Cesarean delivery is now the most commonly performed operation in hospitals in the United States. Approximately 1 to 1.5 million women have this procedure each year. The single most common postoperative complication of cesarean delivery is infection of the endometrium, the surgical site (wound infection), or the urinary tract (UTI).1
This review would not consider UTIs in detail. These infections usually result from placement of a bladder catheter during surgery. The key to prevention of UTIs is use of aseptic technique during placement of the catheter and removal of the catheter as soon as possible after surgery. Rather, I would focus on prevention of endometritis and surgical site (wound) infection.
Endometritis (or endomyometritis) is an entity well recognized by practicing clinicians. It is an infection of the endometrial lining, and often the underlying myometrium, caused by a polymicrobial, mixed aerobic-anaerobic flora. These organisms are introduced into the endometrial cavity during the course of labor and subsequent surgery. The principal risk factors for postcesarean endometritis are extended duration of labor and ruptured membranes, multiple internal examinations after membrane rupture, internal fetal monitoring, and pre-existing group B streptococcal infection or bacterial vaginosis.
The term “surgical site infection” (SSI) can be confusing because it refers to “superficial and deep infections of the wound” and to “deep-seated organ space infections.” In this review, I would usually use the term wound infection to denote either a superficial or deep incisional abscess (ie, drainage of pus from the wound) or wound cellulitis, and I would refer specifically to endometritis or pelvic abscess rather than “organ space infection.”
Wound infections are typically caused by the polymicrobial pelvic flora and by skin organisms such as aerobic streptococci and staphylococci. Obesity is, by far, the principal risk factor for wound infection. Other risk factors include diabetes, smoking, connective tissue diseases, and poor surgical technique (eg, incorrect choice of suture or poor hemostasis that leads to a wound hematoma that becomes infected).
The purpose of this article is to describe the essential evidence-based steps necessary to reduce the frequency of endometritis and surgical site (wound) infection. I would outline these steps in the same sequence that we approach the operation itself, that is, from preparation of the surgical site to closure of the surgical wound.
VAGINAL CLEANSING BEFORE SURGERY
Multiple studies have sought to determine if cleansing the vagina before surgery with an antiseptic solution further reduced the incidence of postcesarean infection beyond what could be achieved with systemic antibiotic prophylaxis. In a randomized, placebo-controlled trial of 308 women having a nonemergency cesarean delivery, Starr et al2 reported a decreased incidence of postoperative endometritis in women who received a 30-second vaginal scrub with povidone-iodine compared with women who received only an abdominal scrub (7% vs. 14.5%; P<0.05). The groups did not differ in the frequency of wound infection. In another randomized controlled trial, Haas et al3 found that preoperative vaginal cleansing with povidone-iodine, compared with an abdominal scrub alone, was associated with a decreased incidence of a composite measure of postoperative morbidity [6.5% vs. 11.7%; relative risk (RR), 0.55; 95% confidence interval (CI), 0.26-1.11; P=0.11]. The postoperative composite included fever, endometritis, sepsis, readmission, and wound infection.
Subsequently, Asghania et al4 conducted a double-blind, nonrandomized study of 568 women having cesarean delivery who received an abdominal scrub plus a 30-second vaginal scrub with povidone-iodine or an abdominal scrub alone. They documented a decreased incidence of postoperative endometritis in women who received the combined scrub [1.4% vs. 2.5%; P=0.03; adjusted odds ratio (OR), 0.03; 95% CI, 0.008-0.7].
In a similar study, Yeldrim et al5 conducted a randomized controlled trial comparing rates of infection in 334 women who received an abdominal scrub plus vaginal cleansing with povidone-iodine and 336 patients who had only a standard abdominal scrub. They documented a decreased incidence of endometritis in women who received the vaginal scrub (6.9% vs. 11.6%; P=0.04; RR, 1.69; 95% CI, 1.03-2.76.)
In 2014, Haas et al6 published a Cochrane review evaluating the effectiveness of preoperative vaginal cleansing with povidone-iodine. The authors reviewed 7 studies that included 2635 women. They concluded that vaginal preparation with povidone-iodine at the time of cesarean delivery significantly decreased the rate of postoperative endometritis compared with the control group (4.3% vs. 8.3%; RR, 0.45; 95% CI, 0.25-0.81). The most profound effect of vaginal cleansing was in women who were in labor before delivery.
In the most recent systematic review and meta-analysis, Caissutti et al7 reviewed 16 randomized controlled trials that included 4837 patients. The primary endpoint of their study was the frequency of postoperative endometritis. The trials were conducted in several countries and included patients of varying socioeconomic classes. Six trials included only patients having a scheduled cesarean delivery. Nine included both scheduled and unscheduled cesareans, and 1 included only unscheduled cesareans. In 11 of the studies, the antiseptic solution was povidone-iodine. Two trials used chlorhexidine-diacetate 0.2%; 1 used the 0.4% version of this solution. Another trial used metronidazole, 0.5% gel, and the other used the antiseptic cetrimide, a mixture of different quaternary ammonium salts. In 6 of the trials, systemic antibiotics were administered before the surgical incision, and in another 6, antibiotics were given after the umbilical cord was clamped. In 2 trials, systemic antibiotics were given at varying times, and, in the 2 final trials, the timing of the administration was not reported.
Overall, in the 15 trials, women in the treatment group had a significantly lower rate of endometritis (4.5% compared with 8.8%; RR, 0.52; 95% CI, 0.37-0.72). The beneficial effect of vaginal cleansing was statistically significant only in the group of women who were in labor (8.1% vs. 13.8%; RR, 0.52; 95% CI, 0.28-0.97). In the subgroup analysis of the 10 trials that used povidone-iodine, the reduction in the frequency of postcesarean endometritis was statistically significant (2.8% vs. 6.3%; RR, 0.42; 95% CI, 0.25-0.71). However, this same protective effect was not observed in women treated with chlorhexidine.
Bottom line: Preoperative cleansing of the vagina with povidone-iodine reduces the frequency of endometritis beyond that achieved with systemic prophylaxis alone (level I evidence/level A recommendation).
Preparation of the Surgical Site
In a landmark investigation published many years ago, Cruse and Foord8 showed that shaving the hair just before surgery, rather than the night before surgery, significantly decreased the rate of wound infection. Moreover, in patients whose hair was removed using clippers or depilatory creams, rather than by shaving, the subsequent rate of postoperative wound infection was decreased further.
For many years, the standard antiseptic solution for cleansing the surgical site was povidone-iodine. Darouiche et al9 recently reported the results of a well-designed, prospective, randomized multicenter trial comparing chlorhexidine-alcohol with povidone-iodine for skin preparation before surgery. The authors included 849 patients who were having many different types of surgical procedures. They noted that there were fewer superficial wound infections in patients who had a skin preparation with chlorhexidine-alcohol (4.2% compared with 8.6%; P=0.008). Of much greater importance, there were fewer deep wound infections in the individuals who received chlorhexidine-alcohol (1% compared with 3%; P=0.005).
More recently, Tuuli et al10 conducted an excellent prospective randomized trial comparing chlorhexidine-alcohol (2% chlorhexidine gluconate with 70% isopropyl alcohol) to povidone-iodine (8.3% povidone-iodine with 72.5% isopropyl alcohol) in patients having cesarean delivery. The authors included 1082 women in their study. The patients were followed for 30 days after surgery. The primary outcome was the frequency of SSI. In the study, the rate of SSI was significantly lower in the chlorhexidine-alcohol group (4.3%) compared with the povidone-iodine group (7.7%; P=0.02). The beneficial effect of the chlorhexidine-alcohol preparation was not affected by whether the cesarean was scheduled or unscheduled, the presence or absence of obesity, the type of skin closure, the presence of chronic disease, or the presence or absence of diabetes. With respect to secondary outcomes, there were no significant differences between the 2 antiseptic groups regarding rates of endometritis, hospital readmission for infection, length of hospital stay, use of other health care services, and rates of other wound complications such as seroma, hematoma, and cellulitis. Of note, patients in the chlorhexidine-alcohol group were significantly less likely to have physician office visits for assessment of possible wound complications.
One of the major strengths of this investigation was its large size and randomized design. Another important strength was that the authors included emergency cesarean deliveries in their analysis. Emergency procedures represent a substantial portion of cesarean deliveries, and they place the patient at increased risk for SSI because of the limited time available to prepare the skin before surgery.
A recent study by Ngai et al11 compared chlorhexidine-alcohol with iodine-alcohol for skin preparation before cesarean delivery. Unlike Tuuli et al,10 these authors found no difference in the incidence of SSI when comparing the 2 antiseptic solutions, either separately or sequentially, except in morbidly obese patients. In these women, sequential application of both solutions reduced the infection rate. However, this study specifically excluded emergency cesarean deliveries, which makes the generalizability of the results questionable.
Administration of Prophylactic Antibiotics
Table 1 summarizes the principal criteria for administration of perioperative antibiotic prophylaxis. Without a doubt, all 3 of these criteria are fulfilled when both urgent and nonurgent cesarean deliveries are performed. In the era before prophylactic antibiotics were routinely used, the rate of postcesarean endometritis in some highly indigent patient populations approached 85%.1 Moreover, although most cases of postcesarean endometritis are relatively straightforward to manage, some may progress to serious, even potentially life-threatening complications such as pelvic abscess formation, septic pelvic vein thrombophlebitis, and septic shock. Wound infections, although less common than endometritis, are more difficult to treat and may progress to dehiscence or even necrotizing fasciitis.
The classic sequence of basic science experiments that forms the foundation for use of prophylactic antibiotics was conducted by Burke more than 50 years ago.12 Using a guinea pig model, he showed conclusively that prophylactic antibiotics have their most pronounced effect when they are administered before the surgical incision and before the time that bacterial contamination occurs. In Burke’s elegant studies, when prophylaxis was delayed more than 4 hours after the start of surgery, no beneficial effect occurred.
Interestingly, when obstetricians first began to use prophylactic antibiotics for cesarean delivery, some investigators expressed concern about the possible exposure of the neonate to antibiotics just before delivery. Specifically, they questioned whether this exposure would lead to an increased frequency of evaluations for suspected sepsis and would select for resistant organisms that would make neonatal sepsis more difficult to treat.
Gordon et al13 published an important report in 1979 that showed the preoperative administration of ampicillin did not increase the frequency of immediate or delayed neonatal infections. However, delaying the administration of ampicillin until after the umbilical cord was clamped was just as effective in preventing postoperative endometritis. Subsequently, Cunningham et al14 showed that preoperative administration of prophylactic antibiotics significantly increased the frequency of sepsis workups in exposed neonates compared with infants with no preoperative antibiotic exposure (28% vs. 15%; P<0.025). On the basis of these 2 reports, obstetricians adopted a policy of delaying antibiotic administration until after the infant’s umbilical cord was clamped.
In 2007, Sullivan et al15 challenged this longstanding practice. In a carefully designed, prospective, randomized, double-blind trial, they showed that patients who received preoperative cefazolin had a significant reduction in the frequency of endometritis compared with women who received the same antibiotic after cord clamping (1% vs. 5%; RR, 0.2; 95% CI, 0.2-0.94). The rate of wound infection was lower in the preoperative antibiotic group (3% vs. 5%), but this difference did not reach statistical significance. Of particular note, the authors found that there was no increase in the frequency of proven or suspected neonatal infection in infants exposed to antibiotics before delivery. Subsequent to the publication by Sullivan and colleagues, other reports have confirmed that administration of antibiotics before surgery is more effective in preventing endometritis and wound infection compared with administration after umbilical cord clamping. Moreover, preoperative administration of antibiotics does not increase the frequency of proven or suspected neonatal sepsis.16,17
In an older review, the present author18 reviewed the evidence regarding choice of antibiotics and number of doses and concluded that a single dose of a first-generation cephalosporin such as cefazolin was the preferred regimen. A single dose was comparable in effectiveness to 2- and 3-dose regimens and to single or multiple-dose regimens of broader-spectrum agents. Continuing until today, the standard of care for prophylaxis has been a single 1 to 2 g dose of cefazolin.
However, 4 recent papers have raised the question of whether the prophylaxis effect could be enhanced whether the spectrum of activity of the antibiotic regimen was broadened to include an agent that was effective against ureaplasms. In the first report, Tita et al19 evaluated an indigent patient population with an inherently high rate of postoperative infection and showed that the addition of azithromycin, 500 mg, to cefazolin significantly reduced the rate of postcesarean endometritis. In a follow-up report from the same institution, Tita et al20 demonstrated that the addition of azithromycin also resulted in a significant reduction in the frequency of wound infection. In both of these investigations, the antibiotics were administered after cord clamping. In a subsequent report, Ward and Duff21 showed that the combination of azithromycin plus cefazolin administered preoperatively resulted in a very low rate of endometritis (1.3%; 95% CI, 1.0-1.78) in an indigent population similar to that of Tita et al.19,20
In a subsequent investigation, Tita et al22 reported the results of the Cesarean Section Optimal Antibiotic Prophylaxis (C/SOAP) trial. This investigation was conducted at 14 centers in the United States and included 2013 women who had a cesarean delivery during labor or after membrane rupture. The authors randomly assigned 1119 women to receive 500 mg of azithromycin plus conventional single-agent prophylaxis (usually cefazolin) and 994 women to receive a placebo plus conventional prophylaxis. The primary outcome was the composite of endometritis, wound infection, or other infection occurring within 6 weeks of surgery.
The authors observed that the primary outcome occurred in 62 women (6.1%) who received azithromycin plus conventional prophylaxis and in 119 women (12%) who received only single-agent prophylaxis. The RR of developing a postoperative infection was 0.51 in women who received the combined therapy. There were significant differences between the 2 groups in both the rates of endometritis (3.8% vs. 6.1%; P=0.02) and wound infection (2.4% vs. 6.6%; P<0.001).
Harper et al23 subsequently validated the cost-effectiveness of the cefazolin-azithromycin protocol using a decision analytic model. When balancing the added cost of the second antibiotic against the reduced number of infections in women who received combined therapy, the cost savings in women having an unscheduled cesarean was $360 (95% CI, $98-$157). In women having a scheduled cesarean delivery, the cost savings was $143 (95% CI, $98-$157).
Bottom line: All patients having cesarean delivery should receive prophylaxis with cefazolin plus azithromycin before surgery (level I evidence/level A recommendation). Patients who are allergic to β-lactam antibiotics should be treated with a single dose of clindamycin plus gentamicin (level III evidence, level C recommendation).
Antibiotic Prophylaxis—New Twists
MODIFICATION OF DOSE OF CEFAZOLIN
Obesity increases the risk of postcesarean infection 3- to 5-fold.24 Because both pregnancy and obesity increase the total volume of distribution of a drug, achieving adequate tissue concentrations of an antibiotic may be hindered by a potential dilutional effect. Furthermore, pharmacokinetic studies have consistently shown that the tissue concentration of an antibiotic, specifically a concentration above the minimal inhibitory concentration (MIC) for common bacteria, determines the susceptibility of those tissues to infection, regardless of whether the serum concentration of the antibiotic is in the therapeutic range.25
These concerns have led to several recent investigations evaluating different doses of cefazolin for obese patients. Pevzner et al26 conducted a prospective cohort study of 29 women having a scheduled cesarean delivery. The patients were divided into 3 groups: lean [body mass index (BMI)<30], obese (BMI 30 to 39.9) and extremely obese (BMI>40). All women received a 2 g dose of cefazolin 30 to 60 minutes before surgery. Cefazolin concentrations in adipose tissue obtained at the time of skin incision were inversely proportional to maternal BMI (r=−0.67; P<0.001). All specimens demonstrated a therapeutic concentration of cefazolin for gram-positive cocci (1 μg/g of tissue), but only 20% of the obese women and 33% of the extremely obese women achieved the MIC for gram-negative bacilli (4 μg/g of tissue) (P=0.14). Results were similar for cefazolin concentrations in adipose tissue at skin closure.
Swank et al24 conducted a prospective cohort study that included 28 women. They demonstrated that, after a 2 g dose of cefazolin, only 20% of the obese women (BMI 30 to 40) and 0% of extremely obese women (BMI>40) achieved an adipose tissue concentration that exceeded the MIC for gram-negative rods (8 μg/mL). However, 100% and 71.4%, respectively, achieved such a tissue concentration after a 3 g dose (P<0.001 and 0.002, respectively). When the women were stratified by actual weight, there was a statistically significant difference between women who weighed <120 kg versus those who weighed >120 kg. In total, 79% of the former had a tissue concentration of cefazolin >8 μg/mL, compared with 0% of women who weighed >120 kg. On the basis of these observations, the authors recommended a 3 g dose of cefazolin for women who weigh >120 kg.
Young et al27 conducted a double-blind, randomized controlled trial in 26 obese women (BMI≥30) and demonstrated that, at the time of the uterine incision and fascial closure, there were significantly higher concentrations of cefazolin in the adipose tissue of obese women who received a 3 g dose of antibiotic compared with a 2 g dose. However, all concentrations of cefazolin were consistently above the MIC of cefazolin for gram-positive cocci (1 μg/g of tissue) and gram-negative bacilli (4 μg/g of tissue).
Further, Maggio et al28 conducted a double-blind randomized controlled trial comparing a 2 g dose of cefazolin versus a 3 g dose in 57 obese (BMI≥30) women. They showed no statistically significant difference in the percentage of women who had tissue concentrations of cefazolin greater than the MIC for gram-positive cocci (8 μg/g of tissue). All samples were above the MIC for cefazolin for gram-negative bacilli (2 μg/g of tissue). On the basis of these data, Young et al27 and Maggio et al28 did not recommend increasing the dose of cefazolin from 2 to 3 g in obese patients.
The studies presented above are difficult to compare for 3 reasons. First, each study used a different MIC of cefazolin for both gram-positive and gram-negative bacteria. Second, the authors sampled different maternal tissues and/or serum at varying times during the cesarean delivery. Third, the studies did not specifically investigate, or were not powered sufficiently to address, the more important clinical outcome of surgical site (wound) infection. In a recent historical cohort study, Ward and Duff21 were unable to show that weight-based dosing of cefazolin reduced the rate of endometritis and wound infection below the level already achieved with combined prophylaxis with cefazolin plus azithromycin.
Sutton et al29 published a recent report evaluating the pharmacokinetics of azithromycin when used as prophylaxis for cesarean delivery. The authors studied 30 women who had a scheduled cesarean delivery and who received a 500 mg intravenous dose of azithromycin that was initiated 15, 30, or 60 minutes before the surgical incision and then infused over 1 hour. They obtained maternal plasma samples of amniotic fluid, placenta, myometrium, adipose tissue, and umbilical cord blood intraoperatively. The median concentration of azithromycin in adipose tissue was 102 ng/g of tissue, which is below the MIC50 for ureaplasmas (250 ng/mL). The median concentration in myometrial tissue was 402 ng/g of tissue. The concentration in maternal plasma consistently exceeded the MIC50 for ureaplasmas.
Bottom line: All women, regardless of weight, should receive a 2 g dose of cefazolin. When azithromycin is used in combination with cefazolin, an intravenous dose of 500 mg appears to provide adequate concentrations in serum and myometrium, but probably not in adipose tissue. More information is needed before a firm recommendation about weight-based dosing of azithromycin can be made.
Antibiotic Prophylaxis—New Twists
POSTOPERATIVE ADMINISTRATION OF ANTIBIOTICS
Valent et al30 recently published the results of an interesting randomized double-blind trial of postcesarean delivery prophylaxis. They sought to determine whether a short course of oral cephalexin (500 mg every 8 h) plus oral metronidazole (500 mg every 8 h), started within 8 hours of surgery and continued for 48 hours, was effective in decreasing the rate of SSI in obese women who also received intravenous preoperative cefazolin.
Two hundred two women received antibiotics; 201 received placebo. The mean BMI of patients in the trial was 39.7. The surgical technique was well standardized. All patients received intravenous cefazolin (2 g) immediately before surgery. The primary end point was SSI, defined as superficial incisional, deep incisional, and organ space infection.
SSI occurred in 13 women (6.4%) in the antibiotic group and 31 women (15.4%) in the placebo group. The difference in frequency of infection was 9.0% (95% CI, 0.22-0.77; P=0.01). The principal impact of prophylaxis was a reduction in the rate of wound cellulitis.
Of note, vaginal cleansing with povidone-iodine was not routinely performed. Moreover, the antibiotic prophylaxis was cefazolin rather than cefazolin plus azithromycin. Whether the same outcome would have occurred had those interventions taken place remains to be determined.
REMOVAL OF THE PLACENTA
One key measure of proven value in reducing the risk of postcesarean endometritis is removing the placenta by exerting traction on the umbilical cord rather than extracting it manually (Fig. 1). In an early investigation, Yancey et al31 demonstrated that, in women who labored before surgery, the glove of the operating surgeon became heavily contaminated with bacteria from the lower uterine segment after delivery of the infant. They reasoned that, whether the surgeon then placed her or his glove behind the placenta to manually extract it, those bacteria could be introduced into the underlying soft tissue and vascular system.
In a follow-up to this original investigation, Lasley et al32 showed that, in 333 high-risk patients who also received intravenous antibiotic prophylaxis after cord clamping, the rate of postoperative endometritis was 15% in the group that had spontaneous delivery of the placenta compared with 27% in women who had manual extraction (RR, 0.6; 95% CI, 0.3-0.9; P=0.02). A recent Cochrane review33 that included 15 reports (4694 women), confirmed this observation. The RR of endometritis in the manual removal group was 1.64 (95% CI, 1.42-1.90). Manual removal of the placenta also resulted in greater blood loss, greater fall in the hematocrit after delivery, and longer duration of hospital stay.
Bottom line: Whenever possible, the placenta should be removed by traction on the umbilical cord rather than by manual extraction (level 1 evidence/level A recommendation).
Closure of the Subcutaneous Layer of the Abdomen
Del Valle et al34 were among the first investigators to direct clinicians’ attention to the optimal technique for closure of the subcutaneous layer of the abdomen (Fig. 2). They randomly assigned 438 women having cesarean delivery to closure of Camper’s fascia with absorbable suture (usually 3-0 plain gut placed as either continuous or interrupted sutures) versus no closure. The skin was reapproximated with staples. The authors observed a significantly higher incidence of wound disruption because of seroma, hematoma, or infection in the group that did not have closure (16/216, 74% vs. 6/222, 2.7%; P=0.03).
Naumann et al35 reported similar findings in a study of obese women having cesarean delivery. A total of 245 women with a subcutaneous layer of at least 2 cm in thickness were randomized to no closure of the deep subcutaneous layer versus closure with a continuous suture of 3-0 polyglycolic acid. Twenty-eight women (11.4%) developed wound seromas; 17 (7%) developed wound infections. The RR of seroma formation in the subcutaneous closure group was 0.3 (95% CI, 0.1-0.7). There was no difference in the frequency of wound infection. The incidence of wound disruptions from all causes was 14.5% in the closure group versus 26.6% in the no closure group (RR, 0.5; 95% CI, 0.3-0.9).
These reports were soon followed by other randomized trials and then by an excellent meta-analysis. Chelmo et al36 showed that, when the subcutaneous layer exceeded 2 cm in thickness, closure of the bottom portion of the subcutaneous tissue significantly reduced the incidence of wound disruption, primarily as a result of a decrease in the frequency of seroma (RR, 0.66; 95% CI, 0.41-0.91). The number needed to treat to prevent a single case of wound disruption was 16.2.
Ramsey et al37 subsequently questioned whether the addition of a closed suction drain further reduced the frequency of wound disruptions beyond that achieved with subcutaneous closure alone in obese women. In a prospective randomized trial, they assigned women whose subcutaneous layer was >4 cm in thickness, to subcutaneous closure alone (n=149) versus closure plus a closed-system drain (131). Their primary end point was “composite wound morbidity.” Interestingly, the frequency of the composite outcome was actually higher in the closure plus drain group (22.7%) compared with suture closure alone (17.4%; RR, 1.3; 95% CI, 0.8-2.1).
There is no firm consensus on the choice of suture, gauge of suture, or technique for placement of the subcutaneous suture. At our medical center, we use a continuous suture of 3-0 polyglactin 910.
Bottom line: If the subcutaneous layer exceeds 2 cm in thickness, it should be closed (level I evidence/level A recommendation).
Wound Closure—New Twist
Yu et al38 recently sought to determine whether prophylactic use of negative pressure devices was effective in reducing the frequency of SSI after cesarean delivery when compared with standard wound dressings. Their systematic review and meta-analysis included 9 studies. Six were randomized controlled trials, 2 were retrospective cohort studies, 1 was a prospective cohort study. The 2 most commonly used negative pressure devices were the Prevena (KCI, USA, San Antonio, TX) and PICO (Smith & Nephew, St. Petersburg, FL) systems. The majority of patients in all studies were at high risk for wound complications because of obesity.
The absolute risk of SSI in the intervention group was 5% (95% CI, 2.0-7.08) compared with 11% (95% CI, 7.0%-16.0%) in the standard dressing group. The absolute risk reduction was 6% (95% CI, −10.0% to −3.0%), and the number needed to treat was 17.
Although the results of this meta-analysis are impressive, I do not believe that obstetricians should routinely employ these devices in all women having cesarean delivery.
- The investigations included in the present study did not consistently distinguish between scheduled versus unscheduled cesarean deliveries.
- The reports did not systematically consider other risk factors besides obesity.
- The studies did not provide full descriptions of other measures that might influence the rate of SSI, such as timing and selection of prophylactic antibiotics, selection of suture material, preoperative skin preparation, and techniques for closure of the deep subcutaneous tissue and skin.
- None of the studies systematically considered the cost-effectiveness of the devices, an important issue given that each device ranges in price from $200 to $500.
With respect to the issue of cost, Echebiri et al39 recently reported a decision-analytic model based on a third-party payer’s perspective to determine the cost-effectiveness of prophylactic application of negative pressure wound therapy compared with a standard postoperative dressing in all women having cesarean delivery. They concluded that the negative pressure therapy was not cost-effective in patients at low to moderate risk for wound infection (≤14%). In very high-risk patients (risk of infection >14%), the device was cost-effective if priced below $192 per patient.
For the present time, I would reserve the use of these devices to morbidly obese patients having cesarean delivery, particularly after an extended period of labor and ruptured membranes.
Closure of the Skin
For many years the standard of care has been to close the skin with surgical staples. Recently, however, 2 comprehensive meta-analyses have shown an improved outcome when the skin was closed with subcutaneous suture.
Tuuli et al40 conducted a systematic review and meta-analysis that included 5 randomized controlled trials and 1 prospective cohort study that compared staples versus sutures for skin closure in women who had a transverse abdominal incision for cesarean delivery. The staple closure group included 803 patients; the suture closure group included 684 patients. Although staple closure was faster, this technique was associated with a 2-fold higher risk of wound infection or separation (pooled OR, 2.06; 95% CI, 1.43-2.98; number needed to harm=16). The closures were equivalent regarding pain, cosmesis, and patient satisfaction.
Clay et al41 conducted a similar meta-analysis that included 5 randomized controlled trials. The primary end points were incidence of wound separation and a “composite of wound complications,” including dehiscence, infection, seroma, and hematoma. The subcutaneous suture closure group included 492 women; 385 patients received closure with staples. The OR for overall frequency of wound complications was 2.11 in the staples group (P=0.003). The OR for wound separation was 4.01 in the staples group (P<0.001). The operating time was 5 minutes shorter in the staples group.
More recently, Figueroa et al42 presented the results of a randomized controlled trial from a single institution that included 398 patients. The primary end point was wound disruption or infection at discharge and then 4 to 6 weeks later. The primary outcome was observed in 7.1% of patients in the staples group versus 0.5% in the suture group at hospital discharge (RR, 14.1; 95% CI, 1.9-106; P<0.001). At 4 to 6 weeks, the cumulative risk of the primary outcome was 14.5% in the staples group and 5.9% in the suture group (RR, 2.5; 95% CI, 1.2-5.0; P=0.008).
There is imperfect consensus about the optimal suture or gauge of suture for closure of the skin. However, Buresch et al43 recently published an informative report comparing poliglecaprone 25 (Monocryl) to polyglactin 910 (Vicryl) for skin closure. The study included 550 term patients having nonemergency cesarean deliveries. The primary outcome was a composite of SSI and wound separation, hematoma, or seroma within 30 days of surgery. The primary outcome was observed in 8.8% of women in the poliglecaprone 25 group versus 14.4% in the polyglactin 910 group (RR, 0.61; 95% CI, 0.37-0.99; P=0.04).
Bottom line: The skin should be closed with a subcutaneous suture (ideally, poliglecaprone 25) rather than staples (level I evidence, level A recommendation).
The following guidelines for prevention of postcesarean endometritis and SSI (wound infection) are supported by level I evidence and merit an A-level rating.
- Cleanse the vagina for a minimum of 30 seconds using a povidone-iodine solution.
- Clip, but do not shave, the hair in the incision area just before surgery.
- Prepare the skin with a solution of chlorhexidine rather than povidone-iodine.
- Administer systemic antibiotic prophylaxis before making the skin incision. The combination of cefazolin plus azithromycin is superior to cefazolin alone.
- Remove the placenta by exerting traction on the cord rather than by extracting it manually.
- If the subcutaneous layer is >2 cm in thickness, close the bottom half of this layer.
- Close the skin with a subcutaneous suture, preferably poliglecaprone 25, rather than with staples.
The following guidelines for prevention of SSI are based on more limited evidence and merit, at best, a B-level rating.
- In obese patients, consider administering a 2-day postoperative course of metronidazole plus cefazolin.
- In obese patients, particularly morbidly obese patients, consider pre-emptive placement of a wound vacuum dressing over the incision.
1. Duff P. Pathophysiology and management of postcesarean endometritis. Obstet Gyecol. 1986;67:269–275.
2. Starr RV, Zurawski J, Ismail M. Preoperative vaginal preparation with povidone-iodine and the risk of postcesarean endometritis. Obstet Gynecol. 2005;105:1024–1029.
3. Haas DM, Pzouki F, Smith RR, et al. Vaginal cleansing before cesarean delivery to reduce postoperative infectious morbidity. Am J Obstet Gynecol. 2010;202:310.e1–310.e6.
4. Asghania M, Mirblouk F, Shakiba M, et al. Preoperative vaginal preparation with povidone-iodine on post-cesarean infectious morbidity. J Obstet Gynecol. 2011;31:400–403.
5. Yildririm G, Gungorduk K, Asicioglu O, et al. Does vaginal preparation with povidone-iodine prior to cesarean delivery reduce the risk of endometritis? A randomized controlled trial. J Matern Fetal Neonatal Med. 2012;25:2316–2321.
6. Haas DM, Morgan S, Contreras K. Vaginal preparation with antiseptic solution before cesarean section for preventing postoperative infections. Cochrane Database Syst Rev. 2014;12:CD007892.
7. Caissutti C, Saccone G, Zullo F, et al. Vaginal cleansing before cesarean delivery. A systematic review and meta-analysis. Obstet Gynecol. 2017;130:527–538.
8. Cruse PJ, Foord R. A five-year prospective study of 23,649 surgical wounds. Arch Surg. 1973;107:206–209.
9. Darouiche RO, Wall MJ Jr, Itani KM, et al. Chlorhexidine-alcohol versus povidone-iodine for surgical-site antisepsis. N Engl J Med. 2010;362:18–26.
10. Tuuli MG, Liu J, Stout MJ, et al. A randomized trial comparing skin antiseptic agents at cesarean delivery. N Engl J Med. 2016;374:647–655.
11. Ngai IM, Arsdale AV, Govindappagari S, et al. Skin preparation for prevention of surgical site infection after cesarean delivery. Obstet Gynecol. 2015;126:1251–1257.
12. Burke JF. The effective period of preventive antibiotic action in experimental incisions and dermal lesions. Surgery. 1961;161–168.
13. Gordon HR, Phelps D, Blanchard K. Prophylactic cesarean section antibiotics: maternal and neonatal morbidity before or after cord clamping. Obstet Gynecol. 1979;53:151–156.
14. Cunningham FG, Leveno KJ, DePalma RT, et al. Perioperative antimicrobials for cesarean delivery: before or after cord clamping? Obstet Gynecol. 1983;62:151–154.
15. Sullivan SA, Smith T, Chang E, et al. Administration of cefazolin prior to skin incision is superior to cefazolin at cord clamping in preventing postcesarean infectious morbidity: a randomized controlled trial. Am J Obstet Gynecol. 2007;196:455.e1–455.e5.
16. Constantine MM, Rahman M, Ghulmiyah L, et al. Timing of perioperative antibiotics for cesarean delivery: a meta-analysis. Am J Obstet Gynecol. 2008;199:301.e1–301.e6.
17. Owens SM, Brozanski BS, Meyn LA, et al. Antimicrobial prophylaxis for cesarean delivery
before skin incision. Obstet Gynecol. 2009;114:573–579.
18. Duff P. Prophylactic antibiotics for cesarean delivery; a simple cost-effective strategy for prevention of postoperative morbidity. Am J Obstet Gynecol. 1987;157:794–798.
19. Tita ATN, Hauth JC, Grimes A, et al. Decreasing incidence for postcesarean endometritis with extended-spectrum antibiotic prophylaxis. Obstet Gynecol. 2008;111:51–56.
20. Tita ATN, Owen J, Stamm AM, et al. Impact of extended-spectrum antibiotic prophylaxis on incidence of postcesarean surgical wound infection. Am J Obstet Gynecol. 2008;199:303.e1–303.e3.
21. Ward E, Duff P. A comparison of 3 antibiotic regimens for prevention of postcesarean endometritis: an historical cohort study. Am J Obstet Gynecol. 2016;214:751.e1–751.e4.
22. Tita AT, Szchowski JM, Boggess K, et al. Two antibiotics before cesarean delivery reduce infection rates further than one agent. N Engl J Med. 2016;375:1231–1241.
23. Harper LM, Kilgore M, Szychowski JM, et al. Economic evaluation of adjunctive azithromycin prophylaxis for cesarean delivery
. Obstet Gynecol. 2017;130:328–334.
24. Swank ML, Wing DA, Nicolau DP, et al. Increased 3-gram cefazolin dosing for cesarean delivery prophylaxis in obese women. Am J Obstet Gynecol. 2015;213:415.e1–415.e8.
25. Lui P, Derendorf H. Antimicrobial tissue concentrations. Infect Dis Clin North Am. 2003;17:599–613.
26. Pevzner L, Swank M, Krepel C, et al. Effects of maternal obesity on tissue concentrations of prophylactic cefazolin during cesarean delivery. Obstet Gynecol. 2011;117:877–882.
27. Young OM, Shaik IH, Twedt R, et al. Pharmacokinetics of cefazolin prophylaxis in obese gravidae at time of cesarean delivery. Am J Obstet Gynecol. 2015;213:541.e1–541.e7.
28. Maggio L, Nicolau DP, DaCosta M, et al. Cefazolin prophylaxis in obese women undergoing cesarean delivery. A randomized controlled trial. Obstet Gynecol. 2015;125:1205–1210.
29. Sutton AL, Acosta EP, Larson KB, et al. Perinatal pharmacokinetics of azithromycin for cesarean prophylaxis. Am J Obstet Gynecol. 2015;212:812.e1–812.e6.
30. Valent AM, DeArmond C, Houston JM, et al. Effect of post-cesarean delivery oral cephalexin and metronidazole on surgical site infection among obese women. A randomized clinical trial. J Am Med Assoc. 2017;318:1026–1034.
31. Yancey MK, Clark P, Duff P. The frequency of glove contamination during cesarean delivery. Obstet Gynecol. 1994;83:538–542.
32. Lasley DS, Eblen A, Yancey MK, et al. The effect of placental removal method on the incidence of postcesarean infections. Am J Obstet Gynecol. 1997;176:1250–1254.
33. Anorlu RI, Maholwana B, Hofmeyr GJ. Methods delivering the placenta at cesarean section. Cochrane Database Syst Rev. 2008;3:CD004737.
34. Del Valle GO, Combs P, Qualls C, et al. Does closure of camper fascia reduce the incidence of post-cesarean superficial wound disruption? Obstet Gynecol. 1992;80:1013–1016.
35. Naumann RW, Hauth JC, Owen J, et al. Subcutaneous tissue approximation in relation to wound disruption after cesarean delivery in obese women. Obstet Gynecol. 1995;85:412–416.
36. Chelmow D, Rodriguez EJ, Sabatini MM. Suture closure of subcutaneous fat and wound disruption after cesarean delivery: a meta-analysis. Obstet Gynecol. 2004;103:974–980.
37. Ramsey PS, White AM, Guinn DA, et al. Subcutaneous tissue reapproximation, alone or in combination with drain, on obese women undergoing cesarean delivery. Obstet Gynecol. 2005;105:967–973.
38. Yu L, Kronen RJ, Simon LE, et al. Prophylactic negative-pressure wound therapy after cesarean is associated with reduced risk of surgical site infection: a systematic review and meta-analysis. Am J Obstet Gynecol. 2018;218:200–210.
39. Echebiri NC, McDoom MM, Aalto MM, et al. Prophylactic use of negative pressure wound therapy after cesarean delivery. Obstet Gynecol. 2015;125:299–307.
40. Tuuli MG, Rampersod RM, Carbone JF, et al. Staples compared with subcuticular suture for skin closure after cesarean delivery. A systematic review and meta-analysis. Obstet Gynecol. 2011;117:682–690.
41. Clay FS, Walsh CA, Walsh SR. Staples vs. subcuticular sutures for skin closure at cesarean delivery: a meta-analysis of randomized controlled trials. Am J Obstet Gynecol. 2011;204:378–383.
42. Figueroa D, Jauk VC, Szychowski JM, et al. Surgical staples compared with subcuticular suture for skin closure after cesarean delivery. A randomized controlled trial. Obstet Gynecol. 2013;121:33–38.
43. Buresch AM, Arsdale AV, Ferzli M, et al. Comparison of subcuticular suture type for skin closure after cesarean delivery. A randomized controlled trial. Obstet Gynecol. 2017;130:521–526.