Chapter 69. Surgical Site Infections

Surgical site infections (SSI) are infections that occur in tissues or organs that a surgeon has incised or come into contact with during the course of a surgical procedure or operation. Despite advances in the understanding of risk factors, pathogenesis, and prophylaxis, SSI are still a significant source of morbidity for children who undergo operative procedures. Estimated to occur in 2–6% of children who undergo an operation, SSI in some studies account for up to early one quarter of all nosocomial infections in this age group.1–7 For the individual, an SSI can mean prolongation of the hospital stay, additional surgical interventions, the risk of further complications and, most importantly, unnecessary pain and anxiety. For society, the treatment of these largely preventable complications substantially increases the overall cost of health care, as the cost of treatment for each patient with an SSI increases by an average of approximately 36%.8

It should be noted that nearly all of the studies regarding the microbiology, pathogenesis, prevention, and treatment of SSI have been conducted in adults. Because of the present dearth of studies involving children, we have no choice but to apply the same principles used in adults, with modifications where clinical experience and good judgment dictate.

In the past, subjective terms such as “postoperative wound infection” or “wound abscess” were commonly used to describe SSI and, unless defined very clearly by the author who used them, it was often difficult to understand exactly which disease processes were being discussed. In 1970, the Centers for Disease Control and Prevention (CDC) established the National Nosocomial Infections Surveillance (NNIS) System (now the National Healthcare Safety Network [NHSN]), which eventually helped to more precisely define the various types of SSI.9 (Table 69–1 and Figure 69–1). The use of these definitions has allowed more accurate comparisons between studies and, perhaps more importantly, has provided a basis for standardization of the criteria used in surveillance programs throughout the United States. These programs and subsequent studies that resulted have provided an abundance of useful data over the past 15–20 years and have helped to enhance our understanding of the risk factors and pathogenesis of SSI.

Several factors are used to define SSI, including the timing of the infection relative to the creation of the incision, the extent of local tissue or organ involvement, and the clinical signs that are suggestive or diagnostic of an actual infection. The question of timing is based on the idea that after a certain period of time, an infection can no longer be presumed to be causally related to the surgical event. The CDC defines an SSI as an infection that occurs within 30 days of an operative procedure, or up to 1 year from the time of surgery if an implant (a “nonhuman-derived implantable foreign body”) is placed permanently in a patient at the time of surgery. Although inevitably somewhat arbitrary, this is a consensus based on our current understanding of wound healing and empiric evidence derived from decades of cumulative clinical practice.

Surgical site infections are further described as superficial incisional infections, deep incisional infections, and organ/space infections. Superficial incisional infections include tissue infections (cellulitis) and organized purulent infections (abscess) that involve the skin and subcutaneous fat down to but not including the level of the muscle fascia. Deep incisional infections are those that involve one or more layers of muscle or fascia down to the structure that defines the underlying organ space (e.g., peritoneum, pleura, periosteum, or dura). Organ/space infections include infections that involve the underlying body cavity or an organ within that space (Table 69–2).

The NNIS also defines SSI on the basis of specific clinical criteria: (1) the presence of purulent material either draining from the wound or identified by surgical or radiographic means; (2) a positive wound culture; (3) clinical signs or symptoms suggestive of an infection in a wound that has been deliberately opened; or (4) the clinical judgment of a surgeon or attending physician that an infection is present. The criteria therefore vary somewhat depending on the depth of tissues involved and the clinical circumstances. Nevertheless, this practical and reproducible paradigm is used extensively as the standard for defining surgical site infections in the United States and increasingly throughout the world.

Not all infections that occur after a surgical procedure are reportable as SSI. Important exceptions include stitch abscess, which is the result of a very small, localized infection or inflammatory reaction to suture material in the skin, and circumcision, episiotomy, and burn infections, which are not classified as operative procedures by the NNIS.10

The organisms most commonly isolated from infected surgical sites have been remarkably consistent over the past several decades, with Staphylococcus aureus, coagulase-negative staphylococci, Enterococcus spp., and Escherichia coli being the most common.11,12 The distribution of less common isolates usually reflect the proximity of the surgical wound to organs or body sites that are known to be colonized or contaminated with certain microbes. Incisions of the head and neck, for example, are more likely to become infected with nose and mouth flora (Eikenella corrodens, Morraxella catarrhalis, and Peptostreptoccus spp.) while after colorectal surgery, incisions are more likely to become infected with Enterobacteriaceae and Bacteroides spp. Another recent trend is an increase in infections caused by antimicrobial-resistant organisms, especially methicillin-resistant S. aureus (MRSA)13,14 and Candida albicans.15 The increase in infections due to MRSA is a reflection of the well-documented increase in community-acquired soft-tissue infections caused by this virulent pathogen, the emergence of which is thought to be due to the widespread (and perhaps indiscriminate) use of broad-spectrum antibiotics. Finally, implants such as indwelling central venous catheters may become infected with common skin flora such as Staphylococcus epidermidis. Under normal circumstances, these organisms are fairly innocuous but in the presence of a foreign body their virulence is clearly enhanced.

Occasionally, a very unusual organism is isolated from a surgical site infection. When these occur in two or more patients in the form of a local outbreak, it is important to recognize and eliminate the source of contamination and to identify others who might be at risk. Examples include a cluster of postoperative Aspergillus infections that were traced to a contaminated air-handling system in a single operating theater in Texas,16 and an outbreak of Pseudomonas aeruginosa sternal wound infections traced to a surgeon with onychomycosis.17 These miniepidemics underscore the importance of systematic and honest reporting of SSI data and the usefulness of ongoing hospital infection control procedures.

The likelihood of developing a surgical site infection depends on a combination of microbial features and host factors.18–20 The combination of an adequate number of microorganisms of sufficient virulence and a susceptible host under the appropriate local wound conditions is necessary and sufficient to produce a surgical site infection. Certain local factors are known to increase the incidence of infection, in particular the presence of a foreign body such as suture material or a drain.21 Likewise, the risk of infection is higher if the host is immunocompromised. The augmented virulence of particular pathogens has been attributed to their ability to evade local host defenses, usually by the secretion of toxins or local tissue proteases. For example, some strains of coagulase-negative staphylococci produce an extracellular polysaccharide (slime) that create an effective barrier against host defenses and antimicrobial agents,22 while some Bacteroides species secrete short chain fatty acids, which impair neutrophil function, and a procoagulant, which inhibits bacterial clearance.23

The vast majority of SSI are caused by microorganisms that are part of the patient's skin flora or that colonize the mucosa of a hollow organ that is close to the wound or is violated during the operation.24 Pathogens may also be introduced by contaminated surgical instruments, operating room personnel, or frank contamination with stool, secretions, or pus. Although it is assumed that most infections are caused by organisms that are introduced into the wound at the time of the operation, there are clearly some cases in which the contamination occurs in the perioperative period. For example, there are published reports of clusters of infections caused by contaminated surgical bandages25 and contaminated tap water.26 In addition, it is standard surgical practice is to take special precautions when incisions are located near a colostomy or near the anus, where significant postoperative wound contamination is to be expected. Nevertheless, SSI are generally thought to be due to microorganisms that are present before the patient leaves the operating theater.

Host factors that are important in the development of an SSI include systemic factors as well as wound factors that create a local environment conducive to the growth of microorganisms. Patients who are immunodeficient or who are administered drugs that compromise the immune system such as corticosteroids have a higher incidence of SSI and are more susceptible to opportunistic pathogens.27 Conditions associated with poor microvascular blood flow or diminished oxygen delivery such as diabetes mellitus, shock, or nicotine abuse may also increase the risk of infection.9 It was often taught in the past that occlusive surgical dressings should be avoided and that the dressing should be removed on the second postsurgical day to avoid wound infection, however it has since been shown that occlusive dressings pose no risk for infection and may in fact decrease the risk of SSI.28

Risk factor assessment for SSI are complicated by the fact that practical risk factors that a clinician or surgeon might use to identify patients at increased risk do not necessarily satisfy the criteria that an epidemiologist would use to define a risk factor in the strictest sense, namely a factor that is an independent predictor of infection based on rigorous multivariate analysis. Nevertheless, in adults, there are numerous local and systemic factors that have been identified as risk factors for the development of SSI. These serve as the basis for preventative measures, many of which have become the standard of patient care in developed countries. Unfortunately, there are still very little data available regarding risk factor assessment in children.

One of the most important and reliable instruments for the assessment of infection risk is the operative wound classification system as defined by the National Research Council (Table 69–3). It is a practical estimate of the degree of contamination of the surgical wound and historically has correlated consistently with the incidence of postoperative wound infection.29,30 There are four classes: clean, clean contaminated, contaminated, and dirty-infected. Clean wounds result from incisions that are made through intact, uninfected skin, in which the gastrointestinal, respiratory, or genitourinary tract has not been entered, and there has been no break in aseptic technique. Clean-contaminated wounds are those that are associated with entry into the gastrointestinal, respiratory, or genitourinary tract in the absence of infection, or those in which there is more than a minor break in aseptic technique. Wounds that are closed with a passive drain in place and nonperforated appendectomy incisions are also considered to be clean-contaminated. Contaminated wounds are those that are the result of a recent traumatic injury, those in which the incised skin is inflamed but not infected, and those in which there has been a major break in aseptic technique. Incisions exposed to gross gastrointestinal tract spillage or infected bile or urine are also considered contaminated. Dirty-infected wounds are those resulting from neglected traumatic wounds and contain devitalized tissue, foreign material, or fecal matter, and those in which the incised skin is infected. These are technically not at risk for getting infected but are considered already infected. All current wound surveillance programs utilize this wound classification system and documentation of the wound classification is required by major hospital accreditation organizations and most standard operating room protocols.

In addition to the degree of wound contamination, various host factors, mechanical factors, and interventions that increase the risk of SSI have been identified. Host factors include the following: (1) Blood glucose—patients with poorly controlled diabetes are known to be at higher risk for SSI, those with an elevated HgA1c level going into surgery,31 and those with elevated blood glucose levels (>200 mg/dL) in the postoperative period32,33 are at higher risk for postoperative infectious complications. (2) Smoking—some studies have confirmed that smokers have a higher risk of SSI34 but whether this risk can be lessened by having patients abstain from smoking for a period of time prior to their operation is unclear, though overall complication rates are probably lower.35 (3) Immunosuppressive drugs—patients who take exogenous corticosteroids are thought to be at higher risk of developing SSI, but the data are somewhat equivocal. (4) Nutritional status—patients with severe protein malnutrition appear to have an increased risk of SSI, but it has been difficult to prove that improving nutrition by either the enteral or parenteral route preoperatively can decrease the risk. However, improving nutrition before a major operation has benefits well beyond the risk of SSI and current surgical strategies routinely incorporate nutritional therapy. Furthermore, in terms of lowering the risk of SSI (and most other complications), enteral nutrition has been shown to be clearly superior to parenteral nutrition.36 (5) Hospital factors—prolonged hospital stay,37 perioperative blood transfusions,34 and preoperative colonization with S. aureus38 have all been associated with an increased risk of developing SSI, but whether these are independently associated with an increased risk is yet to be determined. (6) Procedural factors—prolonged operative time, poor oxygen delivery, and hypothermia.37 Prolonged operative time is the only factor that appears to correlate independently with an increased risk for SSI.9,39

Based on epidemiologic data from numerous multivariate analyses, the NNIS developed a risk index score that is supposed to identify patients who are at the highest risk for SSI. For any given operation, the score is generated by counting the number of the following risk factors present: (1) American Society of Anesthesiology (ASA) preoperative assessment score of 3, 4, or 5; (2) the operative incision is classified as either contaminated or dirty-infected; and (3) the operation takes longer than expected by a certain factor than expected.30 Although the NNIS risk index score is able to predict SSI risk much better than standard wound classification systems and appears to be applicable to many different types of surgical procedures, its usefulness is largely limited to assessment of adults and may not be an accurate predictor of risk in children.40

Surgical delivery systems and personnel work under the assumption that nearly all SSI are preventable. Preventing SSI involves eliminating or minimizing known risk factors, including both host and environmental factors. Modern surgical practice includes standardized procedures specifically designed to prevent SSI (Table 69–4). As these procedures are based on a large amount of data accumulated over many years as well as many ongoing studies, one can expect changes as practices evolve in response to new information. Likewise, the practices described are based on studies that included predominantly adults and therefore require certain modifications when applied to children.

Environmental Interventions

Modern operating theater systems are designed to control environmental infectious risk factors and are aimed principally at the reduction of the number of microorganisms that come into contact with the surgical wound. First and foremost is the concept of aseptic technique. Most surgical instruments can be sterilized in an autoclave, which uses superheated steam under pressure to kill organisms and destroy spores. Optical equipment and electrical appliances are damaged by autoclaves and need to be sterilized ethylene oxide gas or peracetic acid. All drapery and consumable materials such as gauze sponges and suture material are also sterilized. Modern operation theaters have ventilation systems that create positive pressure within the room and fresh filtered air is brought in using laminar flow technology to prevent air turbulence. Surgical personnel are required to apply an antiseptic on their hands and forearms and to wear surgical caps, surgical scrub garments, sterile gowns, surgical masks, and sterile gloves. A methodical and almost ritualized behavior is displayed by all personnel to prevent contamination of the surgical field, which includes a complete covering of the patient and operating table with sterile drapes and towels and meticulous attention to potential contamination.

Local Skin Preparation

The skin of the patient is prepared with an antiseptic solution, typically applied in concentric circles starting at the incision site and working outward. Betadine (povidone–iodine) is the traditional and still the most widely used operative antiseptic. It has a broad spectrum of activity, including gram-positive and gram-negative bacteria, Mycobacterium tuberculosis, fungi, and viruses, is easy to apply, is relatively inexpensive, and well-tolerated by most individuals.41 Isopropyl (or ethyl) alcohol (70% v/v in water) has a similar spectrum of activity, is easy to apply and dries completely; it is also inexpensive and generally well-tolerated. However the use of alcohol is limited, and in some cases prohibited, due to its flammability. Patient fires have been reported when the alcohol has not been allowed to dry completely before electrocautery is used during surgery. Recently there has been a trend toward the use of chlorhexidine gluconate as a substitute for betadine, especially for skin antisepsis prior to central venous access.42 It is usually supplied in an isopropyl alcohol base and has a broad spectrum of activity and has the advantage of being effective after a single application. It is currently more expensive than betadine, may cause a skin rash in susceptible individuals, and must be allowed to dry completely before starting the operation because of the risk of fire from the alcohol component. In practice, there is no perceptible difference in the incidence of infection regardless of which of the commonly used antiseptics is used. Furthermore, a recent large meta-analysis of available studies failed to show a significant advantage of one antiseptic over another.43

Prior to application of antiseptics, the skin of the patient may need to be prepared in other ways. Excessive hair is usually removed prior to making an incision. It is best to clip the hair close to the skin rather than using a shaving razor, which has been shown to increase the risk of SSI,44 and hair should be removed as close to the time of surgery to avoid the creation of folliculitis. Showering with an antiseptic soap the night before a scheduled operation was formerly a time-honored ritual but has been shown to have no effect on the incidence of SSI.45 Similarly, mechanical bowel preparation with cathartics and oral antibiotics was the inviolable rule before elective colorectal surgery for decades but has been shown not only to be ineffective in preventing SSI but may actually increase the risk of anastomotic leak and organ-space infection.46,47

Antibiotic Prophylaxis

One of the most important interventions in the prevention of SSI is the use of prophylactic antibiotics.48–52 It might seem obvious that the administration of antibiotics to patients felt to be at high risk of SSI would prevent infection, and for the most part this is true. However, there are limitations to the effectiveness of prophylactic antibiotics and there are risks involved as well. Current recommendations are summarized in Table 69–5. Patients with infected wounds are treated with appropriate antibiotics and are technically not given “prophylactic” antibiotics. For clean wounds, it is difficult to show a benefit from antimicrobial prophylaxis and the risks, including anaphylaxis and the generation of resistant organisms, may outweigh any small potential benefit. The exception is a clean operation that involves the placement of an implant, such as a heart valve or bone endoprosthesis, or in certain parts of the body, such as the heart, brain, or bone, in which case even a single infection can have deleterious consequences. Although generally not indicated or recommended in most cases, it is nevertheless common practice in some centers to use prophylactic antibiotics routinely in all clean cases because of a perceived decrease in the albeit small number of SSI that are seen in these patients.

The greatest benefit of antimicrobial prophylaxis is for clean-contaminated and contaminated wounds. The choice of antibiotic should reflect the most likely organisms to be encountered and therefore should cover aerobic gram-positive organisms. Coverage for gram-negative bacteria and/or anaerobes should be provided for operations involving exposure to the aerodigestive tract or bowel flora. For prophylaxis to be effective, there should be adequate tissue levels of the antibiotic at the time the incision is made. Antibiotics should therefore be given within 60 minutes prior to the start of the operation, rather than at the time of incision (too late) or “on call” to the operating room (too soon), as was customary in the past. Nearly every study has shown that a single preoperative dose of an appropriate antibiotic is effective in decreasing the incidence of SSI. There are very little data to support the use of antibiotics for the first 24 hours and no data to support their use for more than 1 day after an operation.48,49

Intraoperative Interventions

In addition to efforts designed to minimize contamination of the surgical wound, surgeons have long been taught that careful management of the wound during an operation can lower the risk of SSI. First and foremost is gentle handling of tissues. Infections are more likely to develop in wounds in which the skin edges are devascularized and in the presence of devitalized tissue. Wound edges are grasped carefully with noncrushing forceps or skin hooks and only when necessary. Prior to skin closure, the wound is irrigated with sterile normal saline to wash away necrotic debris. Efforts should be made to avoid maneuvers that may decrease the blood supply and delivery of oxygen to the operative field, including unnecessary or poorly designed tissue flaps and the application of dressings that create pressure on the wound or generate a tourniquet effect.

Wound Closure and Postoperative Care

Wound closure is also an important concept in the prevention of SSI. An incision that is infected or grossly contaminated may be treated with one of several techniques. Wounds may be closed with the placement of drains to prevent an infected fluid collection from forming. Closed-suction drains (e.g., Jackson–Pratt or Hemovac drains) decrease the risk of infections while passive drains (e.g., Penrose drains) that are brought out through the incision are not recommended as they have been shown to increase the risk of SSI.9,53 Finally, in the case of dirty or frankly infected wounds, it is recommended that the wound not be closed at all, and instead is left open and lightly packed with moistened sterile gauze. Options then include (1) delayed primary closure, in which the surgical dressing is left undisturbed for 2–3 days after which the wound edges are brought together primarily in sterile fashion and (2) wound closure by secondary intention, in which the dressings are changed two to three times daily using sterile technique and the wound is allowed to granulate and heal over the course of 2 to several weeks. Some wounds of intermediate risk for SSI pose something of a dilemma regarding the decision to use closed or open wound management. For example, adults who undergo colostomy closure or appendectomy for perforated appendicitis have historically been treated with open wound management due to an expected 50% or greater risk of wound infection, while pediatric surgeons generally utilize primary closure for such wounds because the true incidence of SSI in these circumstances appears to be much lower.54

Although most SSI are thought to be caused by intraoperative wound contamination, some may be due to postoperative contamination or the use of surgical dressings that create a local environment conducive to the growth of microorganisms. Wounds that are near the anus or a colostomy should be protected from fecal effluent with barrier dressings or left open to heal by secondary intention. In rare cases, complex wound closures of the buttocks or perianal region are protected by temporary fecal diversion. For most other incisions, it was long felt that a surgical dressing should be placed sterilely in the operating room and then removed after a few days to allow the wound to “breathe.” The patient was asked to keep it dry for at least a week. Over the past 10–15 years, these recommendations have all but vanished. Modern dressings are usually “occlusive,” not permitting tissue fluid to escape and protecting the wound from exogenous pathogens, and are not removed for 7–14 days, with no increase in the incidence of wound infection. Most popular dressings are also transparent (Tegaderm, Dermabond) allowing clinicians to monitor the wound for signs of infection.

In general, control of host factors that increase the risk of SSI is less of a concern for surgeons who deal primarily with children. Nevertheless, there are circumstances in which it becomes important. Children who are malnourished may benefit from a period of supplemental nutrition, preferably by the enteral route, as parenteral nutrition may increase the risk of infection. Infections present in other parts of the body should be treated aggressively, as there is a correlation between other infections and SSI.34 Patients who take immunosuppressive medications are usually not able to discontinue them, but the dose of some, such as corticosteroids, can sometimes safely be lowered during the perioperative period. Children with diabetes should have their blood glucose levels monitored frequently and maintained within a narrow range, with insulin administration if necessary, both during the operation and for the first 2–3 days postoperatively. Although the administration of supplemental oxygen via nasal cannula has recently been shown to decrease the incidence of SSI in older patients who undergo colorectal surgery,55 it will need to be investigated more thoroughly in children before its use can be recommended.

At the institutional level, the most effective means of lowering the incidence of SSI is to establish a system of prospective wound surveillance, with honest reporting of outcomes and modification of practices to reflect new data and changes in practice standards.56 Aseptic technique should be monitored and violators (re)educated. Wounds are systematically classified according to their degree of contamination and prophylactic antibiotics should be administered according to current standards. Specific cases of SSI should be discussed at regular morbidity and mortality conferences and worrisome trends should be investigated thoroughly with the help of infection control personnel. Inpatients and outpatients with surgical incisions should be surveyed periodically for the presence of infection in the form of systematic audits. Staff should be frequently instructed as to current standards of practice and informed of the results of audits in the form of both positive and instructive feedback. There is a recent trend for the federal government and some private insurers to refuse payment for the treatment of “preventable” nosocomial infections including SSI,57,58 increasing the visibility of the issue and the importance of careful monitoring and reporting practices.

The clinical picture of a surgical wound infection usually follows a more-or-less typical pattern, with predictable variation depending on the type of organism involved59 (Table 69–6). Most bacterial wound infections become clinically apparent between the fourth and seventh postoperative day (by convention the date of the operation is day zero). Early symptoms include worsening pain and increased swelling. On examination, there is usually erythema, induration of surrounding tissues, and tenderness; the area may be palpably warm. The wound may also be fluctuant, suggesting the presence of liquid beneath the surface; however in children a subcutaneous abscess is most often tense and very tender. In rare cases, especially if there is a significant deep component there may be spreading cellulitis or true lymphangitic streaking away from the wound.

The timing of the development of a wound infection can give clues as to the likely organism. Wound infections that develop within the first 24 to 48 hours of an operation are very rare and usually due to infection with Streptococcus pyogenes (group A Streptococcus). The presentation is often impressive, with sharply demarcated edema and lymphangitis (wound erysipelas) that spreads rapidly, exquisite tenderness and scant, watery drainage. Patients are typically febrile and toxic-appearing. Gram-positive infections are more likely to present within 3–6 days of an operation. They are heralded by sometimes dramatic local findings of a wound infection while systemic signs of illness are relatively minor. Gram-negative wound infections are usually more subtle in presentation, typically arising in the second postoperative week, and are more likely to cause generalized symptoms despite relatively minor signs of a local wound infection.

Most wound infections are easily diagnosed on the basis of history and physical examination. Ultrasound can be useful in some cases to identify a subcutaneous abscess if the physical findings are ambiguous or to exclude the diagnosis before opening a wound unnecessarily. The presence of a deep wound SSI or an organ/space infection is generally confirmed by computed tomography (CT) scan or magnetic resonance imaging (MRI), although ultrasound is sometimes useful, especially in infants and small children.

Unless the physical findings are quite minor, in which case there may be a decision to treat with antibiotics for a wound cellulitis or to simply observe, wounds that are suspected of being infected should be aspirated or opened at the bedside. If pus is encountered, specimens should be sent for Gram stain and culture with sensitivities, and the wound should be adequately drained. It is usually necessary to open the entire length of the wound and to break up loculations of pus. The wound is then allowed to heal by secondary intention. This usually entails packing the wound lightly with saline-moistened gauze, which is changed two to four times daily for 2 weeks or more. However, in some cases, it may be possible to achieve adequate drainage of the entire wound, with or without the use of a passive drain, and to minimize the extent of the wound that is opened.

Deep infections are treated with antibiotics alone if the collection is smaller than approximately 5 cm in diameter, or by image-guided percutaneous aspiration and drainage if the abscess is large and safely accessible. Organ or space infections may require operative intervention if the deep space abscess drains spontaneously through the incision (a “necessitating” infection).

Opening an infected wound can generally be performed at the bedside with minimal discomfort, but requires a great deal of patience, compassion, and skill. The skin edges of an infected wound that is less than 2 weeks old can usually be separated with minimal traction and the release of pus under pressure can provide significant symptomatic relief. Nevertheless, the procedure is rarely completely painless and there is always a great deal of anxiety on the part of the patient and their family members. One hopes that the image of the brusque and busy surgeon ripping open a wound quickly and without warning and then leaving a tearful patient and horrified onlookers in his or her wake is a thing of the past. Though the act of opening a wound takes seconds, in some cases, the entire procedure can take more than 20 minutes. Patients should be given an analgesic and, if necessary, an anxiolytic. It is not uncommon for children to have the procedure done with moderate sedation or even general anesthesia. All equipment, disposables, and culture supplies should be prepared in advance and close at hand. Local anesthetic is generally unnecessary and may not be effective in the presence of acute infection. A gentle separation of the incision can usually be made painlessly if made exactly in the previous incision, which as a rule is insensate. An attempt should be made to break up loculations gently with an examining finger, but vigorous probing with instruments, especially to “test the fascia,” is painful, dangerous, and usually unnecessary. The wound should be packed gently with saline-moistened sterile gauze and covered with a clean, dry dressing.

Purulent drainage from a wound should be sent for Gram stain, culture, and sensitivities, even if the most likely organisms are known. Although drainage alone may be sufficient to treat some postoperative wound infections, antibiotics are usually considered an important component of therapy. Patients with systemic toxicity or significant cellulitis should be treated in hospital with broad-spectrum intravenous antibiotics until systemic signs of infection have resolved. Provided the family is reliable and close follow-up is arranged, outpatients with no signs of systemic illness, only minor cellulitis, and a wound that has been adequately drained can usually be treated safely with oral antibiotics and dressing changes at home. Broad-spectrum antibiotics are administered empirically and then the regimen is tailored to the organisms that grow in culture. Initial therapy should especially include coverage against gram-positive organisms, which are the most commonly seen. Coverage against gram-negative organisms and anaerobes should be included if the gastrointestinal tract was opened during the operation. A typical intravenous regimen would include a first-generation cephalosporin such as cefazolin, or for penicillin-allergic patients, clindamycin. Clindamycin is increasingly being used as first-line therapy due to the recent increase in the prevalence of community-acquired MRSA infections. For uncomplicated wound infections that occur after bowel surgery, options include cefazolin and metronidazole, ampicillin/sulbactam alone, or, in penicillin-allergic patients, clindamycin and gentamicin. Organ/space infections that develop after bowel surgery, such as intra-abdominal abscesses, are traditionally treated with “triple antibiotics” (ampicillin or vancomycin, gentamicin, and metronidazole or clindamycin) though many surgeons use ampicillin/sulbactam and gentamicin with excellent results. Some believe that monotherapy with a broad-spectrum antibiotic such as piperacillin/tazobactam is as effective as traditional triple-antibiotic therapy without the risks associated with aminoglycoside administration.60 Therapy for infections that occur in other types of surgical incisions is based on the organisms most likely to be involved. For example, wounds of the head and neck are more likely to be caused by oral flora and therefore penicillin or clindamycin are ideal choices. Nevertheless, antibiotic choices should be based on culture results in addition to clinical judgment.

Necrotizing Infections

Although exceedingly rare, few surgical site infections are as serious or as terrifying as those that result in tissue necrosis. They can present in the form of necrotizing fasciitis or, less commonly, necrotizing cellulitis. Necrotizing SSI is usually a polymicrobial infection that includes aerobic and anaerobic bacteria and is more common in patients with diabetes and peripheral vascular disease.61 Less commonly, a necrotizing soft tissue infection can result from infection with a single predominant organism such as group A Streptococcus,Clostridium perfringens or Vibrio vulnificus. The characteristic soft tissue necrosis is due to microvascular thrombosis that is induced by a procoagulant exotoxin secreted by the microorganisms involved.62

These infections are rapidly progressive and potentially lethal unless recognized early and treated aggressively. Necrotizing infections present acutely, often within the first 24–48 hours after surgery. Local signs include edema and erythema of the wound edges, which can spread very quickly, and a watery discharge (“dishwater” discharge) may be seen coming from the wound. The edema often striking and has a raised edge noting an abrupt change to more normal appearing skin (“wound erysipelas”). There is typically extreme pain and exquisite tenderness at and near the incision, although anesthesia of the surrounding skin may also occur. Crepitus may be noted due to gas generated within the tissues. Systemic signs of illness are nearly always apparent, including fever and malaise, and can rapidly progress to overt toxicity and shock. The diagnosis must therefore be confirmed very quickly (within 4–6 hours). In the proper clinical setting, the diagnosis may be made on the basis of physical findings only. In less obvious cases, it may be necessary to perform X-rays to identify gas in the soft tissues, MRI to identify necrotic tissue, or wound exploration with fascia inspection and/or biopsy.

Treatment includes intravenous antibiotics and immediate radical surgical excision of all necrotic tissue. Although monotherapy with broad-spectrum antibiotics like imipenem/cilastin or piperacillin/tazobactam may be effective, multidrug regimens are the mainstay of current therapy. This may include high-dose penicillin or clindamycin and either a fluoroquinolone (in older children and adults) or aminoglycoside (in younger children). Vancomycin may be added if MRSA is suspected and antifungal drugs may be added if a fungal infection is thought to be involved. Antibiotics should be started even before the diagnosis is confirmed with certainty. Surgical debridement is performed in the operating room under general anesthesia. All necrotic tissue is excised sharply and wound, which may be gigantic, is left open. The wound is inspected several times a day and debrided again if there are signs of further necrosis. After successful therapy, wound management usually requires skin grafts or soft-tissue flaps for closure. Hyperbaric oxygen therapy is advocated by some as a therapeutic adjunct but it is not a substitute for aggressive surgical debridement.63

A particularly devastating form of postoperative necrotizing soft-tissue infection is necrotizing fasciitis that involves the perineum and pelvic fascia (Fournier's gangrene), which has been reported after inguinal hernia repair and circumcision.64,65 Even in the case of successful management, wound management can be extremely challenging and usually results in significant disfigurement and functional impairment. As with all necrotizing infections, early recognition and immediate aggressive medical and surgical management is the key to patient survival.

• Surgical site infections are classified as superficial incisional infections, deep incisional infections, and organ/space infections.
• Postoperative occlusive dressings pose no risk for infection and may in fact decrease the risk of surgical site infection.
• Necrotizing surgical site infections present acutely, often within the first 24–48 hours after surgery. Characteristic features may include “dishwater” discharge, wound erysipelas, and crepitus.