Update on preventing surgical site infections

By Daniel A. Anaya, MD, and E. Patchen Dellinger, MD

Here's what you need to know to identify patients at higher risk for SSIs and provide effective preventive strategies. These measures include good glucose control and well-timed prophylactic antibiotics.

We've long known that asepsis, antisepsis, and prophylactic antibiotics play a major role in reducing surgical wound infections. Even so, surgical site infections (SSI) will always be around to some degree and call for additional preventive strategies.

The third most common nosocomial infection in the United States, SSIs account for 38% of all infections in the 27 million patients who undergo surgery each year. 1 SSI rates for abdominal hysterectomy range from 0.5% to 5.3% depending on risk category, according to the National Nosocomial Infection Surveillance System (NNIS) data for 1992–2003. 2 Moreover, these infections extend hospital stays by up to 10 days—and that alone can drive up costs significantly. 1-4 But the good news is that specific strategies can significantly decrease these costs. One study that examined the cost-effectiveness of instituting such measures estimated that up to $3 million could be saved over a 10-year period, through reducing SSI rates by up to 56%. 5

The growing interest in SSIs is likely due to the greater complexity of operations, the higher number of surgical patients who are older or immune-compromised (AIDS, cancer, transplants, etc.) with significant co-morbidities, and increasing use of implanted foreign body materials.

Infection rates are one standard for judging a hospital's quality assurance, and lowering them can reduce costs at a time when efficient resource utilization is key to institutional survival. Recognizing the need for gynecologic surgeons to identify patients at higher risk for developing infections and to embrace the best preventive measures, our goal is to assess various approaches for preventing SSIs based on the latest data. We'll also examine risk factors for SSI that help us better understand existing and emerging strategies.

Defining SSIs

SSIs are infections that occur after surgery at any site along the surgical tract. Before 1992, when the term wound infection referred to all infections in the skin or subcutaneous tissues that occurred postoperatively, clinicians focused on preventing infections at these specific sites. In 1992, however, the Centers for Disease Control and Prevention broadened the term to surgical site infection , which now applies to all infections within the tract exposed during the operation. They further classified SSIs into superficial incisional, deep incisional, and organ and space-related ones (Table 1). Up to 80% or more of SSIs are superficial, and their morbidity and mortality are usually controlled once they're appropriately treated. Organ/procedure SSIs (deeper infections in organs and body spaces) are less frequent but more dangerous; in fact, 93% of SSIs that lead to death are of this type. 6

Risk factors for SSI

We can group these into three broad areas: factors related to microorganisms, local/surgical factors, and patient-related factors. Despite the many risk factors and conditions identified in each category, only a few are proven independent risk factors for SSIs. (Examples of independent factors include wound class, remote site infections, preoperative shaving, and hypothermia.) So it's mainly the interaction between these risk factors that determines the risk of SSI and not the sole presence of one of them. 7

Microorganism-related factors. Pathogen-related factors are those linked with virulence and degree of contamination. By virulence, we mean specific bacterial characteristics by which pathogens cause infection in a specific milieu.

Clostridial infections, as well as those caused by group A streptococcus, are more aggressive and spread more rapidly than those caused by other organisms, resulting in their being evaluated within the first few days after an operation. Staphylococcus epidermidis produces slime, a substance that allows it to stick to foreign body materials, which explains why it's often the cause of SSIs associated with foreign bodies (implants, prostheses, vascular grafts, etc.). Specifically, the degree of contamination refers to the number of bacteria present in the wound before, during, and shortly after the operation. The fact that it takes roughly 100,000 CFU of bacteria/gram of tissue to cause SSIs led to the classification of wounds as clean, clean-contaminated, contaminated, and dirty, according to their degree of contamination. Certain conditions increase a wound's risk for contamination and likely risk for SSIs: concomitant remote site infections, a prolonged preoperative hospital stay, a preoperative intensive care unit (ICU) stay, and prolonged preoperative colonization of body surfaces. 8,9

Local/surgical factors. These include all factors related to the surgical site's microenvironment before, during, and shortly after the operation (preoperative shaving, hematoma, seroma, drains, foreign bodies, sutures [multifilament] and spillage of contaminants). Interestingly, most of them can be modified through good surgical practice . For example, we've now abandoned the once-standard practice of preoperative shaving—which actually increases the wound infection rate by allowing colonization of the resulting microlacerations. On the other hand, clipping hair at the surgical site right before the operation—or not removing any hair at all preoperatively—does not increase SSIs. 10-12

Other surgery-specific practices like approach (laparoscopic vs. open), appropriate use of drains, sutures, avoidance of hematoma, seroma, dead spaces, thorough debridement of infected/necrotic tissue, and type of closure all can potentially decrease the SSI rate if the ob/gyn surgeon follows good surgical practices.

Patient-related factors. Newer preventive measures to reduce SSIs are targeting patient-related factors like age, diabetes mellitus, hyperglycemia, obesity, malnutrition, smoking, perioperative transfusions, immune suppression, steroid use, hypothermia, hypoxia, hypovolemia/shock, and multiple co-morbidities. Even though most are not proven independent risk factors for SSI, strategies that address multiple risk factors are starting to reap results. 1,3,5,13-23

Scores for risk assessment. A substantial roadblock to identifying risk factors for SSI is the heterogeneity of both patient types and surgical procedures in most studies. A condition that may seem to increase wound infection in a specific group of patients undergoing a particular surgical procedure may be a minor factor in other patient populations or procedures. To address that problem, risk scores have been created to classify patients into more homogeneous groups and to assess their SSI risk.

The first widely used risk scheme was the degree-of-contamination wound classification we described earlier. The SSI rate increases as the degree of probable contamination rises, with reported SSI rates of: 1% to 5% for clean; 3% to 11% for clean-contaminated; 10% to 17% for contaminated; and more than 27% for dirty wounds. 3,15 But even though degree of contamination is a significant risk factor for SSI, never rely on that alone to determine risk, since it represents only one of the categories we've previously outlined. Within the same class of wound, there's a wide range of SSI rates, depending on the other factors (microorganism, local/surgical, and patient-related).

The CDC's NNIS score, the largest and most heterogeneous of various scores, is the one used most universally. This fulfills one of the primary goals of such scores: use for comparisons in SSI studies. 13-15,23

How the NNIS works. This score uses three independent risk factors, each of which—if present—receives a value of 1, and predicts different SSI rates according to the sum of the points from 0 to 3. The variables are:

• duration of procedure greater than the 75th national percentile for that procedure;
• contaminated or dirty wound; and
• American Society of Anesthesiology score (ASA score) of III or greater (Table 2).

We can expect scores developed in the future to narrow the ranges for SSI for patient groups according to their specific risk factors. In addition, specific procedures (like prior rupture of membranes in a patient having cesarean section) will always have procedure- specific risk factors that will permit more precise estimation of risk.

Which preventive measures are most effective?

Dividing SSIs into "potentially preventable" and "apparently unpreventable" ones, we'll define "potentially preventable" SSIs as infections that occur when a known, effective preventive practice was not used. 24 Alternatively, an "apparently unpreventable" infection is one that occurs despite pulling out all the stops. A goal of any ob/gyn should be to eliminate "potentially preventable" infections. Since ob/gyn surgeons can gauge the risk of infection within a short time of the operation, we can easily document and monitor preventive measures to determine their effectiveness before, during, and shortly after surgery (Table 3).

Strategies aimed at microorganisms

In most SSIs, the infection-causing bacteria come from a patient's endogenous flora at the surgical site. Thus, Staphylococcus spp are the primary pathogens isolated in SSIs after clean procedures (skin flora). Gram-negative enteric bacilli are often found after more "contaminated" procedures (GI flora, genitourinary flora, etc). 1 The surgical team is a less frequent source of infecting microorganisms, as are contaminated substances that come in contact with the surgical site (irrigation solutions, prostheses, etc.), and/or the operating room environment (exogenous sources).

Shortening the preoperative stay, treating remote site infections preoperatively, maintaining asepsis and antisepsis, and using prophylactic antibiotics can help decrease the contamination and virulence of organisms from endogenous sources. Measures to decrease contamination from exogenous sources include asepsis and antisepsis as well as postoperative dressings. 1,3,8,25-37 Approaches targeted at specific groups of patients include preoperative mupirocin for nares colonization (in patients with Staphylococcus aureus nares colonization/cardiac surgery patients) and preoperative antiseptic showers. 9,38,39

Techniques for preventing contamination. Two major advances in SSI prevention over the last 100 years are asepsis/antisepsis and antibiotics. Asepsis and antisepsis refer to all those measures that prevent exogenous and endogenous (skin) contamination, which include: the OR environment, sterilization and disinfection, surgical site preparation, hand/forearm scrubbing, and sterile barriers.

We've described the specific architectural characteristics of the ideal OR. While size and air management (filtered, flow, positive pressure, and air cycles/hour) are the most important factors, traffic policies and equipment handling, and sterilization and disinfection techniques all play important roles. It's beyond the scope of this article to fully explore each recommendation for these areas. But it's important to know that hospitals must establish standard protocols, which should include some measures that, although not epidemiologically proven to decrease SSI risk, are an intrinsic part of the OR discipline that over time has helped maintain aseptic and antiseptic techniques. 40,41 These standards should reflect basic principles established by the CDC and should also take into account local factors.

Proper surgical site preparation and hand/forearm scrubbing can directly prevent the OR team from contaminating the surgical site. Antiseptics like tincture of iodine, povidone-iodine, chlorhexidine gluconate, and combinations of these with isopropyl alcohol are preferable for skin preparation. 29,31-34 Chlorhexidine is better than iodinated solutions for skin prep, particularly for preventing infections during less invasive procedures (specifically central line access). 42 While we can't extrapolate this finding to other "clean" operative procedures, it is plausible. A recent article concluded that although chlorhexidine gluconate has effectively minimized SSI in vaginal surgery, clinicians should consider the potential for adverse reaction. 43

Adequate surgical site preparation begins with cleaning the skin of any gross contamination before applying antiseptics and concentric prepping starting centrally and moving outward over areas in need of prepping. The current recommendation for hand/forearm scrubbing is to scrub for 5 minutes for the initial operation and for 3 minutes for each subsequent procedure. 44 An alternative protocol involving a 30-second alcohol hand rub preparation is just as effective in reducing bacteria on the hands of the surgical team as the traditional 3- to 5-minute scrub. 45 The initial scrub should include a thorough cleaning underneath fingernails, and removal of artificial nails. 36

Sterile barriers. Always use sterile gowns and drapes, since they've been proven to decrease SSIs. 46 Sterile drapes impregnated with antiseptics, however, have no advantage in reducing SSIs. Up to 90% of surgeons accidentally puncture their gloves during an operation. While there are no data connecting glove puncture with SSI, double-gloving will protect ob/gyn surgeons from blood exposure and will probably decrease wound contamination. 47,48

When to use prophylactic antibiotics. Although antibiotics have been around since the late 1950s, subsequent human trials have defined their value in clinical practice. 25,49 Space limitations prohibit an extensive summary of the key issues surrounding their use. However, ob/gyn surgeons should know that—compared with placebo—prophylactic antibiotics definitely reduce risk of infection for any procedure that enters the GI tract from the oropharynx to the rectum, for procedures that enter the female genital tract, and for vascular procedures in the abdomen and leg. (And that's true also for insertion of orthopedic prostheses, open heart procedures, and craniotomies.)

Whether to use prophylactic antibiotics in other clean procedures, however, is controversial. Some studies have shown a decline in infections when such treatment is used before procedures involving areas of higher skin contamination, such as breast surgery. 28 Because the benefit may not outweigh the cost and side effects, antibiotic prophylaxis isn't ordinarily used for procedures with an SSI rate below 1%, unless the morbidity of the infection is severe, such as a joint infection after total joint replacement or mediastinitis after open-heart procedures. The NNIS score aids in determining which clean procedures are likely to present a higher SSI rate, according to the score's other two variables, procedure time >75th percentile and ASA number (Table 4). 29,50 A good rule to determine indications for prophylactic antibiotics in clean procedures is to give them to patients in whom a prosthetic material will be used, the NNIS score is elevated (>1), or immunity is compromised (cancer, AIDS, etc.). They also are useful in cardiovascular patients for whom the consequences of a potential SSI would significantly affect the final outcome and those whose operative areas involve higher bacterial count (groin, axilla, breast, perineum, gluteal folds). 30

Match the drug with the likeliest bug. Your choice of antibiotic should cover the most likely bacteria present at the surgical site, but if possible avoid broad-spectrum antibiotics usually reserved for treating established infections (such as carbapenems and penicillin/beta-lactamase inhibitor combinations). Cefazolin is the most commonly used antibiotic for clean procedures when enteric and anaerobic bacteria are not expected, whereas cefotetan or cefoxitin is commonly used whenever they are . You can zero in on the most appropriate drug by reviewing data on specific hospital bacterial causes of SSI and their specific sensitivities at a given institution. A prolonged half-life is ideal to ensure effective antibiotic levels for the duration of the operation. Prolonged operations require an additional dose, which you should give at intervals of one to two half-lives of the drug being used. 51 Remember, you'll need higher than normal dosages for obese patients. 52

Ideal timing of antibiotics. Give prophylactic antibiotics within 1 hour before you'll make your first incision to reach an effective level at the surgical site. 53,54 Studies show that only 40% to 50% of prophylactic antibiotics are given appropriately and most of the errors involve the time at which they're given. 55 The best way to avoid this is to establish a standard procedure by which they're given. Usually, starting the infusion during anesthesia induction is a good way to achieve ideal timing. 26,27,29,30 No data support giving prophylactic antibiotics postoperatively. A wound becomes contaminated during the operation, so it's not surprising that antibiotics continued after the procedure will not decrease SSIs any more than a single preoperative dose. 56 In fact, their use may mask infections that will later become clinically evident, delaying diagnosis. Even worse, prolonged administration raises the risk that resistant organisms will subsequently recover. 57

Proper mechanical bowel preparation. This approach—used when the bowel will be entered—usually consists of 2 to 4 L of polyethylene glycol or 90 mL of sodium phosphate-based solutions combined with oral antibiotics (neomycin and erythromycin/metronidazole) started on the day before the operation. It's controversial whether this approach is better than IV prophylactic antibiotics alone with or without mechanical cleansing of the colon. Currently more than 80% of US surgeons who practice colorectal surgery use the combined approach of both parenteral antibiotics and mechanical bowel cleansing with oral antibiotics. 58-61

Strategies aimed at local/surgical factors

Good surgical technique is the mainstay for preventing SSIs related to local/surgical factors. Specifically, that means the ob/gyn surgeon should handle tissue appropriately, minimize wound/ surgical site bleeding without compromising tissue perfusion, remove devitalized or necrotic tissue, minimize dead space, avoid spillage of GI contents (or other contaminated secretions), and use sutures and drains appropriately. 1,29

Multifilament (braided) sutures increase the SSI risk more than do monofilament sutures, animal studies show—as does any additional foreign material in the wound. Because sutures act as foreign bodies, use them as sparingly as possible, and particularly avoid using them to obliterate subcutaneous (non-strength) tissues. 62,63

Be judicious in your use of drains. Because studies show rising SSI risk whenever drains are placed through the same incision, make separate incisions far from the surgical wound instead. SSIs also thrive when drains are left in place for too long. When you place a drain, make sure you're clear on the indication as well as the criteria you'll use to remove it. Closed-system drains are preferable to "open" ones (e.g., Penrose). 64

Good postoperative wound management can also lower infection risk. Wounds can be closed primarily, left to heal by secondary intention, or closed 3 to 5 days postoperatively (delayed primary closure). Delayed closure lowers the risk of a subsequent SSI diagnosis—but at the cost of having an open wound for an unpredictable postoperative period. Some researchers postulate that delayed primary closure and closure by secondary intention avoid the risk of having a subsequent necrotizing infection, although this complication is relatively rare. 65,66 When a wound is closed primarily, it's customary to cover it for the next 24 to 48 hours to avoid exogenous contamination. A primarily closed wound is easily infected exogenously during the first 24 hours, according to animal studies, and is very unlikely to be infected after 72 hours. 67

Strategies aimed at patient-related factors

Any patient with diabetes or another disease that alters healing or host resistance should raise a red flag for more frequent postoperative complications, including SSIs. Therefore, take extra precautions to prevent these complications and proactively manage any comorbidities, which also include malignancy, advanced age, obesity, malnutrition, and cirrhosis. 1,3,5,7,13-23,29,30 Your first step for preventing SSIs in these patients is usually to identify—and aggressively manage—any modifiable conditions.

Although malnutrition and low albumin levels have been correlated with poor surgical outcome and increased SSIs, no independent association with higher risk has been proved. 7,19,20,22 While studies comparing perioperative TPN in certain patients haven't universally shown lower SSI rates, recent studies are more promising. They suggest that enteral nutrition given preoperatively or early in the postoperative period (± immunonutrition containing arginine, omega-3 fatty acids, and related nutrients) may decrease not only the number of infectious complications in patients with cancer and the critically ill, but even mortality. Giving these patients optimal nutritional support and treating malnutrition in preoperatively malnourished patients whenever possible should be standard of care in preventing SSIs. 68-76

Regulating body temperature. Pre- and intraoperative warming, either by systemic means (forced-air warming blankets) or by local measures (heating pads) decreases the SSI rate in both clean operations and in colorectal procedures. This practice maintains core temperatures above 36.5°C and optimizes oxygen pressure at the surgical site. Oxygen tension is a key component in neutrophil phagocytosis and killing and in wound healing. Maintaining normal body temperature during the operation is part of the standard of care for preventing SSIs. 77,78

Perioperative oxygen tension in the wound has been shown in animal models and in human observational trials to significantly affect the risk of subsequent SSI. 79 A recent study showed that increasing the intraoperative inspired fraction of oxygen (FIO 2 ) to 80% in the OR and recovery room significantly decreased the SSI rate when compared to using an FIO 2 of 30%. 80 Oxygen tension and tissue perfusion also affect the quality of wound healing and collagen deposition. 81 A more recent paper with less than one third the number of patients, a mixed group of procedures, and less rigorous procedures has come to a contradictory conclusion regarding the value of increasing perioperative inspired oxygen concentrations. 82

Controlling blood sugar levels. Glucose control in the perioperative period can also lower SSI rates. A glucose level above 220 mg/dL increased postoperative infection rates fivefold in a study of diabetic patients who had GI and cardiovascular operations. In another study, any glucose level above 200 mg/dL increased infection rates for patients with diabetes who were having open-heart procedures. 83 However, controlling glucose levels in these patients through continuous insulin infusion reduced deep mediastinal infection rates to those seen in nondiabetic patients.

Another researcher showed that cardiac patients whose postoperative glucose levels were above 200 mg/dL had twice the risk of SSI regardless of a diagnosis of diabetes. Half of all hyperglycemic episodes occurred in nondiabetics and it's been shown that undiagnosed and untreated hyperglycemia is very common in general medical and surgical patients. 84,85 Another recent study shows that tight postoperative glucose control (<110 mg/dL) in critically ill surgical patients—whether diabetic or nondiabetic—decreases mortality and organ failure, particularly in the presence of infectious complications. Although the jury's still out, experimental data seem to support the idea that adequate glucose control—rather than insulin—can take credit for this benefit. 86

SSI prevention has focused on managing patient-related factors. Clearly, pathogens have been on the scene for a long time and will continue to be so. Therefore, the way to minimize SSIs—and surgical infections in general—is by boosting the body's immune response. Great advances over the last century have reduced surgical site contamination considerably. But surely the time has come to go a few steps further: to assure optimal medical support and to study new techniques that will enhance a patient's response to microorganisms and further minimize surgical infections.

Finally, surveillance systems for surgical wound infections have been effective in reducing SSIs and treatment costs. They remind surgical teams of the risk of SSIs and the need to establish protocols for standard practices and up-to-date preventive strategies. Moreover, they're a good framework for identifying "potentially preventable" SSIs and correcting systems errors. And they help us to learn more about "apparently unpreventable" SSIs, and create new practices that may one day reduce infection rates more dramatically. 1,2,4,87

Take-home messages

• Insist that patients quit smoking at least 30 days before surgery, treat all remote infections before the operation, control glucose levels, and don't remove hair preoperatively, unless necessary.
• Give prophylactic antibiotics within 1 hour before the procedure begins and during but not after the operation. Starting the infusion during anesthesia induction helps achieve ideal timing.
• Maintaining normal body temperature during surgery is part of the standard of care for preventing SSIs; the authors feel that treating malnutrition whenever possible in preoperatively malnourished patients should be.
• Don't forget hand/forearm scrubbing (or alcohol hand rub), measures that—along with surgical site preparation—directly prevent the OR team from contaminating the surgical site.
• Increasing the intraoperative inspired fraction of oxygen (FIO 2 ) to 80% and maintaining it for 2 hours postoperatively seem to significantly decrease the SSI rate when compared to supplementation with 30% FIO 2 .