Key words
newborn - Bacterial distribution - Bacterial resistance - National bacterial resistance monitoring network
Newborns are prone to nosocomial infections because of their immature immune systems,
and the risk increases with many factors such as preterm birth, low birth weight,
vertical mother-to-infant transmission, perinatal infection, and the unreasonable
use of antibiotics [1]
[2]. In recent years, with the development of
medical technology, the rescue success rate of premature infants and
low-birth-weight infants has increased. As an important place for rescuing and
treating newborns, the neonatal intensive care unit (NICU) has increased the risk of
nosocomial infection outbreaks owing to the increase of invasive operations, the
extension of hospital stay, and the application of broad-spectrum antibiotics.
Studies have shown that the incidence of nosocomial infection in the NICU is
26.05%, and the location is mainly in the blood and lower respiratory tract
[3]. Studies have also shown that neonatal
sepsis is an important cause of death of children and newborns. Neonatal sepsis
accounts for 7% of children’s deaths and 16% of neonatal
deaths [4].
It is therefore very important to identify the relevant pathogens and assess their
drug resistance to reduce the rate of neonatal infections. In this study, we
statistically analyzed infection-related factors, bacterial distribution, and drug
resistance in neonates treated at our hospital from January 2020 to June 2021. We
compared the findings with the data for children and newborns from the national
bacterial resistance surveillance report to provide a scientific basis for clinical
formulation and evaluation of antimicrobial management policies.
Materials and Methods
Clinical samples
Clinical samples, including blood, sputum, cerebrospinal fluid, nasopharyngeal
swab, and secretion samples, were taken from children admitted to the
neonatology department of our hospital. For the analysis, the neonates were
classified as having early-onset (within three days of birth) or late-onset
(more than three days after birth) infection, being premature (born
at<37 gestational weeks) or term (born at≥37 gestational weeks
and<42 gestational weeks), and having low birth weight
(weight<2500 g) or normal birth weight
(weight≥2500 g).
Bacterial identification and drug resistance testing
Bacterial identification [5] and drug
resistance testing were carried out using a French bioMérieux VITEK 2
Compact automatic bacterial culture identification instrument and its supporting
drug susceptibility card. Antibacterial drug sensitivity was defined according
to CLSI M100–2019.
National drug resistance surveillance data of children and newborn patients
[6]
These data were obtained from “Research on Bacterial Resistance
Surveillance in Children and Newborn Patients in China from 2014 to
2017” Chinese Medical Journal, Vol. 98, No. 40, October 30, 2018.
Statistical analysis
WHONET5.6 and SPSS22.0 were used to analyze the data. Data are expressed as
percentages, and the chi-square test was used for between-group comparisons.
When the frequency of a listed item was<1, Fisher’s exact
probability method was used to compare the data. Statistically significant
differences were defined as P<0.05.
Results
Bacterial survey
Patient characteristics
From January 2020 to June 2021, the Department of Neonatology at our hospital
requested bacterial culture analysis of 4,572 samples from neonates. A total
of 209 strains were isolated, for a positive culture rate of 4.57%.
Among the 209 culture-positive patients, 14 died and 195 were discharged
healthy. Of the 14 deaths, 8 were ultra-premature infants due to respiratory
failure, and 6 were due to neonatal sepsis. In all, 119 were male
(56.9%) and 90 were female (43.1%). The minimum age was 1
day, the maximum age was 3 months, and the average age was 19.3 days. There
were 81 cases of vaginal delivery (38.8%) and 128 cases of cesarean
section (61.2%); 29 cases (13.9%) had early onset of
infection, and 180 cases (86.1%) had late onset of infection; 184 of
the infants (88.0%) were born prematurely, and 25 (12.0%)
were full-term infants. One hundred and seventy-seven infants
(84.7%) had low birth weight, and 32 infants (15.3%) had
normal birth weight ([Table 1]).
Table 1 Basic characteristics of the 209
culture-positive neonates.
|
Project
|
Classification
|
Number of cases
|
Proportion%
|
|
Sex
|
Male
|
119
|
56.9
|
|
Female
|
90
|
43.0
|
|
Delivery method
|
Vaginal delivery
|
81
|
38.8
|
|
Cesarean section
|
128
|
61.2
|
|
Onset time
|
Early onset
|
29
|
13.9
|
|
Late onset
|
180
|
86.1
|
|
Gestational age
|
Premature baby
|
184
|
88.0
|
|
Full-term child
|
25
|
12.0
|
|
Birth weight
|
Low birth weight
|
177
|
84.7
|
|
|
Non-low birth weight
|
32
|
15.3
|
Specimen typesh
The 209 culture-positive cases were identified from 97 blood samples, 64 were
from sputum samples, 15 were from cerebrospinal fluid samples, 25 were from
nasopharyngeal swabs, and 8 were from secretions.
Bacterial types and distribution
Among the 209 bacterial isolates, 90 (43.1%) were gram-positive. The
most commonly isolated bacteria were coagulase-negative Staphylococcus
(42.2%), Staphylococcus aureus (32.1%), Enterococcus
(13.3%), and Streptococcus agalactiae (4.5%). Data from the
children and newborn groups of the national drug resistance surveillance
report indicate [1] that the top
gram-positive bacteria are mainly Staphylococcus aureus (35.6%),
Streptococcus pneumoniae (27.4%), coagulase-negative Staphylococcus
(21.1%), and Enterococcus faecium (4.4%) ([Table 2]).
Table 2 Distribution of the four most common
gram-positive bacterial isolates.
|
Our hospital
|
CARSS
|
|
Bacteria species
|
Ratio (%)
|
Bacteria species
|
Ratio (%)
|
|
1
|
Coagulase-negative Staphylococcus
|
42.2
|
Staphylococcus aureus
|
35.6
|
|
2
|
Staphylococcus aureus
|
32.1
|
Streptococcus pneumoniae
|
27.4
|
|
3
|
Enterococcus
|
13.3
|
Coagulase-negative Staphylococcus
|
21.1
|
|
4
|
Streptococcus agalactiae
|
4.5
|
Enterococcus faecium
|
4.4
|
Gram-negative bacteria were isolated from 119 cases (56.9%). Among
them, the five most common isolates were Klebsiella pneumonia
(56.3%), Acinetobacter baumannii (15.1%), Enterobacter
aerogenes (7.6%), Enterobacter cloacae (5.9%), and Serratia
marcescens (5.9%). Data from the National Drug Resistance
Surveillance Children and Newborn Group report show that the top five
gram-negative bacteria are Escherichia coli (26.8%), Klebsiella
pneumoniae (16.8%), Haemophilus influenzae (15.5%),
Pseudomonas aeruginosa (6.4%), and Acinetobacter baumannii
(6.2%) ([Table 3]).
Table 3 Distribution of the five most common
gram-negative bacterial isolates.
|
Rank
|
Our hospital
|
CARSS
|
|
Bacteria species
|
Ratio (%)
|
Bacteria species
|
Ratio (%)
|
|
1
|
Klebsiella pneumoniae
|
56.3
|
Escherichia coli
|
26.8
|
|
2
|
Acinetobacter baumannii
|
15.1
|
Klebsiella pneumoniae
|
16.8
|
|
3
|
Enterobacter aerogenes
|
7.6
|
Haemophilus influenzae
|
15.5
|
|
4
|
Enterobacter cloacae
|
5.9
|
Pseudomonas aeruginosa
|
6.4
|
|
5
|
Serratia marcescens
|
5.9
|
Acinetobacter baumannii
|
6.2
|
Distribution of pathogenic bacteria in children with early- or late-onset
infection
In total, 27 pathogenic bacteria, seven gram-negative bacteria, and 20
gram-positive bacteria were detected in neonates with early onset of
infection. Also, 182 pathogenic bacteria, 112 gram-negative bacteria, and 70
gram-positive bacteria were detected in the late-onset group. The pathogenic
bacteria distribution between the two groups was not statistically
significant, except for Klebsiella pneumoniae and Streptococcus agalactiae
(P<0.05).
Distribution of pathogenic bacteria in children with different
gestational ages
In total, 184 strains of pathogenic bacteria, 108 gram-negative strains, and
76 gram-positive strains were detected in preterm infants, and 25 strains of
pathogenic bacteria, 11 gram-negative strains, and 14 gram-positive strains,
were detected in the term infant group. The pathogenic bacteria distribution
between the two groups was not statistically significant, except for
Streptococcus agalactiae (P<0.05).
Distribution of pathogenic bacteria in children with different birth
weights
In total, 172 pathogenic bacteria, 100 gram-negative strains and 72
gram-positive strains were detected in children with low birth weight; and
27 pathogenic bacteria, nine gram-negative bacteria and 18 gram-positive
bacteria were detected in children with normal body weight. The pathogenic
bacteria distribution between the two groups was not statistically
significant, except for Streptococcus agalactiae (P<0.05).
Comparison of EOS and LOS infections in preterm and term infants
We conducted a stratified analysis of EOS and LOS in preterm and term
infants. We found that the number of LOS cases in preterm infants was much
larger than that in other categories, with a P-value of<0.01,
indicating a statistically significant difference. We found that premature
infants require long-term parenteral nutrition and long, invasive procedures
owing to immature organ development, poor skin and mucosal barrier function,
and low humoral and cellular immunity, which are risk factors for LOS. Some
studies have found that hospitalization stays≥10 days is a risk
factor for nosocomial infection [3],
and the results of this study are consistent with it ([Table 4]).
Table 4 Comparison of bacterial infection in preterm
infants and term infants.
|
Premature baby (179)
|
Full-term baby (20)
|
P
|
|
EOS
|
LOS
|
EOS
|
LOS
|
|
|
Number of infected cases
|
14
|
165
|
10
|
10
|
<0.01
|
|
The top three in the distribution
|
Staphylococcus aureus
|
Klebsiella pneumoniae
|
Streptococcus agalactiae
|
Klebsiella pneumoniae
|
|
|
Klebsiella pneumoniae
|
Staphylococcus aureus
|
Staphylococcus epidermidis
|
Staphylococcus aureus
|
|
|
Enterococcus faecium
|
Staphylococcus epidermidis
|
Enterococcus faecalis
|
Acinetobacter baumannii
|
|
Antimicrobial resistance of the main bacterial isolates
Staphylococcus aureus resistance
In this study, 35 strains of Staphylococcus aureus were isolated, of
which 22 (62.9%) were methicillin-resistant Staphylococcus
aureus (MRSA). The 22 MRSA strains were sensitive to linezolid,
vancomycin, rifampicin, levofloxacin, and gentamicin and were resistant
to compound trimethoprim, clindamycin, and erythromycin ([Table 5]). The rate of methicillin
resistance that we observed in our hospital was lower than that reported
at the national level.
Table 5 Comparison of MRSA resistance profiles of
isolates in our hospital with resistance rates reported by
the National Anti-Drug Surveillance Network.
|
Drug name
|
Our hospital
|
CARSS
|
|
R
|
S
|
R
|
S
|
|
Linezolid
|
0
|
100
|
0
|
100.0
|
|
Vancomycin
|
0
|
100
|
0
|
100.0
|
|
Rifampin
|
0
|
100
|
4.3
|
88.8
|
|
Levofloxacin
|
0
|
100
|
6.4
|
92.3
|
|
Gentamicin
|
0
|
100
|
6.8
|
92.0
|
|
Compound trimethoprim
|
4.5
|
95.5
|
9
|
90.9
|
|
Clindamycin
|
36.4
|
63.6
|
63.7
|
35.4
|
|
Erythromycin
|
36.4
|
63.6
|
83.3
|
15.7
|
Streptococcus agalactiae resistance
Five Streptococcus agalactiae strains were isolated in this study. The
strains were sensitive to most antibacterial drugs, with a resistance
rate of 100% to clindamycin and 60% to tetracycline.
Klebsiella pneumoniae resistance
In total, 67 Klebsiella pneumoniae strains were isolated in this study,
and the drug sensitivity results ([Table 6]) showed that the rates of sensitivity to amikacin,
ertapenem, and imipenem were all 100%.
Table 6 Resistance profiles of 67 Klebsiella
pneumoniae strains.
|
67 Klebsiella pneumoniae
|
Our hospital
|
CARSS
|
|
R
|
S
|
R
|
S
|
|
Amikacin
|
0
|
100
|
3.2
|
96.5
|
|
Levofloxacin
|
14.9
|
85.1
|
5.6
|
92.9
|
|
Ertapenem
|
0
|
100
|
7.7
|
92.7
|
|
Imipenem
|
0
|
100
|
8.8
|
89.9
|
|
Ciprofloxacin
|
17.9
|
82.1
|
8.7
|
86.9
|
|
Piperacillin/ Tazobactam
|
7.5
|
92.5
|
13.2
|
83.5
|
|
Gentamicin
|
4.5
|
95.5
|
19.7
|
79.7
|
|
Compound trimethoprim
|
35.8
|
64.2
|
32.5
|
67.4
|
|
Cefepime
|
22.3
|
77.7
|
27
|
66.6
|
|
Aztreonam
|
37.3
|
62.7
|
32.4
|
66.3
|
|
Ceftazidime
|
28.3
|
71.7
|
31
|
65.4
|
|
Ceftriaxone
|
34.3
|
65.7
|
45.7
|
53.5
|
|
Ampicillin/Sulbactam
|
32.8
|
67.2
|
46.8
|
46.7
|
|
Cefazolin
|
64.1
|
35.9
|
55.4
|
33.8
|
Acinetobacter baumannii resistance
Eighteen strains of Acinetobacter baumannii were isolated in this study.
The drug sensitivity results were compared with national drug resistance
monitoring data ([Table 7]). One
hundred percent of the bacterial isolates from our hospital were
sensitive to tobramycin, imipenem, and gentamicin.
Table 7 Resistance profiles of 18 Acinetobacter
baumannii strains.
|
18 Acinetobacter baumannii
|
Our hospital
|
CARSS
|
|
R
|
S
|
R
|
S
|
|
Tobramycin
|
0
|
100
|
19.1
|
79.7
|
|
Levofloxacin
|
5.6
|
94.4
|
15
|
79.3
|
|
Ciprofloxacin
|
5.6
|
94.4
|
22.3
|
77
|
|
Imipenem
|
0
|
100
|
24.6
|
74.7
|
|
Gentamicin
|
0
|
100
|
23.5
|
74.4
|
|
Cefepime
|
12.5
|
87.5
|
25.7
|
72.4
|
|
Ampicillin/Sulbactam
|
16.7
|
83.3
|
24.8
|
72.1
|
|
Ceftazidime
|
22.2
|
77.8
|
25.2
|
69.6
|
Discussion
Neonatal infections, especially the likes of neonatal sepsis, pneumonia, or
meningitis, are major diseases that threaten newborn lives. Early diagnosis of these
diseases is difficult. Bacterial culture is the “gold standard” for
diagnosis. Strain identification and drug sensitivity tests provide a scientific
basis for guiding rational drug use and controlling infection [7].
This study shows that, from January 2020 to June 2021, the rate of positive bacterial
cultures in the samples taken by the neonatology department of our hospital was
4.35%, which is lower than that reported in the national drug resistance
monitoring data. This is most likely because of regional differences. Most of the
infections detected in newborns delivered at our hospital were acquired in the
intrauterine environment; the second most common cause of nosocomial infections was
prolonged hospitalization. In contrast, most of the children and neonates reported
by the National Drug Resistance Surveillance Network exhibited community-acquired
infections. Furthermore, gram-positive bacteria accounted for 43.1% and
gram-negative bacteria for 56.9% of isolates in this study, virtually the
same as the 45.5% gram-positive and 54.5% gram-negative rates
reported by the national drug resistance monitoring network.
This study conducted a stratified analysis of EOS and LOS in preterm infants and term
infants, finding that the number of LOS cases in preterm infants was much greater
than the number of cases in other categories, with a P-value of<0.01,
indicating a statistically significant difference. A possible explanation for this
is that premature infants and low birth weight infants undergo more invasive
procedures, such as intravenous catheterization and mechanical respiratory support
during medical treatment, are treated for a longer time, and receive more
antibiotics, increasing their risk of nosocomial infection. This is consistent with
the fact that the rate of late-onset infection is significantly higher than that of
early-onset infection in our study. According to expert consensus on the diagnosis
and management of neonatal sepsis (version 2019), EOS patients were treated with a
broad-spectrum combination of antibiotics before blood culture and other
non-specific test results. Ampicillin (or penicillin) and third-generation
cephalosporin were used as the first-line antimicrobial combination for
gram-positive (G+) and gram-negative (G-) bacteria as early as possible. LOS
patients were treated with phenazacillin and nafcillin for staphylococcus
epidermidis or vancomycin instead of ampicillin combined with third-generation
cephalosporin. The follow-up treatment plan should be adjusted according to the
results of the drug susceptibility test, and in principle, priority is given to
treatment with antibiotics alone rather than in combination. If the antibiotics
selected are not empirically in the range of the drug susceptibility test and the
clinical effect is good, they will continue to be used; otherwise, they will be
changed to the sensitive antibiotics according to the drug susceptibility test
results.
Our study shows that Streptococcus agalactiae (GBS) is one of the top four
gram-positive bacteria isolated at our hospital. This is consistent with the fact
that our hospital is a specialist hospital for obstetrics and gynecology, given that
GBS normally resides in the vagina and intestines, and newborns can acquire the
infection vertically from the mother. Early-onset GBS infections primarily cause
pneumonia, meningitis, and sepsis [8]
[9]. Studies [10]
[11] have shown that GBS is
highly sensitive to penicillin, ampicillin, cephalosporins, and vancomycin, and the
drug sensitivity tests that we performed yielded similar results. Therefore, for
children with early-onset GBS infection, penicillin is the first choice for
treatment; for those with mild or severe allergies to penicillin, cefazolin or
clindamycin can be used, respectively; and for those who are resistant to
clindamycin, vancomycin should be used. Doctors perform skin sensitivity tests on
patients before they are given penicillin drugs to prevent allergic reactions to
penicillin. If the test is negative, penicillin is given to the patient; otherwise,
it is prohibited for those who are positive. Of the 209 patients in this study, 20
did not undergo the penicillin skin sensitivity test, so the results are unknown.
The skin test results of the remaining 189 patients were negative. Among the
gram-negative bacteria isolated in this study, 18 strains were Acinetobacter
baumannii. This is likely because of the frequent use of mechanical ventilation in
neonatal intensive care, which increases the incidence of ventilator-associated
pneumonia. Acinetobacter baumannii is the most common pathogen responsible for
ventilator-associated pneumonia [12]
[13].
Two of the top five gram-positive bacteria and gram-negative bacteria reported by the
national drug resistance surveillance data, Streptococcus pneumoniae and Haemophilus
influenzae, were not identified in this study. This may be related to differences in
the study populations because these bacteria tend to circulate in the community and
are not often the main cause of bacterial infection in neonatal patients.
In this study, 67 strains of Klebsiella pneumoniae were detected, of which 58 strains
were extended-spectrum β-lactamase (ESBL) strains. Strains can acquire ESBL
activity through bacterial plasmids, enabling them to hydrolyze broad-spectrum
penicillins, cephalosporins, and monocyclic antibiotics (aztreonam), leading to
increased drug resistance [14]. The drug
sensitivity results showed a sensitivity rate of 100% to all antimicrobial
agents tested, except for amikacin, ertapenem, and imipenem, to which the isolates
exhibited varying degrees of resistance.
In summary, newborns are susceptible to a wide range of bacterial infections and
complex risk factors. There are differences in bacterial distribution and drug
resistance in different regions and age groups. Therefore, understanding the
distribution and drug resistance of pathogens in our hospital is greatly significant
for guiding the rational selection of antibiotics in clinical practice and reducing
neonatal mortality and nosocomial infections.
