase
Bone RC: Sir Isaac Newton, sepsis, SIRS, and CARS.
Crit Care Med 1996;24:1125.
Bone RC et al: Definitions for sepsis and organ failure and
guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM
Consensus Conference Committee: American College of Chest Physicians/Society
of Critical Care Medicine. Chest 1992;101:1644.
Lederer JA et al: The effects of injury on the adaptive immune
response. Shock 1999;11:153.
O’Grady NP et al: Practice parameters for evaluating
new fever in critically ill adult patients. Task Force of the American
College of Critical Care Medicine of the Society of Critical Care
Medicine in collaboration with the Infectious Disease Society of
America. Crit Care Med 1998;26:392.
Nosocomial Infections & Infection Control
Nosocomial infections affect approximately 2 million patients
annually in the United States and add approximately $3.5
billion to the cost of health care. Patients may acquire infection
in hospital through contact with personnel or from a nonsterile
environment, or infection may develop from bacteria harbored by
the patient before operation.
Hospital Personnel as a Source of Infection
Most nosocomially acquired bacteria are transmitted through human
contact. In order to minimize transmission in hospital, rules made
for behavior, dress, and hygiene should be obeyed.
Unwashed hands are by far the most frequent sources of nosocomial
infections such as pneumonia, intravenous catheter-related sepsis,
burn wound infections, and even pseudomembranous colitis. Therefore,
hand washing is the single most important procedure for preventing nosocomial
infections. Routine hand washing should be a matter of reflex conditioning.
In today’s atmosphere, failure to wash one’s hands
between patient contacts in a hospital is essentially an unethical
The Operating Room as a Source of Infection
Any break in operative technique noted by any member of the operating
team should be corrected immediately. Members of the team should
not operate if they have cutaneous infections or upper respiratory
or viral infections that may cause sneezing or coughing.
Scrub suits should be worn only in the operating room and not
in other areas of the hospital. If they must be worn outside the
operating room, they should be changed before reentering. Physicians and
nurses should always wash their hands between patients. Careful
hand washing should follow all contact with infected patients. For
preoperative preparation, hands and forearms up to the elbows should
be scrubbed for 2–5 minutes with any approved agent if
the surgeon has not scrubbed within the past week. Shorter scrubs
are allowable between operations. Traffic and talking in the operating
room should be minimized.
Though many parts of the operating environment are sterile, the
operative field is not—it is merely as sterile as it can
be made. Attempts to achieve a level of sterility beyond normal
standards have not led to further reductions in wound infection
rates. This reflects the fact that bacteria are also present in the patient, and host defense mechanisms are also important determinants of infection not affected by more aggressive attempts to achieve sterility.
Many special and expensive techniques have been devised to minimize
bacterial contamination in the operating room. Ultraviolet light,
laminar flow ventilation, and elaborate architectural and ventilation
schemes have been advocated, but none have been definitively proved
more effective than observation of current infection control guidelines
and surgical discipline.
The only completely reliable methods for sterilization of surgical
instruments and supplies are steam under pressure (autoclaving),
dry heat, and ethylene oxide gas. Saturated steam at 2 atm pressure
and a temperature of 120 °C destroys all vegetative bacteria and
most resistant dry spores in 13 minutes, but exposure of surgical
instrument packs should usually be extended to 30 minutes to allow
heat and moisture to penetrate to the center of the package. Shorter
times are allowable for unwrapped instruments with the vacuum-cycle
or high-pressure autoclaves now widely used. Continuous dry heat
at 170 °C for 1 hour sterilizes articles that cannot tolerate moist
heat. If grease or oil is present on instruments, safe dry-heat sterilization
requires 4 hours at 160 °C.
Gaseous ethylene oxide destroys bacteria, viruses, fungi, and
various spores. It is used for heat-sensitive materials, including
telescopic instruments, plastic and rubber goods, sharp and delicate instruments,
electrical cords, and sealed ampules. It damages certain plastics
and pharmaceuticals. The technique requires a special pressurized-gas
autoclave, with 12% ethylene oxide and 88% Freon-12 at
55 °C, 8 psi pressure above atmospheric pressure. Most items must
be aerated in sterile packages on the shelf for 24–48 hours before use in order to rid them of the dissolved gas. Implanted plastics should be stored for 7 days before use. Ethylene oxide is toxic
and represents a safety hazard unless it is used according to strict regulations.
Miscellaneous sterilization procedures include soaking in antiseptics
such as 2% glutaraldehyde to remove viruses from instruments
with lenses. Total sterilization by this method requires 10 hours. Chemical
antiseptics are often used to clean operating room surfaces and instruments
that need not be totally sterile. Other disinfectant solutions include synthetic
phenolics, polybrominated salicylanilides, iodophors, alcohols,
other glutaraldehyde preparations, and 6% stabilized hydrogen
peroxide. These agents maintain high potency in the presence of
organic matter and usually leave effective residual antibacterial
activity on surfaces. They are also used to clean anesthetic equipment
that cannot be sterilized. Prepackaged instruments and supplies
can be sterilized with gamma radiation by manufacturers. Synthetic fabrics
have now proved to be superior barriers to bacteria and less costly
than the traditional cotton. They can be used in gowns and drapes.
The Patient as a Source of Infection
When possible, preexisting infections should be treated before
operation. Secretions from patients with a history of respiratory
tract infections should be cultured and appropriate treatment given. The
urinary tract should be cultured and specific antibiotics administered
before instruments are introduced; this precaution has eliminated
septic shock as a complication of urologic surgery. The colon should
be prepared as discussed in Chapter 31. Dental
extractions for caries are imperative prior to cardiac valve replacement.
Bacteria on the patient’s skin are a common cause of
infection. Preoperative showers or baths with antiseptic soap reduce
the infection rate in clean wounds by 50%. Shaving of the
operative field hours prior to incision is associated with a 50% increase
in wound infection rates and should not be done. If the patient has
a heavy growth of hair, an area just large enough to accommodate
the wound and its closure should be clipped rather than shaved immediately
before operation. Razor shaving more than a few minutes before operation
raises the wound infection rate.
The skin to be included in the operative field should be cleansed with antiseptic. Nonirritating agents such as benzalkonium salts
should be used in or around the nose or eyes. For other skin areas, the
iodophors (eg, povidone-iodine) and chlorhexidine are used most
Isolation Procedures: Universal Precautions
Traditionally, patients with infection were individually isolated. Since 1985—partly in response to the HIV epidemic—a more general kind of isolation called “universal precautions” has
been substituted. In this system, any procedure
involving close contact with any patient—and
especially those involving contact with blood—is performed
by hospital personnel wearing gloves and other protective devices.
The concept of universal precautions emphasizes (1) prevention of
needlestick injuries, (2) the use of traditional barriers such as
gloves and gowns, (3) the use of masks and eye coverings to prevent
mucous membrane exposure during procedures, and (4) the use of individual
ventilation devices when the need for resuscitation is predictable.
The CDC recommends that universal precautions apply to blood, semen,
and vaginal secretions; to amniotic, cerebrospinal, pericardial,
peritoneal, pleural, and synovial fluids; and to other body fluids
contaminated with blood. Universal precautions are not recommended
for feces, nasal secretions, sputum, sweat, tears, urine, or vomitus unless
they contain visible blood. The need for hand washing is not diminished by
Antibiotic Prophylaxis Against Surgical Infections
Prophylactic use of antibiotics can decrease the incidence of infections, especially surgical site infections, but at the risk
of toxic and allergic reactions to the drug, drug interactions,
bacterial resistance, and superinfection. The principles of antibiotic
prophylaxis are simple: (1) Choose antibiotics effective against
the expected type of contamination. (2) Use antibiotics only if
the risk of infection justifies doing so. (3) Give antibiotics in
appropriate doses and at appropriate times. (4) Stop dosing before the
risk of side effects outweighs benefits.
Antibiotics for preventive use must not be highly toxic and should
not be “first-line” antibiotics for treatment
of established infection. Because resistance to antibiotics may
develop quickly, agents that have been used frequently for prophylaxis
are likely to lose their effectiveness for later treatment. Prophylactic
agents should be chosen for cost-effectiveness and safety as well
as for efficacy.
Prophylactic antibiotics should be selected to target the organisms
most likely to be encountered in the anticipated operative procedure.
A first-generation cephalosporin (eg, cefazolin) is preferred for
most procedures, since it is effective against common gram-positive and
gram-negative bacteria and has a moderately long serum half-life.
The routine use of vancomycin for prophylaxis is discouraged in
light of the emergence of vancomycin-resistant organisms—especially
enterococcus and staphylococcus. Agents with better gram-negative
and anaerobic bacterial activity (eg, cefoxitin, cefotetan) are
preferred for colorectal and gynecologic procedures. A single dose
of antibiotic given 30 minutes prior to making the skin incision
should provide adequate tissue concentrations for most procedures.
Additional doses are advisable for longer procedures (over 4 hours)
or those that require large volumes of resuscitative fluids (larger
volume of distribution). Postoperative doses of prophylactic antibiotics
are usually not necessary; in general, no prophylactic antibiotics
should be given after wound closure. The American Heart Association recommends
that patients with valvular heart disease or prosthetic heart valves receive
antibiotics prior to procedures that result in bacteremia in order
to prevent endocarditis. A similar argument has been made for patients
with indwelling prosthetic joints.
Antibiotic prophylaxis cannot and is not intended to eliminate
bacteria. Use of multiple antibiotics increases the risk of drug
reactions, diminishes effectiveness in the long run by promoting
the emergence of resistant strains, and increases costs. Antibiotics
should be given only when a significant rate of infection is encountered
without them or when the consequences of infection would be disastrous,
as with placement of vascular, cardiac, or joint prostheses.
The surgeon may be tempted to give every patient antibiotics
in order to have an infection-free record, but this strategy is
inappropriate for several reasons: (1) Clean wounds may become infected
with organisms for which prophylactic antibiotics are ineffective.
(2) Resistant organisms will eventually develop, creating a higher
risk of infection within the hospital. (3) The expense and risks
associated with antibiotics (eg, kidney failure, hearing loss, anaphylaxis,
skin rashes, fungal infections, enterocolitis) overshadow the minimal
beneficial effects of using antibiotics in clean cases. The number
of antibiotic-resistant strains has been correlated with the number
of kilograms of antibiotics used in any given hospital.
Control of Infection Within the Hospital
Considering the cost of hospital-associated infections, infection control is a very sound investment. Data indicate that infection
control programs can prevent approximately one-third of nosocomial infections. Consequently, the Joint Commission on Accreditation of Healthcare Organizations in the United States requires each hospital to have an infection
control committee with established infection control procedures.
This multidisciplinary committee establishes rules for isolation
of infected patients and for protection of hospital personnel exposed to
infection, procedures for disposal of materials contaminated by
bacteria, and guidelines for limiting the spread of infection. Infection
control specialists usually record and analyze patterns of infection. Isolates
of bacteria cultured from patients are routinely analyzed for potential
significance to the hospital environment. Attempts are made to determine the
source of “epidemics.” These efforts are coordinated
at the national level by the CDC to monitor and report nosocomial
infection trends. These data are then used to generate recommendations and
guidelines to improve outcomes.
Antimicrobial prophylaxis in surgery. Med Lett Drugs Ther 1999;41:75.