ARTICLE
Changing Patterns in Neonatal
Escherichia coli
Sepsis
and Ampicillin Resistance in the Era of Intrapartum
Antibiotic Prophylaxis
Matthew J. Bizzarro, MDa, Louise-Marie Dembry, MDb,c,d, Robert S. Baltimore, MDc,d,e, Patrick G. Gallagher, MDa
Divisions ofaPerinatal Medicine andeInfectious Diseases, Department of Pediatrics,bDepartment of Internal Medicine, andcDepartment of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut;dDepartment of Quality Improvement Support Services, Yale-New Haven Hospital, New Haven, Connecticut
The authors have indicated they have no financial relationships relevant to this article to disclose.
What’s Known on This Subject
Early-onset neonatal sepsis due toE coli, including ampicillin-resistantE colibacteremia, is increasing in preterm infants, paralleling an increase in intrapartum antibiotic use.
What This Study Adds
In preterm infants, this study identifies intrapartum ampicillin exposure as an indepen-dent risk factor for ampicillin-resistantE coliearly-onset sepsis and describes a significant increase inE colilate-onset sepsis.
ABSTRACT
OBJECTIVE.The goal was to determine current trends inEscherichia coli-related early- and late-onset sepsis and patterns of ampicillin resistance in relation to institutional changes in the use of intrapartum antibiotic prophylaxis.
METHODS.A retrospective review of data for all infants withE colisepsis at Yale-New Haven Hospital from 1979 to 2006 was performed. Study periods were based on predominant intrapartum antibiotic prophylaxis practices at Yale-New Haven Hos-pital, that is, (1) 1979 to 1992 (no formal intrapartum antibiotic prophylaxis), (2) 1993 to 1996 (risk factor-based), and (3) 1997 to 2006 (screening-based). Sepsis rates and patterns of ampicillin resistance were compared.
RESULTS.Fifty-three cases ofE coliearly-onset sepsis and 129 cases ofE colilate-onset sepsis were identified over 3 eras. In very low birth weight (⬍1500 g) infants, increases in E coli early-onset sepsis (period 1: 2.83 cases per 1000 very low birth weight admissions; period 2: 7.12 cases per 1000 very low birth weight admissions; period 3: 10.22 cases per 1000 very low birth weight admissions), intrapartum ampicillin exposure, and ampicillin-resistantE coliwere observed. Intrapartum am-picillin exposure was determined to be an independent risk factor for amam-picillin- ampicillin-resistant E coli early-onset sepsis. For the first time, a significant increase in E coli
late-onset sepsis was observed in preterm infants (period 1: 10.39 cases per 1000 very low birth weight admissions; period 2: 16.01 cases per 1000 very low birth weight admissions; period 3: 21.66 cases per 1000 very low birth weight admissions) and term infants (period 1: 4.07 cases per 1000 admissions; period 2: 4.22 cases per 1000 admissions; period 3: 8.23 cases per 1000 admissions).
CONCLUSIONS.Studies to provide a better understanding of potential consequences of intrapartum antibiotic exposure and its contribution to evolving trends in neonatal sepsis are urgently needed.
B
LOODSTREAM INFECTIONS ARE a significant cause of neonatal morbidity and death. Early-onset sepsis (EOS) (bloodstream infection atⱕ72 hours of life) is a major contributor to these sequelae. In developed countries, group B streptococcus (GBS) emerged as the leading etiologic agent of EOS in the 1970s, with case fatality rates as high as 55%.1 Beginning in 1992, national guidelines for the use of intrapartum antibiotic prophylaxis (IAP) to prevent neonatal GBS infection were published.2–5The current guidelines recommend universal screening for GBS colonization in pregnant women at 35 to 37 weeks of gestation, with penicillin as the first-line agent for IAP.5 Widespread adoption of these guidelines has resulted in a significant decrease in GBS-related EOS.6Most premature infants are delivered before the gestational age (GA) at which routine GBS screening and IAP are recommended, but an estimated 65% of very low birth weight (VLBW) (⬍1500-g) neonates are exposed to antibiotics before delivery.7Although current guidelines advocate GBS screening and IAP even in the setting of threatened preterm delivery,5compliance with these guidelines is not always achieved.8Furthermore, premature
www.pediatrics.org/cgi/doi/10.1542/ peds.2007-2171
doi:10.1542/peds.2007-2171
Key Words
very low birth weight, bacteremia, group B streptococcus
Abbreviations EOS— early-onset sepsis GBS— group B streptococcus IAP—intrapartum antibiotic prophylaxis GA— gestational age
VLBW—very low birth weight LOS—late-onset sepsis Y-NHH—Yale-New Haven Hospital NBSCU—newborn special care unit PROM—prolonged rupture of membranes BW— birth weight
LR—logistic regression
Accepted for publication Aug 30, 2007
Address correspondence to Patrick G. Gallagher, MD, Department of Pediatrics, Yale University School of Medicine, 333 Cedar St, PO Box 208064, New Haven, CT 06520-8064. E-mail: patrick.gallagher@yale.edu
neonates often receive intrapartum antimicrobial expo-sure for reasons other than maternal GBS colonization. The substantial risks and consequences of infection in the vulnerable preterm population are likely responsible for this practice, although consensus opinion regarding the use of intrapartum antimicrobial therapy, other than to prevent GBS disease, is limited.9,10
The increase in the intrapartum antibiotic exposure of term and preterm neonates has raised significant con-cerns. Concerns regarding the widespread use of anti-microbial agents in the parturient population, a group not previously exposed to significant antimicrobial treat-ment, are based on existing data that suggest that fre-quent prolonged exposure to broad-spectrum antibiotics may alter the normal maternal microbiologic flora. This may result in passage of and subsequent colonization of the neonatal skin and gastrointestinal tract with resistant and potentially pathogenic bacteria.11 Potential unin-tended consequences of intrapartum antibiotic exposure at any GA could include increased rates of neonatal sepsis attributable to virulent and resistant pathogens, development of antibiotic-resistant infections beyond the perinatal period in the mother or her infant, and changes in community patterns of antibiotic resistance.12–14 Some of these fears were realized when several centers reported the emergence ofEscherichia coliover GBS as the predominant organism responsible for EOS, as well as in-creased rates ofE coliampicillin resistance, after initiation of IAP.14–17 Additional investigation determined that these concerning trends were primarily limited to the preterm population,15–19 a group particularly susceptible to E coli infection15,20–22 and, once infected, at increased risk of sepsis-related morbidity and death.15,23,24Almost no atten-tion has focused on the potential impact of IAP on the organisms responsible for late-onset sepsis (LOS) (blood-stream infection at ⬎72 hours of life), a more-prevalent problem with significant morbidity and mortality rates.22,25 As the rate of intrapartum antibiotic exposure contin-ues to increase,20its potential effects, both positive and negative, on the organisms responsible for neonatal sep-sis warrant additional investigation. We sought to eval-uate potential trends in the incidence and antibiotic susceptibility patterns of E coli-related EOS and LOS in conjunction with changes in intrapartum antibiotic use over nearly 3 decades. Intrapartum antibiotic exposure was evaluated as a potential risk factor for E coli anti-biotic resistance over the same time period.
METHODS
Study Design
The Yale-New Haven Hospital (Y-NHH) newborn special care unit (NBSCU) is a 46-bed, tertiary care, referral center for infants with complex medical and surgical conditions. E coli has been the most common Gram-negative pathogen responsible for both EOS and LOS at our NBSCU for⬎75 years and therefore was chosen for review.25Data on allE coli-positive blood cultures from newborns and infants collected between January 1, 1979, and December 31, 2006, were obtained from the microbiology laboratory. These included cultures
col-lected at any day of life from inpatients in the Y-NHH NBSCU, as well as those from neonatesⱕ30 days of age in the emergency department, wards, ICU, or ambula-tory clinic. Subjects included both inborn and outborn infants and, with respect to NBSCU inpatients, included only those with long-term admissions (⬎24 hours). Most referring institutions were local or regional Con-necticut hospitals that are part of the Yale-New Haven Health System. The records of all infants were re-viewed, and maternal charts linked to the infants were examined. Data on intrapartum risk factors for neo-natal infection, data on maternal antibiotic exposure, and neonatal data including demographic characteris-tics, antibiotic exposure,E coliantibiotic susceptibility patterns, risk factors for infection, and outcomes were collected.
The majority of newborns identified (125 of 178 in-fants) were inborn; therefore, the 28-year study period was divided on the basis of the predominant IAP prac-tices at Y-NHH, that is, (1) 1979 to 1992 (no formal IAP guidelines), (2) 1993 to 1996 (risk factor-based IAP), and (3) 1997 to 2006 (screening-based IAP). Predominant screening practices at referring institutions over most of the study period, before the recommendations for uni-versal screening and IAP, were not known. Sepsis rates, patterns of ampicillin resistance, and demographic data for the affected population were compared among eras.
Definitions
Sepsis
Sepsis was defined as a positive blood culture forE coli
and was classified, according to the infant’s age, as EOS (ⱕ72 hours of life) or LOS (⬎72 hours of life). Multiple positive blood cultures from a single patient with iden-tical antibiotic susceptibility patterns were considered a single episode.
Risk Factors, Symptoms, and Outcomes
Intrapartum antibiotic therapy was defined as the use of
ⱖ1 antimicrobial agent from the onset of labor until the time of delivery. Because the duration of labor can vary, no time constraints were placed on this definition. Preterm labor, maternal fever, prolonged rupture of membranes (PROM), chorioamnionitis, and sepsis-related death were defined as described.26Prolonged intubation was defined as the need for mechanical ventilation for greater than 14 days. The presence of a central vascular catheter was in-cluded if the catheter was placed before the onset of infec-tion. Surgery was included as a potential risk factor if it occurredⱕ7 days before the onset of sepsis.
Laboratory Analyses
Blood cultures were assessed by using a carbon dioxide detection system (Bactec II or 9240; Becton Dickinson, Franklin Lakes, NJ). Antibiotic susceptibility patterns were tested by using standard disk diffusion methods.27
Statistical Analyses
popu-lation, given its substantial risk for infection, was ana-lyzed both separately and as part of the entire cohort. We documented and analyzed linear trends in EOS, LOS, ampicillin resistance, and intrapartum antibiotic use over the study periods by using2analysis of trends.
Data pertaining to EOS and LOS were presented as num-ber of cases per 1000 NBSCU admissions for the entire population and number of cases per 1000 VLBW NBSCU admissions for the VLBW population. This denominator was chosen to incorporate both inborn and outborn cases of sepsis.
Intrapartum, demographic, hospital course, and out-come data were compared among periods. For neonates who experienced ⬎1 episode of sepsis (n ⫽ 4), demo-graphic data (eg, GA, birth weight [BW], and gender) pertaining to each were evaluated once, and only the first episode of sepsis was evaluated in logistic regression (LR) modeling. Continuous data were compared by us-ing analysis of variance, and dichotomous data were analyzed by using Fisher’s exact test if 1 cell contained ⬍5 or the2test if cells containedⱖ5.
Factors associated with ampicillin-resistantE coliEOS and LOS were analyzed separately. Maternal, intrapar-tum, and neonatal data were analyzed individually by using ampicillin resistance as a binary variable and the primary outcome of interest. Analyses included the ef-fects of intrapartum antibiotic exposure and its duration, postnatal antibiotic exposure and its duration, demo-graphic data, and the presence and duration of use of support apparatus (eg, central vascular catheter). Unad-justed comparisons were made by using the 2test or
Fisher’s exact test for dichotomous data and Student’st
test or the Wilcoxon rank-sum test for continuous data. Covariates withPvalues of⬍.10 were then included in the multivariate LR analysis using ampicillin resistance as the dependent variable. Adjusted odds ratios were calculated for each covariate when possible.Pvalues of ⬍.05 were considered significant.
Ampicillin-resistant cases of sepsis were also analyzed as an independent variable in analyses assessing the potential effects of ampicillin-resistant E coli sepsis on certain outcomes of interest (eg, death). Unadjusted comparisons were made by using the2test or Fisher’s
exact test. Covariates withP values of⬍.10 were then included in the multivariate LR analysis. Adjusted odds ratios were calculated.Pvalues of⬍.05 were considered significant. This study was approved by the human in-vestigation committee of Yale University.
RESULTS
Demographic Data From the Y-NHH NBSCU for 1979 to 2006 Between January 1, 1979, and December 31, 2006, there were 140 442 live births and 22 755 admissions to the Y-NHH NBSCU, including 4316 VLBW infants. The total number of admissions per year (P ⫽ .164), the number of VLBW admissions per year (P⫽.180), and the mean length of hospital stay for the VLBW popula-tion (P ⫽ .107) did not change significantly over 28 years, but survival rates of VLBW neonates increased significantly (P⬍.0001).
Intrapartum Antibiotic Use/Exposure
The proportions of neonates withE coli sepsis with in-trapartum antibiotic exposure increased significantly over the 3 study periods, both in the entire cohort (pe-riod 1: 16%; pe(pe-riod 2: 22%; pe(pe-riod 3: 66%;P⬍.0001) and in the VLBW population (period 1: 22%; period 2: 36%; period 3: 85%; P⬍.0001), with the majority of exposure (61 of 92 neonates) occurring in conjunction with the institutional adoption of IAP guidelines for GBS and PROM (1997–2006).
Sepsis
There were a total of 1385 cases of neonatal sepsis at Y-NHH during the study period, including 232 cases of EOS (17%) and 1153 cases of LOS (83%).E coliwas the respon-sible organism in 182 (13%) of these cases, which occurred in 178 infants (125 inborn and 53 outborn). Cases were further classified and analyzed as EOS and LOS.
E coliEOS
Trends
There were 53 cases of E coli EOS in 53 newborns, including 27 cases in 27 VLBW neonates. There was a significant decrease in the overall incidence of EOS, attributed to any organism, over the study periods (pe-riod 1: 12.03 cases per 1000 NBSCU admissions; pe(pe-riod 2: 12.02 cases per 1000 NBSCU admissions; period 3: 7.19 cases per 1000 NBSCU admissions; P⫽.007). No significant change was observed in the incidence ofE coli
EOS (period 1: 1.72 cases per 1000 NBSCU admissions; period 2: 2.60 cases per 1000 NBSCU admissions; period 3: 3.13 cases per 1000 NBSCU admissions; P ⫽ .125). The incidence of GBS EOS decreased significantly, par-ticularly after the adoption of screening-based IAP (pe-riod 1: 5.79 cases per 1000 NBSCU admissions; pe(pe-riod 2: 8.45 cases per 1000 NBSCU admissions; period 3: 2.67 cases per 1000 NBSCU admissions;P⬍.001).
For VLBW newborns, no consistent trends (P ⫽
.585) in the overall incidence of EOS were observed (Fig 1A). However, a significant increase (P⬍.001) in the incidence ofE coliEOS was observed (Fig 1B). No significant change in the incidence of GBS-related EOS in VLBW neonates was detected (period 1: 6.14 cases per 1000 VLBW NBSCU admissions; period 2: 1.78 cases per 1000 VLBW NBSCU admissions; period 3: 4.21 cases per 1000 VLBW NBSCU admissions;P⫽
.417).
Characteristics of neonates withE coliEOS were com-pared across study periods (Table 1). The mean GA and BW of E coli-infected infants decreased significantly, whereas the proportions delivered after preterm labor and chorioamnionitis and intrapartum exposure to anti-biotics, including ampicillin, increased significantly. No significant differences in outcomes were detected.
E coli EOS in the VLBW Population
proportions delivered after preterm labor (period 1: 50%; period 2: 75%; period 3: 82%;P⫽.318), PROM (period 1: 50%; period 2: 50%; period 3: 44%; P ⫽
.961), and chorioamnionitis (period 1: 0%; period 2: 50%; period 3: 50%; P⫽.080) did not change signifi-cantly but intrapartum ampicillin exposure did (period 1: 0%; period 2: 50%; period 3: 88%;P⬍.001).
Ampicillin-Resistant E coli EOS
Hospital-wide data, which included data from the NBSCU and obstetric wards, revealed no statistically sig-nificant change in the proportions of isolates of ampicillin-resistant E coli over the study periods (period 1: 30%; period 2: 37%; period 3: 41%; Y-NHH clinical micro-biology data;P⫽.117). In the entire NBSCU population withE coliEOS, 20% of cases ofE coliEOS (4 of 20 cases)
were ampicillin-resistant before recommendations for GBS IAP (1979 –1992), compared with 44% (4 of 9 cases) in the risk factor-based era (1993–1996) and 54% (13 of 24 cases) in the screening-based era (1997–2006;
P⫽.061).
A significant increase in the proportion of ampicillin-resistantE coliEOS was identified in the VLBW popula-tion over the study period (P ⫽.005). No cases of am-picillin-resistantE coliwere observed from 1979 to 1992 (0 of 6 cases), compared with 25% from 1993 to 1996 (1 of 4 cases) and 64% from 1997 to 2006 (12 of 17 cases). VLBW neonates with ampicillin-resistant E coli EOS were compared with those with ampicillin-sensitive E coliEOS and were determined to have a lower GA (25.2 weeks and 27.2 weeks, respectively;P⫽.036) and BW (804 g and 1003 g, respectively;P⫽.043), and a greater
16.05 10.68 18.65
94.9
181.49
265.94
0 50 100 150 200 250 300
1997–2006 1993–1996
1979–1992
1997–2006 1993–1996
1979–1992
Year
Ca
se
s pe
r 1
0
00
VLBW
adm
iss
ions
EOS LOS
2.83
7.12
10.22 10.39
16.01
21.66
0 5 10 15 20 25
Year
Ca
ses
per
10
00
VLBW admiss
ions
E coli EOS E coli LOS
A
B
FIGURE 1
Trends in the numbers of cases of EOS and LOS per 1000 NBSCU VLBW admissions at Y-NHH over 3 study periods.
A, All cases of EOS and LOS attributed to any organism in the VLBW population of neonates at Y-NHH over 3 study periods. Although no significant changes in EOS were ob-served (P⫽.585), the total number of cases of LOS per 1000 VLBW admissions increased significantly (P⬍.001).
proportion had intrapartum ampicillin exposure (92% and 31%, respectively;P⫽.004). The effect of intrapar-tum ampicillin exposure on ampicillin resistance was evaluated in a LR analysis that controlled for the effects of GA and BW. Intrapartum ampicillin exposure was determined to be a significant independent risk factor for antibiotic resistance (adjusted odds ratio: 17.91; 95% confidence interval: 1.59 –202.37;P⫽.020).
Aminoglycoside- and Third-Generation Cephalosporin-Resistant E coli EOS
No cases of gentamicin-resistantE coliEOS occurred in any of the 3 study periods. The only case of third-generation cephalosporin-resistantE coliEOS in the 28-year study period occurred in 2006 in a VLBW neonate.
E coliLOS
Trends
There were 129 cases of E coliLOS in 125 infants, cluding 67 cases in 64 VLBW neonates. Significant in-creases in the overall incidence of LOS (period 1: 28.48 cases per 1000 NBSCU admissions; period 2: 55.90 cases per 1000 NBSCU admissions; period 3: 75.97 cases per 1000 NBSCU admissions;P⬍.001) and the incidence of
E coliLOS (period 1: 4.07 cases per 1000 NBSCU admis-sions; period 2: 4.22 cases per 1000 NBSCU admisadmis-sions; period 3: 8.23 cases per 1000 NBSCU admissions; P ⬍
.001) over 28 years were observed. In the VLBW popu-lation, significant increases in the numbers of all cases of LOS (P ⬍.001) (Fig 1A) and cases of E coli LOS (P ⫽
.006) (Fig 1B) were observed.
There were no statistically significant differences in GA and BW in neonates withE coliLOS over the study period. Statistically significant differences in intrapartum antibiotic exposure, intrapartum ampicillin exposure, and duration of central vascular catheter use were ob-served. No significant differences in use and/or duration
of postnatal antibiotic treatment or outcomes were ob-served (Table 2).
E coli LOS in the VLBW Population
Mean GA (period 1: 27.7 weeks; period 2: 27.7 weeks; period 3: 26.3 weeks;P⫽.103) and BW (period 1: 972 g; period 2: 959 g; period 3: 943 g;P⫽.917) did not change significantly among VLBW neonates with E coli LOS. Significant increases in intrapartum antibiotic exposure (period 1: 20%; period 2: 30%; period 3: 82%; P ⬍
.0001) and intrapartum ampicillin exposure (period 1: 10%; period 2: 20%; period 3: 42%; P ⫽ .028) were observed. No significant differences in use or duration of postnatal antibiotic treatment or outcomes were ob-served.
Ampicillin-Resistant E coli LOS
No statistically significant changes in the proportions of ampicillin-resistantE coliLOS in the entire cohort (P⫽
.188) were observed over the study periods. Similarly, no differences in the VLBW population (P⫽.435) were observed.
Cases of ampicillin-resistant (n⫽29) and ampicillin-sensitive (n ⫽ 38) E coli LOS in VLBW infants were compared. More infants with ampicillin-resistant strains of E coli had been exposed to intrapartum antibiotic treatment (ampicillin-resistant: 66%; ampicillin-sensitive: 47%), but this difference was not statistically signifi-cant (P ⫽.215). Infants with ampicillin-resistant E coli
had higher incidence rates of necrotizing enterocolitis (ampicillin-resistant: 62%; ampicillin-sensitive: 21%;
P⫽.001) and sepsis-related death (ampicillin-resistant: 34%; ampicillin-sensitive: 11%;P⫽.031). Associations between E colibacteremia and necrotizing enterocolitis and between intrapartum antibiotic use and necrotizing enterocolitis in the preterm population have been de-scribed.28,29To address these associations, a multivariate TABLE 1 Characteristics of EOS Attributable toEscherichia coliAccording to Study Period
1979–1992 (N⫽20)
1993–1996 (N⫽9)
1997–2006 (N⫽24)
P
GA, mean⫾SD, wk 34.1⫾4.8 32.8⫾6.2 28.1⫾5.4 .002a
BW, mean⫾SD, g 2234⫾849 1988⫾1176 1360⫾984 .014a
Male gender,n(%) 9 (45) 6 (67) 14 (58) .577
Vaginal delivery,n(%) 9 (45) 5 (56) 8 (33) .498
Preterm labor,n(%) 8 (40) 4 (44) 18 (75) .047a
PROM,n(%) 5 (25) 4 (44) 12 (50) .270
Maternal fever,n(%) 4 (20) 1 (11) 5 (21) .998
Chorioamnionitis,n(%) 1 (5) 2 (22) 10 (42) .013a
Intrapartum antibiotic treatment,n(%) 6 (30) 2 (22) 19 (79) .001a
Intrapartum ampicillin treatment,n(%) 3 (15) 2 (22) 14 (58) .007a
Intrapartum ampicillin treatment for⬎24 h,n(%) 2 (10) 1 (11) 5 (21) .684
Multiple antibiotics,n(%)b 3 (50) 2 (100) 14 (74) .420
Meningitis,n(%) 1 (5) 1 (11) 3 (13) .702
Death,n(%) 4 (20) 1 (11) 9 (38) .291
Sepsis-related death,n(%) 3 (15) 0 8 (33) .094
Continuous data were analyzed with analysis of variance and dichotomous data with the2test or Fisher’s exact test (cell number of⬍5). aStatistically significant value (P⬍.05).
LR analysis controlling for effects of prematurity, use and duration of use of support apparatuses and anti-biotics, and the presence of patent ductus arteriosus was performed. A significantly higher incidence of necrotiz-ing enterocolitis was maintained in the cohort with ampicillin-resistantE coli(adjusted odds ratio: 4.51; 95% confidence interval: 1.23–14.28; P ⫽ .026). No other significant differences were observed.
Aminoglycoside- and Third-Generation Cephalosporin-Resistant E coli LOS
No cases of gentamicin-resistantE coli LOS occurred in any of the 3 study periods. The only case of third-generation cephalosporin-resistantE coliLOS in the 28-year study period occurred in 2006 in a VLBW neonate.
DISCUSSION
At Y-NHH, the total number of cases of EOS per 1000 live births has decreased and LOS has increased over 28 years. The decrease in EOS is likely related to the signif-icant reduction in GBS EOS observed after the adoption of widespread IAP. No changes in the overall incidence or antibiotic resistance patterns of E coli EOS were ob-served in the entire population with adoption of this practice. In contrast, in our VLBW neonatal population, we observed significant increases in both EOS and LOS attributed toE coli, in addition to a significant increase in the proportion of ampicillin-resistant cases ofE coliEOS. The EOS findings support similar trends described pre-viously and extend them over nearly 3 decades at a single center.15–19The documented increases inE coliLOS in the overall and VLBW neonates over such a long
period of time are, to our knowledge, reported for the first time.
The increase in E coli EOS observed at Y-NHH in VLBW neonates is likely related to several factors. Al-though the overall number of VLBW admissions to our NBSCU did not change over the study period, the mean GA and BW of the VLBW cohort decreased significantly. Previous investigations documented a significant inverse relationship between prematurity and the risk of neona-tal sepsis.30 This association is likely related to a more-immature immune system, ineffective skin barrier, and need for support apparatus, in addition to other factors. It is possible that this change in the demographic fea-tures of our VLBW population has created a more-vulnerable cohort with a larger proportion of extremely preterm neonates.
Intrapartum events and maternal factors are also thought to contribute to the risk of EOS. Given the threat and consequences of infections in premature newborns and their mothers, intrapartum antibiotic use has become common. In our cohort of VLBW neonates, intrapartum antibiotic exposure increased significantly over a 28-year period. Given the increased use of anti-biotics in term pregnancies observed after the publica-tion of the GBS IAP guidelines,20 one might speculate that the increase observed in our VLBW population was the result of the adoption of similar guidelines for the use of IAP for threatened preterm delivery and for PROM.5,9 However, no concurrent increase in the pro-portion of neonates delivered after the onset of preterm labor or PROM was observed in this cohort. Intrapartum antibiotic therapy in the preterm population may also be TABLE 2 Characteristics of LOS Attributable toEscherichia coliAccording to Study Period
1979–1992 (N⫽44)
1993–1996 (N⫽14)
1997–2006 (N⫽71)
P
GA, mean⫾SD, wk 32.6⫾5.8a 30.6⫾5.8 31.2⫾6.0b .438
BW, mean⫾SD, g 1988⫾1120a 1691⫾1394 1851⫾1155b .728
Male gender,n(%) 25 (58)a 7 (50) 44 (65)b .536
Intrapartum antibiotic treatment,n(%) 4 (9)a 3 (21) 42 (62)b ⬍.001c Intrapartum ampicillin treatment,n(%) 2 (5)a 2 (14) 20 (29)b .003c Intrapartum ampicillin treatment for⬎24 h,n(%) 2 (5)a 2 (14) 10 (15)b .219 Day of life of culture, mean⫾SD 22.8⫾23.2 31.4⫾18.6 30.0⫾30.1 .343
CVC,n(%) 18 (41) 8 (57) 44 (62) .114
Time with CVC, mean⫾SD, d 2.7⫾6.4 13.4⫾20.1 10.5⫾14.6 .004c
Surgery,n(%) 5 (11) 0 8 (11) .525
Intubation,n(%) 10 (23) 6 (43) 30 (42) .088
Prolonged intubation,n(%) 5 (11) 3 (21) 13 (18) .473
Postnatal antibiotic treatment, mean⫾SD, d 7.3⫾12.2 7.9⫾10.2 11.3⫾17.2 .373 Postnatal ampicillin treatment, mean⫾SD, d 6.3⫾9.5 4.9⫾4.5 5.4⫾6.2 .763
BPD,n(%) 5 (12)a 3 (21) 20 (29)b .095
NEC,n(%) 7 (16)a 3 (21) 25 (37)b .059
Meningitis,n(%) 7 (16) 1 (7) 8 (11) .732
UTI,n(%) 3 (7) 2 (14) 11 (15) .368
Death,n(%) 8 (19)a 5 (36) 16 (23)b .441
Sepsis-related death,n(%) 6 (14)a 4 (29) 12 (18)b .463
Continuous data were analyzed with analysis of variance and dichotomous data with the2test or Fisher’s exact test (cell number of⬍5). CVC indicates central vascular catheter; BPD, bronchopulmonary dysplasia; NEC, necrotizing enterocolitis; UTI, urinary tract infection.
aN⫽43. bN⫽68.
used to treat, among other conditions, chorioamnionitis and maternal infections.9,31,32Under these circumstances, the duration of treatment and the antimicrobial drugs of choice are often at the discretion of the treating physi-cian.10Despite the increase in IAP, we did not observe a similar increase in maternal fever and/or chorioamnio-nitis. Although we cannot determine the exact causes of the trends in EOS observed at our institution, the in-crease in antibiotic exposure in the VLBW population and its association with ampicillin-resistantE coliEOS is one concerning finding that warrants additional investi-gation.
The incidence of LOS attributed to any organism and to E coli in the overall and VLBW populations has in-creased at Y-NHH. Although the cause of these trends is again likely to be multifactorial, the increase in LOS did not coincide with an increase in VLBW admissions or duration of hospitalization. It is possible that the in-creased survival rates in this population created a larger cohort of susceptible neonates requiring prolonged sup-port, increasing their risk of exposure and subsequent infection. Unlike the association observed between in-trapartum antibiotic exposure and ampicillin-resistant
E coliEOS in the VLBW population, no such association was observed between postnatal antibiotic exposure and ampicillin-resistant cases of E coli LOS. The consistent use of fewer combinations of drugs for postnatal anti-microbial treatment, for shorter periods, may be partly responsible for this discrepancy.
Our findings exemplify the advantages and disadvan-tages of single-center, observational studies. The longi-tudinal collection of data over nearly 3 decades allowed us to compare changes in EOS, LOS, andE coliampicillin resistance with changes in obstetric and neonatal prac-tice. This lengthy period of observation helped to control for apparent fluctuations in infection rates observed over shorter periods. Conversely, limitations including sam-ple size and exclusion of unknown factors, such as the number of sepsis episodes prevented through intrapar-tum antibiotic use, limit data interpretation. Trends in the overall use of intrapartum antibiotic therapy would have been useful to this investigation but could not be obtained from Y-NHH or from the multiple regional hospitals that refer neonates to our institution. Although we were able to identify trends, the causality of these findings could not be distinguished. We were unable to identify all potential confounding variables, and limited sample size made the integrity and precision of LR anal-yses with the addition of multiple independent variables increasingly unreliable. We recognize these restrictions and the pitfalls related to conclusions derived from as-sociation-based analyses of data lacking a control popu-lation.20
Neonatal sepsis continues to be a significant cause of morbidity and death. Intrapartum antibiotic exposure may contribute to this risk and to the growing problem of antibiotic-resistant organisms. Implementation of reg-imented guidelines for IAP to prevent GBS infection in term and near-term infants has led to a decrease in EOS.6 Strict compliance with these guidelines and ongoing sur-veillance of potential consequences of IAP are essential
components of this continued success.8Widespread, un-restricted, intrapartum antibiotic use, particularly in the preterm population, may have untoward effects. Addi-tional studies to provide a better understanding of the potential risks and benefits of intrapartum antibiotic ex-posure for neonates are therefore needed.
REFERENCES
1. Anthony BF, Okada DM. The emergence of group B strepto-cocci in infections of the newborn infant.Annu Rev Med.1977; 28:355–369
2. American Academy of Pediatrics, Committee on Infectious Dis-eases and Committee on Fetus and Newborn. Guidelines for prevention of group B streptococcal (GBS) infection by che-moprophylaxis.Pediatrics.1992;90(5):775–778
3. Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: a public health per-spective.MMWR Recomm Rep.1996;45(RR-7):1–24
4. American College of Obstetricians and Gynecologists, Commit-tee on Obstetric Practice. ACOG commitCommit-tee opinion: preven-tion of early-onset group B streptococcal disease in newborns.
Int J Gynaecol Obstet.1996;54(2):197–205
5. Schrag S, Gorwitz R, Fultz-Butts K, Schuchat A. Prevention of perinatal group B streptococcal disease: revised guidelines from CDC.MMWR Recomm Rep.2002;51(RR-11):1–22
6. Centers for Disease Control and Prevention. Early-onset group B streptococcal disease: United States, 1998 –1999. MMWR Morb Mortal Wkly Rep.2000;49(35):793–796
7. Stoll BJ, Hansen N. Infections in VLBW infants: studies from the NICHD Neonatal Research Network.Semin Perinatol.2003; 27(4):293–301
8. Gilbert GL, Hewitt MC, Turner CM, Leeder SR. Compliance with protocols for prevention of neonatal group B streptococcal sepsis: practicalities and limitations. Infect Dis Obstet Gynecol.
2003;11(1):1–9
9. American College of Obstetricians and Gynecologists. ACOG practice bulletin: premature rupture of membranes: clinical management guidelines for obstetrician-gynecologists. Int J Gynaecol Obstet.1998;63(1):75– 84
10. Norman K, Pattinson RC, de Souza J, et al. Ampicillin and metronidazole treatment in preterm labour: a multicentre, ran-domised, controlled trial. Br J Obstet Gynaecol. 1994;101(5): 404 – 408
11. Kotloff KL, Blackmon LR, Tenney JH, et al. Nosocomial sepsis in the neonatal intensive care unit.South Med J.1989;82(6): 699 –704
12. Andrews JI, Diekema DJ, Hunter SK, et al. Group B strepto-cocci causing neonatal bloodstream infection: antimicrobial susceptibility and serotyping results from SENTRY centers in the Western Hemisphere. Am J Obstet Gynecol. 2000;183(4): 859 – 862
13. Bland ML, Vermillion ST, Soper DE, et al. Antibiotic resistance patterns of group B streptococci in late third-trimester recto-vaginal cultures.Am J Obstet Gynecol.2001;184(6):1125–1126 14. Lin FY, Azimi PH, Weisman LE, et al. Antibiotic susceptibility
profiles for group B streptococci isolated from neonates, 1995–1998.Clin Infect Dis.2000;31(1):76 –79
15. Stoll BJ, Hansen N, Fanaroff AA, et al. Changes in the patho-gens causing early-onset sepsis in very-low-birth-weight in-fants.N Engl J Med.2002;347(4):240 –247
16. Baltimore RS, Huie SM, Meek JI, et al. Early-onset sepsis in the era of group B streptococcal prevention. Pediatrics. 2001; 108(5):1094 –1098
18. Towers CV, Briggs GC. Antepartum use of antibiotics and early-onset neonatal sepsis: the next 4 years.Am J Obstet Gy-necol.2002;187(2):495–500
19. Hyde TB, Hilger TM, Reingold A, Farley MM, O’Brien KL, Schuchat A. Trends in incidence and antimicrobial resistance of early-onset sepsis: population-based surveillance in San Fran-cisco and Atlanta.Pediatrics.2002;110(4):690 – 695
20. Moore MR, Schrag SJ, Schuchat A. Effects of intrapartum antimicrobial prophylaxis for prevention of group-B-strepto-coccal disease on the incidence and ecology of early onset neonatal sepsis.Lancet Infect Dis.2003;3(4):201–213
21. Stoll BJ, Hansen NI, Higgins RD, et al. Very low birth weight preterm infants with early onset neonatal sepsis.Pediatr Infect Dis J.2005;24(7):635– 639
22. Stoll BJ, Hansen N, Fanaroff AA, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neo-natal Research Network.Pediatrics.2002;110(2):285–291 23. Jones B, Peake K, Morris AJ, et al.Escherichia coli: a growing
problem in early onset neonatal sepsis.Aust N Z J Obstet Gynae-col.2004;44(6):558 –561
24. Schuchat A, Zywicki SS, Dinsmoor MJ, et al. Risk factors and opportunities for the prevention of early-onset neonatal sepsis: a multicenter case-control study.Pediatrics.2000;105(1):21–26 25. Bizzarro MJ, Raskind C, Baltimore RS, et al. Seventy-five years of neonatal sepsis at Yale: 1928 –2003.Pediatrics.2005;116(3): 595– 602
26. Shah SS, Ehrenkranz RA, Gallagher PG. Increasing incidence of Gram-negative rod bacteremia in a newborn intensive care unit.Pediatr Infect Dis J.1999;18(7):591–595
27. National Committee for Clinical Laboratory Standards. Perfor-mance Standards for Antimicrobial Disk Susceptibility Tests. 5th ed. Approved Standard M2-A5. Villanova, PA: National Commit-tee for Clinical Laboratory Standards; 1993
28. Speer ME, Taber LH, Yow MD, et al. Culminant neonatal sepsis and necrotizing enterocolitis associated with a “nonentero-pathogenetic” strain of Escherichia coli. J Pediatr.1976;89(1): 91–95
29. Krediet TG, van Lelyveld N, Vijlbrief DC, et al. Microbiologic factors associated with neonatal necrotizing enterocolitis: pro-tective effect of early antibiotic treatment.Acta Paediatr.2003; 92(10):1180 –1182
30. Fanaroff AA, Korones SB, Wright LL, et al. Incidence, present-ing features, risk factors, and significance of late onset septice-mia in very low birth weight infants.Pediatr Infect Dis J.1998; 17(7):593–598
31. Mercer BM, Moronic M, Thurgau GR, et al. Antibiotic therapy for reduction of infant morbidity after preterm premature rup-ture of membranes: a randomized controlled trial.JAMA.1997; 278(12):989 –995
32. Klein LL, Gibbs RS. Use of microbial cultures and antibiotics in the prevention of infection-associated preterm birth. Am J Obstet Gynecol.2004;190(6):1493–1502
FOR FDA, A MAJOR BACKLOG OVERSEAS
“Washington—The Food and Drug Administration is so understaffed that, at its current pace, the agency would need at least 27 years to inspect every foreign medical device plant that exports to the United States, 13 years to check every foreign drug plant and 1900 years to examine every foreign food plant, according to government investigators. Computer systems at the drug agency are so inadequate that it can only guess the number of the plants, and it cannot produce a list of those that have not been inspected. The situation is particularly dire in China, which has more drug and device plants than any other foreign nation but where FDA inspections are few. These findings come from a series of reports by the Government Accountability Office— obtained by theNew York Times—scheduled to be released Tuesday at a hearing of the House Energy and Commerce Committee. The reports and a recent assess-ment by the agency’s Science Board conclude that the FDA is so over-whelmed by a flood of imports that it is incapable of protecting the public from unsafe drugs, medical devices and food. The backlog of inspections is even more profound among foreign medical device and food plants. Over a six-year period, the agency inspected 64 of the nearly 700 medical device plants registered in China. Medical devices can include items like stents and spinal screws.”
Harris G.New York Times. January 29, 2008
DOI: 10.1542/peds.2007-2171
2008;121;689
Pediatrics
Gallagher
Matthew J. Bizzarro, Louise-Marie Dembry, Robert S. Baltimore and Patrick G.
in the Era of Intrapartum Antibiotic Prophylaxis
Sepsis and Ampicillin Resistance
Escherichia coli
Changing Patterns in Neonatal
Services
Updated Information &
http://pediatrics.aappublications.org/content/121/4/689
including high resolution figures, can be found at:
References
http://pediatrics.aappublications.org/content/121/4/689#BIBL
This article cites 31 articles, 6 of which you can access for free at:
Subspecialty Collections
b
http://www.aappublications.org/cgi/collection/infectious_diseases_su Infectious Disease
http://www.aappublications.org/cgi/collection/neonatology_sub Neonatology
following collection(s):
This article, along with others on similar topics, appears in the
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml
in its entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or
Reprints
http://www.aappublications.org/site/misc/reprints.xhtml
DOI: 10.1542/peds.2007-2171
2008;121;689
Pediatrics
Gallagher
Matthew J. Bizzarro, Louise-Marie Dembry, Robert S. Baltimore and Patrick G.
in the Era of Intrapartum Antibiotic Prophylaxis
Sepsis and Ampicillin Resistance
Escherichia coli
Changing Patterns in Neonatal
http://pediatrics.aappublications.org/content/121/4/689
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.