Maternal-Neonatal Reports

Beyond Early- and Late-onset Neonatal Sepsis Definitions

What are the Current Causes of Neonatal Sepsis Globally?
A Systematic Review and Meta-analysis of the Evidence
Michelle L. Harrison , MPH,*† Benjamin F.R. Dickson, MPH,*† Mike Sharland , FRCPCH,‡ and
Phoebe C.M. Williams , DPhil*†§¶

Abstract: Sepsis remains a leading cause of neonatal mortality, particu-

larly in low- and lower-middle-income countries (LLMIC). In the con-
text of rising antimicrobial resistance, the etiology of neonatal sepsis is
T he global child mortality rate is declining, yet mortality remains
high for the vulnerable neonatal population, with 47% of deaths
in children under 5 occurring in the first 28 days of life.1,2 The
evolving, potentially making currently-recommended empirical treatment third greatest contributor to neonatal mortality is neonatal infec-
guidelines less effective. We performed a systematic review and meta-­ tion, including sepsis and meningitis, resulting in up to 570,000
analysis to evaluate the contemporary bacterial pathogens responsible for sepsis-attributable deaths each year.3,4
early-onset sepsis (EOS) and late-onset neonatal sepsis (LOS) to ascertain The United Nations Sustainable Development Goals have
if historical classifications—that guide empirical therapy recommendations called for the end of preventable child deaths by 2030.5 The success
based on assumptions around causative pathogens—may be outdated. We of this goal is highly dependent on the effective treatment and pre-
analyzed 48 articles incorporating 757,427 blood and cerebrospinal fluid vention of neonatal infections. In the context of burgeoning global
samples collected from 311,359 neonates across 25 countries, to evaluate antimicrobial resistance (AMR), understanding the evolving epide-
4347 significant bacteria in a random-effects meta-analysis. This revealed miology and etiology of neonatal sepsis is paramount.6
Gram-negative bacteria were now the predominant cause of both EOS (53%, It has long been considered that early-onset neonatal sep-
2301/4347) and LOS (71%, 2765/3894) globally. In LLMICs, the predom- sis (EOS)—that is, sepsis occurring within the first 72 hours after
inant cause of EOS was Klebsiella spp. (31.7%, 95% CI: 24.1–39.7%) fol- birth—is predominantly caused by vertical transmission of bac-
lowed by Staphylococcus aureus (17.5%, 95% CI: 8.5 to 28.4%), in marked teria (typically Streptococcus agalactiae and Escherichia coli)
contrast to the Streptococcus agalactiae burden seen in high-income health- from mother to infant, either in utero or during birth7,8 Conversely,
care settings. Our results reveal clear evidence that the current definitions of late-onset neonatal sepsis (LOS), occurring at >72 hours of life, is
EOS and LOS sepsis are outdated, particularly in LLMICs. These outdated considered to be acquired by nosocomial or community sources of
definitions may be guiding inappropriate empirical antibiotic prescribing infection.7,8 However, recent studies have reported a potential epi-
that inadequately covers the causative pathogens responsible for neonatal demiological shift in the causative pathogens responsible for EOS,
sepsis globally. Harmonizing sepsis definitions across neonates, children with an increase in the proportion of multidrug-resistant (MDR)
and adults will enable a more acurate comparison of the epidemiology of Gram-negative bacteria responsible for systemic neonatal infections
sepsis in each age group and will enhance knowledge regarding the true evident from the first day of life, accompanied by concerning AMR
morbidity and mortality burden of neonatal sepsis. profiles previously evident in infants with hospital-acquired LOS.9,10
The World Health Organization (WHO) suggests the early
Key Words: neonatal sepsis, early-onset sepsis, late-onset sepsis,
administration of empirical antibiotics following a clinical diagno-
bacteremia, antimicrobial resistance
sis of neonatal sepsis, which is not defined in its timing of acquisi-
(Pediatr Infect Dis J 2024;43:1182–1190) tion in their guidelines (by EOS or LOS),11 but rather recommends
ampicillin/benzylpenicillin and gentamicin as first-line treatment
regimens, and third-generation cephalosporins as an alternative
agent, for treating neonatal sepsis.11 However, these empirical rec-
Accepted for publication June 17, 2024 ommendations are based on data on the presumed causative path-
*From the School of Public Health, Faculty of Medicine, University of Sydney, ogens of neonatal sepsis that are largely derived from high-income
Sydney, Australia; †Sydney Infectious Diseases Institute, The University of
Sydney, Australia; ‡Infection and Immunity Research Institute, St Georges
countries (HIC).12 However, with 98% of the neonatal sepsis
University, London, England; §Department of Infectious Diseases, Sydney mortality burden arising from low- and middle-income countries
Children’s Hospital Network, NSW, Australia; and ¶School of Women and (LMICs),13 and with mounting evidence to suggest differing etiolo-
Children’s Health, University of NSW, NSW, Australia gies of the bacterial pathogens causing neonatal infection between
P.C.M.W. is supported by an NHMRC-funded grant 1197335 (“Tackling antimi-
crobial resistance in neonatal sepsis: Australia and Beyond”).
high- and low-resourced healthcare settings,14,15 improving the
The authors have no conflicts of interest to disclose. understanding of the causative pathogens driving neonatal sepsis
Supplemental digital content is available for this article. Direct URL citations is essential.
appear in the printed text and are provided in the HTML and PDF versions of A growing body of evidence suggests the WHO-recommended
this article on the journal’s website (www.pidj.com).
Address for correspondence: Phoebe C.M. Williams, DPhil, Department of
neonatal sepsis antibiotic regimens may be becoming redundant, as
Infectious Diseases, School of Women and Children's Health, University the contemporary causative bacteria responsible for neonatal sepsis
of NSW and Sydney Children's Hospital Network, NSW 2031 Australia. are increasingly less susceptible to the current empirical antibi-
E-mail: phoebe.williams@unsw.edu.au. otic regimens in many settings.13,16 Recent epidemiological stud-
Copyright © 2024 The Author(s). Published by Wolters Kluwer Health, Inc.
This is an open-access article distributed under the terms of the Creative
ies across multiple LMICs suggest poor coverage is provided by
Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY- the currently-recommended empirical regimens, resulting in high
NC-ND), where it is permissible to download and share the work provided it rates of divergent empirical antibiotic prescribing across clinical
is properly cited. The work cannot be changed in any way or used commer- settings, which may propagate AMR.13,17,18
cially without permission from the journal.
ISSN: 0891-3668/24/4312-11821190
An increasing prevalence of AMR is evident in the
DOI: 10.1097/INF.0000000000004485 Gram-negative bacteria causative of neonatal sepsis, with rates of

1182 | www.pidj.com The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024 Current Causes of Neonatal Sepsis

nonsusceptibility of up to 97% (to ampicillin) and 70% (to gen- calculated after stratifying the income status of the study setting.
tamicin) reported.9,12,15,18,19 In fact, 570,000 sepsis-attributable Analyses were performed in Stata version 16 (Stata Corporation,
deaths occur in neonates each year, contributed by a lack of effi- TX, USA) using the metaprop command.21 A random-effects model
cacy to currently available and recommended antibiotics.3,4 Con- was used to account for the expected heterogeneity between study
sequently, many physicians, particularly in LMICs (where the populations, with weighting completed via DerSimonian and Laird
burden of AMR and neonatal mortality is highest),13 are no longer methods.22 Study heterogeneity between EOS/LOS subgroups was
following empirical therapeutic recommendations due to the reali- analyzed using the I2 statistic,22 and χ2 tests were used to evaluate
zation that current empirical treatment guidelines are unlikely to be the strength of evidence for heterogeneity (with a P value of <0.05
efficacious.15,17 Despite this, there is a concerning dearth of quality considered statistically significant). Publication bias was assessed
evidence to support new empirical regimens in neonates, and this using the Luis Furuya–Kanamori index of the Doi plot.23 A double
warrants global attention. arcsine transformation of effect size was used for the Doi plot to
To reduce unnecessary deaths in neonates, it is essential that stabilize the variance of the proportions.
empirical treatment regimens closely align with the contemporary
etiology of neonatal sepsis. Based on the available evidence, we
hypothesized that the classical bacterial pathogens presumed to RESULTS
cause EOS and LOS are evolving, and current empirical treatment The systematic review of the literature identified 8699
guidelines based upon these definitions may be guiding ineffica- records eligible for inclusion (Fig. 1). Before the screening, 2620
cious therapy. duplicates were removed, leaving 6079 records for review. Title and
As the published evidence suggests the causes of neonatal abstract screening excluded a further 5901 studies (Table, Supple-
sepsis may differ between resource-constrained and resource-replete mental Digital Content 4, http://links.lww.com/INF/F654) resulting
healthcare settings, we further aimed to evaluate differences in the in 178 for full-text analysis. Investigators were unable to retrieve 2
etiology of neonatal sepsis in HICs and low- and lower-middle- full-text references after attempting to contact the authors, and these
income countries (LLMICs). We suggest new strategies for defining were subsequently excluded. Therefore, 176 full-text studies were
the causative bacteria responsible for the burden of neonatal sepsis assessed for eligibility and ultimately, 48 were included for analysis.
globally, to emphasize the need to evaluate alternative empirical Of the included studies 29 were prospective studies,13,24–51
antibiotic regimens that may better target the contemporary epide- and 19 involved retrospective data collection.52–70 All studies were
miology of neonatal sepsis, to reduce its unacceptable morbidity conducted in hospital settings, with only 1 paper assessing neonates
and mortality burden globally. in the community.38 Six studies provided moderate quality evi-
dence (GRADE level B),13,28,32,36,38,44 15 were low quality (GRADE
level C)27,30,31,33,35,41,42,46,48,49,51,53,55,61,68 and 27 were very low quality
METHODS (GRADE level D).24–26,29,34,37,40,43,45,47,50,52,54,56–60,62–67,69,70 (Figure, Sup-
We followed preferred reporting items for systematic reviews plemental Digital Content 5, http://links.lww.com/INF/F655)).
and meta-analysis guidelines to conduct systematic searches in The included articles incorporated 757,427 blood and cere-
EMBASE, MEDLINE and Global Health databases using search brospinal fluid samples collected from 311,359 neonates across 25
terms comprised of both MESH terms and keywords relating to countries. Pathogens that may be potential contaminants [including
EOS and LOS (Material, Supplemental Digital Content 1, http:// coagulase-negative Staphylococci (CoNS), Streptococcus viridans
links.lww.com/INF/F651). Results were limited to human studies and other Streptococcus spp., Stenotrophomonas maltophilia and
with primary data published in English between January 2017 and Burkholderia cepacia], were removed from the analysis as clinical
March 2022 to capture contemporaneous antimicrobial susceptibil- data was not always included in the retrieved articles. Of the 10,150
ity profiles and bacteria causative of neonatal sepsis. significant bacteria identified, 4358 were isolated from neonates
Predefined inclusion and exclusion criteria were established with EOS, and 3894 bacterial pathogens were isolated from infants
to assess study eligibility (Table, Supplemental Digital Content 2, with LOS (Table 1).
http://links.lww.com/INF/F652). Studies were excluded if they did The data were unevenly distributed across the WHO
not clearly define the timing of neonatal sepsis if they analyzed only regions.71 Fourteen studies were conducted in World Bank
a single pathogen, or if they included only high-risk populations defined72 HIC,28,30–32,36,42,44–46,56,58,61,62,64 7 in upper-middle-income
(for example, very low-birth-weight infants, premature infants <32 countries (UMICs),27,43,49,53,60,66,70 and 27 were conducted in LLM
weeks’ gestation and infants with HIV, tuberculosis or malaria). ICs.13,24–26,29,33–35,37–41,47,48,50–52,54,55,57,59,63,65,67–69 Most data were attained
Small case series (n < 10) were also excluded. Included studies from neonates in the United States (142,934/311,359, 45.9%).28,36,44
required data to have been collected within the clinical context of Eastern Mediterranean countries had the second highest representa-
suspected neonatal infection and reported data needed to have been tion, with 29.2% (91,133/311,359) of the included neonates from
collected after 2012, with samples collected from normally sterile 6 studies.31,39,41,53,68,70 Only 9.2% (28,773/311,359) of the included
sites (blood and/or cerebrospinal fluid). neonates were from African countries,13,24,27,29,43,47,49–51,60,63,66 and
Abstracts yielded from the above searches were exported 5.1% (15,758/311,359) evaluated neonates from Southeast
and reviewed by authors M.L.H. and P.C.M.W. Full-text articles Asia.13,25,26,33–35,38,48,52,54,55,57,65,67,69
were sourced, and studies were assessed for quality using GRADE In neonates with EOS of the 4347 significant pathogens
methodology20 to determine the risk of bias associated with the isolated, 52.9% (2,301/4347) were Gram-negative and 46.8%
study. A data extraction tool was used to summarize pathogen (2,038/4,347) were Gram-positive bacteria. For neonates with
data alongside data on study design, patient recruitment methods, LOS, of the 3894 significant pathogens reported 71% (2,765/3,894)
publication year, data collection dates and specimen handling and were Gram-negative, and 29% (1,129/3,894) were Gram-positive
collection methods (Table, Supplemental Digital Content 3, http:// (Table 1). Meta-analyses with random-effects weighting estimated
links.lww.com/INF/F653). a pooled prevalence of Gram-negative pathogens of 62.6% (95%
Meta-analyses of collected data were undertaken to generate CI: 52.1–72.3) for EOS, and 71.0% (95% CI: 63.5–78.1%) for LOS
and compare pooled estimates of the relative prevalence of patho- (Fig. 2).
gens causing EOS and LOS with exact binomial confidence inter- Figure 2 and Figures S2–S16, Supplemental Digital Con-
vals (CIs). Pathogen prevalence for LOS and EOS was additionally tent 5, http://links.lww.com/INF/F655) reveal the results of the

© 2024 The Author(s). Published by Wolters Kluwer Health, Inc www.pidj.com | 1183

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Harrison et al The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024

FIGURE 1. PRISMA diagram.

meta-analyses comparing the relative prevalence estimates for (9.7%, 424/4,347) (Table 1, Figure S22, Supplemental Digital
pathogens causing EOS and LOS (Table, Supplemental Digital Content 6, http://links.lww.com/INF/F655). However, when the
Content 6, http://links.lww.com/INF/F716). These revealed no results were stratified for HICs versus LLMICs, the predominant
significant difference in the likelihood of Acinetobacter spp. (P = pathogens causing EOS differed significantly, with Klebsiella spp.
0.656), Citrobacter spp. (P = 0.913), E. coli (P = 0.582), Serratia (31.7%, 95% CI: 24.1–39.7%) and S. aureus (17.5%, 95% CI: 8.5–
spp. (P = 0.711), Enterobacter spp. (P = 0.172), Proteus spp. (P = 28.4%) most prevalent in LLMICs, with S. agalactiae only respon-
0.084), Pseudomonas spp. (P = 0.082), Salmonella spp. (P = 1.62), sible for a small proportion of EOS (13.8%, 95% CI: 3.4–28.7) in
Enterococcus spp. (P = 0.376) and Streptococcus pyogenes (P = LLMICs (Fig. 3).
0.272) to cause EOS or LOS. The predominant bacterial species causative of LOS over-
However, there was a significant difference in the likelihood all were Klebsiella spp. (30.7%, 1,197/3,894), E. coli (16.5%,
of some bacteria causing EOS versus LOS. In particular, Strepto- 643/3,894) and S. aureus (15.5%, 605/3,894) (Table 1 and Fig. 4).
coccus agalactiae (P ≤ 0.001) and Listeria spp. (P = 0.006) pre- Only a small difference was found when stratifying between HIC
dominated in EOS, and Klebsiella spp. (P = 0.018) and S. aureus and LLMIC in the meta-analysis for LOS, with Klebsiella spp. still
(P = 0.026) predominated in LOS. The Luis Furuya–Kanamori the predominant bacteria causing LOS in LLMICs (30.6%, 95% CI:
index classified the risk of bias as none (<|1|) or minor (|1| to |1.99|) 24.9–36.7), followed by S. aureus (25.6%, 95% CI: 17.4–34.8%),
for the majority of analyses, with the exception of Listeria spp., S. and E. coli (9.7%, 95% CI: 7.4–12.1%) (Figure S21 Supplemental
pyogenes, S. pneumoniae and Acinetobacter spp., where the risk Digital Content 5, http://links.lww.com/INF/F655 and Supplemen-
was considered major (>|2|; Figures S17-S21, Supplemental Digital tal Digital Content 7, http://links.lww.com/INF/F717).
Content 6, http://links.lww.com/INF/F655).
Although the overall burden of Gram-negative pathogens DISCUSSION
in EOS was high, the predominant pathogen causing EOS was We provide a comprehensive and systematic review of the
S. agalactiae (23.3%, 1,015/4347), followed by E.coli (18.0%, current bacterial pathogens causing neonatal sepsis. Our results
783/4,347), Klebsiella spp. (14.8%, 644/4,347) and S. aureus revealed Gram-negative bacteria are the most frequent cause

1184 | www.pidj.com © 2024 The Author(s). Published by Wolters Kluwer Health, Inc

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024 Current Causes of Neonatal Sepsis

TABLE 1. Gram-negative and Gram-positive Pathogens Causative of Early-onset Sepsis and Late-Onset Sepsis
Early-onset Sepsis Late-onset Sepsis

Gram-negative n (%) Gram-negative n (%)

Escherichia spp. 783 (18.0) Klebsiella spp. 1197 (30.7)
Klebsiella spp. 644 (14.8) Escherichia spp. 643 (16.5)
Enterobacter spp. 147 (3.4) Acinetobacter spp. 246 (6.3)
Acinetobacter spp. 132 (3.0) Enterobacter spp. 211 (5.4)
Pseudomonas spp. 101 (2.3) Serratia spp. 155 (4.0)
Serratia spp. 100 (2.3) Pseudomonas spp. 118 (3.0)
Listeria spp. 96 (2.2) Citrobacter spp. 26 (0.7)
Haemophilus spp. 48 (1.1) Proteus spp. 13 (0.3)
Citrobacter spp. 28 (0.6) Salmonella spp. 9 (0.2)
Proteus spp. 6 (0.1) Listeria spp. 7 (0.2)
Salmonella spp. 6 (0.1) Neisseria spp. 6 (0.2)
Neisseria spp. 3 (0.1) Other Gram-negative 134 (3.4)
Other Gram-negative 207 (4.8)
Total Gram-negative EOS N = 2301 53% Total Gram-negative LOS N = 2765 71%
Gram-positive Gram-positive
Streptococcus agalactiae 1015 (23.3) Staphylococcus aureus 605 (15.5)
Staphylococcus aureus 424 (9.8) Enterococcus spp. 222 (5.7)
Enterococcus spp. 182 (4.2) Streptococcus agalactiae 171 (4.4)
Streptococcus pyogenes 29 (0.7) Streptococcus pyogenes 13 (0.3)
Streptococcus pneumoniae 23 (0.5) Streptococcus pneumoniae 3 (0.1)
Other Gram-positive 365 (8.4) Other Gram-positive 115 (3.0)
Total Gram-positive EOS N = 2038 47% Total Gram-positive LOS N = 1129 29%

EOS indicates early-onset sepsis; LOS, late-onset neonatal sepsis.

of both EOS and LOS globally; of concern given the plasmid-­ bacteria causative of neonatal sepsis in LLMICs. This is an
mediated resistance these bacteria can easily transmit, contributing important finding given the potential virulence and resistance
to the rising burden of AMR.6 The significant rise in Klebsiella spp. profile commonly seen with this pathogen. Recent whole genome
as a dominant species in not only LOS but also EOS has also been sequencing studies evaluating bacteria causing invasive neonatal
noted in other recent epidemiological studies.10,13,15,17,35,38,67 infections identified an abundance of AMR genes and virulence
Our review revealed Klebsiella spp. were the third most iso- factors in Klebsiella spp., exposing neonates to a high risk of
lated bacteria causative of EOS globally, and the most common mortality.13

FIGURE 2. Meta-analysis of the proportion (%) of Gram-negative bacteria causative of early-onset (A) and late-onset (B)
neonatal sepsis

© 2024 The Author(s). Published by Wolters Kluwer Health, Inc www.pidj.com | 1185

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Harrison et al The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024

FIGURE 3. Meta-analysis of species prevalence causative of early-onset neonatal sepsis (%) in low- and lower-middle-income
versus upper- and high-income countries.

FIGURE 4. Proportion of EOS and LOS caused by these pathogens.*NB: pathogen proportions <1% have been omitted from
graph, see Table 1 for full description of causative pathogens.

While our review found S. agalactiae to be the highest iso- the third most prevalent bacteria causing EOS. Overcrowding, poor
lated single pathogen in EOS overall, this was likely impacted by the sanitation and restricted infection, prevention and control resource
overrepresentation of data from high-income settings. When using a availability may predispose infants in LLMIC to early colonization
meta-analysis to stratify data from LLMICs independently, the most with Gram-negative bacteria, which may explain these findings.
prevalent bacteria was Klebsiella spp. in both early- and late-onset Our data from HIC support the historical assumptions that EOS
sepsis. In these resource-constrained settings, S. agalactiae is only is predominantly caused by S. agalactiae, followed by E. coli. The

1186 | www.pidj.com © 2024 The Author(s). Published by Wolters Kluwer Health, Inc

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024 Current Causes of Neonatal Sepsis

TABLE 2. Suggested Contemporary Classification of Neonatal Sepsis, Likely Causative Pathogens and Potential
Efficacious Empiric Regimens
Historical Classification Likely Pathogens Timing of Acquisition Empiric Antibiotic Regimens to Target Likely Pathogens(s)
Early-onset sepsis Streptococcus agalac- Within 72 hours of Aminopenicillin(s) and gentamicin;
tiae; Escherichia coli birth third-generation cephalosporins
Late-onset sepsis E. coli, S. agalactiae, After 72 hours of birth Aminopenicillin(s) and gentamicin;
Streptococcus aureus, third-generation cephalosporins; flucloxacillin if S. aureus infection
coagulase-negative suspected; vancomycin (where available)
staphylococci

Proposed Contempora-
neous Classification Likely Pathogens Timing of Acquisition Antibiotic Regimens Likely to be Efficacious
Vertically-acquired S. agalactiae (GBS), Within 72 hours after While in many contexts aminopenicillin(s) and gentamicin, or
E. coli., Klebsiella birth life third-generation cephalosporins, remain efficacious, there is increas-
spp., Enterobacter ing evidence that in many high-burden AMR settings, antibiotics with
spp., S. aureus activity against extended-spectrum beta lactamases, metallo-beta
Horizontally-acquired; Klebsiella spp., E. These pathogens may lactamases and carbapenem-resistant Acinetobacter spp. may be
nosocomial setting coli, Acinetobacter be acquired during necessary.
spp. S. aureus, the delivery process,
coagulase-negative very early in the
staphylococci perinatal period or
over the course of a
prolonged hospital-
isation following a
premature delivery.
Horizontally-acquired; S. agalactiae, E coli, Within the first month While community-acquired pathogens tend to be less likely to be multid-
community setting S. aureus of life rug resistant, in some settings, there is a high community prevalence
of ESBL-producing Gram-negative bacteria, which should be consid-
ered when selecting empiric antibiotic regimens.

clear contrast in causative bacteria isolated from resource-­constrained bacteria (and resistance profiles) empiric antibiotic regimens need
settings compared to well-resourced healthcare settings, particularly to target, yet data pertaining to community-acquired infections—
in EOS, warrants further research to consider how LLMICs might particularly in LLMICs—are sparse in the published literature.
reduce the current burden of EOS caused by Gram-negative bacteria— New definitions, with less emphasis on the timing of symptom
pathogens against which currently-recommended empiric antibiotic onset, that instead consider the most likely source of infection, may
regimens are unlikely to be ­effective.15,18 enable a more accurate selection of efficacious empirical antibiotic
Klebsiella spp., E.coli and S. aureus were the pathogens pri- regimens to ensure infants receive appropriate antibiotic therapy-
marily responsible for LOS in our review, supporting the evolving Magiorakos, 2012 #20 targeting the most likely bacteria causative
international literature highlighting the overall burden of Gram-­ of their infection (Table 2). Furthermore, classification defining
negative bacteria in neonatal sepsis.10 In fact, less than one-third infections by community acquired infection and hospital acquired
of all bacteria isolated in neonates with LOS were Gram-­positive infection will harmonize neonatal sepsis definitions with the pedi-
within our comprehensive review. atric and adult definitions currently in use.
In the context of increasing facility-based births globally,73 Our research has a number of limitations. The heterogeneity
alongside prolonged hospital stays following the successful resus- of study designs in included studies, and the poor quality of many
citation of very premature infants, early colonization of infants published observational studies, reduces confidence in the estimate
with Gram-negative bacteria may be one factor driving the altered of the effect observed. Many studies meeting the inclusion crite-
etiology our study reveals. The historical consideration that EOS ria provide insufficient microbiological methodological detail and
is caused by bacteria that are vertically-acquired, while LOS is limited denominator data. Additionally, the published and available
horizontally-acquired, is increasingly becoming less clear—as data includes almost exclusively inborn neonates, with the majority
neonates may be admitted to hospital units with a heavy environ- of studies undertaken in hospital settings and only 1 study recruit-
mental prevalence of multidrug-resistant Gram-negative bacteria. ing neonates from the community.38 This likely biases the results
Selection pressure due to a high burden of MDR infections in toward an overrepresentation of hospital-acquired pathogens. To
resource-constrained hospital facilities propagates AMR, and if enhance the quality of published literature in this space, future
these MDR bacteria are colonizing neonates quickly in the peri- observational studies should follow strengthening the reporting
natal period, it is essential to ensure empirical treatment guidelines of observational studies in epidemiology for newborn infection
address the local contemporary causes of neonatal sepsis to avoid guidelines which have been developed to ensure the quality of
unnecessary neonatal morbidity and mortality. observational research in the neonatal population.77 Ensuring stud-
In keeping with evolving themes emerging from single epi- ies comply with these guidelines will enable high-quality published
demiological studies,10,74,75 and in support of recently published lit- data that can be readily compared across sites and regions.
erature,14 the results of our review query the traditional definitions Geographic bias in the available published literature also
and assumptions behind EOS and LOS (and the bacteria typically limits the generalizability of our findings, with a stark paucity of
associated with these defined clinical syndromes, which guide cur- published data on the causative pathogens responsible for neona-
rent empirical antibiotic regimen recommendations). Our results tal sepsis across Southeast Asia, South America, the Pacific and
confirm that it is clear that EOS and LOS assumptions have poor Africa. The lack of epidemiological data in the regions is con-
utility in LLMICs. In these settings, community-acquired versus cerning, given their high birth rates and infant mortality rates.78 In
hospital-acquired definitions76 may be more useful in predicting the 2017, South Asia and sub-Saharan Africa accounted for 79% of all

© 2024 The Author(s). Published by Wolters Kluwer Health, Inc www.pidj.com | 1187

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Harrison et al The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024

neonatal deaths.79 Yet this review reveals the paucity of representa- drug development programs and clinical trials to enable access to
tive epidemiological data available to inform policy and guideline efficacious agents to treat MDR infections.4 To attain sustainable
change from these regions with the highest burden of infant mortal- development goals, the pressing burden of neonatal sepsis requires
ity, where the need for robust evidence is greatest.16 urgent attention to ensure infant and child health outcomes can be
Moreover, data from the United States of America accounted optimized, and the unacceptable burden of mortality due to neona-
for 46% of the published data in this review, which is likely to tal sepsis can be curtailed.
significantly bias the overall prevalence of Gram-positive organ-
isms more commonly seen in high-resourced healthcare settings.
REFERENCES
We rectified this by stratifying and analyzing our data by country
income status; and despite the high representation of North Amer- 1. World Health Organization. Levels and trends in child mortality report.
Online: World Health Organization; 2021. Available from: https://data.
ican sites in the overall dataset, our review still reveals an almost unicef.org/resources/levels-and-trends-in-child-mortality/. Accessed
equal proportion of Gram-negative and Gram-positive bacteria are January 28, 2022.
responsible for EOS globally. The concerningly high proportions of 2. UNICEF. Neonatal Mortality Online2021. Available from: https://data.
E. coli and Klebsiella spp. causative of EOS reinforces the chang- unicef.org/topic/child-survival/neonatal-mortality/. Accessed August 13,
ing pathogen distribution for this clinical syndrome globally, and 2022.
the urgent need to ensure antibiotic regimens to treat EOS provide 3. Fuchs A, Bielicki J, Mathur S, et al. Reviewing the WHO guidelines for
coverage against these bacteria.10 antibiotic use for sepsis in neonates and children. Paediatr Int Child Health.
2018;38(sup1):S3–S15.
Our meta-analysis supports the hypothesis that the causa-
4. Williams P, Qazi S, Agarwal R, et al. Antibiotics needed to treat
tive bacteria responsible for EOS and LOS in neonates are simi- multidrug-resistant infections in neonates. Bull World Health Organ.
lar. This was particularly evident across Gram-negative organisms 2022;100:797–807.
and stands contrary to prior assumptions regarding the traditional 5. United Nations. 2022. Available from: https://www.un.org/development/
causes of EOS and LOS. The findings support the evolving liter- desa/disabilities/envision2030-goal3.html. Accessed June 29, 2022.
ature revealing the changing etiology of neonatal sepsis and the 6. Murray CJ, Ikuta KS, Sharara F, et al. Global burden of bacterial anti-
importance of considering new definitions to guide efficacious microbial resistance in 2019: a systematic analysis. Lancet. 2022;399:
empiric antibiotic regimens against neonatal sepsis.10 Our findings 629–655.
reflect those revealed by the multicenter NeoOBS study, which 7. Shane ALD, Sánchez PJP, Stoll BJP. Neonatal sepsis. Lancet. 2017;390:
highlighted the presence of common pathogens in both EOS and 1770–1780.
LOS and the low prevalence of S. agalactiae (3.4%) isolated within 8. Bethou A, Bhat BV. Neonatal sepsis-newer insights. Indian J Pediatr.
2022;89:267–273.
their global cohort.15 Additionally, our findings concur with the
Delhi Neonatal Infection Study study,10 and the Neonatal Sepsis 9. Wen SCH, Ezure Y, Rolley L, et al. Gram-negative neonatal sepsis in
low- and lower-middle-income countries and WHO empirical antibiotic
in Southeast Asia and the Pacific study observational study across recommendations: a systematic review and meta-analysis. PLoS Med.
Southeast Asia,17 which reveal a low prevalence of S. agalactiae 2021;18:e1003787.
causing neonatal sepsis after the first day of life in low-resource 10. Agarwal R, Sankar J. Characterisation and antimicrobial resistance of sepsis
healthcare settings.10,15 It remains unclear what is causing this low pathogens in neonates born in tertiary care centres in Delhi, India: a cohort
prevalence, with difficulty isolating this pathogen in low-resourced study. Lancet Glob Health. 2016;4:e752–ee60.
settings likely to impact the findings.14 Further granular data repre- 11. Pocket book of hospital care for children: guidelines for the management
sentative of all regions to provide insight into the timing of acquisi- of common childhood illnesses. Second edition, 2013 edition. ed. Geneva,
Switzerland: World Health Organization; 2013.
tion of particular pathogens are needed with the exploration of the
12. Thomson KM, Dyer C, Liu F, et al; BARNARDS Group. Effects of
causal mechanisms of ultra-early horizontal transmission occurring ­antibiotic resistance, drug target attainment, bacterial pathogenicity and vir-
within the neonatal intensive care unit. These data can ultimately ulence, and antibiotic access and affordability on outcomes in neonatal sep-
ensure a more robust categorization of neonatal sepsis based on the sis: an international microbiology and drug evaluation prospective substudy
likely pathogens and the means of transmission, to ensure empiric (BARNARDS). Lancet Infect Dis. 2021;21:1677–1688.
therapies are optimized to avoid unnecessary neonatal mortality. 13. Sands K, Carvalho MJ, Portal E, et al; BARNARDS Group. Characterization
Our results raise the importance of empirical antibiotic treat- of antimicrobial-resistant Gram-negative bacteria that cause neonatal sep-
sis in seven low- and middle-income countries. Nat Microbiol. 2021;6:
ment recommendations that need to be region- and setting-specific 512–523.
to reflect the local causative pathogens and known resistance pat- 14. Russell N, Barday M, Okomo U, et al. Early-versus late-onset sepsis in neo-
terns responsible for neonatal sepsis more closely. The challenge nates – time to shift the paradigm? Clin Microbiol Infect. 2023;30:38–43.
remains; however, that to derive locally-recommended regimens 15. Russell NJ, Stöhr W, Plakkal N, et al. Patterns of antibiotic use, pathogens,
from sufficient observational data, prospective and systematic sur- and prediction of mortality in hospitalized neonates and young infants with
veillance is required. sepsis: a global neonatal sepsis observational cohort study (NeoOBS). PLoS
The findings from this review, while limited in their direct Med. 2023;20:e1004179.
application, support the evolving literature suggesting a higher 16. Okomo U, Akpalu ENK, Le Doare K, et al. Aetiology of invasive bacterial
infection and antimicrobial resistance in neonates in sub-Saharan Africa: a
proportion of Gram-negative pathogens are causing both EOS systematic review and meta-analysis in line with the STROBE-NI reporting
and LOS and an equitable distribution of bacteria are responsible guidelines. Lancet Infect Dis. 2019;19:1219–1234.
for infections across the early- and late-neonatal period, suggests 17. Dickson BF, Putri ND, Adrizain R, et al. High burden of anti-­microbial
new definitions to define the acquisition of invasive infections to resistance among neonatal blood stream infections in Southeast Asia:
the empirical antibiotic guidelines are needed. The predominance results of the NeoSEAP study. 2023. Preprint: medRxiv. 2023. doi:
of Gram-negative bacteria identified in our review as the leading https://doi.org/10.1101/2023.09.11.23295383.
cause of early-onset sepsis in LLMICs further highlights the need 18. Williams PCM, Jones M, Snelling TL, et al. Coverage gaps in empiric anti-
biotic regimens used to treat serious bacterial infections in neonates and
to update the current empirical antibiotic recommendations in neo- children in Southeast Asia and the Pacific. Lancet Reg Health Southeast
natal sepsis to ensure they reflect the contemporaneous etiology Asia. 2023;22:100291.
and epidemiology of neonatal sepsis. 19. Williams PCM, Isaacs D, Berkley JA. Antimicrobial resistance among
Most importantly, in the context of the rising prevalence ­children in sub-Saharan Africa. Lancet Infect Dis. 2018;18:e33–e44.
of Gram-negative bacteria, and their potential for acquisition of 20. Granholm A, Alhazzani W, Møller MH. Use of the GRADE approach in
resistant mechanisms, it is vital that neonates are prioritized in systematic reviews and guidelines. Br J Anaesth. 2019;123:554–559.

1188 | www.pidj.com © 2024 The Author(s). Published by Wolters Kluwer Health, Inc

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024 Current Causes of Neonatal Sepsis

21. Nyaga VN, Arbyn M, Aerts M. Metaprop: a Stata command to perform Network. Early-onset neonatal sepsis 2015 to 2017, the rise of Escherichia
meta-analysis of binomial data. Arch Public Health. 2014;72:39. coli, and the need for novel prevention strategies. JAMA Pediatr.
22. Deeks JJ, Higgins JP, Altman DG, Group CSM. Analysing data and under- 2020;174:e200593.
taking meta‐analyses. Cochrane handbook for systematic reviews of inter- 45. Takassi OE, Atakouma YD, Desfrere L. Predictors of early-onset neo-
ventions. 2019;6.4:241–84. natal sepsis in premature newborns: case-control study. Arch Pediatr.
23. Furuya-Kanamori L, Barendregt JJ, Doi SAR. A new improved graphical 2022;29:183–187.
and quantitative method for detecting bias in meta-analysis. Int J Evid Based 46. Tessema B, Lippmann N, Knupfer M, et al. Antibiotic resistance patterns
Healthc. 2018;16:195–203. of bacterial isolates from neonatal sepsis patients at University Hospital of
24. Aku FY, Akweongo P, Nyarko K, et al. Bacteriological profile and anti- Leipzig, Germany. Antibiotics. 2021;10:323.
biotic susceptibility pattern of common isolates of neonatal sepsis, Ho 47. Tumuhamyeid J, Sommerfelt H, Bwanga F, et al. Neonatal sepsis at
Municipality, Ghana-2016. Matern Health Neonatol Perinatol. 2018;4:2. Mulago National Referral Hospital in Uganda: etiology, antimicro-
25. Bhattarai S, Basnet S, Pokhrel BM, et al. A prospective study on neona- bial resistance, associated factors and case fatality risk. PLoS One.
tal sepsis in a Tertiary Hospital, Nepal. J Pure Appl Microbiol. 2021;15: 2020;15:e0237085.
2409–2419. 48. Utomo MT, Sumitro KR, Etika R, et al. Current-proven neonatal sepsis in
26. Bhattarai S, Chapagain RH, Mishra D, et al. Bacteriological profile and Indonesian tertiary neonatal intensive care unit: a hematological and micro-
antibiotic sensitivity pattern of neonatal sepsis in central paediatric referral biological profile. Iran J Microbiol. 2021;13:266–273.
hospital in Nepal. J Nepal Paediatr Soc. 2019;39:1–5. 49. Velaphi SC, Westercamp M, Moleleki M, et al. Surveillance for incidence
27. Dramowski A, Bekker A, Cotton MF, et al. Epidemiology of clinically sus- and etiology of early-onset neonatal sepsis in Soweto, South Africa. PLoS
pected and laboratory-confirmed bloodstream infections at a South African One. 2019;14:e0214077.
neonatal unit. J Infect Dev Ctries. 2021;15:943–952. 50. Yohannes Z, Tazeb M, Laway B, et al. Bacterial profile and antimicrobial
28. Flannery DD, Puopolo KM, Hansen NI, et al; for the Eunice Kennedy susceptibility pattern of neonatal sepsis in Felege-Hiwot Referral Hospital,
Shriver National Institute of Child Health and Human Development Bahir Dar, northwest Ethiopia: a cross-sectional study design. Ethiop J
Neonatal Research Network. Antimicrobial susceptibility profiles among Health Devt. 2021;35:18–28.
neonatal early-onset sepsis pathogens. Pediatr Infect Dis J. 2022;41: 51. Zamarano H, Musinguzi B, Kabajulizi I, et al. Bacteriological profile, anti-
263–271. biotic susceptibility and factors associated with neonatal Septicaemia at
29. Geyesus T, Moges F, Eshetie S, et al. Bacterial etiologic agents causing neo- Kilembe mines hospital, Kasese District Western Uganda. BMC Microbiol.
natal sepsis and associated risk factors in Gondar, Northwest Ethiopia. BMC 2021;21:303.
Pediatr. 2017;17:137. 52. Bandyopadhyay T, Kumar A, Saili A, et al. Distribution, antimicrobial resist-
30. Giannoni E, Agyeman PKA, Stocker M, et al; Swiss Pediatric Sepsis Study. ance and predictors of mortality in neonatal sepsis. J Neonatal Perinatal
Neonatal Sepsis of early onset, and hospital-acquired and community-­ Med. 2018;11:145–153.
acquired late onset: a prospective population-based cohort study. J Pediatr. 53. Bozkurt HB. Evaluation of our neonatal sepsis cases in terms of caus-
2018;201:106–114.e4. ing microorganism and antibiotic resistance. Cocuk Enfeksiyon Dergisi.
31. Hammoud MS, Al-Taiar A, Al-Abdi SY, et al. Culture-proven early-onset 2018;12:e99–e104.
neonatal sepsis in Arab states in the Gulf region: two-year prospective study. 54. Chauhan R, Gupta S, Dalal JS, et al. Antimicrobial profile and antibiotic
Int J Infect Dis. 2017;55:11–15. susceptibility of neonatal sepsis in neonatal intensive care unit of tertiary
32. Jiang S, Hong L, Gai J, et al. Early-onset sepsis among preterm neonates in care hospital of north India. Indian J Public Health Res Dev. 2019;10:
China, 2015 to 2018. Pediatr Infect Dis J. 2020;38:1236–1241. 733–737.
33. Johnson J, Robinson ML, Rajput UC, et al. High burden of bloodstream 55. Das T, Saha J, Pal K, et al. Bacterial etiology of neonatal sepsis, antibiotic
infections associated with antimicrobial resistance and mortality in the susceptibility profile, and associated factors at Burdwan Medical College,
Neonatal Intensive Care Unit in Pune, India. Clin Infect Dis. 2021;73: Burdwan, West Bengal, India. J Child Sci. 2021;11:e148–e154.
271–280. 56. Gao K, Fu J, Guan X, et al. Incidence, bacterial profiles, and antimicrobial
34. Kumari M, Dash RK, Behera BK, et al. Aetiological profile, risk factors, resistance of culture-proven neonatal sepsis in South China. Infect Drug
antibiotic sensitivity pattern and outcome of neonatal sepsis in tertiary Resist. 2019;12:3797–3805.
care hospitals- a prospective observational study. Eur J Mol Clin Med. 57. Gayathiri G, Vijayakumar A, Mageswari TMU, et al. Sepsis in neonates:
2021;8:1073–1087. prevalence of micro-organisms and their susceptibility pattern in neonatal
35. Palwinder S, Loveena O, Sonalika M, et al. Bacteriological profile and anti- intensive care unit of a tertiary care hospital - a retrospective study. J Basic
biogram of community acquired neonatal sepsis in a tertiary care hospital; a Clin Pharm. 2020;11:11–16.
prospective study. Ann Int Med Dent Res. 2017;3:2395. 58. Guo J, Luo Y, Wu Y, et al. Clinical characteristic and pathogen spectrum of
36. Polcwiartek LB, Smith PB, Benjamin DK, et al. Early-onset sepsis in term neonatal sepsis in Guangzhou City from June 2011 to June 2017. Med Sci
infants admitted to neonatal intensive care units (2011-2016). J Perinatol. Monit. 2019;25:2296–2304.
2021;41:157–163. 59. Hameed A, Izhar M, Hameed S, et al. Neonatal sepsis in Tertiary care hospi-
37. Purba M, Prativa B, Sanghamitra S, et al. Neonatal sepsis-trends in a periph- tal: bacteriological profile and antibiotic susceptibility patterns. Pak J Med
eral tertiary health care facility of eastern India. J Evol Med Dent Sci. Health Sci. 2018;12:1656–1659.
2019;8:1089–1093. 60. Lebea MM, Davies V. Evaluation of culture-proven neonatal sepsis at a
38. Saha SK, Schrag SJ, El-Arifeen S, et al. Causes and incidence of tertiary care hospital in Johannesburg, South Africa. S Afr J Child Health.
community-acquired serious infections among young children in South Asia 2017;11:170–173.
(ANISA): an observational cohort study. Lancet. 2018;392:145–159. 61. Liu J, Fang Z, Yu Y, et al. Pathogens distribution and antimicrobial resist-
39. Salah A, Al-Subol I, Hudna A, et al. Neonatal sepsis in Sana’a city, Yemen: ance in bloodstream infections in twenty-five neonatal intensive care units
a predominance of Burkholderia cepacia. BMC Infect Dis. 2021;21:1108. in China, 2017-2019. Antimicrob Resist Infect Control. 2021;10:121.
40. Sangita T, Sapkota LB. Changing trend of neonatal septicemia and antibiotic 62. Lu L, Li P, Pan T, et al. Pathogens responsible for early-onset sepsis in
susceptibility pattern of isolates in Nepal. Int J Pediatr. 2019;3784529. Suzhou, China. Jpn J Infect Dis. 2020;73:148–152.
41. Seliem WA, Sultan AM. Etiology of early onset neonatal sepsis in neo- 63. Mulinganya GM, Claeys M, Balolebwami SZ, et al. Etiology of early-onset
natal intensive care unit-Mansoura, Egypt. J Neonatal Perinatal Med. neonatal sepsis and antibiotic resistance in Bukavu, Democratic Republic of
2018;11:323–330. the Congo. Clin Infect Dis. 2021;73:e976–e980.
42. Stordal EH, Solevag AL, Bjornholt JV, et al. Sepsis treatment options 64. Pan T, Zhu Q, Li P, et al. Late-onset neonatal sepsis in Suzhou, China. BMC
identified by 10-year study of microbial isolates and antibiotic susceptibil- Pediatr. 2020;20:261.
ity in a level-four neonatal intensive care unit. Acta Paediatr. 2021;111: 65. Pandit BR, Vyas A. Clinical symptoms, pathogen spectrum, risk factors and
519–526. antibiogram of suspected neonatal sepsis cases in tertiary care hospital of
43. Staaden H, Hendricks C, Spicer K. Bacteraemia and antibiotic sensitivity in southern part of Nepal: a descriptive cross-sectional study. J Nepal Med
a tertiary neonatal intensive care unit. S Afr J Infect Dis. 2021;36:195. Assoc. 2020;58:976–982.
44. Stoll BJ, Puopolo KM, Hansen NI, et al; Eunice Kennedy Shriver National 66. Pillay D, Naidoo L, Swe-Han K, et al. Neonatal sepsis in a tertiary unit in
Institute of Child Health and Human Development Neonatal Research South Africa. BMC Infect Dis. 2021;21:225.

© 2024 The Author(s). Published by Wolters Kluwer Health, Inc www.pidj.com | 1189

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Harrison et al The Pediatric Infectious Disease Journal • Volume 43, Number 12, December 2024

67. Pokhrel B, Koirala T, Shah G, et al. Bacteriological profile and antibiotic 74. BARNARDS. Burden of antibiotic resistance in neonates from developing
susceptibility of neonatal sepsis in neonatal intensive care unit of a tertiary societies (BARNARDS). Online 2022. Available from: https://www.ine-
hospital in Nepal. BMC Pediatr. 2018;18:208. osoxford.ox.ac.uk/barnards. Accessed August 3, 2022.
68. Salama K, Gad A, El Tatawy S. Sepsis profile and outcome of preterm neo- 75. Toan ND, Darton TC, Huong NHT, et al. Clinical and laboratory factors
nates admitted to neonatal intensive care unit of Cairo University Hospital. associated with neonatal sepsis mortality at a major Vietnamese children’s
Egypt Pediatric Association Gaz. 2021;69:8. hospital. PLOS Glob Public Health. 2022;2:e0000875.
69. Shruthi KS, Sanji N, Raghoji CR, et al. A retrospective study on antibi- 76. NeonIN. NeonIN. Online2022. Available from: https://neonin.org.uk/#/
otic sensitivity pattern of pathogens isolated from blood culture in cases of home. Accessed August 3, 2022.
late-onset neonatal sepsis at a neonatal intensive care unit. Natl J Physiol
77. Fitchett EJA, Seale AC, Vergnano S, et al; SPRING (Strengthening
Pharm Pharmacol. 2018;8:709–712.
Publications Reporting Infection in Newborns Globally) Group.
70. Yusef D, Shalakhti T, Awad S, et al. Clinical characteristics and epidemiology Strengthening the reporting of observational studies in epidemiology for
of sepsis in the neonatal intensive care unit in the era of multi-drug resistant newborn infection (STROBE-NI): an extension of the STROBE statement
organisms: a retrospective review. Pediatr Neonatol. 2018;59:35–41. for neonatal infection research. Lancet Infect Dis. 2016;16:e202–e213.
71. World Health Organization World Health Organization Regional offices. 78. The World Bank. Infant mortality rates. Online 2020. Available from:
Online 2023. Available from: https://www.who.int/about/who-we-are/ https://data.worldbank.org/indicator/SP.DYN.IMRT.IN. Accessed August 3,
regional-offices. Accessed 2023 2022.
72. The World Bank. The World by Income and Region. Online2021. Available 79. Hug L, Alexander M, You D, et al; UN Inter-agency Group for Child
from: https://datatopics.worldbank.org/world-development-indicators/the- Mortality Estimation. National, regional, and global levels and trends in
world-by-income-and-region.html. Accessed August 12, 2022. neonatal mortality between 1990 and 2017, with scenario-based projections
73. Newborn Mortality Online: World Health Organization 2022. to 2030: a systematic analysis. Lancet Glob Health. 2019;7:e710–e720.

1190 | www.pidj.com © 2024 The Author(s). Published by Wolters Kluwer Health, Inc

Copyright © 2024 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.