Introduction

Despite the advances of modern medicine, infections remain a major cause of morbidity and mortality in surgical patients. Many surgical diseases have an infectious etiology, with pathogenic bacteria present in situ, prior to a surgical intervention. In addition, any break in skin integrity ,including that of a clean incision, has the potential of introducing an infectious inoculum, increasing the risk for a postoperative infection. Nearly every surgical consent discusses this risks and warns the patient of the potential infectious complication as a result of an operation. As such, efforts to reduce the rate and severity of infections are of prime importance to all surgeons.

Bacteria have long been recognized as source of surgical infections, but this recognition was not initially welcomed by the scientific community. In 1843, Oliver Wendel Holmes published his suspicions that the cause of postpartum infection, then known as puerperal fever, was an unknown "contagion," transmitted by physicians [1]. Around the same time in Vienna, during the late 1840s, Ignaz Semmelweis reported similar findings: that physicians had a much higher rate of puerperal fever than the midwives [2]. Despite the denial by the local medical community, this fact was widely known by the public, leading some women to choose birth in the street if they knew that physicians, rather than midwifes, were admitting at the hospital. Semmelweis worked rigorously to elucidate the cause of the infections by eliminating all possible variables. He finally realized that the physicians, unlike midwives, were coming directly from the cadaver lab and were not cleaning their hands prior to assisting with the delivery [3]. Subsequently, he instituted a systematic washing of hands with a chlorine/lime solution, which resulted in an immediate reduction in the physician-associated infection rate, equivalent to that of midwives [1]. Unfortunately, the medical profession was late in adapting these changes, initially either denouncing Semmelweis’s work or simply ignoring it [1]. Perhaps Ignaz Semmelweis’s method of writing to famous obstetricians and calling them murderers might have had something to do with it [3]. His increasing exasperation with the medical profession eventually led to him being declared insane and placed in an asylum where, ironically, he died of sepsis after being beaten by guards [2].

Joseph Lister is the surgeon credited with establishing sterile procedures as respectable practice. In the 1860s, he instituted cleaning of hands and surgical instruments with phenol. Fortunately, by this point, Louis Pasture had advanced the germ theory of disease so that surgeons could understand the scientific rationale of washing their hands as a means to decrease infection [4]. It is this tradition, carried on and refined in the twentieth century with advances such as antibiotics, that has influenced surgery to this day and we now live in an era of historically low surgical infection rates.

Content in this topic is referenced in SCORE Subacute Bacterial Endocarditis Prophylaxis overview

Epidemiology

Approximately half a million surgical site infections occur per year in the United States [5]. Infection rates differ based on the classification of the surgical wound.

Clean cases carry the least risk of infection at approximately one percent. These are the cases in which the respiratory, gastrointestinal or urinary tracts are not violated [6].

Clean contaminated wounds that represent procedures with controlled entry into these tracts have an approximately three percent risk of infection.

Contaminated wounds (open, fresh traumatic wounds or gross spillage from the gastrointestinal tract) carry a six percent risk of infection while dirty wounds (those with previously documented infection) have a seven percent risk of infection [7]. Even within the broader wound classification, however, there are variations depending on the particular operation. The presence of a surgical site infection (SSI) adversely affects patient outcomes tremendously and those who develop an SSI have at least a two-fold increased risk of death [8].

Basic Science

Much of our modern understanding of surgical infections dates back to the germ theory of disease which states that the risk of surgical site infection (SSI) increases with the number of bacteria that colonize the surgical wound. For an infection to occur, an inoculation itself is not enough. There must be an inoculum sufficient enough for bacteria to reproduce. With a count of 10⁵ bacteria per gram of tissue, a wound has an approximately 80% chance of developing an infection [9]. Lesser quantities of bacteria may cause an infection if a foreign body, such as silk suture, is present in the wound [10].

The virulence of the bacteria present in the wound further contributes to the risk and severity of infection. Aerobic and anaerobic bacteria can work synergistically to cause a greater rate of infection than either one would independently [11]. When bacteria enter the wound from the skin surface, a gram positive infection typically occurs. On the other hand, bacteria from an intraluminal source usually result in a gram negative or anaerobic infection [10].

Both gram positive and gram negative organisms can lead to necrotizing fasciitis, a disseminated infection which spreads rapidly throughout the subcutaneous space. Group A beta hemolytic Streptococcus and Staphylococcus aureus, Clostridia species, other gram negative and anaerobic bacteria have been implicated in necrotizing infections. Most often, the source is polymicrobial but single organisms are implicated approximately 15% of the time [12]. The clinical findings include a violaceous discoloration of the wound with subcutaneous emphysema. On incision, there is often a thin discharge described as “dishwater fluid.” The subcutaneous tissues may separate easily away from the fascia. The inciting event can originate from a break in the skin, extension from dental caries, or simply having a surgical incision. Necrotizing fasciitis requires prompt surgical treatment including wide debridement and drainage [13]. (see Steps of the Procedure)

The host immune system plays an enormous role in the susceptibility to infection and the extent of the host response, both of which ultimately affect treatment [14]. Surgery itself results in an inflammatory reaction which may lend the host more susceptible to infection. To that end, some have postulated that a laparoscopic approach is associated with a decrease in infection rate by decreasing the amount of inflammation incurred during the operation [15]. Immunosuppression further increases the risk of infection, morbidity, and mortality. Along with impaired wound healing, immunocompromised patients have an increased propensity to infection in addition to having an impaired response once the infection has occurred. Vigilance to their overall clinical status is of utmost importance as they may not demonstrate the early signs of infection typically seen in an immunocompetent patient (i.e fever, swelling, or erythema of the wound) [16].

While seroma or hematoma formation can complicate the healing of some wounds, contrary to conventional wisdom, they do not necessarily increase the risk of infection [17]. They can cause pain and discomfort for patients leading some surgeons to place drains preemptively. A recent study in orthopedics examined the drains of patients who underwent joint replacement. The authors cultured the drain tips following removal over different points in time and found that the longer a drain remained in place the greater the chance of a positive wound culture [18].

Pathophysiology

What factors contribute to surgical infection?

The factors that contribute to surgical infections can be divided into four broad categories: patient related, preoperative, operative, and postoperative [10].

Patient factors encompass the pre-existing comorbidities that increase the risk of perioperative infection. Most of the data on these factors have been extrapolated from adult studies, although more has emerged in pediatric literature recently. Hyperglycemia, both stress induced and secondary to diabetes, cigarette smoke, steroid use,immunosuppression, and colonization with resistant organisms have all been implicated in increased rates of surgical infection. Adult data have supported tight glycemic control for some time, starting with recognition that hyperglycemia carries an increased risk of infection following open heart surgery. Initial data demonstrated reduction of deep sternal infections from 2.4% to 1.5% after glucose levels were maintained less than 200 mg/dL for at least 48 hours poststernotomy[19]. These findings have been replicated across surgical sub-specialities. Hyperglycemia has been shown to increase rates following hepatobiliary and pancreas operations in Japan. In a study by Ambiru et al rates of infections were reduced from 52% to 20% when glucose rates were maintained less than 200 mg/dL[20]. In orthopedic patients undergoing spinal surgery, hyperglycemia has been identified as a highest independent risk factor for SSI. While rates of SSI in this population are relatively low (2%), greater than 50% of patients with an infection had poorly controlled diabetes or stress-induced, perioperative hyperglycemia [21].

Benefits of tight glycemic control in children are not as clear. Earlier retrospective studies, which were underpowered, suggested a link between hyperglycemia and increased morbidity and mortality. However, more recent data has failed to reproduce such conclusions. Agus at el evaluated the benefits of tight glycemic control, compared to standard therapy following cardiac surgery in children, and found no difference in mortality or infection rates [22]. Hyperglycemia has been shown to increase infection rates in severely burned children, but only the rates of urinary tract infections reached any statistical significance in this study.[23] And while hypoglycemia carries a risk of adverse neurodevelopmental outcomes [24]. hyperglycemia does not. In fact, recent study found no difference in neurological outcomes between children with perioperative hyperglycemia and those with more physiologic glucose levels [25]. To date, the benefits of tighter glucose control are unclear in children [26].

Exposure to cigarette smoke has been studied extensively in adults but little data exists for pediatric population [10]. Likewise, steroid use has been implicated in higher rates of post-operative infections in adults, but pediatric studies have produced conflicting data. A recent review of National Surgical Quality Improvement Program (NSQIP) identified steroid use as an independent risk factor for SSI in general and vascular operations. The overall rate of SSI was 4.3% among 163,624 patients from the VA and large academic centers, among which 4.55% took steroids perioperative, compared to 2.74% who developed no infections [27]. Similar findings have been reported in neurosurgery [28] and orthopedic adult patients [29][30].

In pediatric trauma patients with spinal cord injuries, administration of high dose methylprednisolone has remained controversial in terms of neurologic outcomes. A recent retrospective review demonstrated no difference in neurological outcomes, but identified a 3.5 times higher rate of surgical infection in children who received high dose steroids.[31] Similar findings were reported in patients undergoing congenital heart surgery, in whom steroids were found to increase rate of SSI (rate ratio of 3.32 and 95% CI of 1.56-7.02) [32]. Very different results were reported by Clarizia et al who studied the benefits of intraoperative steroid use in high risk pediatric heart surgery. Benefits were attributed to blunted inflammatory response to the cardiopulmonary bypass. In this study, rate of infection was equivalent between the patients who received steroids and those who did not at 11 and 10% respectively [33]. Based on the current data, the role of steroids in pediatric surgical infections remains unclear.

Malnutrition can increase both infectious and noninfectious complications and a recent meta-analysis has shown that perioperative nutritional support tends to reduce both [34]. Preoperative albumin levels less than 3 g/dL carry a an increased risk for surgical site infections [35]. Malnutrition suppresses the body’s immune response to stress compromising wound healing [34]. In adult patients, early identification of at risk patients and preoperative nutritional repletion have been associated with improved outcomes following surgery. In addition to decreased infection rates the correction of nutritional status tends to reduce overall postoperative morbidity [36], particularly in patients undergoing esophageal, hepatobiliary and upper gastrointestinal operations [37]. On the other hand, preoperative parenteral nutrition in patients with mild to moderate malnutrition tends to increase infection rates and is not recommended [37].

Other patient factors associated with surgical infections are prolonged preoperative hospital stay, nasal S. aureus colonization, perioperative blood transfusions, infection remote to the surgical site and immunosuppression [10]. Neonates represent a particular at risk population as they fall into many of the categories with high risk of surgical infection. Their humoral and cellular immunity is immature and their surgical pathology often requires extended hospital stays, the need for indwelling catheters and prolonged parenteral nutrition - all of which increase their risk of infection [38].

Several preoperative factors contribute to the development of surgical infections. Preoperative body cleansing with chlorhexidine reduces skin bacterial counts but not necessarily infection rates - particularly for deep space infections [39]. Data have suggested that chlorhexidine skin preparation is superior to iodine in the prevention of superficial and deep incisional infections, but not in the development of organ space infections [40]. However, the most recent data suggest no difference between any particular surgical scrub preparation in terms of infection reduction although some studies show a decrease in hand bacterial counts following a scrub with chlorhexidine [10]. Hair removal with clipping has been shown to be superior to shaving in reducing infection. Perhaps the greatest challenge still remains with the proper use and timing of perioperative antibiotics. Antibiotic prophylaxis is indicated for clean contaminated cases and contaminated cases only [41]. It is not indicated for clean cases unless an infection would be catastrophic such as with cardiac or neurosurgical operations. Finally, patients undergoing a dirty procedure (i.e. infected) should already be receiving antibiotic therapy and do not need additional prophylaxis [10].

Operative factors affecting infection are the surgical attire worn by the operating team, sterility of the instruments, ventilation airflow management and surgical technique [10]. An incisional infection can only occur if the incision is closed. This has led to the occasional practice of allowing wounds to heal by secondary intention if the infectious risk is high (with or without assistance by negative pressure dressings) [42].

Finally, postoperative factors affect the incidence of infections: incisional care, perfusion and hemodynamic stability and the development of complications such as anastomotic leaks and fistulae[10].

Prevention

The risk of a surgical infection can be reduced by mitigating the contributing factors. In order to prevent surgical infection, the Center for Disease Control has made recommendations [10] which were updated in 2011 [43].

In the preoperative period, surgeons should attempt to identify and treat all infections prior to performing an elective operation [10]. Any hair around the surgical site that would impede the operation should be removed with a clipper rather than a razor [44]. Every effort should be expended to cleanse the skin prior to incision [39] and to provide antibiotic prophylaxis for indicated cases [41]. Surgeons should cleanse their hands, use sterile instruments and limit the use of flash sterilization when possible [10].

Appropriate selection and timing of perioperative antibiotics contributes to reduced rates of postoperative infection and antibiotic resistance. [45]. The most recent clinical practice guidelines from the Surgical Infection Society outline the choice of peri-operative antibiotics per surgical site and type of procedure.[46] Low risk, clean cases with no implantation of foreign material require no prophylaxis. Since systemic antibiotics need sufficient time to circulate and reach the peri-incisional tissues [10] they should be administered within sixty minutes of the skin incision in order to maximize and maintain tissue levels. Vancomycin and fluoroquinolones require 120 minutes [46]. For most operations, only a single dose of antibiotics is necessary. However, it is important to redose the antibiotics in long operative cases, particularly agents with short half lives such as cefazolin or cefoxitin [46], They should be targeted to skin flora and any potential organisms that colonize the operative site and the viscus being incised. For instance, during colorectal surgery perioperative antibiotics should cover both gram positive skin flora as well as gram negative and anaerobic organisms present in the colon [11]. Knowing the institutions antibiogram is essential in guiding proper choice of antimicrobial. For most cases, cefazolin is recommended as a single agent or in addition to metronidazole for broader gram negative, anaerobic coverage when indicated, such as the colorectal cases [46]. A single dose of vancomycin can be considered in patients who are colonized with methicillin resistant S. aureus.[46] A guideline is available in Patient Care Guidelines.

Skin incisions performed with electrocautery have been reported to increase infection rates [47] but a systematic review has not confirmed this finding [48]. Intraoperative hypothermia and hyperglycemia have also been shown to increase the rate of surgical infections [43].

Skin closure technique and material can affect the rate of surgical infections. In clean and clean-contaminated case, continuous monofilament sutures provide the least risk of wound infection [43]. The contaminated cases are more debatable, however. Traditionally, contaminated wounds were left open to heal by secondary intention or packed with dressing until debride, followed by by delayed primary closure (DPC). [49] In the era of modern antibiotics, primary closure of contaminated wounds became preferred by many surgeons. Several meta-analyses have shown a trend toward improved outcomes with delayed primary closure, but failed to make definitive recommendations due to concerns about design of individual studies being reviewed. In 2011, Alexander et al published updated recommendations for control of SSIs, in which they advocate for the DPC except for cases of perforated appendicitis [43]. In modern surgery, modification to traditional DPC have been shown to successfully decrease rates of SSI. For example, in obese patients the use of wicks between interrupted sutures decreased infection rates to 0.78% [43]. This technique remains a viable option for management of highly contaminated wounds. However, data does support use of primary closure as well, with careful monitoring and early intervention should an infection occur.

Mechanical bowel preparation had previously been shown to be associated with increased anastomotic leaks without reducing complication rates [50]. More recent analyses have not demonstrated a disadvantage, but also have not revealed a reduction in complications - thus mechanical bowel preparation should generally be avoided [51]. Oral antibiotic bowel preparation has been shown to decrease wound infection rates, but has not been adequately studied in patients without concurrent mechanical preparation [52].

Central line infections, while not typically considered a surgical site infection, are nonetheless quite important to pediatric surgeons and their patients. In neonates, a central line infection carries a 10% risk of death and a 20% association with an intra-abdominal process. Central line infections occur at a rate of approximately 4/1000 central line days [53]. Several interventions can be used to prevent central line infections: central lines should be placed under sterile conditions (or replaced quickly if initially placed emergently) with full barrier precautions for the operator and a full body drape for the patient [54]. Immunocompromised children are even more prone to central line infection and may benefit from increased measures such as a positive pressure environment and dust free rooms [55]. Umbilical artery catheters should be removed within five days of insertion while umbilical venous catheters may remain for up to fourteen days [54]. Central venous catheters no longer needed for therapy should be expeditiously removed. The most common organisms implicated in central line infections are Staphylococcus aureus (20%), Enterococcus sp. (nine percent) and Candida sp. (nine percent) [56]. When infection is suspected, unnecessary central venous catheters should be removed [54]. In children who are dependent on long standing central venous access, catheter infection should be treated with antibiotics for fourteen days. Catheters should be removed in the context of a fungal infection, if the patient demonstrates hemodynamic instability or the cultures remain positive for three consecutive days [57].

How is endocarditis prevented?

In order to prevent endocarditis, patients with surgery planned on the mouth, oropharynx, gastrointestinal or genitourinary tracts should receive antibiotics prior to the procedure if they have a

• prosthetic valve
• history of endocarditis
• heart transplant with abnormal valve function
• unrepaired cyanotic heart defect
• repaired defect with residual leaks or abnormal flow
• cardiac anomaly repaired with prosthetic material in the first six months after cardiac surgery [58]

The American Heart Association recommends amoxicillin for patients who can take oral antibiotics or intravenous ampicillin, cefazolin, or ceftriaxone for those who cannot. Patients who are allergic to penicillins should receive cephalexin, clindamycin or azithromycin orally. Intravenous cefazolin, ceftriaxone or clindamycin is recommended if oral medications are contraindicated [58]. Unfortunately, no data documents the absolute risk of endocarditis or that antibiotic administration decreases this risk.

Classification

Depending on their location, surgical infection can be classified as follows: superficial incisional surgical site infection, deep incisional surgical site infection and organ space infection.

Superficial infection occurs above the fascia while deep infection involves the fascial planes as well. Finally, when infection extends below the fascia, it is termed an organ space infection [10].

When either the infection itself overwhelms the body’s defenses or when the body’s defense mechanism itself proves too robust, the body can enter a state of sepsis. This is a state that is characterized by an active infection and a systemic inflammatory response. When tissue perfusion is compromised and is refractory to fluid resuscitation this is termed septic shock. (see Staging)

video link on wound classification (Twitter)

Staging

What are the stages of sepsis?

The stages of disseminated infection are sepsis, severe sepsis and septic shock. Progression occurs in series and the required intervention and monitoring for each stage differ. The stage of shock affects the overall prognosis with increasing mortality, from ten to 20% in sepsis, to 20 to 50% in severe sepsis and 40 to 80% in septic shock [59]. Re-evaluation times change with increasing frequency as the stages progress (hours to minutes to continuous monitoring). For detailed descriptions refer to guidelines developed by the Society of Critical Care Medicine.

What is the systemic inflammatory response syndrome?

Systemic inflammatory response syndrome (SIRS) describes a constellation of symptoms that occur in response to a stressful insult, which may or may not be infectious in nature. Burns, trauma and pancreatitis may result in noninfectious SIRS for example. When SIRS is precipitated by an infection, however, the definition changes to that of sepsis. In short, the ACCP/SCCM consensus defines sepsis as SIRS secondary to an infectious etiology [60]. SIRS criteria include at least two of the following.

• temperature greater than 38^oC (100.4^oF) or less than 36^oC (96.8^oF)
• heart rate greater than 90 beats/min
• respiratory rate greater 20 breaths/min or PaCO₂ less than 32 torr
• white blood cell count greater than 12,000/mm³, less than 4,000/mm³ or greater than 10% immature (band) forms

As this criteria was developed for adults, it is not always applicable to children, particularly since the normal physiologic parameters change over the age spectrum of the child [59]. In 2005, the criteria were modified to address this issue by the International Pediatric Sepsis Consensus Conference [61]. As children present with tachycardia and tachypnea in setting of many clinical scenarios unrelated to SIRS, these two factors alone are not sufficient to diagnose pediatric SIRS. Instead, leukopenia or leukocytosis and temperature abnormalities are needed for the diagnosis. In addition, tachycardia is defined as elevation in heart rate more than two standard deviations of normal for age and bradycardia is considered a SIRS criteria in children less than one year of age and is defined as sustained, mean heart rate less then tenth percentile for age [61].

Definition of pediatric sepsis requires the presence of an infection in addition to SIRS criteria [61][62]. What constitutes an infection has been debated making this definition difficult to apply to all patients. Initially, an infection was defined as pathologic bacterial invasion of normally sterile host tissue. In clinical practice, many of these sites, including the respiratory and gastrointestinal tracts are colonized with pathologic organisms and distinguishing colonization from infection may be challenging. Furthermore, the actual symptoms may not be caused by the presence of bacteria but by the effects of the endotoxin (as in the case with Clostridium difficile). Finally, an infection and sepsis may be suspected without a microbiologic diagnosis in which case empiric therapy should be initiated promptly [63]. Findings consistent with organ dysfunction may be adequate to suspect sepsis, but it is important to note that "hemodynamic instability, arterial hypoxemia, oliguria, coagulopathy, and altered liver function" [63] may be caused by other conditions. In reality, many practitioners suspect sepsis in the setting of SIRS and start treatment early while awaiting confirmation of infection. Occasionally, the treatment course is completed without ever definitely confirming the presence of an infection.

Sepsis can progress to severe sepsis and shock in a matter of minutes, hours or days. The intensity of monitoring in these patients is dependent on the degree of sepsis and how quickly the patient seems to be worsening. A conservative approach, including transfer to an intensive care unit with continuous monitoring as well as central and arterial access, is generally prudent. When organ dysfunction is present, condition has evolved to severe sepsis which is considered to be the most common cause of death in non-coronary critical care units [63]. Septic shock refers to a state of “acute circulatory failure characterized by persistent arterial hypotension unexplained by other causes.” [63] Hypotension is generally based on blood pressure norms for age but may also be considered as acute changes "relative" to baseline individual norms. In children, hypotension is a late finding and represents an uncompensated state of shock. Other markers of poor perfusion such as tachycardia, delayed capillary refill, altered mental status, decreased urine output and mottled extremities may be harbingers of circulatory collapse [63].

As discussed earlier, worsening outcomes are seen as one progresses through the stages of sepsis but other more detailed scoring systems have been developed as a way of further differentiating and predicting more accurate outcomes [64]. THe Pediatric Multiple Organ Dysfunction Score (P-MODS), Therapeutic Intervention Scoring System (TISS), Sequential Organ Failure Assessment (SOFA) score and PIM (Pediatric Index of Mortality) are some of tools developed with a similar goal. P-MODS and PIM are generally applied on admission and used to predict mortality while SOFA and TISS offer tools for follow-up and intervention. SOFA considers the dysfunction in six organ systems (respiratory, cardiovascular, renal, hepatic, hematologic and neurologic.The partial pressure of xxygen, FiO₂, platelet count, Glasgow coma scale, bilirubin levels, degree of hypotension, level of inotrope need and admitting creatinine levels are used to compile the score which can be used to assess individual organ systems and the patient in general [65]. The score can be reassessed daily during the intensive care unit stay and changes in SOFA have been correlated with changes in mortality in hematology oncology patients [66]. In a recent study evaluating mechanically ventilated oncology patients, a decrease in the SOFA score correlated with significant decrease in mortality from 88.2% to 18.5%. Change of score or lack thereof were felt to correlate well with response to therapy [66].

Presentation

How does a surgical infection typically present?

The typical presentation of a surgical infection depends on the site of infection. These include incisional infections, infections affecting the deep space related to the incision, infections involving a cavity different from surgical site, and systemic sepsis.

An infection at the surgical incision typically presents with the classical tetrad of inflammation described by Celcus: calor (warmth), dolor (pain), tumor (swelling) and rubor (erythema). A fifth was added by Galen: functiolaesa (loss of function).

Deep space infections usually manifest with swelling and increased fluid content such as an abscess or fluid collection after intestinal surgery. In the chest, a pulmonary infection may lead to a pleural effusion or mediastinal inflammation with subsequent development of an empyema or mediastinitis.

Infections commonly affect areas separate from the site of surgery as a result of instrumentation or access during the operative procedure (e.g. endotracheal tube accessing airway or a urinary catheter accessing the bladder). Therefore, the search for an infectious source should include an evaluation for pneumonia, urinary tract infections, central venous catheter infections, and venous thrombophlebitis from a peripheral intravenous catheter. Importantly, post-operative fever may also occur from noninfectious sources such as deep venous thromboses or drugs and these should be excluded.

Neonates and children with developmental delay may present with atypical signs of infection. For example, temperature instability or difficulty maintaining temperature in an isolette is a common finding in neonates with infection. Moreover, typical signs of infection may also occur late (e.g., fever) while systemic signs such as altered mental status, lethargy, and hyper- or hypoglycemia (in the absence of diabetes) may manifest early.

What are the presenting signs and symptoms of necrotizing fasciitis?

The presenting signs and symptoms of necrotizing fasciitis include increasing or worse than expected pain at the incision site, erythema (sometimes a purplish rash or blisters), foul smelling discharge, discoloration, peeling of the skin, and crepitus (or palpable gas below the wound). This is often accompanied by systemic signs of severe infection such as fever, altered mental status, dehydration, and lethargy. The infection can progress quickly to septic shock if rapid intervention with early fluid resuscitation, broad spectrum antibiotics, drainage, and debridement of necrotic tissue is not initiated. Risk factors for necrotizing fasciitis include diabetes, malnutrition, renal failure, underlying malignancy, and obesity [67].

The organisms responsible for necrotizing fasciitis include Group A Streptococcus (most common) and Clostridium sp.; therefore, broad spectrum antibiotics are crucial to therapy. Surgical therapy needs to be supplemented with ongoing resuscitation, adequate nutrition, and appropriate wound care. Patients often require multiple trips to the operating room for further debridement, washouts, and reconstruction. The use of hyperbaric oxygen therapy is controversial as there is insufficient data to make any recommendation. Finally, the use of intravenous immunoglobulin to manage toxic shock syndrome should be considered for severe cases.

What is observed in the setting of peritonitis?

Abdominal pain and distension are classic findings of peritonitis. Commonly, peritonitis is also associated with nausea, emesis, and diarrhea. Pain usually is worsened with movement. Abdominal distension can occur as a result of associated bowel dilation with fluid or air, due to the presence of ascites, or an enlarged spleen and/or liver.

Initially, the pain in the setting of peritonitis may be dull and diffuse as the inflammation starts with the visceral peritoneum. Severe, localized pain follows with the spread of inflammation to the parietal peritoneum. Occasionally, the pain can be generalized from the start as occurs with bowel perforation, spontaneous bacterial peritonitis, or acute pancreatitis. In children with developmental delay, liver failure with ascites, or immunosuppression, the typical findings of peritonitis may be masked.

Peritonitis can quickly progress to severe sepsis or shock and early intervention is warranted. In neonates and young children, peritonitis can also contribute to abdominal compartment syndrome requiring urgent decompression. Patients with peritonitis may require significant fluid resuscitation to maintain cardiac output. Indeed, monitoring is necessary to continually assess for adequate intravascular volume.

see Necrotizing Enterocolitis

see Appendicitis

Assessment

What studies should be performed during the evaluation of a patient with a postoperative infection?

All assessments start with a thorough history and physical exam, evaluating all sources for manifestations of infection, including the surgical site, as well other organ systems. On examination, the surgical wound is assessed for drainage, erythema, fluctuance or severe tenderness. Occasionally, dehiscence of the wound or frank peritonitis may be present. Laboratory investigations include a complete blood count (CBC) and differential looking for a left shift (increase in neutrophils and immature bands) and C-reactive protein (CRP). Depending on the severity of the infection and its effect on the hemodynamic status of the patient, thrombocytopenia, renal dysfunction, liver dysfunction, or multiple organ dysfunction may occur that warrant specific laboratory assessment. Blood gas analysis can identify lactic acidosis or a base deficit, which are helpful in guiding the subsequent resuscitation. Finally, blood and wound cultures can aid in targeting antibiotic therapy, but empiric treatment should begin as soon as infection is suspected. Culture results and sensitivities, once available, can help de-escalate broad spectrum antibiotic therapy.

While plain radiographs may reveal an ileus, ascites, or a pleural effusion as an indirect sign of infection, ultrasound, computerized tomography (CT), or magnetic resonance imaging may be more helpful in characterizing the presence of an infection or an abscess.

Early recognition and initiation of goal directed therapy are essential in improving outcomes of septic patients. The Surviving Sepsis guidelines recommend administration of antibiotics within the hour of presentation of a septic patient [61][62]. Despite this, surveys have shown significant delays in initiation of therapy among adult septic patients. A 2007 study by Kumar et al identified only a 14.5% compliance rate in one Canadian and three United States academic centers [68]. In this retrospective review, 32.5% of patients received antibiotics within three hours of symptoms and 29.8% had received none for up to 12 hours after the onset of hypotension [68]. Moreover, each delay in initiation of antibiotics correlated directly with mortality which increased by an average of 7.6% for each hour delay, dropping to 24.5% after a nine hour delay [68]. This study reviewed data available between 1999 and 2004 and we have learned much since then about management of sepsis. Nonetheless, it highlights well the impact in delay of care, in both outcomes and mortality associated with sepsis. In order to optimize recognition and timely therapy, the Surviving Sepsis Campaign and the Institute for Healthcare Improvement developed an “Evaluation for Severe Sepsis Screening Tool.” This is an easy to follow screen which consists of three components.

• history of a new infection with a checklist of common sites
• signs/symptoms of a new infection
• organ dysfunction criteria

If any of the first two criteria are present, presence of infection is suspected and laboratory investigations are ordered: lactic acid, blood cultures, complete blood count with differential, basic chemistry labs and bilirubin. At the physician’s discretion, the following may also be required: urinalysis, chest radiograph, amylase, lipase, CRP, arterial blood gas and CT.

Another tool was recently validated for bedside use in surgical intensive care units. Moore and colleagues developed a a three level sepsis screening system, utilizing increasing levels of expertise and responsibility [69]. The first level, completed by the bedside nurse twice a day, assesses parameters for SIRS as defined above. If positive, the second level, completed by a resident or physician extender, looks for a source of infection. The final level, done by the faculty intensivist, is the initiation of the sepsis protocol.

In addition to above evaluation, procalcitonin is commonly used to differentiate between inflammation and a bacterial infection. This is a relatively recent marker of infection, shown to correlate well with presence and severity of bacterial infection [70]. Patients with bacterial infection typically have procalcitonin levels greater than 5 ng/dL, which rises as the severity of infection increases and fall as patients respond to antibiotics. Levels greater than 200 ng/dL can be seen in septic shock [70][71]. Procalcitonin appears to be more specific for presence of an actual bacterial infection unlike other markers of inflammation such as CRP [70].

surviving sepsis

Infographic from the APSA Practicing Surgeon’s curriculum (Janice Taylor)

video link Surviving Sepsis APSA 2019 Top Educational Content talk

What should be monitored to evaluate the effectiveness of therapy?

To evaluate the effectiveness of therapy, hemodynamic monitoring is absolutely necessary. Based on the degree of illness, monitoring intensity will vary. For example, in a patient with early sepsis without profound tachycardia or hypotension, hourly monitoring of vital signs, including oxygen saturation and urine output is likely adequate. Conversely, patients with ongoing instability require continuous arterial and central venous monitoring. An echocardiogram can be a useful adjunct in patients with suspected myocardial insufficiency, even if the cardiac output is maintained. Evaluation of the venous oxygen saturation helps determine if adequate tissue perfusion is occurring despite a lower arterial saturation. Clinical status is usually monitored with end organ perfusion and function such as urine output, bowel function, or pulmonary function and gas exchange.

Repeat blood or wound cultures may be useful to identify resistant organisms in the context of ongoing or refractory infection. In patients with culture proven sepsis adequate therapy with antibiotics is targeted using initial cultures. Resistance patterns for institutions are usually followed in an antibiogram. However, when patients fail to improve, re-culturing allows one to check if a super infection has occurred with another organism, if the infection has progressed (e.g. from sputum to blood) or if resistance has changed since implementation of therapy. (see Evaluation for Severe Sepsis Screening Tool)

Medical Treatment

What are the principles of medical management of surgical infections?

The management of surgical infections revolves around the following principles.

• source control
• empiric versus culture driven antibiotic coverage
• prevention of secondary end organ injury

Source control is nonspecific and varies with the initial disease process or surgery. For example, an appendectomy may be considered source control for nonperforated appendicitis, whereas drainage of an abscess may be sufficient for perforated appendicitis. With each infection, the length and type of antibiotic therapy varies. In some infections and under certain instances, antibiotics may not even be necessary once source control has been established (e.g. simple appendectomy, drainage of a simple abscess). On the other hand, prolonged antibiotic therapy may be required with infections such as endocarditis or osteomyelitis. Ideally, broad spectrum antibiotics should be started within the hour once an infection is suspected and de-escalated as soon as culture and sensitivity data are available [62]. Antibiotic stewardship is highlighted in most discussions on surgical infection and remains an important factor in minimizing development of antibiotic-resistant organisms.

antibiotic stewardship

Infographic from the APSA Practicing Surgeon’s Curiculum

video link Antibiotic Stewardship APSA 2019 Top Educational Content talk

However, if the choice of antibiotics fails to cover the source of infection, it can only contribute to evolution of resistant organisms without survival benefit to the patient. In fact, delay in proper therapy carries a high risk of mortality [68][72]. Therefore, broad spectrum empiric antibiotics should be given as soon as an infection is suspected with plans to transition to more narrow coverage once culture data are known [62].

If an infection is suspected, the following should be instituted within three hours of presentation (as defined as the time of triage or presentation from another care venue).

• obtain blood cultures and lactate level
• start broad spectrum antibiotics
• fluid resuscitation up to 40 ml/kg within the first hour if the patient is hypotensive or the lactate level is greater than 4 [61][62]

In addition, the following measures should be added within six hours of presentation.

• addition of vasopressors if hypotension is still present despite fluid resuscitation
• additional measurement of volume status with CVP or SvO₂ if the patient does not respond to earlier therapy or shock persists
• repeated lactate measure [62]

Response to therapy should be frequently. Volume status and tissue perfusion can be assessed with a combination of vital signs,capillary refill, skin findings or two of the following: CVP, ScvO₂, bedside cardiovascular ultrasound or dynamic assessment of fluid responsiveness with passive leg raise or fluid challenge.

Finally, the prevention of secondary injury requires goal directed fluid resuscitation with adequate fluid and vasopressor therapy to allow for end-organ perfusion.

Details of the resuscitation are discussed below as guided by the The Surviving Sepsis Campaign. Therapy is generally escalated as more severe stages of disease are identified or as one progresses from severe sepsis to septic shock. In severe sepsis, infection is associated with presence of single or multiorgan dysfunction. In pediatric population, severe sepsis requires presence of either cardiovascular or respiratory dysfunction or compromise of two or more other organ systems, which are defined as follows [61][62].

Respiratory - hypoxia with PaO₂/FiO₂ less than 300 or hypercarbia with PaCO₂ less than 65 or greater than 20 mm Hg from baseline, respiratory failure necessitating positive pressure ventilation, FiO₂ greater than 0.5

Cardiovascular - hypotension following volume repletion of 40 mL/kg over one hour, presence of any of the following: lactic acidosis, metabolic acidosis (otherwise unexplained), delayed capillary refill greater than five sec, oliguria, gap between peripheral and core temperature greater than 3°C

Renal - creatinine greater than or equal to two times normal for age in previously healthy children or double of baseline for those with chronic kidney injury

Hepatic - total bilirubin greater than 4 mg/dL, AST greater than two times normal for age

Hematologic - leukocytosis, leukopenia or thrombocytopenia (less than 100,000), in patients with chronic thrombocytopenia, a decrease of 50% from baseline over three days is significant

11, if baseline GCS is low a decrease by three points is significant

Successful outcomes in pediatric sepsis depend on initiation of early goal directed therapy and proper escalation of care if the patient fails to respond to initial measures. The ultimate goal of supportive therapy is restoration of normal physiology while source control is achieved with surgery, antibiotics or both. The Surviving Sepsis Campaign has set the following goals [62].

What alternative and/or adjuvant medical therapy is available for patients with surgical infection?

Alternative and occasionally controversial therapies for surgical infection (with variable supportive data) include the use of hydrocortisone, extracorporeal life support (ECLS), intravenous immunoglobulin (IVIG), activated protein C (APC), continuous renal replacement therapy (CRRT) and total plasma exchange (TPE).

Though controversial, the use of hydrocortisone therapy in children is supported by the Surviving Sepsis Guidelines for management of refractory, catecholamine resistant shock or adrenal insufficiency. In approximately 25% of children with septic shock [73] adrenal insufficiency exists concomitantly and requires the administration of stress doses steroids. Hydrocortisone is recommended at 1.5 mg/kg/dose or 50 mg/m²/24 hour for at least 48 hours. Some authors suggest obtaining a serum cortisol level at the time of steroid administration to document if true adrenal insufficiency exists [74]. In the reality of clinical practice, the cortisol level has little value (unless it is profoundly low) and most intensivists, including the Surviving Sepsis Campaign, recommend giving hydrocortisone based on clinical assessment of the patient. At this time, data does not support routine collection of cortisol levels in severe sepsis.

A gradual wean of steroids is started after the resolution of catecholamine resistant shock with maintenance doses (0.5 mg/kg/dose) administered for up to a week. It should be noted that several retrospective studies of septic pediatric patients found no difference in outcomes between those who received steroids and those who did not [75][75]. These studies are limited by their retrospective nature, as well as survey of patients treated prior to the most recent Surviving Sepsis Guidelines. At this time, steroids remain in the recommended armamentarium for management of patients with catecholamine resistant shock [62].

In pediatric patients who are refractory to conventional therapy, the use of extracorporeal life support (ECLS) is recommended as a final option in sepsis management [76][77]. ECLS deployment and success (53% survival) for pediatric respiratory illness has increased since the H1N1 flu epidemic in 2009 [78][79]. ECLS has been used successfully in critically ill H1N1 pediatric patients with refractory respiratory failure all over the world [80][81]. Nearly 75% of newborns and 40% of older children survive when ECLS is used for sepsis [82]. When feasible, venovenous ECLS is used early in the course for severe sepsis and septic shock [82]. Several retrospective studies have indicated higher neurological complications in septic children treated with ECLS compared to those who required extracorporeal therapy for other reasons. In septic patients, the rate of seizures and other neurologic complications was found to have an odd ratio of 4.6 and 7.7 respectively, for children under 24 months of age [83]. The odds ratio for older children was 2.6 and 4.4 [83]. The majority of patients in this study, however, were treated with VA ECMO and this data should be re-examined, prospectively if possible, as VV ECMO use has increased in sepsis.

IVIG is used as a therapeutic adjuvant in sepsis with the potential for providing passive immunity by “enhancing opsonization, phagocytosis, and complement; promoting antibody dependent cytotoxicity (ADCC); modulating T cell and macrophage activity via cytokines; stimulating B cell effector functions; and improving neutrophil-mediated killing.” [84] However, IVIG has been extensively studied and polyclonal IVIG does not reduce mortality in neonates or children in the setting of sepsis. For IgM-enriched IVIG, the trials are small and inconclusive. Adjunctive therapy with monoclonal IVIG is still experimental [85].

Over 80% of patients with sepsis have protein C deficiency and the degree of deficiency correlates with mortality and organ dysfunction [86][87]. In an initial trial evaluating children with severe sepsis, up to 58% were severely deficient (level less than or equal to 40% of normal). However, administration of activated protein C led to serious bleeding events in 28% with central nervous system (CNS) bleeding occurring in 2.7%.[87] A subsequent multicenter, randomized, intention-to-treat trial demonstrated that the rate of bleeding events was similar to placebo but there was no difference in 28 day mortality. It was notable, however, that CNS bleeding events in this trial were greater in infants under sixty days of age that received activated protein C when compared to controls [88]. This negative study was countered by a retrospective, 2:1 propensity-matched, multicenter cohort evaluation of pediatric patients with septic shock in which the early use of recombinant human activated protein C correlated with a reduction in mortality [89]. However, since there were no trials demonstrating a substantial benefit and there was a risk of serious bleeding, the drug was removed from the market [90].

Although a few retrospective trials supported the early initiation of continuous venovenous hemofiltration (CVVH) in sepsis, early initiation of CRRT did not improve survival when the technique was subjected to a randomized control trial [91]. In the recent past, however, the use of CRRT in patients with sepsis has evolved, leading to earlier initiation with the hope that immunomodulation by modified CRRT techniques in septic shock and multiorgan dysfunction would improve patient outcome [92]. An analysis of registry data for pediatric patients with multiorgan dysfunction receiving CRRT (with the most common cause of organ failure being sepsis (39%)) demonstrated an overall survival rate of only 51.7%. All of the following were lower at CRRT initiation in survivors: Pediatric Risk of Mortality (PRISM 2) scores, central venous pressures, and percent fluid overload. Fluid overload, defined as total fluid input minus output over seven days before CRRT, indexed to body weight, was found to be an independent predictor of mortality in critically ill children with multiorgan failure which suggested that early initiation of CRRT might be of value [93][94][95]. More recent studies have shown correlation between percent fluid overload (FO) and mortality, which increases significantly with greater than 20% fluid overload on initiation of CRRT[96][97]. Many intensivists have subsequently advocated for early initiation of CRRT in critically ill children. However, the mortality in children requiring CRRT remains high, especially in children who are less than twelve months of age and septic, most commonly as a result of multiorgan failure or hemodynamic instability [98]. As such, the early initiation of CRRT in pediatric sepsis remains controversial and is not supported by the literature.

Severe infection and sepsis can lead to disorders or coagulation and platelet function which may benefit from total plasma exchange. These include entities such as thrombotic thrombocytopenic purpura, hemolytic uremic syndrome and disseminated intravascular coagulation under the umbrella of thrombocytopenia associated with multiorgan failure (TAMOF) which is associated with acute kidney injury (42 to 100%) and a high mortality rate (up to 80%) [99].

In thrombotic thrombocytopenic purpura (TTP), levels of ADAMTS-13 are low leading to microthrombotic events and platelet aggregation. ADAMTS-13 is a metalloprotease which cleaves von Willebrand Factor limiting its prothrombotic activities to physiologic levels [100]. ADAMTS-13 activity in TTP is less than 10% leading to th3 presence of an ultralarge von Willebrand Factor multimer (ULvWF), which is greatly prothrombotic resulting in TTP [101]. Clinical signs and symptoms include fever, anemia, thrombocytopenia and kidney and neurologic dysfunction. The treatment of TTP is accomplished by replenishing ADAMTS-13 activity with frozen plasma (FP), removing ADAMTS-13 inhibitors using steroids or rituximab and removing ULvWF by therapeutic plasma exchange (TPE). TPE, now the standard for management of TTP, has decreased the mortality from nearly 100% to less than 20%.

Infection induced hemolytic uremic syndrome (HUS) (90% of all cases) caused by Shiga toxin-producing Escherichia coli (STEC) and several other bacteria, develops in 6 to 15% of patients up to ten days after the onset of bloody diarrhea. Supportive care is the current recommendation for HUS with atypical cases, caused by an autoantibody to Factor H or by a complement factor gene mutation, treated with total plasma exchange (TPE). TPEthoretically removes the anti-Factor H antibody along with other large plasma proteins [102]. It is not recommended for standard STEC-HUS management.

The use of TPE in disseminated intravascular coagulation (DIC) is limited. Acute kidney injury (AKI) can occur with DIC, but is more common in HUS and TTP. A smaller number of patients with DIC have a low ADAMTS-13 activity, with associated higher incidence of AKI when that activity is less than 20%. If DIC is associated with bleeding and severe microthrombotic events, supportive therapy includes transfusion of platelets, fibrinogen and prothrombin complex concentrate or frozen plasma. Low molecular weight heparin (LMWH) may also be used if thrombosis predominates. If AKI develops despite these measues, and in setting of low ADAMTS-12 levels TPE may be considered but the data on this is limited.

To summarize, in thrombocytopenic pediatric patients with sepsis, multiple organ failure, TAMOF, and/or evidence of a thrombotic microangiopathic process, TPE should be considered [99].

What is the role of interventional radiology in managing surgical infection?

The role of interventional radiology is to provide image-guided percutaneous drainage of an abscess or fluid collection for diagnosis and/or management of a surgical infection, potentially eliminating the need for "open" surgical drainage. While percutaneous drainage of an intra-abdominal abscess can address an acute bacterial infection, it is often a temporizing measure since source control is often required [103]. For example, in perforated diverticulitis, or an ileo-psoas abscess secondary to Crohn disease, percutaneous drainage is a valuable adjunct in preparation for definitive surgery (i.e. bowel resection).

Postoperative abscesses secondary to an enteric source is another common indication for percutaneous drainage. For example, abdominal and pelvic abscesses may occur from solid organs such as the spleen, kidney, or pancreas or an enteric source [104]. The abscess can be simple or it can be complex and multiloculated. The approach to drainage can be through the abdominal wall, trans-gastric, transrectal, transvaginal, trans-perineal or trans-gluteal [104]. Occasionally, a catheter cannot be placed and only aspiration with sampling of the fluid is performed especially if the abscess is not large or loculated [105]. The optimal approach for children with a fluid collection in the abdomen involves the use of ultrasound guided drainage for diagnosis or for therapy with a computerized tomography/magnetic resonance image guided approach if needed or preferred [106].

Medical Decision Making

The medical options for treating surgical infections include early goal directed resuscitation [76], frequent monitoring and early administration of antibiotics. See section on Assessment and Medical Treatment

Ideally, cultures should be obtained prior to the initiation of antibiotics, but should not delay their administration. Venous access is a priority for resuscitation and monitoring. Except for antibiotic administration and proper tailoring for suspected infection, most of medical therapy remains supportive. Antibiotics themselves can provide source control in smaller abscess, for example, but majority of surgical infections ultimately require an intervention. Supportive care should be initiated promptly, as outlined in Medical Treatment, starting with adequate fluid resuscitation. While inotropes are indicated following appropriate volume repletion, they should be used in conjunction with plans for prompt control of infection source.

The surgical options for infection control may be as simple as washout or drainage alone. Cultures should be obtained when possible to target antibiotics. In a patient with intra-abdominal sepsis from perforated viscus, who remains unstable, prompt source control may require diversion with an ostomy or temporary abdominal closure until further stabilization. Source control can also be obtained by opening a wound to allow adequate drainage of an infection which may prevent the development of a deep surgical site infection, fasciitis, or progression to wound dehiscence. Finally, removal of media for bacteria such as foreign bodies or a hematoma at the initial operation may prevent the development of infection. The indications for surgical intervention include a progressive infection or a lack of source control. Decision-making is important in the patient who has an imminent need for operation, but is inadequately resuscitated, perhaps with coagulopathy, hypothermia, or acidosis, and who could potentially become unstable intraoperatively. If a patient fails to respond appropriately to resuscitation (e.g. increasing base deficit or lactate levels), surgical options should be considered.

Occasionally, in the course of resuscitation, abdominal compartment syndrome (ACS) develops leading to inadequate venous return, decreased cardiac output, reduced kidney perfusion and low lung volumes with high peak inspiratory pressures and inadequate ventilation. In critically ill patients at risk for ACS (those requiring high volume resuscitation, patients with pancreatitis or following recent abdominal operation), serial measures of bladder pressure may help identify this entity early. A recent consensus defined ACS as sustained bladder pressure greater than 20 mm Hg in the setting of new organ dysfunction [107]. A single measure may be indicative of ACS, but serial measurements are more helpful in discerning the need for an intervention such as a decompressive laparotomy.

Preoperative Preparation

While minimally invasive approaches continue to evolve and may be the only intervention needed in some infections, open operative options remain at the forefront of surgical infection management. As a result, interventions intending to provide source control span a broad range of options, from image-guided drain placement or simple incision and drainage to wide debridement and extensive exploration. Depending on the presentation, an operation may be needed immediately to manage the source, as in the case of pneumoperitoneum, anastomotic disruption, or severe necrotizing enterocolitis. Often, however, operation can be delayed until nonoperative management has been deemed inadequate.

More importantly, an operation may need to be deferred for a period of hours (and sometime days) until the patient has been optimally resuscitated. As much as a delay in operation can be suboptimal, so can an operation in a poorly optimized pediatric patient. This is particularly true in the setting of septic shock when patients may experience irreversible cardiovascular collapse and disseminated intravascular coagulation in the operating room [108][109]. In patients who require emergent operation, resuscitation may need to continue simultaneously with an operation, meeting similar goals to those in established sepsis.[62] In unstable children, care mirrors that of the bleeding, coagulopathic trauma patient, in which damage control surgery is followed by a delayed definitive operation. The goal of the initial operation, therefore, is prompt source control followed by transfer to an intensive care unit.

Patients with simple surgical site infections, such as superficial wound infections, may only require an incision and drainage of the wound. They are generally well appearing and need minimal preoperative optimization. However, since the magnitude of the septic response depends more on the host than on the severity of infection, even patients with “minor” infections may develop significant instability. As such, reliable intravenous access and the availability of fluids and blood products should be considered in all patients coming to the operating room for management of infection.

The optimal preparation of an infected, septic patient includes all the following

• timely initiation of antibiotics
• early mechanical ventilation (or at least some form of respiratory support), when indicated, to minimize further decompensation
• ample intravenous access
• availability of type-specific blood products
• US (or echocardiogram if available) to guide volume resuscitation
• preparation of inotropic/chronotropic medications to optimize cardiac function and tissue perfusion (should they become necessary)

Many of these steps are often done simultaneously, thereby preparing the patient for optimal outcome both in and out of the operating room [76]. Finally, children in severe septic shock may need escalation of support, including the initiation of ECLS and/or CRRT. (see Medical Treatment). If these resources are not available, early transfer should be considered.

Antibiotics should be administered as soon as the infection is suspected. If culture and sensitivity data are available, the choice of antimicrobials should be culture driven and based on an individual institution’s antibiogram. More often, however, broad spectrum coverage is initially required with plans to de-escalate based on intra-operative cultures.(see Medical Treatment) In an era of multidrug resistance, antimicrobial stewardship is something each of us should practice and de-escalation should be considered as soon as is possible.

Neonates, in particular, have immature thermoregulation mechanisms and often present with temperature instability as the initial sign of sepsis. (see Care of the Surgical Neonate) Transport to the operating room may further contribute to hypothermia and instability. A bedside operation should be considered, as long as it is safe, for many of those infants [110][111]. If transport is needed, measures should be taken to maintain normothermia including transport in closed isolettes or those with active radiant warmers. In either case, the transport and operating environment should be thermo-regulated [112].

In patients who are critically ill, the treatment algorithm is focused on maintaining proper oxygen delivery and end organ perfusion while addressing the source of sepsis. As such, being able to deliver crystalloid, blood products, and life saving medications is essential. Optimal vascular access includes large bore peripheral intravenous catheters and central venous lines. If obtaining initial access is challenging, an intraosseous line provides immediate access for fluid resuscitation and medication while working on additional access. In addition, arterial lines, both umbilical and peripheral, provide a means of accurate blood pressure monitoring and blood sampling for ongoing patient assessment.

Infection, particularly in neonates who are mainly dependent on granulopoiesis as a defense mechanism, leads to relative suppression of the bone marrow. Not only are the neonatal neutrophils immature, production increases multifold at the expense of other cell lines.This often renders the patient anemic, thrombocytopenic and coagulopathic[113]. While this is a “normal” physiologic response to infection and can be monitored for recovery, it does increase the risk of bleeding and the potential need for transfusion during operative interventions. Therefore, type-specific, Rh compatible blood products should be prepared for the operating room. Many institutions use irradiated, group O, Rh negative or Rh compatible red cells for neonatal transfusions, but type specific blood can be used, if available.[114] That being said, as more data emerges on the increase in morbidity and mortality associated with transfusion, blood products should be administered judiciously [115][116].

Pulmonary support with mechanical ventilation in the setting of sepsis decreases the cardiac output required to support work of breathing, enhances oxygen delivery via reduction in cardiac afterload, allows for an increase in sedation (therefore decreasing O₂ consumption), and improves the V/Q shunt which is often worsened by aggressive fluid resuscitation [76]. As such, early intubation and mechanical ventilation should be considered for all critically ill patients with surgical infections. In addition, large volume crystalloid administration, while needed to maintain intravascular volume, may lead to pulmonary edema and ARDS, further worsening respiratory status. Mechanical ventilation should be considered for children who receive more than 60 ml/kg of fluid within the first six hours of admission.

Steps of the Procedure

see Surgical Infection Procedures

Postoperative Care

The immediate postoperative care of patients with surgical infection focuses on restoring stability and minimizing secondary injury. Hypovolemia and hypoxia lead to further tissue disruption and ischemia and may precipitate a vicious cycle leading to septic shock. Restoring intravascular volume, optimizing perfusion, and providing adequate gas exchange dictates much of the early care. (see Cardiovascular Physiology and Shock)

Antibiotics should be initiated as soon as an infection is suspected. Postoperatively, appropriate therapeutic levels should be attained and maintained [117]. If aminoglycosides or vancomycin are used, the concern for toxicity often leads to subtherapeutic or infrequent dosing leading not only to treatment failure, but also bacterial resistance [72]. Once intraoperative cultures are returned (up to 48 hours), antibiotic therapy should be de-escalated based on the sensitivity data. Unless there is evidence of treatment failure, antimicrobial duration should be limited to five to seven days [118]. Ongoing hemodynamic instability, systemic inflammatory response syndrome, fevers, leukocytosis, an upward trend in inflammatory markers such as C-reactive protein and erythrocyte sedimentation rate, disseminated intravascular coagulation, ongoing capillary leak, and multiorgan failure all suggest treatment failure. In this case, further investigation is required including ultrasound and/or cross-sectional imaging. Additional interventions are likely needed, potentially including percutaneous abscess drainage, in order to effect better source control.

Antifungal therapy, particularly for gastric and proximal small bowel perforations, has recently been reviewed. In adolescent and adult patients, antifungal coverage is typically not indicated. Cultures that grow Candida albicans typically reflect colonization rather than infection. Fungi are typically pathogenic in immunosuppressed hosts, including those with malignancy, autoimmune disease, and following transplant. In addition, neonates and pediatric patients with recurrent or hospital acquired infections have an increased mortality rate due to fungal infections and should receive timely coverage [118][119]. Current guidelines recommend the use of fluconazole as the first line agent for C. albicans. Unfortunately, the empiric use of fluconazole has selected for C. glabrata resistance in some intensive care units, the treatment of which requires escalation to echinocandins.

Many patients with surgical infection present with hypoalbuminemia and a poor nutritional state which is only worsened by the hypermetabolism induced by sepsis. Early nutrition is essential in restoring nitrogen balance, promoting wound healing, and facilitating full recovery.[120] While enteral feeds remain the standard of care for optimal nutrition, many surgical patients (particularly those with intestinal discontinuity or requiring high-dose vasoconstrictors) do not tolerate enteral nutrition early in their recovery. Neonates and children must also sustain growth while recovering from critical illness. As such, if enteral nutrition cannot be initiated, they should be started on parenteral nutrition (PN) until their gastrointestinal tract is functional. In addition to maintaining the patient’s nutrition, PN promotes pro-inflammatory states and contributes to further immunosuppression of the child by inhibiting physiologic cytokine release. The end result renders the host more susceptible to bacterial infection [121]. Therefore, the transition to enteral nutrition should be as expeditious as possible.

How long should surgical drains be left in place?

Another concern in the postoperative care of infected patients is the duration of surgical drainage. The routine use of drains are not encouraged as they may lead to more complications such as the inoculation of sterile compartments or drain erosion. The benefits in most uncomplicated operations are few [17][122]. However, there is still a role for drains in modern pediatric surgery. Some operations are based on wide drainage including necrotizing pancreatitis or an anastomotic leak in a ’’frozen" abdomen. Postoperative abscesses are often managed with percutaneous drains and some empyemas are controlled with chest tubes. In all of these situations, the duration of drainage depends on patient factors. Non-functional drains should be removed. Others are followed until they are no longer draining. If the patient is improving clinically, additional imaging is generally not necessary prior to removing the drain. Otherwise, repeat ultrasound, computerized tomography, and magnetic resonance imaging may guide drain replacement or removal.

Complications

Surgical infections often start as complications. Surgical site infections, deep space abscesses, and anastomotic disruptions can all results from elective, planned operations.

The development of these infections leads to prolonged hospitalization and increased morbidity and mortality. One of the feared complications is development of a necrotizing infection due to Streptococcus or Clostridium. These can be rapidly progressive and lead to severe shock even when addressed promptly. Primary management of these infections is surgical, including wide debridement of all involved tissues. This often leaves a large wound that ultimately requires complex reconstruction as well as frequent and/or prolonged hospitalization. In the more immediate period, adjuncts to standard therapy have been considered to both minimize further tissue damage and optimize the host’s native immune response. Both IVIG and hyperbaric therapy have been considered in these cases. They are not widely available and the data on their efficacy is still emerging [123][124][125].

Surgical site infection prophylaxis can also lead to complications. Even a single dose of antibiotics can predispose to antibiotic associated colitis, the rates of which are rising alarmingly in the United States [126]. A recent review of patients hospitalized in children’s hospitals found an increase in Clostridium difficile infection from 7.24/10,000 admissions to 12.8 between 1997 and 2006 [127]. The overuse of antibiotics along with the improper choice and timing of administration, augments resistance and increases the chance for subsequent infection [128].

Outcomes

Pediatric data on the cost, morbidity and mortality of surgical infection is tied to that of pediatric sepsis. Even in severe sepsis, outcomes are better for children than adult patients. As such, a more aggressive approach is used along with earlier consideration of adjunct therapies such as ECLS. According to a query of the PHIS database in 2014, the prevalence of pediatric sepsis was 7.7% in the US for the period between 2004 and 2012.[129] Despite improved outcomes over time, mortality from in pediatric sepsis still approaches 14.5% and sepsis remains the fourth leading cause of death in children. [129] Premature and neonatal populations, as well as children with congenital heart disease, have the highest rate of mortality.[130]

Prevention is a cornerstone of infection management, starting with proper selection and optimization of patients for elective procedures to minimize risk of surgical site infections (SSI). A recent Joint Commission review found that nearly half of SSIs were preventable [5]. Compared to adult population, rates of SSI in children are much lower, with two percent identified in a recent review of pediatric NSQIP database [131]. However, the cost of each infection is high, costing nearly $30,000 per infection, and contributing to 500,000 SSIs in the United States each year.[5][132] Once an SSI occurs, the risk of mortality increases up to eleven fold [5]. At that point, aggressive early treatment, source control and appropriate antibiotic coverage lead to improved survival and overall outcomes [129].

Research and Future Directions

Emergence of multidrug resistant organisms has become a serious global health issue, necessitating development of both new antibiotics, but also adjunct therapies. MRSA, VRE, and multidrug resistant Acinetobacter baumannii have all resulted in rapidly spreading intra-facility infections, at times necessitating closure of entire ICUs and hospital wards. As current antibiotics have become less effective, we are starting to see organisms resistant to all modern antibiotics, which is a rather frightening concept. Development of newer antibiotics is ongoing and seems to be a natural step in research. However, it is likely a question of time before resistance develops to these agents as well. As such, research into novel therapies is needed urgently.

Lantibiotics are bacterial peptides, produced by many Gram positive organisms. They are showing promise in treatment of multi-drug resistant organisms with high activity against MRSA in recent animal studies. Peptides such as nicin, clausin, and amyloliquecidin bind to lipids within the bacterial cell membranes, inhibiting membrane biosynthesis and leading to bacterial cell death [133]. In addition, nicin has immunomodulatory properties, enhancing the host defenses against the organism [133]. As such, they are being actively pursued as alternatives to antibiotic therapies.

Developments in molecular biology and genetics of bacterial species are further contributing to our understanding of resistance patterns. Recent studies of sporulation mechanisms of Bacillus species offer an insight in their ability to survive extreme conditions of heat and nutrient depletion, due to a mobile genetic element, spoVA^2mob. Deletion of spoVA^2mob renders the spores heat-sensitive. Manipulation of such elements has a potential for developing better control of other spore-based species, such as C. dificile.

In addition to development of new therapies, a search for rapid markers of sepsis is underway and to date, more than 180 bio-markers have been identified [134]. Procalcitonin is one such marker. And while it is shown to be more specific for presence of a bacterial infection, procalcitonin does not help identify organisms responsible for the infection. Newer insights into molecular biology are being directed towards rapid identification of bacteria and fungi within two to six hours of specimen collection, as compared to 24 hours typically needed for standard techniques. These tests have the ability to detect presence of genes that code for antibiotic resistance, for example the mecAgene in MRSA [134]. Their specificity and rapid results make them very relevant in clinical medicine, but they are commercially available at this time.

Patient Care Guidelines

guideline from Bratzler and the Surgical Infection Society [46]

Antibiotic Prophylaxis 1

Antibiotic Prophylaxis 2

Perspectives and Commentary

To submit commentary on this topic contact us at think@apsapedsurg.org.

Additional Resources

For a list of quality improvement projects see the Antibiotic Stewardshipand Surgical Site infection toolkits.

Also see the COVID-19 for Pediatric Surgeons page

video link Surviving sepsis video from the 2019 APSA Annual Meeting Journal of Pediatric Surgery Top Educational Content talks

video link Antibiotic stewardshhip from the 2019 APSA Annual Meeting Journal of Pediatric Surgery Top Educational Content talks

video link Sepsis management discussion from the Surgical Critical Care committee session during the APSA 2020 virtual meeting

Discussion Questions and Cases

To submit interesting cases which display thoughtful patient management please contact the editors at think@apsapedsurg.org.

Additonal questions are in SCORE Subacute Bacterial Endocarditis Prophylaxis conference prep

References

Media

antibiotic prophylaxis 1

antibiotic prophylaxis 2