The n e w e ng l a n d j o u r na l of m e dic i n e

Original Article

Adjunctive Azithromycin Prophylaxis
for Cesarean Delivery
Alan T.N. Tita, M.D., Ph.D., Jeff M. Szychowski, Ph.D., Kim Boggess, M.D.,
George Saade, M.D., Sherri Longo, M.D., Erin Clark, M.D., Sean Esplin, M.D.,
Kirsten Cleary, M.D., Ron Wapner, M.D., Kellett Letson, M.D., Michelle Owens, M.D.,
Adi Abramovici, M.D., Namasivayam Ambalavanan, M.D., Gary Cutter, Ph.D.,
and William Andrews, M.D., Ph.D., for the C/SOAP Trial Consortium*​​


The addition of azithromycin to standard regimens for antibiotic prophylaxis be- From the Departments of Obstetrics and
fore cesarean delivery may further reduce the rate of postoperative infection. We Gynecology (A.T.N.T., J.M.S., W.A.), Bio-
statistics (J.M.S., G.C.), and Pediatrics
evaluated the benefits and safety of azithromycin-based extended-spectrum pro- (N.A.), University of Alabama at Birming-
phylaxis in women undergoing nonelective cesarean section. ham, Birmingham; the Departments of
Obstetrics and Gynecology, University of
METHODS North Carolina, Chapel Hill, Chapel Hill
(K.B.), and Mission Hospital, Asheville
In this trial conducted at 14 centers in the United States, we studied 2013 women (K.L.) — both in North Carolina; the Uni-
who had a singleton pregnancy with a gestation of 24 weeks or more and who were versity of Texas Medical Branch, Galves-
undergoing cesarean delivery during labor or after membrane rupture. We randomly ton (G.S.), and the University of Texas
Health Sciences Center, Houston (A.A.);
assigned 1019 to receive 500 mg of intravenous azithromycin and 994 to receive Ochsner Health System, New Orleans
placebo. All the women were also scheduled to receive standard antibiotic prophy- (S.L.); the University of Utah (E.C., S.E.)
laxis. The primary outcome was a composite of endometritis, wound infection, or and Intermountain Health Care (E.C., S.E.),
Salt Lake City; Columbia University, New
other infection occurring within 6 weeks. York (K.C., R.W.); and the University of
Mississippi, Jackson (M.O.). Address re-
RESULTS print requests to Dr. Tita at the Mater-
The primary outcome occurred in 62 women (6.1%) who received azithromycin and nal–Fetal Medicine Division and Center
in 119 (12.0%) who received placebo (relative risk, 0.51; 95% confidence interval [CI], for Women’s Reproductive Health, De-
partment of Obstetrics and Gynecology,
0.38 to 0.68; P<0.001). There were significant differences between the azithromycin University of Alabama at Birmingham,
group and the placebo group in rates of endometritis (3.8% vs. 6.1%, P = 0.02), 619 19th St. S., Birmingham, AL 35249, or
wound infection (2.4% vs. 6.6%, P<0.001), and serious maternal adverse events at ­atita@​­uabmc​.­edu.

(1.5% vs. 2.9%, P = 0.03). There was no significant between-group difference in a *A complete list of investigators in the
secondary neonatal composite outcome that included neonatal death and serious Cesarean Section Optimal Antibiotic
Prophylaxis (C/SOAP) Trial Consortium
neonatal complications (14.3% vs. 13.6%, P = 0.63). is provided in the Supplementary Ap-
pendix, available at
N Engl J Med 2016;375:1231-41.
Among women undergoing nonelective cesarean delivery who were all receiving DOI: 10.1056/NEJMoa1602044
standard antibiotic prophylaxis, extended-spectrum prophylaxis with adjunctive Copyright © 2016 Massachusetts Medical Society.
azithromycin was more effective than placebo in reducing the risk of postoperative
infection. (Funded by the Eunice Kennedy Shriver National Institute of Child Health
and Human Development; C/SOAP number, NCT01235546.)

n engl j med 375;13  September 29, 2016 1231
The New England Journal of Medicine
Downloaded from on January 8, 2017. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.

subse- It has been thought that the efficacy of such quent vaginal delivery. Patients who were allergic to cephalo- tained from all the patients. went the consent procedure at admission for elective cesarean delivery with standard preinci. donated the azithromycin that was used in the fection is a major cause of maternal trial but did not participate in the design. is associated with a rate of surgical-site infection (including endometritis and wound infection) Trial Design that is 5 to 10 times the rate for vaginal delivery. tion of 24 weeks or more who were undergoing monly. regardless deliveries are nonelective. chorioamnionitis species. hypo- in the United States. All the women were to receive standard prophy- which is available with the full text of this article laxis (cefazolin) according to the protocol at each at NEJM. The institutional review board calcemia. Antibiotic prophylaxis was adminis- 1232 n engl j med 375.0 mg per deciliter (177 μmol per liter) or the need for dialysis. Most women under- occur in up to 12% of women undergoing non. For personal use only. which are commonly associated with in.14 ity after the decision was made to proceed to Studies (including a single-center randomized cesarean delivery. a known allergy to azithromycin. at each study site approved the trial protocol. cal change of at least 1 cm of dilation or at least brane rupture. or other infection requiring postpartum antibi- fections after cesarean section.13. 2016 The New England Journal of Medicine Downloaded from nejm. We also excluded patients who had substan- of infection after cesarean section without increas. postoperative infections of whether labor had started. pragmatic. Pfizer gentamicin). azithromycin to standard antibiotic prophylaxis and fetal death or known major congenital anom- before skin incision would reduce the incidence aly. a serum creatinine level of more than 2. use of medications known to pro- laxis (C/SOAP) trial was a double-blind. Exclusion criteria were an inability to provide sarean section than standard prophylaxis alone.2. Written informed consent was ob. trial center. 50% effacement.15 consent. a cephalosporin given before skin  September 29.13 nejm. particularly among women fined as regular contractions with cervical dila- who undergo nonelective procedures (i. long the QT interval. tial liver disease (cirrhosis or an aminotransfer- ing the risk of other adverse maternal and perina. All rights reserved. Women with membrane rup- cies). unsched. azithromycin use within prophylaxis was due to coverage for ureaplasma 7 days before randomization. Funding was provided sporin or penicillin received the local alternative by the Eunice Kennedy Shriver National Institute of medication (clindamycin alone or clindamycin plus Child Health and Human Development. diarrhea at the time of Me thods planned randomization.5 Women with a singleton pregnancy with a gesta- Despite routine use of antibiotic prophylaxis (com.e. The Cesarean Section Optimal Antibiotic Prophy. such as hypokalemia. nonelective cesarean delivery during labor or after sion6). cardiomyopathy or pul- Trial Design and Oversight monary edema. on January 8. pregnancy-associated in..22 — may result in a lower risk of infection after ce. death and is the fourth most common or reporting of the trial. infection after cesarean section remains membrane rupture were eligible. or hypomagnesemia. tion of 4 cm or more or with documented cervi- uled cesarean section during labor. The also associated with a prolonged hospital stay and first two authors take responsibility for the ac- increased health care costs. delivery and were rescreened to confirm eligibil- sion prophylaxis. maternal structural heart disease. An independent data and cause in the United States.1 Maternal infection is safety monitoring board oversaw the trial.3 Cesarean delivery is curacy and completeness of the reporting and the the most common major surgical procedure4 and fidelity of the report to the trial protocol. or for maternal or fetal emergen. The n e w e ng l a n d j o u r na l of m e dic i n e G lobally. conduct. after mem. ase level at least three times the upper limit of the tal outcomes.6-12 As many as 60 to 70% of all cesarean ture for at least 4 hours were eligible. in accordance with the guidelines of the Ameri- spectrum prophylaxis — a single dose of azithro. Gestational age was estimated trial) suggest that azithromycin-based extended.16-21 We performed otic therapy (although patients receiving antibi- this study to assess whether the addition of otics for group B streptococcus were eligible). can College of Obstetricians and Gynecolo- mycin plus standard cephalosporin prophylaxis gists. arrhythmias. Copyright © 2016 Massachusetts Medical or known substantial elec- randomized clinical trial conducted at 14 hospitals trolyte abnormalities. . normal range). No other uses without permission. Labor was de- an important concern.

4°F]). Neonatal serious adverse events included the dominopelvic abscess.complications. a maternal safety composite outcome cording to Food and Drug Administration guide. maternal fever. available at NEJM. periventricular leukomalacia. 2016 1233 The New England Journal of Medicine Downloaded from nejm.(defined below as maternal serious adverse events). induration around the incision or purulent dis. pneumonia. Detailed trial saline) or an identical-appearing saline placebo.adverse complication. grade III or members retrieved the next sequentially numbered higher intraventricular hemorrhage.facility). are provided in the access to the randomization algorithm through Supplementary Appendix. Azithromycin Prophylaxis for Cesarean Delivery tered before surgical incision or as soon as pos. or transfer to a long-term care cation was infused over a period of 1 hour. necrotizing entero- recycling the randomization sequence. or purulent drainage from the uterus. lines for azithromycin. . Maternal serious adverse gery. maternal sepsis. wound infection. and bron- typically once the decision was made to proceed to chopulmonary on January 8. including pulmonary em- tion characterized by cellulitis or erythema and bolism. any seri- 38°C [100. Among such outcomes were specific mater- nal postoperative infections.listed in Table S1 in the Supplementary Appen- search staff. Trial Outcomes unscheduled visits and readmissions.) Study staff colitis.neonatal safety composite. or other infections (ab. n engl j med 375. The computer-generated surgical site infections are provided in the Sup- block-designed randomization plan was produced plementary Appendix. Primary outcomes were a dedicated password-protected website. seroma. including the re- tion.ous adverse event leading to the discontinuation ness. Only the investigational clinical diagnosis leading to therapy with antibi- pharmacists who prepared the study drug had otics and additional criteria. and infection with resistant organisms. All rights reserved. or placebo were sequentially numbered and kept A major secondary neonatal outcome was a in a secure refrigerator (7-day shelf life). Criteria for other infections. Wound hema- toma.comes that were specified in the statistical analysis sion was connected.and cardiac events. dix. suspected allergic reactions (includ- recognized cause: fever (temperature of at least ing anaphylaxis or generalized skin rash). (Expired study bags were discarded without spiratory distress syndrome. which included a fied according to site. Endometritis was defined as the presence of events (maternal safety composite outcome) in- at least two of the following signs with no other cluded death. or breakdown alone in the absence Interventions of the preceding signs did not constitute infection. At the time that the study infu. Cesarean procedures and care at each center Other secondary maternal and neonatal out- followed providers’ usual practices. the system- study drug bag up to 1 hour before incision and ic inflammatory response syndrome. and length of hospital stay. For personal use only. Copyright © 2016 Massachusetts Medical Society. criteria consistent with the recommendations of Clinical and research staff members other than the National Healthcare Safety Network of the the investigational pharmacist were unaware of Centers for Disease Control and Prevention for treatment assignments. ac. the patient was considered plan included a neonatal safety composite (death.sepsis. and included necrotizing  September 29. admission to an intensive care unit (ICU). neonatal The primary outcome was a composite of endo. uterine tender.charge from the incision site with or without fever sible thereafter. to have undergone randomization. sepsis.reported serious events. Study medi. and other or meningitis) occurring up to 6 weeks after sur. which composite of death. Patients were randomly assigned to receive either Diagnosis of abdominal or pelvic abscess required azithromycin (at a dose of 500 mg in 250 ml of radiologic or surgical confirmation. intraventricular hemorrhage. Wound of a study medication or suspected to be due to infection was defined as the presence of either the medication. and any other reported serious superficial or deep incisional surgical-site infec. 2017. or other complications.allergic reaction. Other secondary out- cesarean section. centrally adjudicated by investigators who were The 250-ml bags containing the azithromycin unaware of treatment assignments. grade III or higher tic thrombophlebitis. Trial outcomes comes that were specified in the protocol are and other data were abstracted by certified re. pelvic sep. metritis.23 by the data coordinating center and was strati.13 nejm. suspected or confirmed allowed for rapid administration after randomiza. No other uses without permission. abdominal pain.

associated with significantly lower rates of en- teraction in multivariable logistic-regression mod. who were unaware of treatment assignments as- certained maternal and infant outcomes by re. 0. R e sult s viewing medical records from the delivery hospi- talization. The specific char- outcomes.001) (Table 3).51.38 to 0. 0.34. 6.0%) who received placebo (relative risk.55). P<0. In secondary analy. mem.6%. assuming a base. For personal use only. includ- ing indications for cesarean delivery. 17 to 41) for wound infections. P = 0. 0. The number of patients who would meters) of less than 30 versus 30 or more. Tests of in. and initiation of study medica.02) and wound infec- treatment effect on the primary outcome across tions (2.4% vs. els were used to test the homogeneity of the 95% CI. acteristics related to the cesarean delivery. 0.40 to 0. The primary composite outcome occurred in 62 ables. accord.001).61.62.22 to 0.42 to 0. and 24 tion before versus after skin incision. The use of azithromycin was sion models for the primary outcome. 24 to 245) for endometritis.68.1%) who received azithromycin and in were calculated for outcomes. 95% CI.048 level of significance.48. relative risk. of the primary outcome using O’Brien–Fleming 0. We calcu. 0. P<0. No other uses without permission.56.13 nejm.05.05 level of significance. body-mass index (the weight in tions were low and did not differ significantly be- kilograms divided by the square of the height in tween groups. We used the chi-square test or Fisher’s exact test to analyze categorical Primary Outcome variables and Student’s t-test for continuous vari.14 able for 1961 of the 2013 women (97. 0. The n e w e ng l a n d j o u r na l of m e dic i n e Outcome Ascertainment and Follow-up 0.35. Azithromy- reduction from a baseline risk of 8%.790 women who were screened at the 14 missions. (adjusted odds ratio. The risks of other infec- ing to trial site. All rights reserved. timing of receipt of study We determined that a sample size of 2000  September 29. 2017.35 to 0. We performed one planned interim analysis endometritis (adjusted odds ratio. 1). intention-to-treat principle. The lated the number of patients who would need to results were similar after planned adjustment for be treated to prevent one primary outcome event smoking with respect to the primary outcome and 95% confidence intervals. composite neonatal outcome. need to be treated to prevent one study outcome brane rupture before randomization versus after was 17 (95% CI.92. balanced at randomization using logistic-regres.66). medication.4%) who All analyses were performed according to the underwent randomization (Fig. the final analysis was evaluated at a odds ratio. 12 to 30) for the primary outcome. The characteristics of the patients at 3 months to identify infant deaths and adverse at baseline were similar in the two groups. 0. Maternal this sample size would provide a power of 80% and neonatal outcome data were available for all or more to assess a 30% relative reduction in the the patients at the time of hospital discharge. (95% CI. and type of surgical skin preparation. and from hospital ad. dometritis (3. we adjusted for characteristics that were not 0. 119 (12. . 1). at a two. receipt of Statistical Analysis standard prophylaxis. from visits to a postpartum clinic or Characteristics of the Patients emergency department. ses. Patients were scheduled for a 6-week clinical sites from April 2011 through November postpartum visit (or were contacted by tele. 2016 The New England Journal of Medicine Downloaded from on January 8. All secondary outcomes Trained and certified research staff members were evaluated at a 0. Relative risks and 95% confidence intervals women (6. subgroups in four prespecified analyses. Of 17. We also calculated that in 88% of the women in each group. 43 (95% CI. 95% CI. and wound infections (adjusted boundaries. Postpartum follow-up within 6 weeks was avail- line risk of 16%. 95% confidence interval [CI]. relative risk. 0. a total of 1019 were randomly assigned to phone) to ascertain maternal and infant medical the azithromycin group and 994 to the placebo events and visits and were contacted by telephone group (Fig. Copyright © 2016 Massachusetts Medical Society. tients would provide a power of 80% to detect a were similar in the two groups (Table 2). except events. 6. 0.1%. randomization.8% vs. 95% CI.21 to 0. Medical records (including those at other that smoking was slightly less prevalent in the health facilities) were required to verify study azithromycin group (Table 1). 0. 95% CI. cin or placebo was administered before incision sided alpha level of 0. More 33% relative reduction in the primary outcome than 99% of the patients in each group received from a baseline risk of 12% or a 40% relative the standard antibiotic prophylaxis. Results 1234 n engl j med 375.91). 2014.

but data were missing on the timing of admin- istration in 9. azithromycin group and in 135 (13. 0. 1018 patients received the assigned drug. No other uses without permission. Copyright © 2016 Massachusetts Medical Society. group B streptococcal status. 2017. from survival analyses were also similar to the greater reduction in infections for women re- findings in the primary analysis (Table S2 and ceiving staples than for those receiving sutures.3%) in the tion was detected in a post hoc analysis of skin. 2016 1235 The New England Journal of Medicine Downloaded from nejm.6%) in the closure methods (P = September 29.02). Enrollment and Outcomes. mycin was not detected in prespecified sub.790 Patients were screened for eligibility 10. 95% CI.716 Were ineligible 6149 Underwent vaginal delivery 1382 Did not provide consent 848 Underwent a scheduled or elective cesarean section 534 Had chorioamnionitis or an active bacterial infection 387 Delivered a dead fetus or an infant with a major abnormality 1416 Had other reasons 3017 Declined to participate 4057 Patients were screened and gave consent 2044 Were ineligible at time of delivery 1623 Underwent vaginal delivery 28 Had chorioamnionitis 14 Declined after initial consent 379 Had other reasons 2013 Patients underwent randomization 1019 Were assigned to receive azithromycin 994 Were assigned to receive placebo 1018 Received intervention 992 Received intervention 1 Did not receive intervention 2 Did not receive intervention 25 Did not have 6-wk maternal 27 Did not have 6-wk maternal follow-up follow-up 54 Did not have 3-mo child 55 Did not have 3-mo child follow-up follow-up 1019 Were included in the primary analysis 994 Were included in the primary analysis Figure 1. but the timing was not docu- mented in 11. 1. and timing Secondary Neonatal and Maternal Outcomes of medication administration (Table S3 in the The composite neonatal outcome of death or com- Supplementary Appendix). S1 through S4 in the Supplementary Ap. plications occurred in 146 infants (14. No heterogeneity in treatment effect was detected pendix). . Azithromycin Prophylaxis for Cesarean Delivery 17. In the azithromycin group. diabetes status. All rights reserved. which suggested a placebo group (relative risk. including Heterogeneity of the effect of adjunctive azithro. groups. Figs.13 nejm.85 n engl j med on January 8. For personal use only. in other post hoc subgroup analyses.05. A significant interac. according to study site. In the placebo group. and preterm delivery. membrane status at randomization. vaginal preparation. 992 patients received the assigned saline infusion. obesity status.

Other maternal outcomes. 2017. All rights reserved.0) 47 (4.047). and ferent between groups. † Race or ethnic group was self-reported.4) Other 109 (10.4) 312/983 (31.5±7.8) 249 (25.5 Race or ethnic group — no.4) Smoking during pregnancy — no.2) 2 (0.0) 114 (11.0±2. (%)† Non-Hispanic black 351 (34. For personal use only.3±7.1) Missing data 2 (0. (%) 35 (3.4) 266 (26.4) 478 (48. infection.9) 342 (34. The n e w e ng l a n d j o u r na l of m e dic i n e Table 1.2) 30 to <40 503 (49.3) 221 (22.8) Positive for group B streptococcus — no.7) Use of illegal drugs during pregnancy — no.9) Non-Hispanic white 356 (34.31. (%) Any 552 (54. and scheduled visits for any reason or specifically for at 72 days from the sudden infant death syndrome.3) Alcohol use during pregnancy — no. (%)† 317/1008 (31.7) Chronic hypertension — no.9) 146 (14.1 28.3) ≥20 wk of gestation 416 (40.7) 103 (10.0) 54 (5.3) Hispanic 203 (19. three deaths in the azithromycin group.8) Gestational age At randomization — wk  September 29.9 Category — no.4) Diabetes mellitus — no. There were no significant differences between the groups except for smoking dur- ing pregnancy (P = 0.8) 402 (40. which including rates of postpartum fever. (%) Any 142 (13.4±6.3 39.2±6. to 1. .4) Body-mass index‡ Mean 35. treatment occurred at 15 days from fulminant herpes sim. (%) 249 (24.5 to <25 53 (5. 2016 The New England Journal of Medicine Downloaded from nejm. (%) <18.4) 28 (2.13 nejm. at 42 days from uncertain cause.5) 122 (12. azithromycin group (Table 4). 1236 n engl j med 375. which occurred 5 days tion after discharge. ‡ The body-mass index is the weight in kilograms divided by the square of the height in meters.7 35.7) Gestational only 99 (9. and need for readmission or un- plex virus. (%) 41 (4.* Azithromycin Placebo Characteristic (N = 1019) (N = 994) Age — yr 28.1) 1 (0. were significantly less common in the The frequencies of other neonatal outcomes. were not significantly dif- after birth as a result of extreme prematurity.3 <37 wk at delivery — no.2) 560 (56./total no.3) 25 to <30 217 (21. (%) 97 (9.2) Private insurance — no.1) 18.9±2.5 1 (0. in. with antibiotics. There was one neonatal cluding neonatal ICU admission or hospitaliza- death in the placebo group.7) Previous pregnancy — no. Percentages may not total 100 because of rounding.1) ≥40 243 ( on January 8.2) 43 (4.9) 208 (20. Copyright © 2016 Massachusetts Medical Society.63) (Table 4).5) * Plus–minus values are means ±SD. (%) 51 (5. Characteristics of the Patients at Baseline.4) 341 (34.7) 106 (10. No other uses without permission. P = 0. (%) 112 (11.

0) 0.7) 0.3) 342/993 (34. Azithromycin Prophylaxis for Cesarean Delivery Table 2.2) 95/993 (9.2) Vaginal preparation Any 265/1019 (26.9) 0.7) 949/992 (95.8) Chlorhexidine–alcohol plus iodine 218/1019 (21.1) 12/992 (1. No other uses without permission.78 Chlorhexidine–alcohol 340/1019 (33.9) 243/994 (24.6) Abnormal presentation 59/1019 (5.1) >60 min before 51/1009 (5.6) 0./total no. Copyright © 2016 Massachusetts Medical Society. .99 Low transverse 975/1019 (95.10 Pfannenstiel 987/1019 (96.4) 316/994 (31.1) 15/994 (1.1) 45/992 (4.5) None 754/1019 (74.95 Skin-incision type 0. (%) Primary indication for cesarean delivery*  September 29.4) Iodine–alcohol 92/1019 (9.4) 121/981 (12.0) * One patient in the placebo group did not have a primary indication for cesarean delivery.4) Nonreassuring fetal heart tones 268/1019 (26.98 Type Iodine 254/1019 (24. † The P value for this category is for the between-group comparison for administration of the study drug before the inci- sion versus administration after the incision.0) 258/994 (26.6) After incision 125/1009 (12.3) 43/992 (4.3) 103/993 (10.9) 947/992 (95.3) 258/993 (26. All rights reserved. 2017. Characteristics of Cesarean Procedures.1) 45/981 (4.6) 860/981 (87.45 Timing of study-drug administration Before skin incision† 884/1009 (87. n engl j med 375.4) 213/994 (21.9) Receipt of standard antibiotic prophylaxis 1017/1019 (99.4) Dermabond 11/1019 (1. For personal use only.4) 0.7) Other reason 133/1019 (13.0) 101/994 (10.0) 736/994 (74.68 Chlorhexidine 11/1019 (1.4) Elective repeat procedure meeting study criteria 94/1019 (9.1) 128/993 (12.0) Failed induction 105/1019 (10. 2016 1237 The New England Journal of Medicine Downloaded from on January 8.7) Other 44/1019 (4.2) 364/994 (36.7) 411/992 (41.6) 0.8) 990/994 (99.8) 67/993 (6.97 Failure to progress 360/1019 (35.97 0 to 60 min before 833/1009 (82.4) Suture 593/1019 (58.2) 569/992 (57.6) 815/981 (83.5) Closure method 0.91 Staples 415/1019 (40.3) Membrane rupture before skin incision 889/1012 (87.2) Uterine incision 0. Azithromycin Placebo Characteristic (N = 1019) (N = 994) P Value no.8) 868/987 (87.3) Skin preparation Chlorhexidine 369/1019 (36.13 nejm.5) Vertical 32/1019 (3.

The number of eli- (0.5% vs. we gram-negative bacilli and staphylococcus and en.5) >0.06 Deep wound infection 6 (  September 29. Adverse Events samples).4) 66 (6.4) NA 0.3) 6 (0. 2017. Overall.1) 2 (0.62 (0.2) 1 (0.6) 0.001 Necrotizing fasciitis 0 4 (0.and S8 in the Supplementary Appendix). wound cultures in the azithromycin group and and the risks of adverse neonatal outcomes were four in the placebo group. 19 newborns not increased in this group. 2.12–1.92) 0.001) and bacteria resistant to at least ternal events and several other maternal out- one antibiotic (1. Other maternal or neonatal adverse events did Sensitivity Analyses not differ significantly between groups (Tables S4 We conducted sensitivity analyses that excluded and S5 in the Supplementary Appendix).2) 0.9%.94) 0. Fifty Discussion women (2.42).8) 0.01). The azithromycin group had a dard antibiotic prophylaxis significantly reduced significantly lower prevalence than the placebo the frequency of infection after nonelective ce- group with respect to positive cultures (1.22–0.68) <0. the results were simi- Resistance lar to those in the primary analyses (Tables S7 We examined results of all clinical maternal post.51 (0.4%.34 Abdominal or pelvic ab. sarean section.1%] in the pla- cebo group (1.found that the addition of azithromycin to stan- terococcus species.99 Pneumonia 1 (0.1) 0 NA >0.5%) had cultures that were positive for at least one bacterial organism. 2.50]) or in the prevalence of bac- cant between-group difference was observed in teria resistant to at least one antibiotic (0. most commonly In this large.02 Wound infection 24 (2.13 nejm.5% vs. randomized trial. no signifi. partum cultures in those with wound infections (Table S6 in the Supplementary Appendix). including readmissions.6) 0.* Azithromycin Placebo Relative Risk Outcome (N = 1019) (N = 994) (95% CI) P Value no. 0 4 (0.4) NA 0.56 Other infection 3 (0. P = 0.73 (0. (%) Primary composite outcome 62 (6. For personal use only.49 (0. 0 0 NA NA bitis Maternal sepsis 2 (0.25–2. All rights reserved.8) 61 (6. with no significant between-group dif- Maternal serious adverse events were less com.1) 119 (12. cluding the safety composite outcome (Table 4). the rates of neonatal serious adverse events. .42–0. P = 0.06 scess Septic pelvic thrombophle.03).6) 8 (0. Bacteria comes.99 Pyelonephritis 1 (0.95 (0.10) on January 8.8%] in mon in the azithromycin group than in the pla.37) 0. in. ference in the prevalence (8 newborns [0.04–5.56) <0. the azithromycin group and 11 [1. Copyright © 2016 Massachusetts Medical Society.0% vs.4% vs.1) 1. No other uses without permission. were lower in the resistant to azithromycin were identified in three azithromycin group than in the placebo group.8%. The risks of serious adverse ma- 3.62 Meningitis 0 0 NA NA * NA denotes not applicable. cebo group [P = 0.0) 0. P = 0. The n e w e ng l a n d j o u r na l of m e dic i n e Table 3.35 (0. multicenter.9%) had positive culture results (mainly in blood gible women who would need to be treated to 1238 n engl j med 375. patients with protocol violations or were restrict- ed to women with complete postpartum follow- Bacterial Cultures and Antimicrobial up data.001 Endometritis 39 (3. 2016 The New England Journal of Medicine Downloaded from nejm. Primary Composite Outcome and Its Components. In these analyses.38–0.18–21.1) 0.6%. P = 0. 0.49 (0.

3) 1.86) 0.94) 0.001 Postpartum use of antibiotics 126 (12.006 Composite serious adverse events† Neonatal serious adverse events Any 7 (0.29) 0.38–0.1) 2.9) 0.6) 49 (4. ‡ This category is a composite of perinatal death. 0 1 (0.79–1.04–5.44–0.85–1.99 Death Within 3 mo 3 (0.8) 43 (4.00) 0.0) 81 (8. 43 for endometritis.2) 0. † Details about serious adverse events are provided in Table S4 in the Supplementary Appendix. Our findings are consistent with those of n engl j med 375. Copyright © 2016 Massachusetts Medical Society.60–0.8) 124 (12.493 drome NICU admission 171 (16. and NICU neonatal intensive care unit.58) <0.002 visit Clinic visit 32 (3.34–0. § This category is the same as the maternal safety composite.0) 0.88 (0.1) 0 NA >0.4) 0. prevent one study outcome was 17 for the primary cording to prespecified subgroup.80) >0.1) 33 (3.63 Confirmed 1 (0. .1) 0.1) 1 (0.36 (0.3) 0.004 Any postpartum readmission or unscheduled 83 (8. Secondary Neonatal and Maternal Outcomes. All rights reserved.4) 34 (3.91) 0. perinatal allergic reaction.1) 0.93 (0.43–4. of patients (%) Neonatal Composite neonatal outcome 146 (14.19) 0.23–0.62 Within 28 days 1 (0.06–15.1) 53 (5.98 (0. and neonatal transfer to a chronic care facility.35) 0.99 (0.99 Periventricular leukomalacia 0 0 NA NA Intraventricular hemorrhage of grade III or 1 (  September 29.6) >0. 2016 1239 The New England Journal of Medicine Downloaded from nejm.1) 123 (12.92) 0.62 more Bronchopulmonary dysplasia 4 (0.58–1.3) 1 ( on January 8.6) 1.29–5.87) 0.99 Composite neonatal complications 45 (4. including clini- outcome.3) 84 (8.1) 1.1) 0.57 Maternal Postpartum fever 51 (5.30 (0.30–28.13 nejm.29 (0.3) 0.99 Systemic inflammatory response syn.6) >0.5) 1.93 (0.05 (0.85) 0.25 Respiratory distress syndrome 42 (4.11 Readmission 39 (3.5) 0.27–0. the benefit of the inter.1) NA 0.3) 1.54 (0.06–15.9) 0.81–1.005 Readmission 27 (2.94 (0.34 Necrotizing enterocolitis 1 (0.51–0.02 Emergency department visit 54 (5.1) 1 (0.4) 3 (0.4) 166 (16.3) 1 (0.96–1.86) 0.31) 0.18 (0.89 Readmission or unscheduled visit 170 (16. and 24 for wound cal site and timing of administration of the infections.5) 29 (2.45–0.30–28.20) 0. In addition.4) 1.59 (0.50 (0.8) 169 (17. Azithromycin Prophylaxis for Cesarean Delivery Table 4.3) 62 (6. 2017.63 (0.37) 0.98 (0.62 All maternal serious adverse events§ 15 (1.24 (0. No other uses without permission.61 (0.1) 2 (0.77 Safety composite‡ 3 (0.5) 0.3) 135 (13.1) 0.94) 0.75–1.63 Sepsis Suspected 120 (11.74 (0.83–2.* Azithromycin Placebo Relative Risk Outcome (N = 1019) (N = 994) (95% CI) P Value no.7) 140 (14.medication in relation to skin incision.49 (0.007 Because of infection 23 (2. For personal use only.2) 0.7) 5 (0.46) 0.7) 0.37 (0.03 * NA denotes not applicable.1) 0.66 (0.1) 2. vention did not appear to vary significantly ac.

rates of infection and maternal use of health care A limitation of our study is the exclusion of resources without increasing the risk of neonatal women undergoing a scheduled cesarean section adverse outcomes. and those with a diagnosis of chorio. Sciences. population of young and middle-aged healthy women in the azithromycin group also had a adults. In some reports.22-24 but further investigation is war. Horizon Pharma. and PTC Therapeutics. of azithromycin. Receptos. 2016 The New England Journal of Medicine Downloaded from nejm. and grant support to his institution from Teva Neu- population. cultures are reassuring. receiving have a scheduled cesarean delivery have a low risk fees for serving on data and safety monitoring boards from of infection. 1240 n engl j med 375.30 shorter hospital stay. Pfizer. . Transparency Life Sciences. serving as a member of scientific advisory Such factors are important. No other potential conflict of interest relevant to this that the beneficial effect of azithromycin prob. but ongoing monitoring ganisms. The mechanism by Pharma. Previous studies Supported by a grant (#HD64729) from the Eunice Kennedy of azithromycin-based extended prophylaxis have Shriver National Institute of Child Health and Human Develop- suggested potential benefits for these two groups ment (NICHD). The available culture results suggest roscience. excluded women with a history of arrhythmia or aerobes when specific cultures are performed. findings in these two groups.13 nejm. ganisms. Genentech. Specific boards from EMD Serono. phylactic extended-spectrum coverage than with The selection of resistant organisms is a po- standard antibiotic prophylaxis. oral doses of azithromycin over a period of at tis. ably extends beyond coverage of ureaplasma or- tion after cesarean section with the use of pro. tential concern regarding azithromycin-based fewer infections were reported with the addition prophylaxis. Ono/Merck. Roche. which are more commonly associated for changes in resistance profiles is needed. Genentech. AirStrip. and those with intrapartum chorioamnionitis. which covers anaerobes. older patient cohort with underly- studies from the same center indicated that women ing coexisting conditions. Cutter. Somah- plasma species are not routinely performed in lution. show any safety signal involving cardiac events or trum antibiotic prophylaxis administered after maternal death with the single intravenous dose umbilical-cord clamping had a rate of postop. The azithromycin used in the study was provid- ed by Pfizer through an investigator-initiated grant. Neuren Pharmaceuticals. with demon. For personal use only. because women who boards of Sera Prognostics and Clinical Innovations and having an equity interest in Sera Prognostics. and Janssen.29 Our data did not who received azithromycin-based extended-spec. and Nivalis. However.24-26 A single-center randomized trial involv. No other uses without  September 29. article was reported.11 Our findings indicate cision approach. Esplin. given a previous observational and because it has been associated with reduced study reporting an association between multiple risks of both wound infections and endometri. Teva Neuroscience. Genzyme. Gilead amnionitis are treated with broad-spectrum anti. Opko. fees for serving on advisory after cesarean section remains unclear. GlaxoSmithKline. Clark reports receiving travel and grant support from ranted to assess efficacy and cost-effectiveness. Sanofi. Atlanta. 2016. Our findings from clinical maternal azithromycin because it covers ureaplasma or. 2017. on January 8. least 5 days and the risk of cardiac death in a ing 597 women and subsequent observational nonpregnant. Apotex/Modigenetech. this is consistent with reassur- erative infection that was at least 30% lower ing findings subsequently reported in a general than did women receiving standard prophylaxis. The vast majority of patients that extended-spectrum prophylaxis with adjunc- received antibiotics before incision. tests for the presence of ureaplasma or myco. Questcor. These exclusions limit the generalizability of our Presented in part at the 36th annual meeting of the Society for Maternal–Fetal Medicine. we tested a prein. fees from a law firm for providing a practice and were not available for this study legal opinion regarding a patent infringement case on behalf of Galderma.24-26 Contrary to previous Standard antibiotic prophylaxis has been shown studies in which prophylactic extended-spectrum to reduce rates of surgical-site infection after cesar- antibiotics were administered after skin incision ean section. Reata biotics after cesarean section. tive azithromycin for cesarean delivery in women strated maternal benefits and no evidence of at increased risk for infection safely reduces the neonatal harm. fees for serving on a steering which azithromycin reduces the rate of infection committee from MedImmune.15. consulting fees from Teva Neuroscience. it is unlikely that the single of metronidazole. We with infections after cesarean section than an.24-28 We focused on resistance. Novartis. than dose of antibiotic would significantly increase with standard prophylaxis. All rights reserved. along with rates of serious maternal and umbilical-cord clamping. of women. Copyright © 2016 Massachusetts Medical Society. complications and death. Opexa. Savara. The n e w e ng l a n d j o u r na l of m e dic i n e previous studies supporting a lower risk of infec. cardiomyopathy. Dr. February 1–6. Biogen Idec. and Dr.

Watts DH. ACOG practice bulle- phylaxis with antibiotics for hospital re. Costantine MM.​179:​1254-60. Hauth​provgovpart/​ 203. initiatives/​nqi/​Documents/​ 4. Kingston J. Obstet Gynecol 2001. for oversight. early postpartum endometritis with fac. PG. All rights reserved. stet Gynecol 2009. 24.​184:​ cy. The National Healthcare Safety Net- SE. Faro S. 120: use of prophylactic antibiotics 16. Hauth JC. Orange. Hood WA. Pitt C.​75:​52-8. Kirkinen P. Tita AT. Adjunctive intravaginal metronida- microbial prophylaxis in surgery: Ameri. Robert Schelonka. Obstet Gynecol 2008.. 2017. Am J Obstet Gynecol 1998. Tita AT. Kaunitz 8. Impact of ex- vention of surgical site infection. Callaghan WM.N.H. B. Evid Rep Technol infections at term. Rebecca M. Owen J.. Horan TC. CD007482. 28. Scott K. . 81:​383-6. Grimes Silver LC. ASHP therapeutic guidelines on anti. For personal use only. Pharm. Stein CM. M. Berg CJ. M. Cullen KA.S. and Chlamydia trachoma.. Svanström H. National Hospital Discharge Survey: 2005 yah L.S.​111:​51-6. Maccato M. Saade plasma urealyticum to reduce post-cesar- erperal infection. M.D. Duncan BW.​199(3):​303. Romero R. venting infection after cesarean section. Hauth JC. tin no. 2008 Physician Quality Report. Wachter RM. mycoplasmas. Hauth JC. Gynecol 2003. Mugford M.. Cliver SP. Robin Steele. Tamhane.N.. protocol.​153:​1028-37. B.A. A. Mangram AJ. for oversight and outcome tween clinical research centers and the data coordinating cen. concepts in antibiotic prophylaxis for ce.​299:​1003-6. Bacterial vaginosis as a 2008. Obstet Gynecol 2009.​ Assess (Summ) 2001.​ diovascular causes. M. Emerg Infect Dis 2003. P. crobial invasion of the amniotic cavity Whitehead SJ. Emerging ean delivery endometritis. Meara E. Reddy. tis. their services.​9:​196. e1-6. W.. Goldenberg RL. Obstet Gynecol 2005. annual summary with detailed diagnosis biotics for cesarean delivery: a metaanaly. Yoon BH. Timing of perioperative anti. risk factor for post-cesarean endometri. 2016 1241 The New England Journal of Medicine Downloaded from nejm. M.. M. Huang E. Pregnancy-related mortali. and maternal compartments..D. Andrews WW. RL. wound infection.​101:​1183-9. Reducing the incidence of infection after Prophylactic use of antibiotics for nonla.​ 10:​ ed with a robust host response in fetal.. et al. Tita AT. for data management and statistical tance with protocol development and overall coordination be. genital 26. 25. Ob. Andrews WW. Sue Cliver. Cliver SP. Varner. Shojania KG. boring patients undergoing cesarean de.. the members of the data and safety monitoring board ter. Obstet Cochrane Database Syst Rev 2014. Chalmers I.​43:​1-668. tion: significance for maternal peripartal Use of azithromycin and death from car- tient safety practices. ultative and anaerobic bacteria. Arbogast -Instruments/PQRS/downloads/2008PQRI Obstet Gynecol 1995. Eschenbach DA. Katila ML. metritis with colonization of the chorio.H. Roberts S. McDonald Ollikainen M. 3.S. M. N Engl J Med 2013.. 6.. Clinical risk factors for pu. Krohn MA. N Engl J Med 10. Chauhan S. Gibbs RS. M.. phylaxis. macy services at the University of Alabama at​... 7. Chang J. Copyright © 2016 Massachusetts Medical Society. Shah SR. Obstet Gynecol 1980. Stamm AM. and Richard rics and gynecology residents and fellows at the trial sites for Mailhot. Andrews WW. for coordinating investigational-drug phar.P. fect Control 1999. Rahman M. The microbiology of post-cesarean endometritis: a randomized controlled cists. 14:​151-6.. Guadagnoli E.cms. Cosgrove 656-61.. 20.pdf). KM. Watkins H. the full text of this article at NEJM. Timing of prophylactic antibiotic ad. R. 18. prophylaxis versus no prophylaxis for pre.. Obstet Gynecol LS. Smaill FM.​ trial. with Ureaplasma urealyticum is associat- ty in the United States. Christopher Parks. Obstet Gynecol 55:​5 Suppl:​178S-184S. Ph. No other uses without permission.​96:​992-5. Chelmow D.​85:​509-14. 2. Pinell zole for the prevention of postcesarean can Society of Health-System Pharma. January 2008 discharge. 368:​1704-12. Grivell RM. 22.. Hillier SL. and Sharon on January 8. Platt R. Lip- 9. 2012. Meyer NL. 12. Owen J. Grimes A. Hviid A. 101: ultrasonography in pregnan- sources. 27. Stamm AM.D. AM. 1991-1997. on incidence of postcesarean surgical tion (CDC) Hospital Infection Control 17. Azithromycin and the risk MeasureSpecifications123107.​113:​451-61.D. Rouse DJ. Am J Obstet Gynecol 2001. Goldenberg scomb GH.​ GR. Pasternak B. for assis. Ph. sociation of post-cesarean delivery endo. amniotic. otic prophylaxis with coverage for Urea- 5.​199(3):​301. analysis. gravida: a randomized clinical trial.D. 13. Obstet Gynecol 1990. Joseph Biggio. John Hauth. Quality-Initiatives-Patient-Assessment amnion by Ureaplasma urealyticum. and Gynecologists.. South Med J ment (https:​//www.H.  September 29. of cardiovascular death. Markowitz AJ. Am J Obstet Gynecol Practices Advisory Committee. Hall K. M.. 23.. Jarvis WR. Savage K.D. for the development and maintenance of their assistance with recruitment. B. Cefazolin versus cefazolin Services. Obstet Gynecol 1993. NHSNManPSPCurr​. and Gynecologists. — for Quinn. Mi. Tompkins incidence of postcesarean endometritis in labor and delivery. clinical trial of extended spectrum antibi- 2007. Ghulmi. Rosene K.​56:​ wound morbidity. Park JS. Mark Landon. Copyright © 2016 Massachusetts Medical Society. 11. Am J Obstet Gynecol 2008.​192:​1864-8. tin. Victoria Jauk. DeFrances CJ. n engl j med 375. American College of Obstetricians caesarean section: implications of pro. Making Amniotic fluid tended-spectrum antibiotic prophylaxis Centers for Disease Control and Preven.B. M. Guideline for pre. Ray WA. Pearson ML. Suonio S. et work manual: patient safety component Health and economic impact of surgical al. Blackwell S. urealyticum coloniza. Eschenbach DA. Andrews WW. Cassell GH. Thigpen BD. M. 2003.N.S. review. 14. livery with intact membranes: a meta. Atlanta:​Centers for Diseases site infections diagnosed after hospital ministration in the uninfected laboring Control and Prevention. Murray KT. Am J Perinatol 1997. Randomized and procedure data.​113:​675-82. M.D. and the obstet- Yukiko N.D. References 1.D. 1839-88. Ph. for their assistance as NICHD program officers. Am J Health Syst Pharm 1999. Spong CY. Hosier KV. et al. R. Kozak LJ. and Ashutosh We thank Rachel LeDuke. Michael W.​27:​97-132. Polymicrobial with extended-spectrum antibiotic pro- 2011. Saarikoski S. for assistance with conception and — Janet Wittes. oversight of the trial. BMJ 1989.​366:​1881-90. Azithromycin Prophylaxis for Cesarean Delivery Disclosure forms provided by the authors are available with the trial database and data-entry system. Vital Health Stat 13 sis. Uma M.​165:​1-209.N.​98:​745-50.​101:​289-96. M. Ruehli MS. health care safer: a critical analysis of pa. Centers for Medicare and Medicaid 19. Decreasing tin no. 15. Sanchez-Ramos L. Am J (http://www​. Rouse.. plus metronidazole for antibiotic prophy- ing Initiative (PQRI) specifications docu.D. 29.e1-3.13 nejm. Kenny GE. American College of Obstetricians sarean delivery: a systematic review. laxis at cesarean section. 30. 1999: tis: treatment with piperacillin or cefoxi. ACOG practice bulle.P. Keski-Nisula L.Sc. As. Am J In. Dwight J. Perencevich EN.D.pdf). Antibiotic 21. J Infect Dis 1986.D. and Caroline Signore. Andrews WW.dhcs​.​117:​1472-83. Sands KE.