Critical Care Medicine

Infection in the Post-operative patient

Perioperative infections


Surgical site infection, mediastinitis, post-operative peritonitis

1. Description of the problem

What every physician needs to know

A postoperative infection is defined as any infection that occurs within 30 days of operation and may be related to the operation itself or the postoperative course. The incidence of postoperative infections varies greatly depending on the definition used, surveillance methods used to detect infection, type of procedure performed and the patient's inherent risk factors for development of infection. Some sources refer to all postoperative infections as surgical site infections, whereas others use this term to refer to infections involving the skin or subcutaneous tissues only.

Postoperative infections can involve the wound itself, deeper infections within body cavities, or more distant infections such as pneumonia or catheter associated urinary infections. The timing from operation to manifestation of signs and symptoms suggesting infection can help elucidate the source of the infection. Whereas urinary tract infections and superficial wound infections can manifest as early as 72 hours following the operation, deep seated abscesses and pneumonia tend to manifest 5-7 days following operation.

Mediastinitis refers to an infection of the central structures in the chest excluding the lungs and pleural space. In the post-operative setting, it is most commonly seen following cardiac surgery and generally presents within 1-2 weeks following operation, although presentation can be delayed upwards of 4-6 months. Its incidence varies from less than 1-3%.

Diagnosing a surgical site infection can be challenging, as the signs and symptoms are frequently nonspecific and may lack sensitivity depending on the clinical scenario.

Therefore, it is critical that the clinician maintain a high index of suspicion for infection based on risk factors noted below. Frequently, deviation from the expected postoperative course is one of the first signs that can be used to suggest the presence of a postoperative infection. Generally, patients should be expected to manifest a diuresis (referred to as mobilization of third space fluids) 48-72 hours following a major operation. Although non-specific, ongoing inflammation is one of several causes for failure to manifest this finding and may intimate the presence of infection.

Clinical features

The clinical manifestation and significance of the infection varies based on the patient's immune status and depth and severity of infection. Superficial infections of the skin can manifest as cellulitis alone with little to no discharge. Deeper or more severe infections can manifest as purulent discharge in addition to cellulitis. Rarely, this can result in wound dehiscence which can be especially dramatic following operations involving sternotomy, where a deep infection can result in sternal dehiscence; or laparotomy where fascial dehiscence can result in organ evisceration.

Deeper infections, such as abscesses in a body cavity or parenchymal infections such as pneumonia, can present with nonspecific systemic findings such as fever and leukocytosis or more organ specific findings such as a prolonged ileus in a patient with an intraperitoneal abscess or oxygenation failure in a patient with severe pneumonia.

Persistent serous discharge from the wound, although nonspecific, suggests infection which has resulted in fascial dehiscence. Sternal dehiscence can be difficult to diagnose, as sternal instability is present in fewer than 50% of patients with mediastinitis, systemic signs are frequently nonspecific and the most common finding is serous discharge from the wound.

Patients with advanced immunosuppression either due to pharmacotherapy or disease (such as HIV) can have remarkably few signs and symptoms of infection. Commonly, this is seen in patients who are pharmacologically immunosuppressed following transplantation. In this cohort, the sensitivity of the physical and laboratory exams decreases significantly such that these patients may have a normal physical and laboratory exam despite the presence of catastrophic infection. As such, the diagnosis of infection in this cohort requires a high index of suspicion and liberal use of imaging and microbiological assay.

Key management points

As with any infection, the priority in addressing a postoperative infection should be identifying the source and controlling ongoing contamination. This is particularly relevant in a patient who has had gastrointestinal surgery because source control may necessitate repeat laparotomy. In this instance, a CT scan is frequently needed to evaluate for intestinal leak and peritoneal abscess.

Source control may also require removal of embedded prosthetic material such as a vascular graft or non-dissolvable mesh. Concomitant with this effort, empiric antibiotics therapy should be strongly considered and fluid resuscitation should be instituted as clinically indicated.

2. Emergency Management

A superficial surgical site infection is rarely the cause of septic shock. Management priorities, as with other forms of infection, involve source control and appropriate use of antibiotics. In some instances, source control can be obtained easily by opening the incision itself to allow the infection to drain or, in other instances, may require percutaneous drainage of deeper infections. Rarely, source control may require repeat operation to evacuate a large pocket of infection or control an ongoing leak as may occur following operation on the gastrointestinal tract.

Removal of embedded prosthetic material is rarely an emergency unless the patient is noted to be recurrently bacteremic or septic from this source. As will be further discussed later, antibiotic therapy should target the most likely organisms responsible for the infection. In most instances, infection is due to gram-positive cocci, frequently including MRSA, and less commonly gram-negative rods.

3. Diagnosis

Diagnostic criteria and methods

The diagnostic criteria and tests used to confirm the diagnosis of the surgical site infection differ based on its location, depth and severity. Ultimately, a high index of suspicion is needed to diagnose a surgical site infection due to the lack of specific signs and symptoms associated with this condition.

Superficial wound infections manifest as cellulitis alone and are diagnosed clinically based on the characteristics of the wound/incision. Such infection may be associated with pain or tenderness, swelling, redness or heat at the site of the wound, or the patient may manifest a mild fever or leukocytosis. Deeper wound infections will manifest as purulent discharge from the incision, frequently (but not always) with fever and/or leukocytosis. Although the diagnosis of a wound infection remains clinical, culture of the discharge is helpful in determining the optimal antibiotic therapy.

CT scan which is ideally performed with intravenous contrast remains a diagnostic modality of choice for most deeper infections regardless of location. Postoperative infections of the thoracic cavity can be difficult to diagnose. This is the particularly true following cardiac or thoracic operations, as the oxygenation index and chest x-ray become predictably abnormal in the postoperative setting in this patient population (Figure 1).

Figure 1.

Normal chest X-ray following lobectomy

Reports on the utility of radiographic imaging of the chest to diagnose either pneumonia or mediastinitis (Figure 2 and Figure 3) following thoracic operation find a sensitivity that is less than 50% in the early postoperative period. Rarely, sternal dehiscence (Figure 4) can be noted as a separation of the wires used to close the sternum. However, the sensitivity of CT scan to diagnose infection approaches 100% approximately 2 weeks following operation when the underlying inflammatory process resulting from the operation itself should have subsided.

Figure 2.

Chest X-ray following lobectomy with probable underlying pneumonia

Figure 3.

CT scan showing mediastinitis

Figure 4.

Sternal dehiscence

How do I know this is what the patient has?

Wound dehiscence can be a complication of surgical site infection and should be suspected in instances where there is ongoing leakage of fluid from the wound. Frequently, the fluid will be either serous or serosanguineous in nature. This characteristic can help differentiate a wound dehiscence from an abscess which has broken throughto the surface and will be purulent in its appearance.

Sternal dehiscence can be noted as instability of the sternum with or without crepitus on palpation. However, this finding is present in less than half of all cases. A more consistent and less specific finding is ongoing discharge of serous fluid from the incision. Laboratory testing is of limited value in diagnosing a wound infection and/or wound dehiscence. Most patients may manifest a mild leukocytosis or fever.

Confirmatory tests

In instances where wound dehiscence is a concern but the diagnosis is not definite, the base of the wound can be gently probed to assess fascial integrity. Ability to penetrate the fascia or sternum suggests wound dehiscence.

4. Specific Treatment

Treatment of any infection begins with strategies to prevent the event from occurring. Wound infections can be prevented or minimized by leaving the skin open. A fundamental principle of surgery is that an open wound has a much lower incidence of infection than a closed wound. Therefore, when the probability of infection is high, the surgeon may opt to leave the skin open while closing the underlying fascia to protect the deeper tissues.

While chronic malnutrition and hyperglycemia are known risk factors for development of infection, there is little data to support the notion that preoperative nutritional supplementation or tight glycemic control decrease the incidence of infection, and a recent study found an increase in mortality in critically ill patients randomized to tight glucose control. Next, although there is preliminary evidence that chlorhexidine gluconate baths result in a lower bacterial count on the skin relative to simple soap and water or povidone-iodine, studies performed to date have not found a consistent decrease in the incidence of surgical site infection.

Some studies, however, suggest a decrease in the incidence of catheter-related blood stream infection and pneumonia in critically ill patients who are routinely bathed with chlorhexidine gluconate. Thus, the role of chlorhexidine in prevention of perioperative infection remains uncertain and further study is needed. Selective decontamination of the digestive tract (SDD) or the oropharynx alone (SOD) has been shown to decrease the incidence of pneumonia and bacteremia and possibility decrease mortality in perioperative critically ill patients.

Most SDD regimens involve use of non-absorbable oral antibiotics (e.g. polymyxin E, colistin, tobramycin, amphotericin B) and parenteral administration of a second generation cephalosporin. Regimens for SOD involve an antimicrobial oral paste that may be supplemented with chlorhexidine mouth wash. Although effective, concern for emergence of resistant bacteria has tempered adoption of SDD in the United States.

Surgical site infections are unique in that frequently the most effective therapy is opening of the wound and expression of the infection. Antibiotics are of limited utility if there is an underlying collection of purulent material. Once the wound is opened, wet-to-dry dressing changes with saline are the most common method of treating the infection locally and antibiotic therapy may not be necessary in all cases.

The exception to this involves infections that have resulted in fascial dehiscence. These infections may require surgical intervention for fascial reconstruction. Expert surgical consultation should be obtained in these instances. This is particularly true for mediastinitis where antibiotic therapy alone is inadequate and should be viewed as a supplement to surgical treatment. Surgical options in the treatment of postoperative mediastinal infection range from debridement with repeat closure of the sternum, percutaneous drainage of fluid collections or chest wall reconstruction with myocutaneous or omental flaps.

In choosing a specific antibiotic to be administered empirically, one must remember that more than 50% of infections are caused by gram-positive pathogens and the incidence of methicillin-resistant Staphylococcal Aureus (MRSA) as the causative organism in surgical site infections has increased to more than 50% in some reports. In addition, the clinician should consider the likely causative organism(s) in light of the operation performed and tailor the antibiotic regimen appropriately.

Whereas infection in patients who underwent non-alimentary tract operation are most likely to be gram-positive in nature, the incidence of gram-negative infection increases following gastrointestinal operation. Ultimately, antibiotic therapy should be tailored to microbiological culture results. However, it is possible to obtain a false sterile sample if the patient has received a prolonged course of antibiotics. This is the reason that, ultimately, determination of a postoperative infection remains a clinical diagnosis.

In all cases, the patient's nutritional status must be optimized to allow for healing.

Drugs and dosages

Pending final culture results, an empiric antibiotic regimen consisting of intravenous vancomycin dosed to obtain serum trough levels of 10-20 mg/dL and a second agent to provide gram-negative coverage is prudent in most cases. More often than not, the second agent should consist of an advanced generation penicillin plus beta-lactamase inhibitor or cephalosporin rather than a quinolone. This is because excessive use of quinolones is more associated with development of antimicrobial resistance than use of other antibiotic classes.

Specific options for gram-negative coverage include ampicillin/sulbactam, piperacillin/tazobactam or cefepime. When needed, metronidazole can be added for anaerobic coverage if the spectrum is on already covered in the anabolic regimen prescribed. Carbapenem agents should be considered in instances where there is concern for the presence of extended spectrum beta lactamase gram negative organisms, such as patients who have received numerous antibiotic courses.

Vancomycin resistant enterococci are rarely the cause of infection in the post-operative period but should also be considered in patients with a prolonged hospital stay and in those with multiple previous exposures to antibiotics.

Rarely, a topical antimicrobial agent may be needed to help eradicate the infection. Such agents include dilute sodium hydrochloride (Dakin's) solution or boric acid. However, these agents also impede wound healing and therefore should not be used for an extended period of time. These agents are most commonly used to initially treat severe wound infections particularly those involving Pseudomonas Aeruginosa. In these instances, consideration should also be given to sharp debridement of the wound to remove all necrotic tissue, and surgical consultation should be obtained.

There is no consensus on the duration of antibiotic therapy to treat a particular postoperative infection. However, most authors recommend a 2 week course of therapy for the treatment of a deep abscess following drainage and a 5-7 day course of therapy for cellulitis, which responds well to antibiotics and local wound care. Due to poor tissue penetration, mediastinitis requires a prolonged course of antibiotics with most cases requiring 4-6 weeks of intravenous therapy. Duration of therapy may need to be extended in cases of sternal osteomyelitis.

Refractory cases

Consider deeper infection or atypical/resistant organisms in instances where the infection does not improve following implementation of adequate local wound care and antibiotics. A CT scan should be obtained as clinically indicated to evaluate for deep infection. Based on culture data, the antibiotic regimen should be altered to include other organisms such as vancomycin resistant enterococci and/or resistant gram-negative organisms.

Based on the severity of infection, an alternate antibiotic regimen may include linezolid to cover vancomycin resistant enterococci and/or a carbapenem to cover resistant gram-negative organisms. Although extremely rare, consideration should also be given to the possibility of fungal infection, particularly in the immunosuppressive population with pneumonia or urinary tract infection.

5. Disease monitoring, follow-up and disposition

Expected response to treatment

In most instances, a local wound infection should be controlled, purulent discharge should decrease in amount and character, and cellulitis should begin to resolve within 36-48 hours after the start of therapy. Similarly, following percutaneous drainage of an abscess, systemic signs of infection and illness should begin to improve within the same time period. Because of its characteristics, mediastinitis remains an exception in that it requires a prolonged period of therapy to control and eradicate infection and therefore the patient's clinical condition may not improve for many days to weeks following institution of therapy.

Incorrect diagnosis

Lack of clinically discernible improvement in the characteristics of the wound or persistent signs of infection 2-3 days following the start of local wound and antibiotic therapy suggests inadequacy of treatment. An alternate diagnosis for the cause of the patient's deranged condition should be made or an alternate treatment regimen considered.


In general, wounds cannot be closed after they have been opened for infection. Therefore, patients will need ongoing dressing changes to allow the wound to heal by secondary intention. This process can take days to months depending on the nature of the wound and the patient's ability to heal.

Postoperative mediastinitis is associated with bacteremia in 40-60% of cases. Therefore, these patients should be observed closely for signs of systemic infection and some authors recommend routine surveillance blood cultures. Moreover, deep sternal wound infections can be identified in up to two thirds of patients who manifest bacteremia after cardiac operation.




Surgical site infections are the third most frequently reported healthcare associated infection. The incidence of surgical site infection varies widely amongst different centers and also based on the operation being performed. For example, lower extremity arterial bypass procedures have one of the highest incidence of infection, with the range varying from 5-20%. Although the exact cause for this is not known, it is most likely related to chronically impaired blood flow with microvascular disease.

As would be expected, the incidence of surgical site infection increases as the nature of the wound changes from clean to dirty.

The most common cause of wound infection regardless of procedure performed remains gram-positive cocci which comprise more than 50% of all infections. Specifically, Staphylococcal Aureus and coagulase-negative staphylococci are the most common organisms isolated from a wound infection.

There has been an increasing incidence of MRSA infection reported in centers throughout the world and there appears to be an association between nasal and skin colonization with this organism and subsequent postoperative infection. However, reports on the utility of eradication of chronic colonization have not consistently found a concomitant decrease in surgical site infections.

Gram-negative and anaerobic organisms may be more prevalent following operation on the alimentary tract or in cases of necrotizing soft tissue infection, but have also been noted in up to 40% of mediastinal infections following cardiac surgery. The most common gram-negative organisms associated with surgical site infections are: Escherichia coli, Enterobacter, and Pseudomonas Aeruginosa. Anaerobic surgical site infection is rare in the absence of pre-existing perforated viscus, but when it does occur Bacteroides fragilis and Peptostreptococcus are common pathogen. The incidence of fungi, particularly Candida, as a cause for postoperative infection increases if the patient has received prolonged preoperative antibiotics or TPN.

As opposed to infection following perforation of a hollow viscus, postoperative infections tend to be mono-bacterial in nature. For example, postoperative mediastinitis is polymicrobial and only 10% of cases.

Risk factors for surgical site infections

Surgical site infections can be caused for a variety of factors. Common pathophysiologic factors to all surgical site infections can be broken down into 3 general categories: 1) immune dysfunction (intrinsic factors); 2) environmental and external factors related to the operation itself (extrinsic factors); and 3) surveillance of the microbial organism.

See Table I. Intrinsic and extrinsic factors

Table I.

Intrinsic and extrinsic factors
Intrinsic factors Extrinsic factors
Age Skin preparation
Nutritional status Hair removal and method
Diabetes mellitus Perioperative antibiotic prophylaxis
Smoking Use of drains
Immunocompromise Surgical technique
Obesity Dead space
Microbial colonization Presence of foreign body
Presence of other Infections Transfusion
Intraoperative hypothermia (<36°C) Duration of operation
Shock in the perioperative setting Gastrointestinal surgery
  • Although an increasing percentage of the population is colonized with MRSA, most routine preoperative antimicrobial prophylaxis does not target this organism.

  • Patients with preoperative risk factors such as cachexia, wasting or malnourished state should be considered to be compromised and therefore at high risk for infection. In addition, hemodynamic instability prior to operation and the need for massive resuscitation and transfusion results in an immunosuppressed state which increases the probability for infection.

  • The presence of foreign bodies such as central venous catheters or surgical implants increases the risk for infection.

  • When needed, hair should be removed at the time of operation using an electric razor because use of a sharp blade is associated with an increased risk for infection.

  • Antibiotic prophylaxis should be administered within 60 minutes of skin incision.

  • Surgical technique resulting in excessive handling of tissues, hemorrhage, excessive use of cautery, and duration of operation exceeding 2 hours are all independently associated with an increased risk for infection.


Although the overall prognosis of surgical site infection is quite good, it is associated with increased morbidity, increased hospital length of stay and significantly increased cost. Certain infections, however, are associated with severe morbidity and increased mortality; a common example being postoperative mediastinitis which carries a mortality risk of 15-40%. In addition, this complication is associated with empyema and infectious pericarditis from direct extension of the infection. Patients requiring abdominal or chest wall reconstruction for infection involving the fascia are at increased risk for chronic pain, paresthesia and overall weakness.

Special considerations for nursing and allied health professionals.

Please comment on considerations when packing wounds with wet-to-dry dressings stressing the need to pack a wound adequately without excessive pressure and also stressing that wounds heal from the base out and that it is expected that the wound will become progressively more superficial and shallow as it heals. The nurse provider is to consider this and allow the wound to heal by decreasing the amount of packing material as this process progresses.

What's the evidence?

Description of the problem

Mangram, AJ, Horan, TC, Pearson, ML. "Guidelines for prevention of surgical site infection". Am J Infect Control. vol. 27. 1999. pp. 97-132.

(Reporting of common microbiological isolates causing surgical site infection over two time periods published by the Center for Disease Control and Prevention.)

Fridkin, SK, Hill, HA, Volkova, NV. "Temporal changes in prevalence of antimicrobial resistance and 23 US hospitals". Emerg Infect Dis. vol. 8. 2002. pp. 697-701.

(A discussion of reasons underlying the development of antimicrobial resistance in hospitalized patients and how to assess for the presence of extended spectrum beta lactamase [ESBL] organisms.)

El Oakley, RM, Wright, JE. "Postoperative mediastinitis: Classification and management". Ann Thoracic Surg.. vol. 61. 1996. pp. 1030-6.

(This is an excellent review with proposed diagnostic and treatment schema for postoperative mediastinitis.)

El Oakley. "Mediastinitis in patients undergoing cardiopulmonary bypass: risk analysis and mid term results". J Cardiovasc Surg. vol. 38. 1997. pp. 595-600.


Akman, C, Kantarci, F, Cetinkaya, S. "Imaging in mediastinitis: a systematic review based on aetiology". Clin Radiol. vol. 59. 2004. pp. 573-85.

(An evidence based review of optimal radiographic modalities in the evaluation of mediastinitis.)

Specific treatment

Webster, J, Osborne, S. "Preoperative bathing or showering with skin antiseptics to prevent surgical site infection". Cochran database Syst Rev. 2007. pp. CD004985.

(This is a meta-analysis of studies examining the use of chlorhexidine gluconate in the prevention of surgical site infection.)

Darouiche, R, Wall, M, Itani, K. "Chlorhexidine-alcohol versus povidone-iodine for surgical site antisepsis". N Engl J Med. vol. 7. 2010. pp. 18-26.

(This is one of two recent studies suggesting that use of a chlorhexidine-alcohol skin preparation in the preoperative setting reduces the incidence of surgical site infections.)

Bratzler, DW, Houck, PM. "Antimicrobial prophylaxis for surgery: an advisory statement from the National Surgical Infection Prevention Project". Am J Surg. vol. 189. 2005. pp. 395-404.

(This is an excellent review of the evidentiary basis and recommendations for perioperative antibiotic prophylaxis.)

Robicsek, A, Jacoby, G, Hooper, D. "The worldwide emergence of plasmid-mediated quinolone resistance". Lancet Infect Dis. vol. 6. 2006. pp. 629-40.

(Review of the mechanisms of vertical and horizontal transmission of resistance between quinolone and other classes of antibiotics.)

Nathens, A, Marshall, J. "Selective decontamination of the digestive tract in surgical patients: A systematic review of the evidence". Arch Surg.. vol. 134. 1999. pp. 170-6.

(Review of studies in surgical patients showing SDD is associated with a decreased incidence of bacteremia and pneumonia.)

De Smet, A, Kluytmans, J, Cooper, B. "Decontamination of the digestive tract and oropharynx in ICU patients". N Engl J Med.. vol. 360. 2009. pp. 20-31.

(Largest study performed to date comparing SDD to SOD. The study finds that there is no difference in mortality between the two arms but there was a significant cost associated with SDD.)

Al, N, Heddema, E, Bart, A. "Emergence of multidrug resistant gram-negative bacteria during selective decontamination of the digestive tract on an intensive care unit". J Antimicrob Chemother. vol. 58. 2006. pp. 853-6.

(This study found that the incidence of MRSA increased significantly following adoption of SDD.)

"Intensive versus conventional glucose control in critically ill patients". New Engl J Med. vol. 360. 2009. pp. 1283-97.

(Randomized controlled study showing an increase in mortality in critically ill patients randomized to tight glucose control.)

Disease monitoring, follow-up and disposition

Quinn, A, Hill, A, Humphreys, H. "Evolving issues in the prevention of surgical site infections". Surgeon. vol. 7. 2009. pp. 170-2.

(This article reviews the current literature in relation to three aspects of surgical site infection: compliance with antibiotic prophylaxis, post discharge wound surveillance and methods to prevent surgical site infections.)

Broex, E, Van Asselt, A, Bruggeman, C. "Surgical site infections: How high are the costs?". J Hosp Infect. vol. 72. 2009. pp. 193-201.

(The study reviews the literature in an attempt to discern cost associated with surgical site infections. It takes into account differences in methodology to assess cost amongst various reports.)


Leaper, DJ. "Risk factors for an active etiology of surgical site infections". Surg Infect. vol. 11. 2010. pp. 283-7.

(A review of the literature regarding risk factors for an acute etiology of surgical site infections.)

Brandyk, D. "Vascular surgical site infection: risk factors and preventive measures". Semin Vasc Surg. vol. 21. 2008. pp. 119-23.

(A review of the literature focusing on postoperative infection in patients with prosthetic vascular implants. The article discusses methods to prevent and treat such infections.)

Owens, C, Stoessel, K. "Surgical site infections: Etiology, microbiology and prevention". J Hosp Infect. vol. 70 Suppl 2. 2008. pp. 3-10.

(An up-to-date article describing the definition, incidence and cause of surgical site infection. The article also discussed discusses methods to prevent infection.)

Related Resources

You must be a registered member of Cancer Therapy Advisor to post a comment.

Regimen and Drug Listings


Bone Cancer Regimens Drugs
Brain Cancer Regimens Drugs
Breast Cancer Regimens Drugs
Endocrine Cancer Regimens Drugs
Gastrointestinal Cancer Regimens Drugs
Gynecologic Cancer Regimens Drugs
Head and Neck Cancer Regimens Drugs
Hematologic Cancer Regimens Drugs
Lung Cancer Regimens Drugs
Other Cancers Regimens
Prostate Cancer Regimens Drugs
Rare Cancers Regimens
Renal Cell Carcinoma Regimens Drugs
Skin Cancer Regimens Drugs
Urologic Cancers Regimens Drugs

Sign Up for Free e-newsletters