PEDIATRICS Vol. 108 No. 5 November 2001, pp. 1094-1098

Early-Onset Neonatal Sepsis in the Era of Group B Streptococcal Prevention

Robert S. Baltimore, MD^{*, , §}, Sharon M. Huie, BS, MPH^{, §}, James I. Meek, MPH^{, §}, Anne Schuchat, MD, and Katherine L. O'Brien, MD, MPH

From the Departments of ^* Pediatrics and  Epidemiology and Public Health and the ^§ Emerging Infections Program, Yale University School of Medicine, New Haven, Connecticut; and  Centers for Disease Control and Prevention, Atlanta, Georgia.

	ABSTRACT

Top Abstract Methods Results Discussion References

Objective.  To determine whether intrapartum antibiotic prophylaxis for neonatal group B streptococcal (GBS) disease has resulted in an increased rate of non-GBS or antibiotic-resistant early-onset invasive neonatal disease.

Methods.  Maternal and infant chart review of all infants with bacteria other than GBS isolated from blood or spinal fluid in 1996 through 1999 in 19 hospitals (representing 81% of in-state births to state residents) throughout Connecticut. Suspected cases were identified through clinical microbiology laboratory records or through International Classification of Diseases, Ninth Revision codes when microbiology records were incomplete.

Results.  Ninety-four cases of non-GBS early-onset sepsis or meningitis were detected between 1996 and 1999. The rate of GBS-related early-onset infection (days 0-6 of life) dropped from 0.61/1000 to 0.23/1000 births, but the annual rate of non-GBS sepsis remained steady, ranging from 0.65 to 0.68/1000 during the surveillance period. There was an increase in the proportion of Escherichia coli infections that were ampicillin resistant between 1996 and 1998, but the proportion decreased. in 1999

Conclusion.  There was no increase in the incidence of non-GBS early-onset neonatal infections between 1996 and 1999. Fluctuations in the annual incidence of E coli infections, including ampicillin-resistant infections, suggest the need for continuation of surveillance in Connecticut and expansion to monitor larger populations.  Key words:  neonatal infections, sepsis, bacterial infections, bacterial infections/cause, bacterial infections/prevention and control, infant, premature, newborn.

Neonatal sepsis occurs at an estimated rate of 1 to 2 cases per 1000 live births in the United States. Blood cultures may be positive in up to 20% of infants in neonatal intensive care units (NICUs).^1-4 The organisms that cause early-onset (first week of life) sepsis include group B Streptococcus (GBS) and Escherichia coli, which together may account for 70% to 80% of blood and cerebrospinal fluid cultures. Although less common, enterococci, Listeria monocytogenes, and species of Gram-negative enteric bacilli other than E coli are known to cause disease in neonates.^2,3 Studies in the 1980s demonstrated that treatment of mothers intrapartum^5,6 or infants immediately postpartum with penicillin or ampicillin^7,8 could reduce the incidence of early-onset GBS infections in infants. In addition, it was noted that women who are at high risk of delivering infants who develop early-onset GBS infections could be identified either by detection of GBS carriers through collection of rectovaginal cultures at 35 to 37 weeks of gestation or by recognition of certain obstetrical risk factors, such as having a threatened premature delivery or prolonged rupture of the membranes. As a result of these and other studies, in 1996, the Centers for Disease Control and Prevention (CDC), in collaboration with the American College of Obstetricians and Gynecologists and the American Academy of Pediatrics, developed GBS prevention guidelines that advocated intrapartum antibiotics for women who are at high risk for delivering an infant with GBS disease.^1,9,10 Using a GBS screening-based approach, it is estimated that an 80% reduction of early-onset neonatal GBS infections would be possible, with up to 40% reduction possible using the risk-based prophylaxis approach.¹¹ On the basis of data from epidemiologic surveillance¹² and results of additional studies,¹³ partial adoption of the recommendations has reduced the incidence of early-onset GBS disease by 65% in geographic areas where population-based surveillance is conducted. One concern, however, is whether adoption of prophylaxis recommendations on a community-wide basis will have an adverse impact on the incidence, on the distribution of species that cause infections, or on the antibiotic susceptibility patterns of isolates that cause early-onset sepsis. In fact, in a few reports of neonatal sepsis in the setting of GBS prophylaxis, the occurrence of Gram-negative sepsis might have increased after intrapartum and postnatal prophylaxis.^514-17 These reports, typically from single hospital centers, are limited by the absence of denominators or small sample size.

We undertook a statewide surveillance program to determine whether adoption of the GBS prophylaxis guidelines would affect the incidence and antibiotic susceptibilities of non-GBS organisms that cause early-onset neonatal sepsis and meningitis in Connecticut. We gathered data on early-onset neonatal sepsis and meningitis cases that occurred from 1996 to 1999 at 19 hospitals in Connecticut. This period includes the year during which the GBS prophylaxis standards were issued (1996) and the 4 years of surveillance after the publication of the recommendations.

	METHODS

Top Abstract Methods Results Discussion References

Data Collection

We conducted surveillance for cases of early-onset neonatal sepsis and meningitis that occurred from 1996 to 1999 at 19 of 34 acute care hospitals in Connecticut. These 19 hospitals included all 18 hospitals with >1000 deliveries per year and all 11 hospitals with NICUs and inpatient pediatric facilities. Using 1994 data from the state vital records, the 19 hospitals surveyed accounted for 83.5% of live births (38 236/45 795) to state residents. Thus, we estimated that we would survey approximately 83% of infants with infections arising in the hospital of birth and virtually all infants seen in emergency rooms, pediatric inpatient facilities, or NICUs. Cases of neonatal sepsis that occurred between 1996 and 1998 were identified retrospectively, whereas cases that occurred in 1999 were identified prospectively.

Data collection took place between December 1998 and July 2000 using a standard record abstraction form. Data were collected on suspected maternal risk factors for neonatal sepsis, including maternal infections, obstetrical history and delivery, and whether there was intrapartum antibiotic use. Data collected on infants included demographic data, microbiology of infection, antibiotic susceptibility of all isolates, and outcome of the infection. Data were entered into EpiInfo 6.0 (CDC, Atlanta, GA), which also was used for data analysis.

Entry Criteria

To capture infants whose infections might be influenced by maternal antibiotic use, we surveyed only early-onset disease. A case of early-onset sepsis was defined as an infant who was hospitalized at 1 of the 19 Connecticut surveillance hospitals and who was 0 to 6 days of age when a blood or cerebrospinal fluid culture that grew a bacterial or fungal pathogen was obtained. Infants with cultures that grew only coagulase-negative staphylococcus, GBS, or diphtheroids were excluded. Surveillance for early-onset GBS infections was reported separately.^12,18

Cases were identified either through audit of hospital microbiology records or by review of hospital discharge data for specific International Classification of Diseases, Ninth Revision codes. A list of these codes and the number of cases identified by each code or by microbiology records can be obtained from the corresponding author. In hospitals with adequate microbiology record keeping, the microbiology source of data was preferred. Hospital charts for both the infant and the mother were reviewed.

State Demographic Data and GBS Incidence Data

Data on GBS infections for the surveillance years were obtained from the Active Bacterial Core surveillance of the Connecticut Emerging Infections Program. Annual live birth data were obtained from the 19 participating hospitals.

Definitions

Premature rupture of the membranes was defined as rupture of the membranes at <37 weeks' gestation, and prolonged rupture of the membranes was defined as occurring when membranes ruptured 18 hours before delivery.

Approval

This study was approved by the Human Investigation Committee at Yale University School of Medicine and at each of the hospitals surveyed. The CDC considered this surveillance of neonatal sepsis exempt from CDC institutional review board approval.

	RESULTS

Top Abstract Methods Results Discussion References

Cases and Incidence

We identified 94 cases of non-GBS early-onset sepsis that met the entry criteria during 1996 through 1999. Of these, 3 were not born at 1 of the 19 surveillance hospitals (2 in 1996, 1 in 1998). One was born at home, 1 was born at a nonsurveillance hospital, and for 1 the birth hospital information was not in the medical records but the child was born in Connecticut.

During this period, there were 174 535 in-state live births to Connecticut residents; of these, 140 923 live births occurred at the 19 surveillance hospitals. The average annual incidence of early-onset disease caused by organisms other than GBS at the surveillance hospitals was 0.67 per 1000 live births. During these 4 years, GBS caused 76 cases of early-onset sepsis, resulting in an average annual incidence of 0.44 early-onset GBS cases/1000 live births in the state (Table 1). Table 1 also shows that the size of the annual birth cohort at the 19 surveillance hospitals was similar for each of the 4 years of surveillance. Approximately 81% of the Connecticut births occurred at the 19 surveillance hospitals. There was no change in the rate of sepsis caused by non-GBS pathogens over time, whereas early-onset GBS disease declined from 0.61 to 0.23 per 1000 births between 1996 and 1999. 

Date of disease onset was available on all 94 cases. Onset of disease was the day of birth (day 0) in 46% of cases, and the mean age at onset was 1.7 days (Fig 1). During the 4 years, the proportion presenting with infection on the day of birth increased and was 38% in 1996, 43% in 1997, 42% in 1998, and 61% in 1999. All 4 of the cases of meningitis were diagnosed on the day of birth.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Age of onset of early-onset invasive neonatal infections attributable to species other than GBS (n = 90), Connecticut, 1996 to 1999 (0 = day of birth).

Clinical Characteristics Of the 94 cases, 80 had both maternal and infant charts reviewed, 8 had only infant charts available for review, and 6 had neither infant nor maternal charts available for review. Thus, microbiologic data were known for all 94 cases but, clinical and demographic data could be assessed only for 88 cases. Of the 88 cases for which infant charts were reviewed, 61 (69%) had symptomatic bacteremia (septicemia), 13 (15%) had asymptomatic bacteremia, and for 2 cases the charts did not indicate the clinical presentation. The remaining 12 cases included various focal findings, including 5 (6%) with pneumonia, 4 (5%) with meningitis, 2 (2%) with respiratory distress syndrome, and 5 (6%) with abnormal cerebrospinal fluid analyses but no bacterial or fungal isolate. Some had more than 1 focus listed. Boys accounted for 61 (69%) cases. Of the 88 cases, 55 (63%) were white, 15 (17%) were Hispanic, 13 (15%) were black, 1 (1%) was Asian, and 4 (5%) were of unknown race or ethnicity. Among the 88 infants, 17 (19%) died, 59 (67%) survived without known sequelae, 8 (9%) survived with significant sequelae, and 4 (5%) were transferred to other facilities, where outcome could not be assessed.

Microbiology of Early-Onset Neonatal Sepsis The distribution of species isolated from blood or spinal fluid during 1996 through 1999 is shown in Table 2. Only 1 isolate of a particular species per infant was counted. When more than 1 species was isolated during the first 6 days of life, each was counted as a separate infection. There were no discordant blood and cerebrospinal fluid species cultured. E coli was the predominant pathogen encountered in this study; other species caused relatively few cases. No fungal isolates were obtained during the first week of life in these infants. During the 4 years of the study, 35 (37%) of the 94 non-GBS sepsis cases were attributable to Gram-positive species and 59 (63%) were attributable to Gram-negative species, 39 of which (41% of all isolates, 66% of Gram-negatives) were E coli. E coli accounted for 21% (5 of 24) of the non-GBS cases in 1996, 52% (12 of 23) in 1997, 58% (14 of 24) in 1998, and 35%, (8 of 23) in 1999 (exact test for linear trend, P = .27).

Antibiotic Susceptibility Figure 2 shows the ampicillin susceptibility patterns of E coli isolates for which antibiotic susceptibility data were available. The rate of resistance varied by year from 0% (0 of 3) in 1996% to 25% (3 of 12) in 1997, 63% (7 of 11) in 1998, and 57% (4 of 7) in 1999 (exact test for linear trend, 2-tailed, P = .04). The number of cases in which the antibiotic susceptibility data were missing were 2 in 1996, 0 in 1997, 3 in 1998, and 1 in 1999. During the 4 years of surveillance, the E coli isolates were uniformly susceptible to cephalosporins.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Susceptibility to ampicillin of E coli isolated from the blood or spinal fluid of neonates with early-onset infections, by year, Connecticut, 1996 to 1999 (gray = susceptible; white = intermediate; black = resistant).

Maternal Features Eighty (85%) maternal charts were available for review. Of these 80 mothers, 40 (50%) had received intrapartum antibiotics, which in 26 cases (65%) was ampicillin. The reasons for intrapartum antibiotics as stated on these 40 maternal charts were as follows (many charts listed more than 1 reason): premature rupture of membranes, 24 (59%); prematurity, 24 (59%); prolonged rupture of membranes, 14 (36%); fever, 13 (33%), GBS prophylaxis, 9 (23%); chorioamnionitis, 8 (20%);and miscellaneous others, 4 (10%).

	DISCUSSION

Top Abstract Methods Results Discussion References

If a substantial number of parturient women are likely to be exposed to penicillin for prevention of early-onset GBS disease, then the consequence of such a mass exposure needs to be a concern of medical providers and the lay population as well. If the organisms that cause neonatal sepsis merely shift from GBS to other species, especially to Gram-negative rod species, then the value of recommending perinatal prophylaxis should be questioned. If exposure to penicillin results in neonatal sepsis being caused by organisms that are resistant to first-line antibiotics, then the same questions should arise, and there could be danger to the public as a result of dissemination of antibiotic-resistant organisms. The degree of antimicrobial exposure of parturient women in Connecticut was estimated recently through a review of 992 charts of mothers who delivered in 1996. Overall, 15% received intrapartum antibiotic prophylaxis (J. Hadler, Connecticut state epidemiologist, personal communication, March 2001). Although similar data for later years are not yet available, studies in other locales, including a large West Coast health maintenance organization,¹⁹ suggest that perinatal antibiotic exposure approximately doubled from 1996 to 1999.

The data presented here should provide some assurance that antibiotic treatment of parturient women for GBS prevention has not resulted in an increased rate of early-onset neonatal sepsis as a result of non-GBS species during a 4-year period. During this period, the rate of neonatal sepsis attributable to GBS declined; therefore, the total number of early-onset neonatal sepsis cases of all causes also declined.

One of the main aims of this study was to determine whether antibiotic susceptibility of non-GBS early-onset sepsis would be affected by the antibiotics given to mothers as intrapartum prophylaxis for GBS. Because the number of isolates of species other than E coli was very small (Table 2), we were restricted in the ability to evaluate species-specific trends. Although there is no clear trend toward an increase in the proportion of antibiotic-resistant non-GBS species, the data on E coli may raise some concerns. E coli clearly was the most common non-GBS species encountered, and for the first 3 years of the study, the number and rate of E coli infections increased while the proportion of isolates that were resistant to ampicillin also increased. Concern about these results is mitigated by the 1999 data that showed a drop in both the number and rate of E coli infections and the rate of resistance among the isolates. In addition, it is important to note that the total number and rate of non-GBS isolates did not change, and thus intrinsically resistant speciessuch as Enterococcus species and Enterobacter speciesalthough uncommon, as a group did not increase either.

These E coli data also highlight that monitoring trends in disease cause and antimicrobial resistance should be undertaken in population-based settings over extended periods to prevent misinterpretation of normal random fluctuations and drawing of inferences that may not hold up under additional evaluation.

The conclusions that can be drawn from this study are limited by the relatively small number of cases of non-GBS neonatal sepsis or meningitis in Connecticut. The advantage of performing this study in Connecticut is that through surveillance in 19 hospitals, it was possible to review the neonatal sepsis experience for 81% of infants who were born in the state and virtually all infants who were hospitalized in a pediatric inpatient facility or NICU. Balancing this is the low rate of early-onset neonatal sepsis and a relatively small state population. Surveillance is continuing in this and other states as part of the Emerging Infections Program, Active Bacterial Core Surveillance Program to determine whether with a larger number of cases and with increased use of preventive antibiotics there will be significant evidence of increasing non-GBS infection rates and resistance of E coli-associated neonatal sepsis or meningitis.

	ACKOWLEDGMENTS

The work presented in this article was supported by the Connecticut Emerging Infections Program, a Cooperative Agreement (U50/CCU111188-07) from the Centers for Disease Control and Prevention.

	FOOTNOTES

Dr O'Brien's current affiliation is the Center for American Indian and Alaskan Native Health, Johns Hopkins School of Public Health, Baltimore, Maryland.

Received for publication Mar 22, 2001; accepted Jun 5, 2001.

Reprint requests to (R.S.B.) Department of Pediatrics, Yale University School of Medicine, Box 208064, New Haven, CT 06520-8064. E-mail robert.baltimore{at}yale.edu

	ABBREVIATIONS

NICU, neonatal intensive care unit; GBS, group B Streptococcus (streptococcal); CDC, Centers for Disease Control and Prevention.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: a public health perspective. MMWR Morb Mortal Wkly Rep. 1996;45(RR-7):1-24
2. Baltimore RS. Perinatal bacterial and fungal infections. In: Jenson HB, Baltimore RS, eds. Pediatric Infectious Diseases: Principles and Practice. Norwalk, CT: Appleton and Lange; 1995:1387-1406
3. Gladstone IM, Ehrenkranz RA, Edberg SC, Baltimore R A ten-year review of neonatal sepsis and comparison with the previous fifty-year experience. Pediatr Infect Dis J 1990; 9:819-825 [Medline]
4. Beck-Sague CM, Azimi P, Fonseca SN, Bloodstream infections in neonatal intensive care unit patients: results of a multicenter study. Pediatr Infect Dis J 1994; 13:1110-1116 [Medline]
5. Boyer KM, Gadzala CA, Burd LI, Fisher DE, Paton JB, Gotoff SP Selective intrapartum chemoprophylaxis of neonatal group B streptococcal early-onset disease. I. Epidemiologic rationale. J Infect Dis 1983; 5:795-801
6. Boyer KM, Gotoff SP Prevention of early-onset neonatal group B streptococcal disease with selective intrapartum chemoprophylaxis. N Engl J Med 1986; 314:1665-1669 [Abstract]
7. Siegel JD, McCracken GH Jr, Threlkeld N, Milvenan B, Rosenfeld CR Single-dose penicillin prophylaxis against group B streptococcal infections: a controlled trial in 18,738 newborn infants. N Engl J Med 1980; 303:769-775 [Abstract]
8. Siegel JD, McCracken GH, Threlkeld N, DePasse BM, Rosenfeld CR Single-dose penicillin prophylaxis of neonatal group B streptococcal disease. Lancet 1982; 1:1426-1430 [Medline]
9. American Academy of Pediatrics, Committee on Infectious Diseases and Committee on Fetus and Newborn Revised guidelines for prevention of early-onset group B streptococcal (GBS) disease. Pediatrics 1997; 99:489-496 [Abstract/Free Full Text]
10. American College of Obstetricians and Gynecologists, Committee on Obstetric Practice. Prevention of Early-Onset Group B Streptococcal Disease in Newborns. Washington, DC: American College of Obstetricians and Gynecologists; 1996
11. Rosenstein NE, Schuchat A, the Neonatal Group B Streptococcal Disease Study Group Opportunities for prevention of perinatal group B streptococcal disease: a multi-state surveillance analysis. Obstet Gynecol 1997; 90:901-906 [CrossRef][Medline]
12. Schrag SJ, Zywicki S, Farley MM, Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis. N Engl J Med 2000; 342:15-20 [Abstract/Free Full Text]
13. Factor SH, Whitney CG, Zywicki S, Schuchat A, the ABC Surveillance Team Effects of hospital policies on the 1996 group B streptococcal consensus guidelines. Obstet Gynecol 2000; 95:377-382 [CrossRef][Medline]
14. Towers CV, Carr MH, Padilla G, Asrat T Potential consequences of widespread antepartal use of ampicillin. Am J Obstet Gynecol 1998; 179:879-883 [CrossRef][Medline]
15. McDuffie RS, McGregor JA, Gibbs RS Adverse perinatal outcome and resistant Enterobacteriaceae after antibiotic usage for premature rupture of the membranes and group B streptococcus carriage. Obstet Gynecol 1993; 82:487-489 [Medline]
16. Terrone DA, Rinehart BK, Einstein MH, Britt LB, Martin JN Jr, Perry KG Neonatal sepsis and death caused by resistant Escherichia coli: possible consequences of extended maternal ampicillin administration. Am J Obstet Gynecol 1999; 180:1345-1348 [CrossRef][Medline]
17. Joseph TA, Pyati SP, Jacobs N Neonatal early-onset Escherichia coli disease. The effect of intrapartum ampicillin. Arch Pediatr Adolesc Med 1998; 152:35-40 [Abstract/Free Full Text]
18. Centers for Disease Control and Prevention Early-onset group B streptococcal diseaseUnited States, 1998-1999. MMWR Morb Mortal Wkly Rep 2000; 49:793-796
19. Davis RL, Hasselquist MB, Cardenas V, Introduction of the new Centers for Disease Control and Prevention group B streptococcal prevention guideline at a large West Coast health maintenance organization. Am J Obstet Gynecol 2001; 184:603-610 [CrossRef][Medline]