An#bio#c

Stewardship
Why we do it and How weve done
it
C. Michael Co,en MD MHS

CM Cotten: Disclosures
Disclosures:

DSMB for rEVO pharmaceuticals (preeclampsia)

Consultant Windtree Pharma (respiratory
intervention)

Advisory Board for GW Pharma (neuroprotection)

The use of antibiotics in neonates

without infection is off-label.

 Are an3bio3cs an inevitable part of NICU/

Neonatal life, especially for preterm infants?
If they are, why should we care (microbiome?)
and what can we do to use them most
eec3vely and eciently?

What is Our Microbiome?

Microbiota: Microbial community
Microbiome: collective genomes and
gene products of resident microbes
living within and on humans
Metagenome: Collection of genomes
within complex microbial
communities and human DNA

WE ARE MOSTLY MICROBIAL

Human microbiome
1,000,000+ genes

Human genome
23,000 genes

What do the microbes do for us?

Provide the ability to harvest nutrients
Produce additional energy otherwise
inaccessible to the host.
Produce vitamins
Metabolize xenobiotics (including drugs)
Provide resistance to tumor and cancer
leading neoplasms
Assist in developing a mature immune
system

EARLY LIFE LESSONS

Inflammation/Immunity

Tolerance

Antibiotics influence this

balance..What else tips the
balance????
(Tribute to Nancy Fawcett, MD)

Body weight - microora con#nued

WT and obese mouse experiments
Stool calorimetry:
energy lost from similar diets

Increase in body fat 2 wks aLer obese mouse

ora is transplanted into germ free (clear bar)
vs. increase when lean mouse ora transplanted
into germ free (dark bar)

Turnbaugh PJ et al Nature 2006; 444: 1027-131

Bacterial succession in term

babies
Aerobes

Faculta3ve anaerobes
Streptococci
Staphylococci
Coliforms
Lactobacilli** Birth Wean
Obligate Anaerobes**
Bidobacteria**

Obligate anaerobes
Bacteroides
Clostridia

Adulthood

**In infants receiving breast milk

Modied from Hooper, Trends in Microbiology, 2004

An#bio#cs
ADer Fleming and Penicillin (1928),,,,The word an3bio3c was coined by soil
microbiologist Selman Waksman, the Nobel Prize-winning co-discoverer of
streptomycin. He and his dis3nguished team (including Albert Schatz) isolated
hundreds of ac3nomycetes from dierent soils and subsequently iden3ed
compounds with an3bio3c ac3vity in the laboratory (streptothricin-1943,
neomycin, ac3nomycin D, etc.). Schatz demonstrated that streptomycin was
ac3ve against the bacterium that causes tuberculosis. These discoveries were
the genesis of the an3bio3c industry.

Waksman described an an#bio#c as a compound produced by a

microbe that kills or inhibits the growth of another microbe.

Penicillins: inhibit bacterial cell wall synthesis and are bactericidal.

Aminoglycosides: bactericidal, inhibit protein synthesis and cell wall integrity through
binding to ribosomes.
Cephalosporins: like penicillins, b-lactam antibiotics, bactericidal and interfere with
synthesis of the bacterial cell wall.
Glycopeptides: bactericidal, inhibit bacterial cell wall synthesis.
Carbepenems: bactericidal, resistant to B-lactamases
de Hoog M, Mouton JW, van den Anker JN. New dosing strategies for an3bacterial agents in the neonate. Pediatrics. 2011;127:e367-74
Davies J, Davies D. Origins and Evolu3on of An3bio3c Resistance Microbiol Mol Biol Rev. 2010; 74: 417433.
h,p://urwebsrv.rutgers.edu/focus/ar3cle/Albert%20Schatz,%20co-discoverer%20of%20streptomycin,%20dies%20at%2084/1504

An#bio#c Use Prevalence

1/3rd of U.S. Mothers receive intrapartum
an3bio3cs
10% of U.S. newborns receive an3bio3cs
100% of Very Low Birthweight (VLBW; < 1500
grams BWgt)/Extremely Low Gesta3onal Age
(ELGANS; < 28 weeks), receive an3bio3cs
during their hospital stay.
Van Dyke MK, Phares CR, Lyneld R, Thomas AR, Arnold KE, Craig AS, Mohle-Boetani J, Gershman K, Schaner W, Pe3t S,
Zansky SM, Morin CA, Spina NL, Wymore K, Harrison LH, Shu, KA, Bareta J, Bulens SN, Zell ER, Schuchat A, Schrag SJ.
Evalua3on of universal antenatal screening for group B streptococcus. N Engl J Med. 2009 Jun 18;360(25):2626-36.
Weston EJ, Pondo T, Lewis MM et al. The burden of invasive early-onset neonatal sepsis in the United States, 2005-2008.
Pediatr Infect Dis J 2011; 30: 937-41.

An#microbials ARE Commonly Used in NICUs

1,425,992 records of 409 unique medica#ons
covering 1/1996 4/2005.
Among the top 10 most used:
Ampicillin #1
Gentamicin #2
Cefotaxime #5
Vanco #8

Clark RH et al. Pediatrics 2006; 117:1979-1987

Rate of Early- and Late-Onset GBS in the U.S.,

1990-2008
Early-onset GBS

Late-onset GBS

Before national prevention policy Transition

Universal screening

Source: Active Bacterial Core surveillance / Emerging Infections Program

Preterm Infants
Characteris3cs for risk of acquired infec3on
Imperfect, leaky barriers
Rely heavily of innate immune mechanisms and
neutrophil ac3vity
Exposed to hospital environment for weeks
Poor early nutri3on
Dependence on intravenous access and mul3ple
invasive procedures

Cuenca AG, Wynn JL, Moldawer LL, Levy O. Role of innate immunity in neonatal
infec3on. Am J Perinatol 2013; 30:10512.

Likelihood of Finding A Posi#ve Culture By Day It Was

Drawn

Data source: Mednax/Pediatrix medical group, by way of P. Brian Smith

Necrotizing Enterocolitis
7-10% of ELBW infants/1-5% of all neonates admits
20-50% mortality/leading cause of morbidity in survivors
Quadruples risk of neurodevelopmental impairment
30% with positive cultures
Surgical survival about 50%

Ancient Connec#on
NEC happens in the intes3ne, where microbes
are at home, and necessary, and compete for
space
Humans did not invent an#bio#cs; we merely
observedoLen by accidentthat bacteria
and other microorganisms produced biological
compounds capable of killing or suppressing
growth and reproduc3on of other
bacteria.and we use them a lot!!!
Chones ER, et al. An3bio3c Resistance: Implica3ons for Global Health and Novel Interven3on
Strategies: Workshop Summary h,p://www.nap.edu/catalog/12925.html

Neonatal An#bio#cs and wheeze and asthma risk:

systema#c review and metaanalysis
18 studies were eligible for meta-
analysis
pooled OR 1.27 (95% CI 1.12-1.43)

9 Studies without design bias

pooled OR 1.12 (95% CI 0.98-1.26).

3 studies focused on wheeze/

asthma beyond 5-6 yrs of age
pooled OR 1.08, 95% CI 0.93-1.23;
dominated by one study
Penders J et al. Infant an3bio3c use and wheeze and asthma risk: a systema3c
review and meta-analysis. Eur Respir J. 2011 Aug;38(2):295-302.

A robust and dose-dependent associa#on was found between

an#bio#c use in the rst 2 yr of life and asthma at age 7.5 yr but
did not appear to be mediated through an associa#on with
atopy.
Hoskin-Parr L et al. An3bio3c exposure in the rst two years of life and

development of asthma and other allergic diseases by 7.5 yr: a dose-dependent

rela3onship. Pediatr Allergy Immunol. 2013 ;24:762-71

Duke and Network Med+Surg NEC

GDB/ELBW inborn
For our rst four years in the NICHD Neonatal Research Network,
Duke had the highest rate of medical+surgical NEC.
35
30
25

ALL
Low
Duke
alone

20
15
10
5
0
2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

# of Bacterial Species vs. # of days on

an#bio#cs of the rst 30 postnatal days

Culture-based evaluation;
Lower diversity with higher number of antimicrobial days

Gewolb, I. H et al. Arch. Dis. Child. Fetal Neonatal Ed. 1999;80:167-F173

Innate Immunity and Commensal Bacteria

Madara, J. N Engl J Med 2004;351:1685-1686

21

Experiment 1A: Intact Innate Immune response is

important to Gut Integrity post injury
DSS + WT Intact immune response:
100% survival

DSS + gene#cally decient immune

response, MyD88 -/-0% survival

Rako-Nahoum et al. Cell 2004;118:229-241.

Experiment 1B: Remove microbes with

An#bio#cs and screw up immune homeostasis

All 4 abx combined, no ora = No survival.

No produc#on of IL-6, TNF, or KC-1
Rako-Nahoum et al. Cell 2004;118:229-241

Slide courtesy of Patrick Seed (peds ID Duke/now Northwestern)

Clamoxyl aects intes#nal gene expression

genes related to an#gen presenta#on downregulated,

MHC class II and nonclassical class Ib genes. expression of CD1d,

a gene associated with the presenta3on of lipid-containing
an3gens, was signicantly downregulated in the proximal small
intes3ne.

genes coding for an#-microbial products downregulated.

Genes associated with Paneth cell secre3on products were
dieren3ally regulated especially in proximal small intes3ne at
D17, including an3-microbial pep3des such as defensins,
lysozyme, and phospholipase A2.

upregulated expression of mast cell proteases

in the distal small intes3ne.

Schumann A et al. Phys Genomics 2005;23:235-245

Delayed Coloniza3on: Development of

Intes3nal Angiogenesis with Microbes
Germ - Free

Ex germ-free
conventionalized

Ex germ free with B.

thetha

Green marker for developing endothelial cells

Stapperbeck, et al. PNAS 99:154-51, 2002

26

Metagenome

RNA metabolism
Virulence
Phosphorous metabolism
Protein metabolism
DNA metabolism
Cell wall and capsule
Clustering-based subsystems
Carbohydrates
Amino acids and deriva3ves
Stress response
Respira3on

Adult

Baby 1

Latuga MS, et al. 2011

Baby 2

Results: Scaolds-Baby 1

Sequence Size Distribu3on

(n=52 scaolds)

GC (%) Distribu3on)

Virulence

An3bio3c resistance

Cell Cycle
Amino Acids
Fa,y Acids/Lipid
Carbohydrate

Regula3on and Cell Signaling

Virulence
Cofactors, Vitamins, Pigments
Clustering-based subsystems
Stress Response
RNA Metabolism

Iron Scavenging
TIII,IV,ESAT Secre3on
An3bio3c Resist, Toxins
Prophage, Mobile Element

Methicillin Resistance
Teicoplanin Resistance
Mercuric Reductase

Latuga MS et al. Beyond bacteria: a study of the enteric microbial consor@um in extremely low
birth weight infants. PLoS One. 2011;6(12):e27858. doi: 10.1371/journal.pone.0027858.

Results: Scaolds-Baby 2

Sequence Size Distribu3on

(n=56 scaolds)

GC (%) Distribu3on)

virulence
An3bio3c resistance

Cell Cycle
Amino Acids
Fa,y Acids-Lipids
Phosph Metabol
Carbohydrates

DNA Metabolism
Cofactors, Vitamins, Pigments
Virulence
Clustering-based subsystems
RNA Metabolism
Protein Metabolism

Iron Scavenging
An3bio3c Resist, Toxins
Pathogenicity Islands
Prophage, Mobile Element

Methicillin Resistance
Teicoplanin Resistance
Fluoroquinolone Resistance
Beta-lactamase

Latuga MS et al. Beyond bacteria: a study of the enteric microbial consor@um in extremely low
birth weight infants. PLoS One. 2011;6(12):e27858. doi: 10.1371/journal.pone.0027858

Microbiome NEC vs no NEC

Weekly stool specimens
Gesta3onal ages 32 completed weeks or birth
weights 1250 g.
High throughput 16S rRNA sequencing
Compare diversity of microbiota and prevalence
of specic bacterial signatures in 9 NEC infants
and 9 matched controls.

Same maturity/chronologic age

454 sequences
Calculated diversity indices and rarefac3on curves.
Opera3onal Taxonomic Units (OTUs)

Mai V, Young CM, Ukhanova M, Wang X, Sun Y, Casella G, Theriaque D, Li N, Sharma R, Hudak M, Neu J. Fecal microbiota in premature
infants prior to necro3zing enterocoli3s. PLoS One. 2011;6(6):e20647

Microbiome NEC vs non NEC

Microbiota composi3on diered in the
matched samples collected 1 week but not
<72 hours prior to NEC.
Detected a bloom (34% increase) of
Proteobacteria and a decrease (32%) in
Firmicutes in NEC cases between the 1 week
and <72 hour samples.
No signicant change was iden3ed in the
controls.
Mai V, Young CM, Ukhanova M, Wang X, Sun Y, Casella G, Theriaque D, Li N, Sharma R,
Hudak M, Neu J. Fecal microbiota in premature infants prior to necro3zing enterocoli3s.
PLoS One. 2011;6(6):e20647

Chao rarefaction diversity.

Mai V, Young CM, Ukhanova M, Wang X, et al. (2011) Fecal Microbiota in Premature Infants Prior to Necrotizing Enterocolitis. PLoS
ONE 6(6): e20647. doi:10.1371/journal.pone.0020647
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0020647

Changes in propor#on of bacterial phyla.

Proteobacteria
Firmicutes
Bacteroidetes
Actinobacteria

Cases increase proteobacteria from 36% of microbiome to 70%

Mai V, Young CM, Ukhanova M, Wang X, et al. (2011) Fecal Microbiota in Premature Infants Prior to
Necrotizing Enterocolitis. PLoS ONE 6(6): e20647. doi:10.1371/journal.pone.0020647
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0020647

Early-onset Sepsis
We choose:
Who we treat
What we use
How long we treat

Duration of Initial Antibiotic Course in

Extremely Low Birthweight Infants:
Association with Necrotizing
Enterocolitis and Death

Results
Median Duration of Empirical Initial Antibiotics
10
9

Median Days

8
7
6
5
4
3
2
1
0
1

10 11 12 13 14 15 16 17 18 19 Total
20

Centers
Cotten CM et al. Prolonged Initial Empirical Antibiotics are Associated with Increased Necrotizing Enterocolitis
and Mortality in Extremely Low Birthweight Infants. Pediatrics.2009; 123: 58-66.

Duke

Results
% with Prolonged Empirical Initial Antibiotics: > 5
days
90

80

Range 27 85%

% 5 or more days

70
60
50
40
30
20
10
0
1

10 11 12 13 14 15 16 17 18 19 20
Total

Centers

DUKE
Now < 70%

FIGURE 1 Numbers of study infants according to duration of initial empirical antibiotic

treatment

Cotten, C. M. et al. Pediatrics 2009;123:58-66

Copyright 2009 American Academy of Pediatrics

Results
Risk factor adjusted associations between NEC or Death and
1. Duration of empirical initial antibiotics
2. Prolonged empirical initial antibiotics
Outcome

Duration empirical
antibiotics

Prolonged empirical initial

antibiotics (> 5 days)

Odds ratio (95% CI)

p-value Odds ratio (95% CI) p-value
(each day)
NEC or
Death
NEC
Death

1.04 (1.02, 1.06)

< 0.01

1.30 (1.10, 1.55)

< 0.01

1.07 (1.04, 1.10)

<0.01

1.21 (0.98, 1.51)

0.08

1.18 (1.08, 1.28)

< 0.01

1.46 (1.19, 1.78)

< 0.01

Cotten CM et al. Prolonged Initial Empirical Antibiotics are Associated with Increased Necrotizing Enterocolitis and Mortality in
Extremely Low Birthweight Infants. Pediatrics.2009; 123: 58-66.

Additional Finding: late onset sepsis

Both 4 and 5 days of initial empirical antibiotic
treatment associated with increased risk of the
combined outcome LOS caused by organisms
other than CONS or death
4 days: OR: 1.32 [95% CI: 1.111.58]
5 days: OR: 1.24 [95% CI: 1.061.44]

Cotten CM et al. Prolonged Initial Empirical Antibiotics are Associated with Increased Necrotizing Enterocolitis and Mortality in
Extremely Low Birthweight Infants. Pediatrics.2009; 123: 58-66.

An#microbial Exposure and NEC

Yale cohort replicates NRN associa#on

2000 -2008
124 cases; 248
controls (1:2)
GA mean: 28 wk
BW mean: 1100 g
Total days on
ABX, not just EOS
empirical therapy
Human milk was
protec3ve

Alexander VN, Northrup V, Bizzarro MJ. An3bio3c exposure in the newborn intensive care unit
and the risk of necro3zing enterocoli3s. J Pediatr. 2011;159(3):392-7.

3rd Ini#al Empirical An#microbial Study

CCHMC

3 hospitals in Cincinna3; 2000 - 2004

< 32 weeks gesta3onal age
< 1500 g birth weight
Survived free of sepsis and NEC for 7 days.
Mul3variable logis3c regression

Outcome variables: NEC, LOS, death + combina3ons

LOS: posi3ve blood, cerebrospinal uid, urine, or sterile site culture aLer 3 postnatal days.
NEC II/III

Prolonged ini3al an3bio3cs dened as abx > 5 days, beginning day 1

Almost universally Amp/Gent

Covariables:

birth weight,
gesta3onal age,
race,
prolonged premature rupture of membranes,
days HFV in 7 days,
amount of breast milk received in the rst 14 days of life.

Kuppala VS. Prolonged Initial Empirical Antibiotic Treatment is Associated with Adverse Outcomes in Premature Infants.
J Pediatr 2011;159:720-5

Prolonged An#bio#cs (III)

Outcome

N (total =
365

OR*

CI

NNH

Late onset
sepsis

76 (21%)

2.45

1.28-4.67

NEC

17 (4.6%)

1.28

0.42-3.93

Death

20 (5.5%)

1.12

0.40-3.10

Composite

91 (25%)

2.66

1.12-6.30

LOS+NEC+Death

*Odds Ra#o for categorical variable, ini#al empirical an#bio#cs > 5 days

Kuppala VS. Prolonged Ini3al Empirical An3bio3c Treatment is Associated with Adverse Outcomes in Premature Infants.
J Pediatr 2011;159:720-5

The Crisis in An#bio#c Resistance

The synthesis of large numbers of an#bio#cs over the past three decades
has caused complacency about the threat of bacterial resistance.
Bacteria have become resistant to an3microbial agents as a result of
chromosomal changes or the exchange of gene3c material via plasmids
and transposons. Streptococcus pneumoniae, Streptococcus pyogenes,
and staphylococci, organisms that cause respiratory and cutaneous
infec3ons, and members of the Enterobacteriaceae and Pseudomonas
families, organisms that cause diarrhea, urinary infec3on, and sepsis, are
now resistant to virtually all of the older an3bio3cs.
The extensive use of an#bio#cs in the community and hospitals has
fueled this crisis.
Mechanisms such as an3bio3c control programs, be,er hygiene, and
synthesis of agents with improved an3microbial ac3vity need to be
adopted in order to limit bacterial resistance.

Harold Neu. The crisis in an3bio3c resistance. Science 1992;257:1064-1073

Pathophysiology of Necro#zing Enterocoli#s.

Neu J, Walker WA. N Engl J Med 2011;364:255-264.

Average Duration of Initial Course of Antibiotic Therapy (Days)

Monitoring An#bio#c Dura#on

6

Target: Less than 5 days

2006

2007

2008

2009

2010

Years:

2011

P <0.001

2012

Individual Prac##oner ABX Dura#on

1.0
Group
CC
EE
FF
GG
HH
II
JJ
KK
OO
QQ
Combined

Percentage of treated patients

0.9
0.8
0.7
0.6

Mean
3.94884
4.52657
4.09881
4.27485
4.56832
5.17241
4.29944
3.61658
4.04233
4.4006
4.30318

Std Error
0.18974
0.22577
0.17986
0.2131
0.17349
0.24473
0.20656
0.18201
0.20254
0.16811
0.06264

Over this period of 3me, one

provider (II) has a tendency to
treat for more than the ve day
target. Provider-specic counseling
in 2011 has resulted in a reduc3on
of an3bio3c days in 2012 (see
following graphic)

0.5
0.4
0.3
0.2
0.1
0
1

10

11 12

Duration of initial antibiotic course (days)

13

14

15

Average Duration of Initial Course of Antibiotic Therapy (Days)

Prac##oner Modies Prac#ce

7
6
5
4
3
2
1
0

2006

2007

2008

2009
Years

2010

2011

2012

An#bio#cs for > 5 days 1st An#bio#c Course

100
90
80
70
ALL

60

DUKE
High

50

Low

40
30
20
10
0

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

Pediatrix Data
An#bio#c Dura#on Nega#ve Cultures
0.6

1997

1998

1999

2000

2001

2002

2003

2005

2006

2007

2008

2009

2010

2011

2011

2004

1997
Peaks in an#bio#c dura#on are
shiDing to the leD shorter
dura#on of empiric use

0.5
0.4
0.3
0.2
0.1
0
0

7
Days

10 11 12 13 14

Pediatrix Slide courtesy of A. Spitzer

2016 Choosing Antibiotics Wisely

70

NICHD NRN ELBW: Late onset sepsis

ALL
High

60

Low
Duke alone

50
40
30
20

36%
19%

10
0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

ELBWAll NEC
NRN overall, high, low, Duke site and Duke alone
as of 9/30/2015
35

ALL
High
Low

30

Duke alone

25

20

15

10

0
2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Extreme Preterm Mortality

70

Duke ELBW inborn infants and NICHD NRN GDB: all sites combined, and high
and low site % mortality for GDB infants
Duke Inborn
NRN GDB

60

NRN GDB High Ctr

NRN GDB Low Ctr

50

40

30

20

10

0
1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

NICHD NRN: Late Onset Sepsis: Gram Posi#ve Organisms

Era 1 (2000-05)
(N = 1896)

Era 2 (2006-11)
(N = 1728)

P-value

1504 (78%)

1386 (79%)

.55

CoNS

973 (50%)

1007 (57%)

Staphylococcus aureus

217 (11%)

212 (12%)

Staphylococcus spp.
(unspecied)

136 (7%)

0 (0%)

Enterococcus spp.

61 (3%)

92 (5%)

Streptococcus spp.
(unspecied)

52 (3%)

0 (0%)

Group B Streptococcus

30 (2%)

45 (3%)

Streptococcus
pneumoniae

1 (0.1%)

1 (0.1%)

Group A Streptococcus

0 (0%)

6 (0.3%)

Clostridia spp.

0 (0%)

1 (0.1%)

Possible
contaminantsa

34 (2%)

22 (1%)

Gram-posi#ve, N (%)

Greenberg R. et al. Late-onset sepsis in extremely premature infants: trends over 3me and
correla3on with mortality. EPAS2014:1540.641

NICHD NRN: ELBW Late Onset Sepsis: Gram nega#ve organisms

Era 1 (2000-05)
(N = 1896)

Era 2 (2006-11)
(N = 1728)

P-value

Gram-nega#ve, N (%)

371 (19%)

329 (19%)

.74

Escherichia coli

103 (5%)

117 (7%)

Klebsiella spp.

92 (5%)

90 (5%)

Enterobacter spp.

Pseudomonas
spp.

74 (4%)

45 (3%)

58 (3%)

30 (2%)

Serra@a spp.

30 (2%)

24 (1%)

Citrobacter spp.

7 (0.4%)

10 (0.6%)

Proteus spp.

4 (0.2%)

6 (0.3%)

Acinetobacter spp.

2 (0.1%)

6 (0.3%)

Stenotrophomonas
maltophilia

0 (0%)

1 (0.1%)

1 (0.1%)

0 (0%)

Possible
contaminantsa

Greenberg R. et al. Late-onset sepsis in extremely premature infants: trends over 3me and correla3on with mortality. EPAS2014:1540.641

2016 Choosing Antibiotics Wisely

Candidiasis and An#bio#c Stewardship

Infants <1000 g birth weight
Candida - 73% death or neurodevelopmental impairment

Pediatrix Cohort of infants <1250 g (N = 21,233 cultures; 6,172 babies) > 3

days old

Variable

Gestational age

Thrombocytopenia
Cephalosporin or
Carbepenem

Category

OR

95% CI

28 wk

Referrent

25-27 wk

2.02

1.52 3.05

< 25 wk

4.15

3.12 6.12

Plt ct > 150

Value <150

Referrent
3.56

2.684.74

no

Referrent

yes

1.77

1.33 2.29

Benjamin DK Jr et al. Empirical therapy for neonatal candidemia in very low birth weight infants. Pediatrics. 2003;112:543-7.

2016 Choosing Antibiotics Wisely

Mortality Following Posi#ve Blood Culture

Pediatrix Database
< 1250 g birth weight
6172 with blood culture aLer
day of life 3
1299 had a posi3ve culture
Candida and Pseudomonas
mortality (33% vs. 13% other
G neg) the worst

Benjamin, DK Jr et al. Mortality following blood culture in premature infants: increased with gram-nega3ve bacteremia and
Candidemia, but not gram-posi3ve bacteremia. Journal of Perinatology 2004;24:175-180.
Benjamin DK Jr et al. Postconcep3on age and other risk factors associated with mortality following Gram-nega3ve rod bacteremia. J
Perinatol. 2004;24:169-74.

An#bio#c risks: First days of life

Empirical Ini3al Broad Spectrum use varies by center

15 % of 165 sites that reported data for >100 patients administered amp and
cefotaxime concurrently for >50% of their patients
Proxies for selection bias may not reflect adequately
the true severity of illness or the therapeutic approach.

Clark RH et al. Pediatrics 2006;117: 67-74

Adjusted Mortality Odds Ratios within gestational-age groups

Cefotaxime use associated odds of death

adjusted for: need for assisted ventilation, anomalies, birth depression, and
estimated gestational age [EGA] within each estimated gestational-age group)
Clark, R. H. et al. Pediatrics 2006;117:67-74
Copyright 2006 American Academy of Pediatrics

2016 Choosing Antibiotics Wisely

Candidiasis and An#bio#c Stewardship

NICHD NRN cohort study: 2,888 inborn ELBWs; 1998-2001
Variable

Correlation with Candidiasis

Correlation coefficient
(p-value)

Candida Rate
Center Range: 2.4% - 20.4%

1
N/A

Average Days of Broad Spectrum Antibiotics

(BSAs) per infant
Range: 0.9 15.42 days

0.6925
(0.0126)

Average Days of BSAs with negative cultures

per infant
Range: 0.86 13.66 days

0.7048
(0.0105)

Co,en CM et al. The associa3on of third-genera3on cephalosporin use and invasive candidiasis in extremely low birth-weight
infants. Pediatrics 2006;118:717-22.

2016 Choosing Antibiotics Wisely

Candidiasis is decreasing

Broad-spectrum an#bio#c use

Fluconazole prophylaxis

Aliaga S et al. Changes in the incidence of candidiasis in neonatal intensive care units. Pediatrics. 2014 Feb;133(2):236-42.

NICHD NRN: Late onset sepsis: Fungal Infec#ons

Era 1 (2000-05)
(N = 1896)

Era 2 (2006-11)
(N = 1728)

P-value

188 (10%)

111 (6%)

<.001

Candida spp.b

182 (9%)

96 (6%)

Other fungic

6 (0.3%)

15 (0.9%)

Fungus, N (%)

Greenberg R. et al. Late-onset sepsis in extremely premature infants: trends over 3me and correla3on with mortality.
EPAS2014:1540.641

NI C H D

N EONATAL R ESEARCH N ETWORK

RG Greenberg, PB Smith, D Chowdhury, NI Hansen, PJ Sanchez,

BJ Stoll, RD Higgins, CM Co,en, and the GDB Subcommi,ee,
on behalf of the Eunice Kennedy Shriver NICHD Neonatal
Research Network (NRN)

Background
In 20092, the NICHD NRN reported that 53% of infants

(1998-2001) had empiric an3bio3c therapy con3nued for

5 days for culture-nega3ve early-onset sepsis
Associated with:
OR (95% CI)
Death

1.46 (1.19-1.78)

Necro#zing enterocoli#s (NEC) or death

1.30 (1.10-1.54)

Late-onset sepsis or death

1.24 (1.06-1.44)

2Cotten

et al., Pediatrics 2009;123:5866

50%
36%

Adjusted OR (95% CI) for birth year: 0.87 (0.85-90)

69%
First report: range was 27 85%

30%

Mul#variable logis#c regression

Outcome
Death, N
Prolonged initial
therapy, n (%)
Death or NEC, N
Prolonged initial
therapy, n (%)
NEC, N
Prolonged initial
therapy, n (%)
Death, NEC, or LOS, N
Prolonged initial
therapy, N (%)
Fungal LOS, N
Prolonged initial
therapy, N (%)

No
4934

Yes
919

2105 (43%)

496 (54%)

4528

1325

1930 (43%)

670 (51%)

5196

668

2301 (44%)

303 (45%)

3457

2400

1447 (42%)

1155 (48%)

1531

120

681 (45%)

71 (59%)

Adjusted OR (95% CI) P-value

1.20 (1.02, 1.42)

0.030

1.10 (0.95, 1.26)

0.196

1.01 (0.84, 1.20)

0.955

0.99 (0.87, 1.12)

0.837

1.50 (0.99, 2.26)

0.054

43% of survivors had prolonged abx compared with 54% of those that died
43 % of survivors without NEC had prolonged abx compared with 51% of those that died
or had NEC
45% of those that did not have fungal sepsis had prolonged abx compared with 59% of
those that had fungal infections

Conclusions
The propor3on of infants receiving prolonged ini3al

empiric an3bio3c therapy

Varied greatly by center
Decreased over 3me
The associa3on of this prac3ce with adverse outcome
remains somewhat consistent with previous reports
These ndings support the con3nued need for emphasis
on an3bio3c stewardship in the NICU

2016 Choosing Antibiotics Wisely

An#microbials: Commonly Used in NICUs

1,425,992 records: 1996 2005.

Ampicillin #1
Gentamicin #2
Cefotaxime #5
Vanco #8

450,000
NICU records: 2005 2010
Ampicillin #1
Gentamicin #2
Vancomycin #4
Cefotaxime #15
Fluconazole #27
CeDaz #38
Pip/Tazo #41

ELBW Infants 2005 - 2010

Gent #1
Amp #2
Vanc # 4
Cefotax #14
Flucon #18

Clark RH et al. Pediatrics 2006; 117:1979-1987

Hsieh EM et al. Am J Perinatology 2014;31:811-2

2016 Choosing Antibiotics Wisely

An#microbials increase in use

From 2005 report to 2010 report, all NICU

admits

70% increase in use of cefepime (4th gen)

97% increase in use of piperacillin-tazobactam
100% increase in use of meropenem
350% increase in use of cefoxi3n
500% increase in use of linezolid
2900% increase in use of azithromycin
Clark RH et al. Pediatrics 2006; 117:1979-1987
Hsieh EM et al. Am J Perinatology 2014;31:811-2

How can we limit how many we

treat?

Management of neonates with suspected or proven early-onset neonatal sepsis

Evaluation of asymptomatic infants <37 weeks gestation with risk factors for sepsis.
The diagnosis of chorioamnionitis is problematic and has important implications for the management of the newborn
infant.

Polin R A , and the COMMITTEE ON FETUS AND

NEWBORN Pediatrics 2012;129:1006-1015
2012 by American Academy of Pediatrics

Management of neonates with suspected or proven early-onset neonatal sepsis

Evaluation of asymptomatic infants 37 weeks gestation with risk factors for sepsis.
The diagnosis of chorioamnionitis is problematic and has important implications for the
management of the newborn infant.

Polin R A , and the COMMITTEE ON FETUS AND

NEWBORN Pediatrics 2012;129:1006-1015
2012 by American Academy of Pediatrics

On the basis of the available data, Polin et

al conclude the following:
Symptoma3c neonates without risk factors for infec3on (who improve
over the rst 6 hours of life) may not require treatment, but must be
monitored closely.
Chorioamnioni#s signicantly increases the risk of early-onset sepsis;
however, the likelihood of sepsis in an infant who appears well at birth
is low.
The risk of sepsis is reduced in infants born to mothers with
chorioamnioni3s who receive intrapartum an3bio3cs, but an3bio3cs may
be less eec3ve once chorioamnioni3s is established.
The intrapartum use of an3bio3cs decreases the sensi3vity of postnatal
blood cultures.
Commonly used laboratory tests have a limited posi3ve predic3ve
accuracy and should never be used as a ra3onale to con3nue treatment in
an otherwise healthy term infant at 48 to 72 hours of life.
Physical examina#on is as good or bever than most laboratory tests in
ruling in or ruling out sepsis.

Polin et al. suggest:

An3bio3cs may be discon3nued in well-appearing term
newborn infants born to women with chorioamnioni3s
by 48 hours of life; treatment of 72 hours might be
considered for infants with greater degrees of
prematurity or abnormal screening studies
A lumbar puncture should be performed in
(1) infants with a posi3ve blood culture,
(2) infants with a high probability of sepsis on the basis of
clinical signs or abnormal laboratory data,
(3) infants who do not clinically improve when treated
with appropriate an3microbial therapy

Conundrum of Early Onset Sepsis: NEW ALGORITHM, Labs, what labs?

Evaluation of asymptomatic infants (any gestational age).

Revised leaves out the continue

antibiotics if labs abnormal
Polin R A et al. Pediatrics 2014;133:1122-1123

2014 by American Academy of Pediatrics

Escobar GJ, Puopolo KM, Wi S et al. Stra#ca#on of risk of early-onset

sepsis in newborns >/= 34 weeks' gesta#on. Pediatrics 2014;133(1):30-36.

It is possible to combine objec3ve maternal data

with evolving objecCve neonatal clinical ndings to
dene more ecient strategies for the evalua3on
and treatment of EOS in term and late preterm
infants.
Judicious applica3on of our scheme could result in
decreased an3bio3c treatment in 80,000 to 240,000
US newborns each year

Slide from Reese Clark

Escobar GJ, Puopolo KM, Wi S et al. Stra#ca#on of risk of

early-onset sepsis in newborns >/= 34 weeks' gesta#on.
Pediatrics 2014;133(1):30-36.

OBJECTIVE: To dene a quan3ta3ve stra3ca3on

algorithm for the risk of early-onset sepsis (EOS) in
newborns >/= 34 weeks' gesta3on.
A retrospec3ve nested case-control study that used
split valida3on. Using a combina3on of recursive
par33oning and logis3c regression, we developed a
risk classica3on scheme for EOS on the deriva3on
dataset. This scheme was then applied to the
valida3on dataset.

Neonatal Sepsis Calculator

Results are quan#ta#ve:

EOS risk at birth

Risks dierent per clinical
appearance
Well appearing
Equivocal
Clinical Illness
Clinical recommenda#on for
each clinical exam scenario

Recommenda#ons include
Culture y/n
An3bio3cs y/n
Vital sign schedule

Clinical Exam

Description

Clinical Illness

1.Persistent need for CPAP / HFNC / IMV (outside the

delivery room)
2.Hemodynamic instability requiring vasoactive drugs
3.Neonatal encephalopathy /Perinatal depression
1. Seizure
2. Apgar Score @ 5 minutes < 5
4.Need for supplemental O2 > 2 hours to maintain oxygen
saturations > 90% (outside of the delivery room)

Equivocal

1.Persistent physiologic abnormality > 4 hrs

1. Tachycardia (HR > 160)
2. Tachypnea (RR > 60)
3. Temperature instability (> 100.4F or < 97.5F)
4. Respiratory distress (grunting, flaring, or
retracting) not requiring supplemental O2
2.Two or more physiologic abnormalities lasting for > 2 hrs
1. Tachycardia (HR > 160)
2. Tachypnea (RR > 60)
3. Temperature instability (> 100.4F or < 97.5F)
4. Respiratory distress (grunting, flaring, or
retracting) not requiring supplemental O2
Note: abnormality can be intermittent

Well Appearing

No persistent physiologic abnormalities

hvp://www.dor.kaiser.org/external/DORExternal/research/Infec#onProbabilityCalculator.aspx

Reducing Reliance on Diagnos#c Tests

Swiss Health System//U of Lausanne
Lab guided tests leads to lots of false posi3ves
treated with ABX.
Vast majority of infants with EOS develop
signs in rst 24 postnatal hours
GBS moms all got CDC guided prophylaxis.
Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk
factors for early-onset neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun
25;144:w13981.

Reducing Reliance on Diagnos#c Tests

Old Guideline: December 2006 September 2009
CBC w/ manual di and CRP in all infants born to mothers
with at least one risk factor for neonatal infec3on:

inadequate GBS prophylaxis,

rupture of membranes >18 hours [PROM],
maternal fever,
prematurity <37 weeks of gesta3on

No specic instruc3ons regarding the 3ming of diagnos3c

tests.
Vital signs were checked by midwives every 4 hours during
the rst 24 hours and every 8 hours during the next 24
hours in all infants with risk factors for EOS.
Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for
early-onset neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Reducing Reliance on Diagnos#c Tests

New Guideline: October 2009 December 2011
Monitoring of vital signs by midwives as before, Plus:
Infants with risk factors for EOS were examined by
paediatric residents every 8 hours during the rst 24 hours.
A CBC was performed only in infants exposed to maternal
chorioamnioni3s,
Chorio: dened as maternal fever > 38 C plus at least two of
the following signs:

maternal heart rate >100/min,

fetal heart rate >160/min,
uterine tenderness,
purulent amnio3c uid,
maternal leucocytosis 15,000/mL

Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for early-onset
neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Reducing Reliance on Diagnos#c Tests

Neonatal infec#on was dened as:
(A) culture-proven infec3on (posi3ve blood and/or cerebrospinal
uid culture) or
(B) probable infec3on with 2 signs of sepsis within the rst 7
postnatal days

temperature instability,
irritability or lethargy,
feeding dicul3es,
capillary rell >2 seconds,
apnoea,
tachycardia and/or tachypnoea),
and elevated CRP >20 mg/l,
and the decision of the avending neonatologist to treat for at least 7
days with intravenous an#bio#cs.

Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for
early-onset neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Reducing Reliance on Diagnos#c Tests

Common to both:
Decision for ABX y/n up to clinician
Blood cultures only in treated pa3ents
CBC and CRP measurements performed during
both study periods in all treated pa@ents to
assist clinicians in determining the dura3on of
an3bio3c treatment.
Lumbar punctures were performed on an
individual basis
Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for early-
onset neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Reducing Reliance on Diagnos#c Tests

N = 11,613 35 weeks (6183 period 1; 5320 period 2)
Maternal characteris3cs in periods 1 and 2 similar.
CRP measurements and CBCs** performed :

26.8 (CRP) and 51.3 (CBC) per 100 live births during Period 1
2.3 (CRP) and 36.1 (CBC) per 100 live births during Period 2 (p
<0.0001).

222 total infants total treated for suspected EOS.

2.1% (132/6183) treated in Period 1
1.7% (90/5,320) treated in Period 2.

27% (36/132) treated longer for infec#on in Period 1

18% (16/90) treated longer for infec#on in Period 2
Period 1: three w/ posi3ve blood cultures, 1 w/ CSF
Period 2: 0 with posi3ve blood or CSF culture.
**CBCs commonly obtained with hyperbilirubinemia work up
Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for early-
onset neonatal sepsis does not delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Reducing Reliance on Diagnos#c Tests

Time between birth and rst dose shorter in Period 2, (Figure)

median 4.5 hours; Q1Q3 2.610.9 hours) compared with Period 1 (median 10.7
hours; Q1Q3 4.126.9 hours; table 2).

Dura#on of ABX in non-infected pa#ents shorter in Period 2 than Period 1

(median 66.6 hours; Q1Q3 62.4127.1 during Period 1 and 64.0 hours; Q1Q3
60.666.8 during Period 2, p = 0.003),

all infants treated for

suspected EOS
Full line: period 1

all infants with

signs of infec#on
Full line: period 1

Duvoisin G, Fischer C, Maucort-Boulch D, Giannoni E. Reduc3on in the use of diagnos3c tests in infants with risk factors for early-onset neonatal sepsis does not
delay an3bio3c treatment. Swiss Med Wkly. 2014 Jun 25;144:w13981.

Summary
More than 95% of the pa3ents we treat with an3bio3cs
have nega3ve cultures,,, even in high risk pa3ents
We use lots of an3bio3cs for lots of days and lots of doses.
With every extra day and every extra dose there is the
poten3al for error
Accumula3ng evidence from several dierent sources
suggest prolonged exposure to an3bio3cs is associated
with NEC, mortality, and a higher likelihood of subsequent
development of infec3ons with resistant organisms
Some hopeful strategies to reduce exposures

From: Variation in Performance of Neonatal Intensive Care Units in the United States
JAMA Pediatr. Published online January 09, 2017.e164396 doi:10.1001/jamapediatrics.2016.4396
Mortality

LOS in VLBWS
Drops from unadjusted rate of 22% to 10%
678 of 695 NICUs (97.6%; 95% CI,
95.8%-99.6%) achieved shrunken adjusted
rates of late-onset infection as low as or lower
than the rate of the best quartile in 2005
Worst performers in 2014 (90th percentile)
were better than best performers (10th
percentile) of 2005!!
ALSO SIMILAR RESULTS FOR MORTALITY

CLD

Late onset sepsis

Severe IVH

NEC

Severe ROP

Figure Legend:
Risk-Adjusted Rates of Outcomes in the Neonatal Intensive Care Unit at the 10th, 25th, 50th, 75th, and 90th Percentiles, 2005-2014These charts illustrate
percentiles of risk-adjusted rates for mortality and neonatal morbidities by year. A, Mortality. B, Chronic lung disease. C, Late-onset infection. D, Necrotizing
enterocolitis. E, Severe intraventricular hemorrhage. F, Severe retinopathy of prematurity.
Copyright 2017 American Medical Association. All
Date of download: 1/9/2017
rights reserved.

Closing thoughts
An3bio3cs save lives
Preven3ng infec3ons is most eec3ve way to
avoid consequences of bacterial overuse
An3bio3c exposures are unavoidable
Follow local epidemiology, but look for best
prac3ces
Con3nue learning to use the EHR and your
clinical skills
QC for prac3ce
Epidemiology of infec3ons
Combining mul3center data to iden3fy highest risk infants

Thanks!
PQCNC Teams and
Leaders
Duke colleagues

Ron Goldberg
Patrick Seed (now a W-cat)
Brian Smith
Danny Benjamin
Susan Latuga (now in NYC)
Noelle Younge
Rachel Greenberg
Eric Benner
Jim Wynn (now a Gator)

NRN colleagues and

mentors
Barbara Stoll
Pablo Sanchez
Rich Polin

Non-NRN colleague and

mentors
Joe Neu

Study team at Duke

Kim Fisher BSN PhD

Joanne Finkle BSN JD
Mandy Marion