- Email: [email protected]
Intrapartum treatment of acute chorioamnionitis: Impact on neonatal sepsis
Intrapartum treatment of acute chorioamnionitis: Impact on neonatal sepsis Larry C. Gilstrap III, MD, Kenneth J. Leveno, MD, Susan M. Cox, MD, Jody S. Burris, RN, BSN, Melinda Mashburn, RN, BSN, and Charles R. Rosenfeld, MD Dallas, Texas In a study of 312 women with acute chorioamnionitis, 152 women received antibiotics before delivery, 90 received antibiotics after cord clamping, and 70 did not receive antibiotics. Antibiotics were administered during labor rather than after cord clamping if delivery was not imminent. Although endometritis developed more frequently in the patients receiving antibiotics after cord clamping, the difference was not statistically significant (5.6% versus 3.9%, difference not significant). There were two cases of verified sepsis in the group of infants (35 weeks) born to mothers receiving intrapartum antibiotics and there were eight cases in the no antibiotics group (p = 0.06). More importantly, in neonates ;;.35 weeks' gestational age, there was a significant difference in the frequency of positive blood cultures for group 8 streptococci (0/133 versus 8/140, p < 0.05). We conclude that administration of antibiotics to the mother during labor may result in a decreased incidence of neonatal sepsis. (AM J OesTET GYNECOL 1988;159:579-83.)
Key words: Chorioamnionitis, intrapartum antibiotics, group B streptococci
Acute chorioamnionitis is a relatively common complication of pregnancy, occurring in 0.5% to 1.3% of all pregnancies ... 2 Moreover, it is well established that intraamniotic infection may result in significant maternal and neonatal morbidity., .• Although there is general agreement that therapy basically should consist of antimicrobial therapy and delivery, there is no unanimity of opinion regarding the initiation and timing of antibiotic therapy. The major reason for this is the uncertainty of the implications associated with antibiotic therapy on the unborn fetus and neonate. In fact, in a previous study from our institution, 5 it was reported that antimicrobials for prophylaxis administered after cesarean section but before cord clamping were associated with an increase in the number of infants evaluated for sepsis over those whose mothers received antibiotics after cord clamping, although there was no difference in the frequency of neonatal sepsis. There is some concern that intrapartum administration of antibiotics to the mother may interfere with establishing the diagnosis and with the duration of treatment of infection in the neonate. Thus it is not uncommon for the obstetrician to withhold antibiotic therapy in
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, and the Division of Neonatal-Perinatal Medicine, Department of Pediatrics, University of Texas Southwestern Medical School Presented at the Eighth Annual Meeting of the Society of Perinatal Obstetricians, Las Vegas, Nevada, February 4-6, 1988. Reprint requests: Larry C. Gilstrap 111, MD, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, TX 75235-9032.
women with acute chorioamnionitis until after the infant is delivered and the cord is clamped, especially if delivery is expected to occur in a relatively short period of time. Unfortunately, there is a paucity of data regarding the impact of the intrapartum administration of antibiotics to the mother on the fetus and newborn infant. To date, there has been only one study that addressed this particular aspect of management of chorioamnionitis.6 In this study the~e was a significantly lower incidence of neonatal sepsis in infants born to mothers who received intrapartum antibiotic therapy. The purpose of the present study therefore was to evaluate the neonatal and maternal effects, if any, of intrapartum antibiotic therapy for acute chorioamnionitis.
Material and methods The obstetric records of all women with intrapartum fever and a diagnosis of acute chorioamnionitis from January 1986 through june 1987 were reviewed. The diagnosis of intraamniotic infection was based primarily on the presence of maternal fever during labor. Secondary criteria included fetal tachycardia, maternal tachycardia, uterine tenderness, and foul-smelling amniotic fluid. Only women with an oral temperature of ;;.38° C and no other obvious cause of infection were included for analysis. The patients were divided into three groups regarding antibiotic therapy: (1) those who received antibiotics before delivery, (2) those who received antibiotics immediately after cord clamping, and (3) those who did not receive antibiotics either before delivery or in the period immediately postpartum.
579
580
Gilstrap et al.
September 1988 Am .J Obstet Gynecol
Table I. Associated factors and timing of antibiotic administration Antibiotic administration Before delivery (n = 152) Factor
Labor Spontaneous Induced Augmented Amniotomy Not in labor Membrane rupture Spontaneous Artificial Before admission Electronic monitoring Internal External None
no.
I
After cord clamping (n = 90)
%
no.
45 26 79
30 17 52 0.6 0.6
44
90 62 76 132
I I
11
10
I
None (n = 70)
I
%
no.
II
35
49 12 39
34 12 24
49 17 34
59 41 50
46 44 34
51 49 38
35 35 28
50 50 40
87 7 7
74 8 8
82 9 9
55 7 9
79 10 13
These latter two groups were combined into a no antibiotics group to evaluate the effects of maternal antibiotic therapy on neonatal sepsis and outcome. The neonatal records of all infants born to these mothers were also reviewed with regard to neonatal sepsis, necrotizing enterocolitis, intracranial hemorrhage, respiratory distress, and neonatal death. The diagnosis of definite neonatal sepsis required a positive blood culture, whereas "suspected sepsis" required the presence of one or more abnormal peripheral neutrophil counts in the first 12 to 24 hours, antibiotic therapy for ""3 days, and the presence of negative blood cultures. The diagnosis of necrotizing enterocolitis was based on abdominal distention, bloody stools, and a characteristic abdominal radiographic pattern. Intracranial hemorrhage was diagnosed by cranial ultrasound and respiratory distress was based on the need for mechanical ventilation. For purposes of analysis the patients were divided into two groups according to gestational age (""35 weeks' gestation and <35 weeks' gestation). Gestational age was based primarily on obstetric assessment. All neonates in the group ,35 weeks' gestation weighed >2500 gm. Statistical analysis of data was by Student's t test, X2 contingency tables, or Fisher's test of exact probability. The computer statistical program of the Birth Defects and Genetic Diseases Branch of the Centers for Disease Control, Atlanta, GA 30333, was used.
Results There were 21,270 deliveries from January 1, 1986, to June 30, 1987; 312 (1.5%) of the women were diagnosed clinically as having acute chorioamnionitis. Forty-seven percent were Latin-American and 17%
%
were white. Forty percent of the patients were teenagers (aged 13 to 19 years) and 52% were between the ages of 20 and 29 years. The majority (74%) of the patients were nulliparous and 84% had attended a prenatal clinic. All patients had fever during labor; 66 (21%) had a fever at the time of admission. Of the total number of women evaluated, 284 had electronic fetal heart monitoring and 153 (54%) had a fetus who manifested tachycardia(> 160 beats/min) during labor. There were no significant differences with regard to demographic characteristics of the patients in the three antibiotic groups. Associated factors and timing of antibiotic administration are summarized in Table I. There were no significant differences in the type of labor, membrane status, or use of electronic fetal montioring among groups. Almost half (44%) of all patients had rupture of the fetal membranes before admission. The methods of delivery are summarized in Table II. As might be expected, when compared with the number of women who received antibiotics before delivery (65%) or immediately after cord clamping (57%), significantly more women who did not receive antibiotics in the period immediately post partum were delivered vaginally (96%). Of the 242 women who received antibiotics either before delivery or immediately postpartum, 114 (4 7%) received triple antibiotic therapy with either ampicillin or penicillin in combination with clindamycin and gentamicin, 54 (22%) received dual therapy with either ampicillin or penicillin in combination with gentamicin, and 49 (20%) received single-agent therapy with cefoxitin. The remaining 25 patients received a variety of other antibiotics. The interval from onset of fever to delivery was sig-
Intrapartum treatment of acute chorioamnionitis
Volume 159 Number 3
581
Table II. Method of delivery and timing of antibiotic administration Antibiotic administration Before delivery (n = 152) Method of delivery
Vaginal Spontaneous Low forceps Mid forceps Total Cesarean section Dystocia Fetal distress Breech Repeat Other Total
no.
73 18 8 99
I
After cord clamping (n = 90)
%
no.
48 12 5 65
37 12
2 l l
53
35
I
None (n = 70)
I
%
no.
28 21
31 23
55 10
77 14
51
57
67
96*
_1 24 5 3 3 4 39
2
%
3
2
2 l
4*
43
*p < 0.05. nificantly longer in those women who received antibiotics before delivery than in either of the other two groups (Table III). There were no significant differences in the interval of ruptured membranes to delivery or of oxytocin use to delivery according to the timing of antibiotic administration. Maternal complications are summarized in Table IV. The incidence of endometritis after vaginal delivery in women who did not receive antibiotics in the period immediately post partum was 6%. Although this is higher than that in our general population (approximately 2%) and in the other two groups, it was not significantly greater than that in the other two groups. Of the 95 patients who were delivered by cesarean section, 92 (97%) received antibiotics either before delivery or immediately after cord clamping. Of these 92 women, eight (9%) developed postpartum endometritis. The outcomes of neonates who were ~35 weeks' gestational age are summarized in Table V. There were 10 neonates who had "definite" sepsis (defined as a positive blood culture), two in the antibiotics group and eight in the no antibiotics group (p "" 0.06). Of the 10 patients with positive blood cultures, eight had group B streptococci, and five of these also had positive urine latex test results for group B streptococci. Urine specimens were not sent for three of the eight patients. There was a significant difference in the frequency of positive blood cultures for group B streptococci between the two groups (0/133 versus 8/140, p < 0.05), which may reflect the type of antibiotics used to treat the mother. The two remaining positive blood cultures were for Streptococcus pneumoniae and Bacteroides fragilis. There were eight additional neonates with positive urine latex test results for group B streptococci who had negative blood cultures, five in the antibiotics
Table III. Interval (hours, mean ± SD) from onset of fever to delivery, rupture of membranes to delivery, and oxytocin use to delivery Antibiotic administration lnterval
Fever to delivery Rupture of membranes to delivery Oxytocin use to delivery
None
Before delivery
5.8 (n 28.0 (n
± = ± =
4.2*t 120) 89.8 153)
10.7 ± 7.6 (n = 105)
2.5 (n 20.6 (n
± = ± =
2.0* 68) 49 88)
12.2 ± 7.3 (n = 47)
3.0 (n 12.1 (n
± 3.8t =58) ± 8.7 = 70)
10.1 ± 7.2 (n = 36)
p < 0.0001 compared with after cord clamping. tt = 6.23; p < 0.001 compared with before delivery.
*t = 10.2;
group and three in the no antibiotics group. Eightyeight percent of neonates in the antibiotics group and 65% of the neonates in the no antibiotics group received antibiotics sometime during the neonatal period (p < 0.05). There were four neonates in the antibiotics group and one in the no antibiotics group with a diagnosis of pneumonia based on x-ray findings. However, only two neonates (one in each group) had a positive culture. There was no significant difference in the frequency of neonates diagnosed as having "suspected" sepsis between the two groups (64/133 versus 50/140). There were four perinatal deaths in neonates who were ~35 weeks' gestational age; all four of these were stillbirths in women who had received antibiotics before delivery. All four fetuses were dead at the time of admission of the mother. The mean birth weights were similar in both groups (3313 and 3473 gm). The neonatal outcomes of neonates ~34 weeks' ges-
582 Gilstrap et al.
September 1988 Am J Obstet Gynecol
Table IV. Maternal complications according to timing of antibiotic administration Antibiotic administration Before delivery Complication
l
no.
Endometritis Vaginal delivery Cesarean section Total Average hospital charges
After cord clamping
2/99 4/53 6/152
%
no.
2.0 7.5 3.9
l/51 4/39 5/90
$3287*
l
None
%
no.
2.0 10.2 5.6
4/67 l/3 5170
$3529
l
% 5.8 33 7.1
$2595*
*Bill not available for two patients.
Table V. Neonatal outcome in antibiotics group versus no antibiotics group 2::35 weeks' gestation Antibiotcs group* (n = 133)
Sepsis Positive blood culture for GBS Positive blood culture Positive urine latex test result Suspected Perinatal death Morbidity Apgar score :56 at 5 min Intraventricular hemorrhage Necrotizing enterocolitis Ventilation required Seizures Average hospital charges
Table VI. Neonatal outcome in antibiotics group versus no antibiotics group :534 weeks' gestation
No antibiotics group
Antibiotics group*
(n = 140)
0
8t
2
8:j:
5
8
64 6
50 0 0
0
0
0
0
2
2
0 $1499
2 $1148
CBS, Group B streptococci. *Intrapartum antibiotics only. No antibiotics group includes women who did not receive antibiotics at all or who received antibiotics after cord clamping. tp = 0.004. :j:p = 0.06.
tation are summarized in Table VI. There were only 39 infants in this category and none of these had sepsis verified, although 23 were diagnosed by the neonatologist as having suspected sepsis. There were nine perinatal deaths, four in the antibiotics group and five in the no antibiotics group (difference not significant). None of these deaths were felt to be secondary to neonatal sepsis. The mean birth weights were similar in the antibiotics and no antibiotics groups (1586 and 1523 gm).
(n = 19)
Sepsis Positive blood culture Positive urine latex test result Suspected Perinatal death Perinatal morbidity Apgar score :56 at 5 min Intraventricular hemorrhage Necrotizing enterocolitis Ventilation required Seizures Average hospital charges
No antibiotics group (n = 20)
0
0
2
2
14 4
13 5
4
10
2
5
2 6 I
$17,121
7 0 $16,022
*Intrapartum antibiotics.
Comment Recently, Sperling et al. 6 reported a significantly greater incidence of neonatal sepsis in neonates whose mothers received antibiotics after cord clamping than in those whose mothers received antibiotics before delivery, approximately 17% versus 3%. These authors suggested that intrapartum treatment of mothers may, in turn, provide "early in utero therapy for the infected fetus." In the present study the difference between the antibiotics groups and the no antibiotics group with regard to positive blood cultures was of borderline significance (p = 0.06). However, positive blood cultures for group B strep.tococci were not found in any of the neonates whose mothers received antibiotics before delivery but were found in six neonates born to mothers who did not receive antibiotics during labor (p < 0.05).
Volume 159 Number 3
This would lend support to this premise of Sperling et al. 6 of in utero therapy being of benefit to the fetus. Of the two positive blood cultures in the antibiotics group, one was positive for B. fragilis and occurred in a neonate whose mother had received ampicillin and gentamicin. The other culture was positive for S. pneumoniae in a neonate whose mother had received cefoxitin approximately 20 to 30 minutes before delivery. None of the neonates in our study had a positive cerebrospinal fluid culture although nine neonates had a diagnosis of pneumonia based primarily on x-ray findings. Only two of these had a positive culture (S. pneumoniae and group B streptococci), one in each group. An additional neonate with the x-ray diagnosis of pneumonia had a positive urine latex test result for group B streptococci. However, accurately identifying neonatal sepsis, especially in the absence of positive bacterial cultures, remains a m,Yor diagnostic dilemma. Respiratory distress syndrome often may be confused with pneumonia and vice versa. Because of this, we chose to use a positive blood culture as a marker for "verified sepsis," realizing that we may have missed some cases of actual sepsis with negative cultures. This may explain why we had a lower incidence of proved sepsis than reported by Sperling et a!. Of paramount importance is the fact that the eight neonates with positive blood cultures for group B streptococci were born to mothers who were not considered to be "sick", i.e., they had shorter intervals from fever to the time of delivery and from ruptured membranes to delivery, and the majority were delivered vaginally soon after diagnosis. Like Sperling et al., 6 we also found no difference in maternal outcome according to timing of antibiotic administration. This may represent a type II error. More likely, however, any apparent difference is probably "masked" by the fact that almost all patients who were delivered by cesarean section also received antibiotics and almost all of these received broad spectrum antibiotics that provided coverage against the majority of aerobic and anaerobic organisms encountered in postcesarean section infections. 7 The 9% incidence of endometritis after cesarean section in this study is significantly lower than that previously reported from our institution for women undergoing cesarean section, both with and without antibiotic prophylaxis." Obviously, prospective randomized studies that address broad spectrum antibiotic therapy for acute chorioamnionitis are needed before the impact on maternal morbidity can be fully addressed.
Intrapartum treatment of acute chorioamnionitis
583
The interval from onset of fever to delivery was significantly greater in women who received intrapartum antibiotics, as expected, and is similar to that found for the interval from diagnosis to delivery in the study by Sperling et al. 6 The mean hospital charges were significantly higher in women who had received antibiotics than in those who did not; this is probably related to the fact that more women in the antibiotic group were delivered by cesarean section. This probably also accounts for the differences in hospital cost in the neonates. In summary, data from the present study indicate that intrapartum antibiotic therapy for the mother with acute chorioamnionitis may result in decreased neonatal sepsis, at least as defined by a positive blood culture. Although a randomized, prospective study is needed to truly answer this question, there are several major drawbacks with performing such a study, paramount of which is withholding antibiotics in a mother who has significant infection and is remote from delivery. Acute chorioamnionitis may still result in maternal death. 9 Randomized, prospective studies comparing the combination of penicillin-ampicillin, plus an aminoglycoside, with regimens providing broad coverage against anaerobes, as well as aerobes, are needed, especially with regard to maternal morbidity. REFERENCES 1. Yoder PR, Gibbs RS, Blanco JD, et al. A prospective, controlled study of maternal and perinatal outcome after intraamniotic infection at term. AM J 0BSTET GYNECOL 1983; 145:695. 2. Hauth JC, Gilstrap LC Ill, Hankins GDV, Connor KD. Term maternal and neonatal complications of acute chorioamnionitis. Obstet Gynecol1983;66:59. 3. Koh KS, Chan FH, Monfared AH, et al. The changing perinatal and maternal outcome in chorioamnionitis. Obstet Gynecol 1979;53:730. 4. Gibbs RS, Castillo MS, Rodgers PJ. Management of acute chorioamnionitis. AM J 0BSTET GYNECOL 1980;136:709. 5. Cunningham FG, Leveno Iq, DePalma RT, et al. Perioperative antimicrobials for cesarean delivery: before or after cord clamping? Obstet Gynecol 1983;62: 151. 6. Sperling RS, Ramamurthy RS, Gibbs RS. A comparison of intrapartum of intrapartum versus immediate postpartum treatment of intraamniotic infection. Obstet Gynecol 198 7; 70:861. 7. Gilstrap LC, Cunningham FG. The bacterial pathogenesis of infection following cesarean section. Obstet Gynecol 1979;53:545. 8. Cunningham FG, Hauth JC, Strong JD, et al. Infectious morbidity following cesarean section. Comparison of two treatment regimens. Obstet Gynecol 1978;52:656. 9. Gibbs CE, Locke WE. Maternal deaths in Texas, 1969 to 1973: a report of 501 consecutive maternal deaths from the Texas Medical Association's Committee on Maternal Heatlh. AM j 0BSTET GYNECOL 1976; 126:687.
Key words: Chorioamnionitis, intrapartum antibiotics, group B streptococci
Acute chorioamnionitis is a relatively common complication of pregnancy, occurring in 0.5% to 1.3% of all pregnancies ... 2 Moreover, it is well established that intraamniotic infection may result in significant maternal and neonatal morbidity., .• Although there is general agreement that therapy basically should consist of antimicrobial therapy and delivery, there is no unanimity of opinion regarding the initiation and timing of antibiotic therapy. The major reason for this is the uncertainty of the implications associated with antibiotic therapy on the unborn fetus and neonate. In fact, in a previous study from our institution, 5 it was reported that antimicrobials for prophylaxis administered after cesarean section but before cord clamping were associated with an increase in the number of infants evaluated for sepsis over those whose mothers received antibiotics after cord clamping, although there was no difference in the frequency of neonatal sepsis. There is some concern that intrapartum administration of antibiotics to the mother may interfere with establishing the diagnosis and with the duration of treatment of infection in the neonate. Thus it is not uncommon for the obstetrician to withhold antibiotic therapy in
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, and the Division of Neonatal-Perinatal Medicine, Department of Pediatrics, University of Texas Southwestern Medical School Presented at the Eighth Annual Meeting of the Society of Perinatal Obstetricians, Las Vegas, Nevada, February 4-6, 1988. Reprint requests: Larry C. Gilstrap 111, MD, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, TX 75235-9032.
women with acute chorioamnionitis until after the infant is delivered and the cord is clamped, especially if delivery is expected to occur in a relatively short period of time. Unfortunately, there is a paucity of data regarding the impact of the intrapartum administration of antibiotics to the mother on the fetus and newborn infant. To date, there has been only one study that addressed this particular aspect of management of chorioamnionitis.6 In this study the~e was a significantly lower incidence of neonatal sepsis in infants born to mothers who received intrapartum antibiotic therapy. The purpose of the present study therefore was to evaluate the neonatal and maternal effects, if any, of intrapartum antibiotic therapy for acute chorioamnionitis.
Material and methods The obstetric records of all women with intrapartum fever and a diagnosis of acute chorioamnionitis from January 1986 through june 1987 were reviewed. The diagnosis of intraamniotic infection was based primarily on the presence of maternal fever during labor. Secondary criteria included fetal tachycardia, maternal tachycardia, uterine tenderness, and foul-smelling amniotic fluid. Only women with an oral temperature of ;;.38° C and no other obvious cause of infection were included for analysis. The patients were divided into three groups regarding antibiotic therapy: (1) those who received antibiotics before delivery, (2) those who received antibiotics immediately after cord clamping, and (3) those who did not receive antibiotics either before delivery or in the period immediately postpartum.
579
580
Gilstrap et al.
September 1988 Am .J Obstet Gynecol
Table I. Associated factors and timing of antibiotic administration Antibiotic administration Before delivery (n = 152) Factor
Labor Spontaneous Induced Augmented Amniotomy Not in labor Membrane rupture Spontaneous Artificial Before admission Electronic monitoring Internal External None
no.
I
After cord clamping (n = 90)
%
no.
45 26 79
30 17 52 0.6 0.6
44
90 62 76 132
I I
11
10
I
None (n = 70)
I
%
no.
II
35
49 12 39
34 12 24
49 17 34
59 41 50
46 44 34
51 49 38
35 35 28
50 50 40
87 7 7
74 8 8
82 9 9
55 7 9
79 10 13
These latter two groups were combined into a no antibiotics group to evaluate the effects of maternal antibiotic therapy on neonatal sepsis and outcome. The neonatal records of all infants born to these mothers were also reviewed with regard to neonatal sepsis, necrotizing enterocolitis, intracranial hemorrhage, respiratory distress, and neonatal death. The diagnosis of definite neonatal sepsis required a positive blood culture, whereas "suspected sepsis" required the presence of one or more abnormal peripheral neutrophil counts in the first 12 to 24 hours, antibiotic therapy for ""3 days, and the presence of negative blood cultures. The diagnosis of necrotizing enterocolitis was based on abdominal distention, bloody stools, and a characteristic abdominal radiographic pattern. Intracranial hemorrhage was diagnosed by cranial ultrasound and respiratory distress was based on the need for mechanical ventilation. For purposes of analysis the patients were divided into two groups according to gestational age (""35 weeks' gestation and <35 weeks' gestation). Gestational age was based primarily on obstetric assessment. All neonates in the group ,35 weeks' gestation weighed >2500 gm. Statistical analysis of data was by Student's t test, X2 contingency tables, or Fisher's test of exact probability. The computer statistical program of the Birth Defects and Genetic Diseases Branch of the Centers for Disease Control, Atlanta, GA 30333, was used.
Results There were 21,270 deliveries from January 1, 1986, to June 30, 1987; 312 (1.5%) of the women were diagnosed clinically as having acute chorioamnionitis. Forty-seven percent were Latin-American and 17%
%
were white. Forty percent of the patients were teenagers (aged 13 to 19 years) and 52% were between the ages of 20 and 29 years. The majority (74%) of the patients were nulliparous and 84% had attended a prenatal clinic. All patients had fever during labor; 66 (21%) had a fever at the time of admission. Of the total number of women evaluated, 284 had electronic fetal heart monitoring and 153 (54%) had a fetus who manifested tachycardia(> 160 beats/min) during labor. There were no significant differences with regard to demographic characteristics of the patients in the three antibiotic groups. Associated factors and timing of antibiotic administration are summarized in Table I. There were no significant differences in the type of labor, membrane status, or use of electronic fetal montioring among groups. Almost half (44%) of all patients had rupture of the fetal membranes before admission. The methods of delivery are summarized in Table II. As might be expected, when compared with the number of women who received antibiotics before delivery (65%) or immediately after cord clamping (57%), significantly more women who did not receive antibiotics in the period immediately post partum were delivered vaginally (96%). Of the 242 women who received antibiotics either before delivery or immediately postpartum, 114 (4 7%) received triple antibiotic therapy with either ampicillin or penicillin in combination with clindamycin and gentamicin, 54 (22%) received dual therapy with either ampicillin or penicillin in combination with gentamicin, and 49 (20%) received single-agent therapy with cefoxitin. The remaining 25 patients received a variety of other antibiotics. The interval from onset of fever to delivery was sig-
Intrapartum treatment of acute chorioamnionitis
Volume 159 Number 3
581
Table II. Method of delivery and timing of antibiotic administration Antibiotic administration Before delivery (n = 152) Method of delivery
Vaginal Spontaneous Low forceps Mid forceps Total Cesarean section Dystocia Fetal distress Breech Repeat Other Total
no.
73 18 8 99
I
After cord clamping (n = 90)
%
no.
48 12 5 65
37 12
2 l l
53
35
I
None (n = 70)
I
%
no.
28 21
31 23
55 10
77 14
51
57
67
96*
_1 24 5 3 3 4 39
2
%
3
2
2 l
4*
43
*p < 0.05. nificantly longer in those women who received antibiotics before delivery than in either of the other two groups (Table III). There were no significant differences in the interval of ruptured membranes to delivery or of oxytocin use to delivery according to the timing of antibiotic administration. Maternal complications are summarized in Table IV. The incidence of endometritis after vaginal delivery in women who did not receive antibiotics in the period immediately post partum was 6%. Although this is higher than that in our general population (approximately 2%) and in the other two groups, it was not significantly greater than that in the other two groups. Of the 95 patients who were delivered by cesarean section, 92 (97%) received antibiotics either before delivery or immediately after cord clamping. Of these 92 women, eight (9%) developed postpartum endometritis. The outcomes of neonates who were ~35 weeks' gestational age are summarized in Table V. There were 10 neonates who had "definite" sepsis (defined as a positive blood culture), two in the antibiotics group and eight in the no antibiotics group (p "" 0.06). Of the 10 patients with positive blood cultures, eight had group B streptococci, and five of these also had positive urine latex test results for group B streptococci. Urine specimens were not sent for three of the eight patients. There was a significant difference in the frequency of positive blood cultures for group B streptococci between the two groups (0/133 versus 8/140, p < 0.05), which may reflect the type of antibiotics used to treat the mother. The two remaining positive blood cultures were for Streptococcus pneumoniae and Bacteroides fragilis. There were eight additional neonates with positive urine latex test results for group B streptococci who had negative blood cultures, five in the antibiotics
Table III. Interval (hours, mean ± SD) from onset of fever to delivery, rupture of membranes to delivery, and oxytocin use to delivery Antibiotic administration lnterval
Fever to delivery Rupture of membranes to delivery Oxytocin use to delivery
None
Before delivery
5.8 (n 28.0 (n
± = ± =
4.2*t 120) 89.8 153)
10.7 ± 7.6 (n = 105)
2.5 (n 20.6 (n
± = ± =
2.0* 68) 49 88)
12.2 ± 7.3 (n = 47)
3.0 (n 12.1 (n
± 3.8t =58) ± 8.7 = 70)
10.1 ± 7.2 (n = 36)
p < 0.0001 compared with after cord clamping. tt = 6.23; p < 0.001 compared with before delivery.
*t = 10.2;
group and three in the no antibiotics group. Eightyeight percent of neonates in the antibiotics group and 65% of the neonates in the no antibiotics group received antibiotics sometime during the neonatal period (p < 0.05). There were four neonates in the antibiotics group and one in the no antibiotics group with a diagnosis of pneumonia based on x-ray findings. However, only two neonates (one in each group) had a positive culture. There was no significant difference in the frequency of neonates diagnosed as having "suspected" sepsis between the two groups (64/133 versus 50/140). There were four perinatal deaths in neonates who were ~35 weeks' gestational age; all four of these were stillbirths in women who had received antibiotics before delivery. All four fetuses were dead at the time of admission of the mother. The mean birth weights were similar in both groups (3313 and 3473 gm). The neonatal outcomes of neonates ~34 weeks' ges-
582 Gilstrap et al.
September 1988 Am J Obstet Gynecol
Table IV. Maternal complications according to timing of antibiotic administration Antibiotic administration Before delivery Complication
l
no.
Endometritis Vaginal delivery Cesarean section Total Average hospital charges
After cord clamping
2/99 4/53 6/152
%
no.
2.0 7.5 3.9
l/51 4/39 5/90
$3287*
l
None
%
no.
2.0 10.2 5.6
4/67 l/3 5170
$3529
l
% 5.8 33 7.1
$2595*
*Bill not available for two patients.
Table V. Neonatal outcome in antibiotics group versus no antibiotics group 2::35 weeks' gestation Antibiotcs group* (n = 133)
Sepsis Positive blood culture for GBS Positive blood culture Positive urine latex test result Suspected Perinatal death Morbidity Apgar score :56 at 5 min Intraventricular hemorrhage Necrotizing enterocolitis Ventilation required Seizures Average hospital charges
Table VI. Neonatal outcome in antibiotics group versus no antibiotics group :534 weeks' gestation
No antibiotics group
Antibiotics group*
(n = 140)
0
8t
2
8:j:
5
8
64 6
50 0 0
0
0
0
0
2
2
0 $1499
2 $1148
CBS, Group B streptococci. *Intrapartum antibiotics only. No antibiotics group includes women who did not receive antibiotics at all or who received antibiotics after cord clamping. tp = 0.004. :j:p = 0.06.
tation are summarized in Table VI. There were only 39 infants in this category and none of these had sepsis verified, although 23 were diagnosed by the neonatologist as having suspected sepsis. There were nine perinatal deaths, four in the antibiotics group and five in the no antibiotics group (difference not significant). None of these deaths were felt to be secondary to neonatal sepsis. The mean birth weights were similar in the antibiotics and no antibiotics groups (1586 and 1523 gm).
(n = 19)
Sepsis Positive blood culture Positive urine latex test result Suspected Perinatal death Perinatal morbidity Apgar score :56 at 5 min Intraventricular hemorrhage Necrotizing enterocolitis Ventilation required Seizures Average hospital charges
No antibiotics group (n = 20)
0
0
2
2
14 4
13 5
4
10
2
5
2 6 I
$17,121
7 0 $16,022
*Intrapartum antibiotics.
Comment Recently, Sperling et al. 6 reported a significantly greater incidence of neonatal sepsis in neonates whose mothers received antibiotics after cord clamping than in those whose mothers received antibiotics before delivery, approximately 17% versus 3%. These authors suggested that intrapartum treatment of mothers may, in turn, provide "early in utero therapy for the infected fetus." In the present study the difference between the antibiotics groups and the no antibiotics group with regard to positive blood cultures was of borderline significance (p = 0.06). However, positive blood cultures for group B strep.tococci were not found in any of the neonates whose mothers received antibiotics before delivery but were found in six neonates born to mothers who did not receive antibiotics during labor (p < 0.05).
Volume 159 Number 3
This would lend support to this premise of Sperling et al. 6 of in utero therapy being of benefit to the fetus. Of the two positive blood cultures in the antibiotics group, one was positive for B. fragilis and occurred in a neonate whose mother had received ampicillin and gentamicin. The other culture was positive for S. pneumoniae in a neonate whose mother had received cefoxitin approximately 20 to 30 minutes before delivery. None of the neonates in our study had a positive cerebrospinal fluid culture although nine neonates had a diagnosis of pneumonia based primarily on x-ray findings. Only two of these had a positive culture (S. pneumoniae and group B streptococci), one in each group. An additional neonate with the x-ray diagnosis of pneumonia had a positive urine latex test result for group B streptococci. However, accurately identifying neonatal sepsis, especially in the absence of positive bacterial cultures, remains a m,Yor diagnostic dilemma. Respiratory distress syndrome often may be confused with pneumonia and vice versa. Because of this, we chose to use a positive blood culture as a marker for "verified sepsis," realizing that we may have missed some cases of actual sepsis with negative cultures. This may explain why we had a lower incidence of proved sepsis than reported by Sperling et a!. Of paramount importance is the fact that the eight neonates with positive blood cultures for group B streptococci were born to mothers who were not considered to be "sick", i.e., they had shorter intervals from fever to the time of delivery and from ruptured membranes to delivery, and the majority were delivered vaginally soon after diagnosis. Like Sperling et al., 6 we also found no difference in maternal outcome according to timing of antibiotic administration. This may represent a type II error. More likely, however, any apparent difference is probably "masked" by the fact that almost all patients who were delivered by cesarean section also received antibiotics and almost all of these received broad spectrum antibiotics that provided coverage against the majority of aerobic and anaerobic organisms encountered in postcesarean section infections. 7 The 9% incidence of endometritis after cesarean section in this study is significantly lower than that previously reported from our institution for women undergoing cesarean section, both with and without antibiotic prophylaxis." Obviously, prospective randomized studies that address broad spectrum antibiotic therapy for acute chorioamnionitis are needed before the impact on maternal morbidity can be fully addressed.
Intrapartum treatment of acute chorioamnionitis
583
The interval from onset of fever to delivery was significantly greater in women who received intrapartum antibiotics, as expected, and is similar to that found for the interval from diagnosis to delivery in the study by Sperling et al. 6 The mean hospital charges were significantly higher in women who had received antibiotics than in those who did not; this is probably related to the fact that more women in the antibiotic group were delivered by cesarean section. This probably also accounts for the differences in hospital cost in the neonates. In summary, data from the present study indicate that intrapartum antibiotic therapy for the mother with acute chorioamnionitis may result in decreased neonatal sepsis, at least as defined by a positive blood culture. Although a randomized, prospective study is needed to truly answer this question, there are several major drawbacks with performing such a study, paramount of which is withholding antibiotics in a mother who has significant infection and is remote from delivery. Acute chorioamnionitis may still result in maternal death. 9 Randomized, prospective studies comparing the combination of penicillin-ampicillin, plus an aminoglycoside, with regimens providing broad coverage against anaerobes, as well as aerobes, are needed, especially with regard to maternal morbidity. REFERENCES 1. Yoder PR, Gibbs RS, Blanco JD, et al. A prospective, controlled study of maternal and perinatal outcome after intraamniotic infection at term. AM J 0BSTET GYNECOL 1983; 145:695. 2. Hauth JC, Gilstrap LC Ill, Hankins GDV, Connor KD. Term maternal and neonatal complications of acute chorioamnionitis. Obstet Gynecol1983;66:59. 3. Koh KS, Chan FH, Monfared AH, et al. The changing perinatal and maternal outcome in chorioamnionitis. Obstet Gynecol 1979;53:730. 4. Gibbs RS, Castillo MS, Rodgers PJ. Management of acute chorioamnionitis. AM J 0BSTET GYNECOL 1980;136:709. 5. Cunningham FG, Leveno Iq, DePalma RT, et al. Perioperative antimicrobials for cesarean delivery: before or after cord clamping? Obstet Gynecol 1983;62: 151. 6. Sperling RS, Ramamurthy RS, Gibbs RS. A comparison of intrapartum of intrapartum versus immediate postpartum treatment of intraamniotic infection. Obstet Gynecol 198 7; 70:861. 7. Gilstrap LC, Cunningham FG. The bacterial pathogenesis of infection following cesarean section. Obstet Gynecol 1979;53:545. 8. Cunningham FG, Hauth JC, Strong JD, et al. Infectious morbidity following cesarean section. Comparison of two treatment regimens. Obstet Gynecol 1978;52:656. 9. Gibbs CE, Locke WE. Maternal deaths in Texas, 1969 to 1973: a report of 501 consecutive maternal deaths from the Texas Medical Association's Committee on Maternal Heatlh. AM j 0BSTET GYNECOL 1976; 126:687.