Infectious Disease UPDATE

Infectious disease UPDATE

Patrick Duff, MD Dr. Duff is Professor of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Florida College of Medicine, Gainesville.

The author reports no financial relationships relevant to this article.

Recent news and expert perspective on therapy for cesarean incision wounds, vaginal cleansing, managing skin abscesses, C difficile infection in obstetric patients, and risks of maternal Zika virus infection

n this Update I highlight 5 interesting inves- third investigation examines the role of sys- tigations on infectious diseases. The first temic antibiotics, combined with surgical Iaddresses the value of applying prophylac- drainage, for patients who have subcutane- tically a negative-pressure wound dressing to ous abscesses ranging in size up to 5 cm. The prevent surgical site infection (SSI) in obese fourth study presents new information about women having cesarean delivery (CD). The the major risk factors for Clostridium difficile second report assesses the effectiveness of infections in obstetric patients. The final study a preoperative vaginal wash in reducing the presents valuable sobering new data about the frequency of postcesarean endometritis. The risks of congenital Zika virus infection. IN THIS ARTICLE

Pre-CD vaginal Negative-pressure wound cleansing therapy after CD shows some page 33 C difficile benefit in preventing SSI infection in Yu L, Kronen RJ, Simon LE, Stoll CR, Colditz GA, Tuuli Six studies were ran- pregnancy MG. Prophylactic negative-pressure wound therapy af- domized controlled page 36 ter cesarean is associated with reduced risk of surgical trials (RCTs), 2 were retro- site infection: a systematic review and meta-analysis. spective cohort studies, and 1 Congenital Am J Obstet Gynecol. 2018;218(2):200–210.e1. was a prospective cohort study. Five studies risks of were considered high quality; 4 were of low Zika by u and colleagues sought to determine quality. trimester if the prophylactic use of negative- page 37 Y pressure devices, compared with Details of the study standard wound dressing, was effective in Several types of negative-pressure devices reducing the frequency of SSI after CD. were used, but the 2 most common were the The authors searched multiple databases Prevena incision management system (KCI, and initially identified 161 randomized con- San Antonio, Texas) and PICO negative- trolled trials and cohort studies for further pressure wound therapy (Smith & Nephew, assessment. After applying rigorous exclu- St. Petersburg, Florida). The majority of sion criteria, they ultimately selected 9 stud- patients in all groups were at high risk for

ILLUSTRATION: USED WITH PERMISSION. COURTESY OF KCI, AN ACELITY COMPANY USED WITH PERMISSION. COURTESY ILLUSTRATION: ies for systematic review and meta-analysis. wound complications because of obesity. CONTINUED ON PAGE 32

mdedge.com/obgmanagement Vol. 30 No. 7 | July 2018 | OBG Management 31 UPDATE infectious disease CONTINUED FROM PAGE 31

accounted for by the difference in the rate of SSI.

How negative-pressure devices aid wound healing Yu and colleagues explained that negative- pressure devices exert their beneficial effects in various ways, including: • shrinking the wound • inducing cellular stretch • removing extracellular fluids Passive wound closure (left) compared with negative-pressure wound therapy • creating a favorable environment for with the Prevena incision management system (right). healing • promoting angiogenesis and neurogenesis. The primary outcome of interest was Multiple studies in nonobstetric patients the frequency of SSI. Secondary outcomes have shown that prophylactic use of negative- included dehiscence, seroma, endometritis, pressure devices is beneficial in reducing the a composite measure for all wound compli- rate of SSI.1 Yu and colleagues’ systematic cations, and hospital readmission. review and meta-analysis confirms those The absolute risk of SSI in the interven- findings in a high-risk population of women tion group was 5% (95% confidence interval having CD. [CI], 2.0%–7.0%) compared with 11% (95% CI, 7.0%–16.0%) in the standard dressing group. Study limitations The pooled risk ratio was 0.45 (95% CI, 0.31– Before routinely adopting the use of negative- 0.66). The absolute risk reduction was 6% pressure devices for all women having CD, (95% CI, -10.0% to -3.0%), and the number however, obstetricians should consider the needed to treat was 17. following caveats: There were no significant differences in • The investigations included in the study the rate of any of the secondary outcomes by Yu and colleagues did not consistently other than the composite of all wound distinguish between scheduled versus complications. This difference was largely unscheduled CDs. • The reports did not systematically consider WHAT THIS EVIDENCE MEANS FOR PRACTICE other major risk factors for wound complications besides obesity, and they did not Results of the systematic review and meta-analysis by Yu and col- control for these confounders in the statis- leagues suggest that prophylactic negative-pressure wound therapy tical analyses. in high-risk mostly obese women after CD reduces SSI and overall • The studies included in the meta-analysis wound complications. The study’s limitations, however, must be kept did not provide full descriptions of other in mind, and more data are needed. It would be most helpful if a measures that might influence the rate of large, well-designed RCT was conducted and included 2 groups with SSIs, such as timing and selection of pro- comparable multiple major risk factors for wound complications, and in which all women received the following important interventions2–4: phylactic antibiotics, selection of suture • removal of hair in the surgical site with a clipper, not a razor material, preoperative skin preparation, • cleansing of the skin with a chlorhexidine rather than an iodophor and closure techniques for the deep subcu- solution taneous tissue and skin. • closure of the deep subcutaneous tissue if the total subcutaneous • None of the included studies systemati- layer exceeds 2 cm in depth cally considered the cost-effectiveness of • closure of the skin with suture rather than staples the negative-pressure devices. This is an • administration of antibiotic prophylaxis, ideally with a combination important consideration given that the of cefazolin plus azithromycin, prior to the surgical incision. acquisition cost of these devices ranges

from $200 to $500. OF KCI, AN ACELITY COMPANY USED WITH PERMISSION. COURTESY ILLUSTRATION:

32 OBG Management | July 2018 | Vol. 30 No. 7 mdedge.com/obgmanagement Vaginal cleansing before CD lowers risk of postop endometritis

Caissutti C, Saccone G, Zullo F, et al. Vaginal cleansing placebo or with no treat- before cesarean delivery: a systematic review and meta- ment, women in the treat- analysis. Obstet Gynecol. 2017;130(3):527–538. ment group had a significantly lower rate of endometritis (4.5% aissutti and colleagues aimed to compared with 8.8%; relative risk [RR], 0.52; determine if cleansing the vagina 95% CI, 0.37–0.72). When only women in C with an antiseptic solution prior to labor were considered, the frequency of surgery reduced the frequency of postcesar- endometritis was 8.1% in the intervention ean endometritis. They included 16 RCTs group compared with 13.8% in the control (4,837 patients) in their systematic review group (RR, 0.52; 95% CI, 0.28–0.97). In the and meta-analysis. The primary outcome was women who were not in labor, the difference the frequency of postoperative endometritis. in the incidence of endometritis was not statistically significant (3.5% vs 6.6%; RR, 0.62; Details of the study 95% CI, 0.34–1.15). The studies were conducted in several coun- In the subgroup analysis of women with tries and included patients of various socio- ruptured membranes at the time of surgery, economic classes. Six trials included only the incidence of endometritis was 4.3% in FAST patients having a scheduled CD; 9 included the treatment group compared with 20.1% TRACK both scheduled and unscheduled cesar- in the control group (RR, 0.23; 95% CI, 0.10– eans; and 1 included only unscheduled 0.52). In women with intact membranes at Women treated cesareans. In 11 studies, povidone-iodine the time of surgery, the incidence of endo- with vaginal was the antiseptic solution used. Two tri- metritis was not significantly reduced in the cleansing before als used chlorhexidine diacetate 0.2%, and treatment group. cesarean had a 1 used chlorhexidine diacetate 0.4%. One Interestingly, in the subgroup analysis significantly lower trial used metronidazole 0.5% gel, and of the 10 trials that used povidone-iodine, rate of endometritis another used the antiseptic cetrimide, which the reduction in the frequency of postcesar- compared with is a mixture of different quaternary ammo- ean endometritis was statistically significant those who received nium salts, including cetrimonium bromide. (2.8% vs 6.3%; RR, 0.42; 95% CI, 0.25–0.71). placebo or no In all trials, patients received prophylac- However, this same protective effect was treatment—4.5% tic antibiotics. The antibiotics were adminis- not observed in the women treated with vs 8.8% (RR, 0.52; tered prior to the surgical incision in 6 trials; chlorhexidine. In the 1 trial that directly com- 95% CI, 0.37–0.72) they were given after the umbilical cord was pared povidone-iodine with chlorhexidine, clamped in 6 trials. In 2 trials, the antibiotics there was no statistically significant differ- were given at varying times, and in the final ence in outcome. 2 trials, the timing of antibiotic administration was not reported. Of note, no trials described Simple intervention, solid benefit the method of placenta removal, a factor of Endometritis is the most common complica- considerable significance in influencing the tion following CD. The infection is polymi- rate of postoperative endometritis.5,6 crobial, with mixed aerobic and anaerobic Endometritis frequency reduced with organisms. The principal risk factors for post- vaginal cleansing; benefit greater in cesarean endometritis are low socioeco- certain groups. Overall, in the 15 trials in nomic status, extended duration of labor

PHOTO: SHUTTERSTOCK which vaginal cleansing was compared with and ruptured membranes, multiple vaginal

mdedge.com/obgmanagement Vol. 30 No. 7 | July 2018 | OBG Management 33 UPDATE infectious disease

in reducing the incidence of endometritis: WHAT THIS EVIDENCE MEANS FOR PRACTICE administration of prophylactic antibiotics prior to the surgical incision and removal From my perspective, the interesting unanswered question is why a chlorhexidine solution with low alcohol content was not more effec- of the placenta by traction on the cord as 5,6 tive than povidone-iodine, given that a chlorhexidine abdominal wash opposed to manual extraction. is superior to povidone-iodine in preventing wound infection after The assessment by Caissutti and col- cesarean delivery.7 Until additional studies confirm the effectiveness leagues confirms that a third measure— of vaginal cleansing with chlorhexidine, I recommend the routine use preoperative vaginal cleansing—also helps of the povidone-iodine solution in all women having CD. reduce the incidence of postcesarean endometritis. The principal benefit is seen in women who have been in labor with rup- examinations, internal fetal monitoring, and tured membranes, although certainly it pre-existing vaginal infections (principally, is not harmful in lower-risk patients. The bacterial vaginosis and group B streptococcal intervention is simple and straightforward: a colonization). 30-second vaginal wash with a povidone- Two interventions are clearly of value iodine solution just prior to surgery.

Treat smaller skin abscesses with antibiotics after surgical drainage? Yes.

Daum RS, Miller LG, Immergluck isolated from 388 (49.4%). The cure rate L, et al; for the DMID 07-0051 Team. was similar in patients in the clindamy- A placebo-controlled trial of antibiot- cin group (83.1%) and the trimethoprim- ics for smaller skin abscesses. N Engl J Med. sulfamethoxazole group (81.7%), and the 2017;376(26):2545–2555. cure rate in each antibiotic group was significantly higher than that in the placebo group or treatment of subcutaneous (68.9%; P<.001 for both comparisons). The abscesses that were 5 cm or smaller difference in treatment effect was specifically F in diameter, investigators sought to limited to patients who had S aureus isolated determine if surgical drainage alone was from their lesions. equivalent to surgical drainage plus systemic Findings at follow-up. At 1 month of antibiotics. After their abscess was drained, follow-up, new infections were less com- patients were randomly assigned to receive mon in the clindamycin group (6.8%) than either clindamycin (300 mg 3 times daily) in the trimethoprim-sulfamethoxazole group or trimethoprim-sulfamethoxazole (80 mg/ (13.5%; P = .03) or the placebo group (12.4%; 400 mg twice daily) or placebo for 10 days. P = .06). However, the highest frequency of The primary outcome was clinical cure 7 to adverse effects occurred in the patients who 10 days after treatment. received clindamycin (21.9% vs 11.1% vs 12.5%). No adverse effects were judged to be Details of the study serious, and all resolved without sequela. Daum and colleagues enrolled 786 partici- pants (505 adults, 281 children) in the pro- Controversy remains on antibiotic use spective double-blind study. Staphylococcus after drainage aureus was isolated from 527 patients (67.0%); This study is important for 2 major reasons.

methicillin-resistant S aureus (MRSA) was First, soft tissue infections are quite common PHOTO: SHUTTERSTOCK

CONTINUED ON PAGE 36

34 OBG Management | July 2018 | Vol. 30 No. 7 mdedge.com/obgmanagement UPDATE infectious disease CONTINUED FROM PAGE 34

and can evolve into serious problems, WHAT THIS EVIDENCE MEANS FOR PRACTICE especially when the offending pathogen is MRSA. Second, controversy exists about In my opinion, this investigation by Daum and colleagues supports whether systemic antibiotics are indicated a role for consistent use of systemic antibiotics following surgical if the subcutaneous abscess is relatively drainage of clinically significant subcutaneous abscesses that have a small and is adequately drained. For exam- 5 cm or smaller diameter. Several oral antibiotics are effective against S aureus, including MRSA.10 These drugs include trimethoprim- ple, Talan and colleagues demonstrated sulfamethoxazole (1 double-strength tablet orally twice daily), that, in settings with a high prevalence of clindamycin (300–450 mg 3 times daily), doxycycline (100 mg twice MRSA, surgical drainage combined with daily), and minocycline (200 mg initially, then 100 mg every 12 hours). trimethoprim-sulfamethoxazole (1 double- Of these drugs, I prefer trimethoprim-sulfamethoxazole, provided that strength tablet orally twice daily) was supe- the patient does not have an allergy to sulfonamides. Trimethoprim- rior to drainage plus placebo.8 However, sulfamethoxazole is significantly less expensive than the other 3 drugs Daum and Gold recently debated the issue and usually is better tolerated. In particular, compared with clindamy- of drainage plus antibiotics in a case vignette cin, trimethoprim-sulfamethoxazole is less likely to cause antibiotic- and reached opposite conclusions.9 associated diarrhea, including Clostridium difficile infection. Trimethoprim-sulfamethoxazole should not be used in the first trimester of pregnancy because of concerns about fetal teratogenicity.

Antibiotic use, common in the obstetric population, raises risk for C difficile infection

Ruiter-Ligeti J, Vincent S, Czuzoj- 13,881,592 births during 1999–2013 and Shulman N, Abenhaim HA. Risk fac- identified 2,757 (0.02%) admissions for deliv- tors, incidence, and morbidity associated ery complicated by C difficile infection, a with obstetric Clostridium difficile infection. rate of 20 admissions per 100,000 deliveries Obstet Gynecol. 2018;131(2):387–391. per year (95% CI, 19.13–20.62). The rate of admissions with this diagnosis doubled from he objective of this investigation was 1999 (15 per 100,000) to 2013 (30 per 100,000, to identify risk factors for Clostridium P<.001). T difficile infection (previously termed Among these obstetric patients, the pseudomembranous enterocolitis) in obstet- principal risk factors for C difficile infection ric patients. The authors performed a retro- were older age, multiple gestation, long-term spective cohort study using information from antibiotic use (not precisely defined), and a large database maintained by the Agency concurrent diagnosis of inflammatory bowel for Healthcare Research and Quality. This disease. In addition, patients with pyelone- database provides information about inpa- phritis, perineal or cesarean wound infec- tient hospital stays in the United States, and it tions, or pneumonia also were at increased is the largest repository of its kind. It includes risk, presumably because those patients data from a sample of 1,000 US hospitals. required longer courses of broad-spectrum antibiotics. Details of the study Of additional note, when compared with

Ruiter-Ligeti and colleagues reviewed women who did not have C difficile infection, PHOTO: SHUTTERSTOCK

CONTINUED ON PAGE 37

36 OBG Management | July 2018 | Vol. 30 No. 7 mdedge.com/obgmanagement infectious disease UPDATE CONTINUED FROM PAGE 36

patients with infection were more likely to WHAT THIS EVIDENCE MEANS FOR PRACTICE develop a thromboembolic event (38.4 per 1,000), paralytic ileus (58.0 per 1,000), sepsis Clearly, clinicians should make every effort to prevent C difficile (46.4 per 1,000), and death (8.0 per 1,000). infection in the first place. The following preventive measures are essential: Be on guard for C difficile infection in • Avoid the use of extremely broad-spectrum antibiotics for antibiotic-treated obstetric patients prophylaxis for CD. C difficile infection is an uncommon but • When using therapeutic antibiotics, keep the spectrum as narrow potentially very serious complication of as possible, consistent with adequately treating the pathogens antibiotic therapy. Given that approximately causing the infection. half of all women admitted for delivery are • Administer antibiotics for the shortest time possible, consistent exposed to antibiotics because of prophy- with achieving a clinical cure or providing appropriate prophylaxis laxis for group B streptococcus infection, for surgical procedures (usually, a maximum of 3 doses). • If a patient receiving antibiotics experiences more than 3 loose prophylaxis for CD, and treatment of chorio- stools in 24 hours, either discontinue all antibiotics or substitute amnionitis and puerperal endometritis, cli- another drug for the most likely offending agent, depending on the nicians constantly need to be vigilant for this clinical situation. 11 complication. • If, after stopping or changing antibiotics, the clinical findings do Affected patients typically present with not resolve promptly, perform a culture or PCR assay for C difficile frequent loose, watery stools and lower and assays for the C difficile toxin. Treat as outlined above if these abdominal cramping. In severe cases, blood tests are positive. may be present in the stool, and signs of intestinal distention and even acute peritonitis may be evident. The diagnosis can most likely to be the causative agent of C dif- be established by documenting a positive ficile infection. Patients with relatively mild culture or polymerase chain reaction (PCR) clinical findings should be treated with oral assay for C difficile and a positive cytotoxin metronidazole, 500 mg every 8 hours for assay for toxins A and/or B. In addition, if 10 to 14 days. Patients with severe findings endoscopy is performed, the characteristic should be treated with oral vancomycin, gray membranous plaques can be visualized 500 mg every 6 hours, plus IV metronidazole, on the rectal and colonic mucosa.11 500 mg every 8 hours. The more seriously ill Discontinue antibiotic therapy. The first patient must be observed carefully for signs step in managing affected patients is to stop of bowel obstruction, intestinal perforation, all antibiotics, if possible, or at least the one peritonitis, and sepsis.

Danger for birth defects with maternal Zika infection present in all trimesters, but greatest in first

Hoen B, Schaub B, Funk AL, et al. Pregnancy outcomes o estimate the risk of con- after ZIKV infection in French territories in the Ameri- genital neurologic defects asso- cas. N Engl J Med. 2018;378(11):985–994. T ciated with Zika virus infection, Hoen PHOTO: SHUTTERSTOCK and colleagues conducted a prospective

mdedge.com/obgmanagement Vol. 30 No. 7 | July 2018 | OBG Management 37 UPDATE infectious disease

cohort study of pregnant women with symp- may not appear until the child is older. tomatic Zika virus infection who were enrolled The present study is an excellent comple- during March through November 2016 in ment to 2 recent reports that defined the risk French Guiana, Guadeloupe, and Martinique. of Zika virus–related fetal injury in patients All women had Zika virus infection confirmed in the United States and its territories. Based by PCR assay. on an analysis of data from the US Zika Pregnancy Registry, Honein and colleagues Details of the study reported an overall rate of congenital infec- The investigators reviewed 546 pregnancies, tion of 6%.12 The rate of fetal injury was 11% which resulted in the birth of 555 fetuses and when the mother was infected in the first tri- infants. Thirty-nine fetuses and neonates mester and 0% when the infection occurred (7%; 95% CI, 5.0–9.5) had neurologic and in the second or third trimester. The overall ocular findings known to be associated with rate of infection and the first trimester rate of Zika virus infection. Of these, 10 pregnancies infection were similar to those reported by were terminated, 1 fetus was stillborn, and 28 Hoen and colleagues. were live-born. Conversely, Shapiro-Mendoza and col- Microcephaly (defined as head circum- leagues evaluated rates of infection in US ter- ference more than 2 SD below the mean) ritories (American Samoa, Puerto Rico, and was present in 32 fetuses and infants (5.8%); the US Virgin Islands) and observed cases of 9 had severe microcephaly, defined as head fetal injury associated with second- and circumference more than 3 SD below the third-trimester maternal infection.13 These mean. Neurologic and ocular abnormalities authors reported an overall rate of infec- were more common when maternal infection of 5% and an 8% rate of infection with FAST tion occurred during the first trimester (24 first-trimester maternal infection. When TRACK of 189 fetuses and infants, 12.7%) compared maternal infection occurred in the second with infection during the second trimester and third trimesters, the rates of fetal injury The Zika virus is (9 of 252, 3.6%) or third trimester (6 of 114, were 5% and 4%, respectively, figures almost quite pathogenic 5.3%) (P = .001). identical to those reported by Hoen and col- and can cause leagues. Of note, the investigations by Honein debilitating injury Studies report similar rates and Shapiro-Mendoza included women to the developing of fetal injury with both symptomatic and asymptomatic fetus at any stage Zika virus infection primarily is caused by infection. of gestation a bite from the Aedes aegypti mosquito. The infection also can be transmitted by sexual contact, laboratory accident, and blood transfusion. Eighty percent of infected per- WHAT THIS EVIDENCE sons are asymptomatic. In symptomatic MEANS FOR PRACTICE patients, the most common clinical manifestations are low-grade fever, a disseminated Taken together, the studies discussed pro- maculopapular rash, arthralgias, swelling vide 2 clear take-home messages: of the hands and feet, and nonpurulent • Both symptomatic and asymptomatic conjunctivitis. maternal infection pose a significant risk of injury to the fetus and neonate. The most ominous manifestation of con- • Although the risk of fetal injury is great- genital Zika virus infection is microcephaly. est when maternal infection occurs in the Other important manifestations include lis- first trimester, exposure in the second sencephaly, pachygyria, cortical atrophy, and third trimesters is still dangerous. ventriculomegaly, subcortical calcifications, The Zika virus is quite pathogenic and ocular abnormalities, and arthrogryposis. can cause debilitating injury to the devel- Although most of these abnormalities are oping fetus at any stage of gestation. immediately visible in the neonate, some

38 OBG Management | July 2018 | Vol. 30 No. 7 mdedge.com/obgmanagement References 1. Hyldig N, Birke-Sorensen H, Kruse M, et al. Meta-analysis sulfamethoxazole versus placebo for uncomplicated skin of negative-pressure wound therapy for closed surgical inci- abscess. N Engl J Med. 2016;374(9):823–832. sions. Br J Surg. 2016;103(5):477–486. 9. Wilbur MB, Daum RS, Gold HS. Skin abscess. N Engl J Med. 2. Duff P. A simple checklist for preventing major complica- 2016;374(9): 882–884. tions associated with cesarean delivery. Obstet Gynecol. 10. Singer AJ, Talan DA. Management of skin abscesses in the era 2010;116(6):1393–1396. of methicillin-resistant Staphylococcus aureus. N Engl J Med. 3. Patrick KE, Deatsman SL, Duff P. Preventing infection after 2014;370(11):1039–1047. cesarean delivery: evidence-based guidance. OBG Manag. 11. Unger JA, Whimbey E, Gravett MG, Eschenbach DA. The 2016;28(11):41–47. emergence of Clostridium difficile infection among peripar- 4. Patrick KE, Deatsman SL, Duff P. Preventing infection after tum women: a case-control study of a C difficile outbreak on cesarean delivery: 5 more evidence-based measures to con- an obstetrical service. Infect Dis Obstet Gynecol. 2011;267249. sider. OBG Manag. 2016;28(12):18–22. doi:10.1155/2011/267249. 5. Lasley DS, Eblen A, Yancey MK, Duff P. The effect of placental 12. Honein MA, Dawson AL, Petersen EE, et al; US Zika Preg- removal method on the incidence of postcesarean infections. nancy Registry Collaboration. Birth defects among fetuses Am J Obstet Gynecol. 1997;176(6):1250–1254. and infants of US women with evidence of possible Zika 6. Duff P. A simple checklist for preventing major complica- virus infection during pregnancy. JAMA. 2017;317(1): tions associated with cesarean delivery. Obstet Gynecol. 59–68. 2010;116(6):1393–1396. 13. Shapiro-Mendoza CK, Rice ME, Galang RR, et al; Zika Preg- 7. Tuuli MG, Liu J, Stout MJ, et al. A randomized trial compar- nancy and Infant Registries Working Group. Pregnancy out- ing skin antiseptic agents at cesarean delivery. N Engl J Med. comes after maternal Zika virus infection during pregnancy— 2016;374(7):647–655. US territories, January 1, 2016–April 25, 2017. MMWR Morb 8. Talan DA, Mower WR, Krishnadasan A, et al. Trimethoprim- Mortal Wkly Rep. 2017;66(23):615–621.

mdedge.com/obgmanagement Vol. 30 No. 7 | July 2018 | OBG Management 39