Neonatal septicemia is defined as generalized
systemic infection of the newborn,associated with
pure growth of bacteria from one or more sites.
It is one of the most important causes of mortality
and morbidity in newborn.
Among the low birth weight & pre term
babies,neonatal septicemia is the most important
cause of mortality,especially in developing countries
According to NNPD-2,the incidence ranges from 0.1
to 4.5 percent with an overall mortality rate varying
With an estimated 25 million births occuring in India
and approximately 2% incidence of sepsis,the figure
for neonatal sepsis in India would be 5,00,000 babies
every year with about one-third,i.e.170,000 babies
succumbing to the disease.
When pathogenic bacteria gain access into the blood
they may cause overwhelming infection without much
may get predominantly localized to the lung
the meninges (meningitis).
Neonatal sepsis is the single most important cause of
neonatal deaths in the community, accounting for
over half of them.
If diagnosed early and treated aggressively with
antibiotics and good supportive care, it is possible to
save most cases of neonatal sepsis.
Neonatal sepsis can be classified into two sub-types
depending upon whether the onset of symptoms is
before 72 hours of life (early onset) or later (late
Early onset sepsis
Early onset sepsis (EOS) often presents as a
fulminant, multi-system illness within 72 hours of
delivery and is mainly due to bacteria acquired before
and during delivery
Early onset sepsis manifests frequently as pneumonia
and less commonly as septicemia or meningitis.
The associated factors for early-onset sepsis include
low birth weight,
prolonged rupture of membranes > 24 hours,
foul smelling liquor,
multiple per vaginum examinations,
difficult or prolonged labour,
and aspiration of meconium.
Risk factors for neonatal sepsis.
Urinary tract infection,diarrhea,
Prolonged rupture of the membrane>24hrs,
Foul smelling amniotic fluid,
Maternal genital tract infection,& genital colonization,
Socio economic factors:poor socio economic
status,poor nutrition and hygiene.
Prolonged rupture of membranes(PROM):the risk is
reported to be 1% percent compared to a baseline
incidence of 0.1 to 0.5 percent.
Chorioamnionitis:increases the risk of sepsis by 2 to 3
If PROM is associated with chorioamnionitis ,the risk
of sepsis increases by four fold.
Prematurity & low birth weight:pre term babies are
deficient in immunoglobulin
concentration,complement function & phagocytic
They have 3 to 10 times higher risk of developing
sepsis than term infants.
Chorioamnionitis may coexist & may trigger for
Association of chorioamnionitis is and low birth
weight increases the risk of sepsis to 16 percent
compared to association with normal weight babies.
Perinatal asphyxia:asphyxia is associated with
depressed immune function.in addition,several
interventional procedures increase the risk of
Presence of low apgar score (6 or less at 5
minutes)along with prolonged rupture of membranes
has shown to increase the risk of infection by 4% &
Male gender:boys have 2 to 6 times higher risk of
development of neonatal sepsis.
Other factors:maternal fever,G.U.T.infection,poor
socio economic condition & feeding artificial milk.
Late onset sepsis.
Late-onset septicemia is caused by the organisms
thriving in the external environment of the home or
late onset sepsis (LOS) can present as either a
fulminant or a smoldering infection.
The infection is often transmitted through the hands
of the care-providers.
The onset of symptoms is usually delayed beyond 72
hours after birth and the presentation is that of
septicemia, pneumonia or meningitis.
The associated factors of late-onset sepsis include:
low birth weight,
lack of breastfeeding,
superficial infections (pyoderma, umbilical sepsis),
aspiration of feeds,
disruption of skin integrity with needle pricks and use
of intravenous fluids.
These factors enhance the chances of entry of
organisms into the blood stream of the neonates
whose immune defences are poor as compared to
older children and adults.
National neonatology forum of India defines neonatal
sepsis as follows;
Proven sepsis:the baby presents with clinical picture of
sepsis & isolation of pathogens from blood,csf.,urine or
other body fluids or autopsy evidence of sepsis.
Probable sepsis:newborn with clinical picture
suggestive of sepsis with one or more of the following
Existence of predisposing factors,e.g.maternal fever foul
smelling liquor or prolonged rupture of the membrane (>12
hours) or gastric polymorphs more than 6/high power field.
Positive sepsis screen(2 of the 4 parameters to be present).
Total leukocytes count<5,000/mm3,immature to total
neutrophil count ratio>0.2,c-reactive protein positive &
micro ESR > 15 mm/1st hour or >age in days + 3.
Radiological evidence of pneumonia.
Sepsis syndrome:when septicemia is associated with
altered organ perfusion(hypoxia,increased blood
lactate,oliguria & altered mental state)it is termed as
If untreated,this condition leads to early septic shock
with decreased capillary refilling and low blood
pressure,which can be reversible with appropriate
If untreated,this state progresses quickly into
refractory shock & leads to multi organ dysfunction.
The common organisms reported by NNPD Survey
showed 3.8 percent incidence of neonatal sepsis from
pooled hospital data with
Coagulase negative staphylococcus,
Candida as the predominant organisms.
In the Indian subcontinent, the distinction between
EOS and LOS is somewhat blurred (Sundaram V et al).
The clinical presentations of EOS and LOS are
different and the risk factors are different, but the
organisms causing the EOS and LOS are similar and
so are their antibiograms (Zaidi AK et al,NNPD NETWORK).
The early & efficient diagnosis of neonatal bacterial
sepsis remains a difficult task since most of the
symptoms & signs of sepsis in the neonatal period are
of non specific nature.
The spectrum & severity of symptoms required to
decide in the evaluation of sepsis is a matter of clnical
It is estimated that between 11 & 23 non infected
newborns are treated in neonatal intensive care
nurseries for every one with documented infection.
The manifestations of neonatal septicemia are often vague
and therefore demand a high index of suspicion for early
The possibility of sepsis must be considered with any
clinical deterioration unless the event is readily explained
by other causes.
The most common and characteristic manifestation is
An alteration in the established feeding behavior in late
Respiratory distress in early onset sepsis.
The baby, who had been active and sucking well,
gradually or suddenly, becomes lethargic, inactive or
unresponsive and refuses to suckle.
Hypothermia is a common manifestation of sepsis,
whilst fever is infrequent.
Diarrhea, vomiting and abdominal distension may
Episodes of apneic spells or gasping may be the only
manifestation of septicemia.
In sick neonates, the skin may become tight giving a
hide-bound feel (sclerema) and the perfusion
becomes poor (capillary refill time of over 3 seconds).
Cyanosis may appear.
The additional features of pneumonia or meningitis
may be present depending upon the localization of
infection in different systems and organs of the body.
The evidence of pneumonia includes
apneic spells in addition to inactivity and poor feeding.
Cough is unusual.
Findings on auscultation of the chest are non-specific
and non- contributory.
Meningitis is often silent, the clinical picture being
dominated by manifestations of associated
the appearance of excessive or high-pitched crying,
bulging anterior fontanel are highly suggestive of
Classification of sick young infants
for bacterial infection(WHO).
Severe chest indrawing,
Pus draining from ear,
Umbilical redness extending to skin,
Fever or hypothermia,
Many or severe skin pustules,
Lethargic or unconscious,or less than normal movement.
Red umbilicus or draining pus,
local bacteria infection.
Sensitivity:if infection is present,how often is the test
Specificty:if infection is absent,how often is the test
Positive predictive value:if the test result is
abnormal,how often is the infection present?
Negative predictive value:if the test result is
normal,how often is infection absent?
Diagnostic tests with maximal(100%)sensitivity &
negative predictive value are desirable for diagnosis of
This indicates if infection were present,the result
would always be abnormal & if results were
normal,infection would always be absent.
A good specificity & positive predictive value is also
Isolation of organism from blood or a central body
fluid is the standard & most specific method to
confirm the diagnosis of neonatal sepsis.
The main draw back is that,the results are available
after 48 to 72 hrs.
Another dilemma often faced by the clinicians is the
high percentage of cases(upto 40%)where no bacteria
can be isolated from blood even in highly suspected
cses of sepsis.
The standard is to collect 1 ml of blood by sterile
venepuncture in a blood culture bottle.
Umblical artery catheter is preferred over umbilical
vein catheter for sampling.
The blood culture should be incubated for at least 72
hours before being considered negative.
In a study conducted by Geme JW et al,13 percent of
positive blood culture did not grow until beyond 72
The late growing organisms were anaerobes &
coagulase negative staphylococci.
Blood culture should not be considered as the final
arbiter for the clinical diagnosis of neonatal infections
It is negative in 20% cases where infection was proven
by immedite post mortem culture and autopsy.
In the mildly symptomatic neonate and at the earliest
onset of septicemia,the false negative blood culture rate
would be very high.
In upto 50% of cases with congenital bacterial
pneumonia proved by tracheal aspirate culture ,blood
culture is negative.
In a prospective observational study from India, 101
cases of suspected neonatal sepsis were used to
compare the manual method of blood culture with an
automated BacT/Alert system for detection of
The mean times to positivity with the manual and
BacT/Alert 3D systems were 53.1 h and 14.3 h,
respectively (p <0.001)(Hasana AS,et al)
With conventional methods the detection rate is
89.1% by day 2 and 99.5% by day 4.
Cerebro-spinal fluid (CSF)
In an unstable neonate, the LP can be deferred, until
stabilization is achieved.
The cellular and biochemical abnormalities in the
CSF of older patients with bacterial meningitis persist
for up to 3 days.
Gram positive bacteria clear in 36 hours of
appropriate therapy whereas gram negative bacteria
may take up to 5 days.
Apart from culture and gram stain, 4 parameters are
total WBC count (per micro L),
percentage neutrophil count,
Traditionally, the following cut-offs have been used:
more than 60% of polymorphs,
glucose less than 50% of blood glucose,
protein more than 150 mg/dL in term babies and 180
mg/dL in preterm babies.
These cut-offs are no longer acceptable as they are based
on old normative data, and represent an over-simplified
approach based on single cut-off points derived from 2
standard deviation values.
Preterm infants: Treat if CSF WBC count ≥10 OR
glucose <24 OR protein >170.
Do not treat if “CSF WBC count <10 AND glucose ≥25
AND protein <170”.
For in-between results, clinical judgment will have to
be used, keeping in mind clinical features (seizures,
degree of altered sensorium, fullness of fontanelles)
and prematurity (the lower the gestation, lower
should be the threshold for diagnosis).
Term infants: Treat if CSF WBC count >8 OR glucose
<20 OR protein >120.
There is no safe cut-off at which one can recommend
“do not treat”. Clinical judgment as above would have
to be used.(The WHO Young Infant Study Group, Ahmed AS et al,
Smith PB et al, Garges HP et al)
In a neonate with meningitis not showing clinical
recovery after institution of antibiotics, LP should be
repeated after 48 hours.
If the LP is traumatic, the CSF should be sent for
gram stain and culture.
The concentration of glucose is not significantly
altered by a traumatic lumbar puncture.
Therefore a low CSF glucose in the setting of a
traumatic LP is abnormal.
Ideally, the WBC cell count must be performed
within 30 minutes of drawing the sample.
It must be noted that CSF WBC and glucose rapidly
fall with time, giving spurious results(Rajesh NT,et al).
Useful in first 12 hours of life.
A positive tracheal culture may be found in 44% cases
with pneumonia were blood culture was sterile.
There is a positive co-relation of tracheal aspirate
gram stain for bacteria with clinical or pathological
pneumonia and has a 47% predictive accuracy in
Latex particle agglutination tests:a relatively simple
test to perform but is not very sensitive.
The standard kits for group B
streptococcus,pneumococcus and meningococcus are
available & hence cannot be used routinely.
Polymerase chain reaction:a definitive
advancement,but the prohibitive cost & difficult
technique precludes its use in the routine evaluation
of neonatal sepsis.
Nonspecific diagnostic septic
These tests indicate infection without identifying the
infecting micro organisms.
Because of the severity of the disease,it is essential for
these indirect tests not to miss any case(have 100%
To rule out convincingly sepsis when it is not
present(have a high negative value).
White blood cell count.
A total leukocyte count<5000/mm3,
Total neutrophil count<1750/mm3,
An immature to total neutrophil count ratio(I/T)of
>0.2 are suggestive of sepsis.
It has been well documented that the CBC depends
upon the infant’s age,on whether the sample is
arterial or venous,and on whether the infant is crying
This means that for a given infant,a test value is not a
static value;there is considerable intraindividual
In addition,maternal hypertension,perinatal asphyxia
& intra ventricular h’age may cause neutropenia.
Non specific stresses such as asphyxia,maternal fever
or stressful labor can elevate the I/T ratio.
The presence of neutrophil vacuolization or toxic
granulation are also good indicators of neonatal
CRP is a rapidly responsive acute phase
reactant,synthesizd by the liver within 6-8 hrs of
stimulus of inflammatory process.
A positive CRP latex agglutination test corresponds to
plasma CRP concentration of 0.8 to 1.0 mg/dl.
A single value of negative CRP done at the outset may
not be of much significance.
Therefore,n suspected cases,a repeat CRP test is done
12 hours later is more significant.
If the test is negative negative-it almost excludes
Normalization of CRP is a helpful tool in determining
the response to antimicrobial therapy and duration of
Failure to mount a CRP response is a poor prognostic
Micro ESR is an inexpensive & easy bedside screenig
test for neonatal sepsis.
Normal values increase with postnatal age & are equal
to the day of life plus 3mm/hour upto a maximum of
ESR is less sensitive but more specific than CRP or I/T
Other acute phase reactants
Haptoglobin & orosomucoid are acute phase
reactants that have been evaluated for diagnostic aids
for neonatal sepsis,are of limited utility because of
their slower response to infection.
A low plasma fibronectin concentration is
suggestive of neonatal sepsis,
It is also depressed by RDS & perinatal asphyxia,thus
limiting its use.
complex:raises rapidly,100% sensitive but not highly
Direct visualization of bacteria in neutrophils
stained by acridine orange,
NBT reduction by neutrophils,
Gastric aspirate of polymorph count & culture:has
limited value,indicative of high risk because of
exposure to chorioamnionitis.
Newer diagnostic modalities.
Procalcitonin:a pro-peptide,released into the blood
3-6 hrs after endotoxin injection & increses upto 24
Very high serum procalcitonin levels are present in
neonates with proven or clinically diagnosed bacterial
infection;early decrease of these concentrations
reflects appropriate antibiotic therapy.
Compared with CRP,proclcitonin has the advantage
that it increases more rapidly.
A recent meta analysis showed that procalcitonin
showed better accuracy than the CRP test for LOS.
Cytokines and receptors.
Interleukin-1 receptor antagonist(IL-1ra)
The level of these two markers increases two days
before clinical diagnosis of sepsis.
IL-6 although a very early marker,the level becomes
normal later even if infection continues.
The simultaneous determination of CRP can obviate
this problem because the rise in plasma CRP levels
occurs 12 to 48 hours after the onset of infection,at a
time when IL-6 level would have fallen.
Specific leukocyte surface
CD11b-promising marker for diagnosing early onset
CD64-late onset sepsis.
Preterm express CD 64 to the same degree as those
from term infants,children & adults.
G-CSF;a higher level has been found to be associated
with a sensitivity of 95 percent for prediction of sepsis
& a specificity of 73 percent if the level is >200pg/ml.
The best combination for newer markers for sepsis
would be estimation of
IL-6,IL-1ra:1-2 days before the onset of symptoms,
IL-6,IL-1ra,IL-8,CD 11b,G-CSF,TNF,CRP & hematological
indices on day 1,
CRP on the following days to monitor response to
CD 64 is probably one of the most useful infection
markers for diagnosis of late onset nosocomial sepsis.
However,none of the current diagnostic markers are
sensitive and specific enough to influence the
judgement to withhold antimicrobial treatment
independent of the clinical findings.
Approach to a neonate with
There is no rationale for performing a “sepsis screen”
(i.e. CRP,hematological parameters, micro ESR) in
The negative predictive value (NPV) of various sepsis
screen parameters is too low to confidently rule out
EOS (Benitz WE et al, Mahale R et al,).
Procalcitonin and IL-6 are more promising than the
standard screen for the diagnosis of EOS, but they are
currently not easily available on the bedside and are
not considered standards of care.
Neonates who turn symptomatic within 72 hours
must be clinically assessed for probability of sepsis.
Twenty percent of symptomatic neonates in India
suspected to have EOS are blood culture positive
(Mahale R et al).
The following neonates need not be immediately
started on antibiotics but their clinical course must be
Those who are born without any of the known risk
factors of sepsis
premature rupture of membranes (pPROM),
prolonged rupture of membranes (PROM) >18 hrs,
spontaneous preterm onset of labor (SPTOL),
foul smelling liquor,
unclean vaginal examinations,
maternal urinary or other systemic infections,
frequent (>3) per vaginal examinations in labor,
and maternal recto-vaginal group B Streptococcus
Chest X ray is not suggestive of pneumonia,
Have alternative reasons to explain the symptoms.
Those symptomatic neonates with any of the known
risk factors or who have a chest X-ray suggestive of
pneumonia or do not have any alternate explanation
for the signs, must be immediately started on
antibiotics after drawing a blood culture.
Lumbar puncture (LP) for CSF examination must be
performed in all symptomatic neonates, with the
exception of premature neonates presenting with
respiratory distress at birth with no risk factors for
sepsis (Eldadah M et al, Weiss MG et al,)
The decision for performing LP should not be based
on sepsis screen results or blood culture results.
Approach to a neonate with
Neonates who become symptomatic after 72 hours
must be evaluated for LOS.
30% neonates clinically suspected to have LOS in an
NICU setting have positive blood culture (Singh SA et al).
A single episode or transient presence of one of the
above signs may not warrant any action. The more
persistent the sign the more likely it is associated with
LOS (Kudawla M et al,).
Based on clinical assessment the neonate must be
categorized into those with low probability of sepsis
or high probability of sepsis.
The rule of thumb is “low probability” represents
situations where the clinician would be willing to
withhold antibiotics if the sepsis screen is negative.
Those assessed to have a low probability of sepsis (eg.
single episode of apnea or vomiting, but otherwise
well)should undergo a sepsis screen.
The purpose of the sepsis screen is to rule out sepsis
rather than to rule in sepsis.
Traditionally, the sepsis screen consists of 4 items:
C-reactive protein (CRP),
absolute neutrophil count (ANC),
immature to total neutrophil ratio (ITR)
and micro-erythrocyte sedimentation rate (μ-ESR).
CRP: Quantitative CRP assayed by nephelometry is
superior to CRP by ELISA and semiquantitative CRP
by a latex agglutination kit.
Cut-off value for quantitative assay is 10 mg/L.
ANC: It must be read off Manroe’s charts or
Mouzinho’s chart, depending on whether it is a term
baby or a preterm baby respectively(Kudawla M, Mouzinho
ITR: Value above 27% in term babies is considered
positive.18 For preterms, it is considered to be 20%.
ITR is defined as Immature neutrophils (band forms,
metamyelocytes, myelocytes) (Mature + immature
μ-ESR:Value (in mm in first hour) of more than 3+
age in days in the first week of life or more than 10
thereafter is considered positive.
Two systematic reviews on sepsis screens reached the
same conclusions- that there is no ideal test or
combination of tests which achieves an LR+ ≥10 or
LR- ≤0.1, which are the benchmarks of an excellent
test (Schelonka RL et al, Da Silva O et al, Fowlie PW et al).
Among the various tests, quantitative CRP is the best,
followed by qualitative CRP and immature to total
If all the parameters of the sepsis screen are negative
in a neonate assessed to have low probability of LOS,
antibiotics may not be started and the neonate must
be monitored clinically.
The screen must be repeated after 12-24 hours. Two
consecutive completely negative screens are
suggestive of no sepsis.
The practice of designating the screen positive if ≥2
parameters are positive finds its origin in these
Sen Spe LR LR
studies. Year Test
Any 2 + ve of:
>8 mg/l, ESR
>15 mm/1st hr,
Any 2 + ve of:
>0.2, and CRP
> 1 mg/dL
Since we now realize that CRP is the key parameter in
the sepsis screen, a pragmatic approach would be that
if the quantitative CRP alone is positive or any two
parameters of the sepsis screen are positive, a blood
culture must be drawn and empirical antibiotics must
A CSF examination must be performed.
Meningitis occurs in 3.4% cases of suspected LOS and
25% cases of culture positive LOS (Fielkow S et al,Kumar et
Neonates assessed to have a high clinical probability
of sepsis (for which the clinician is convinced that
antibiotics must be started) may not be subjected to a
sepsis screen, because a negative screen would not
alter the decision to start antibiotics.
A CSF examination must be performed.
In recent years, procalcitonin has attracted interest.
Head-to-head comparisons with CRP have shown
that procalcitonin is superior.(Malik A et al, Philip AG et al,
Gerdes JS et al,)
Urinary tract infections (UTI) in
The signs of UTI in neonates are nonspecific and
The common clinical signs in cases of UTI in
neonates are –
failure to thrive (50%),
jaundice (18%) and
irritability & lethargy (17%).
The yield of a routine urine culture is very less ( DiGeronimo
RJ. Tamim MM et al, Lin DS et al.).
Urine culture must always be performed on a sample
obtained by a supra-pubic puncture or by a fresh bladder
In neonates, use of ultrasound guidance simplifies supra-
pubic aspiration and improves the diagnostic yield of
obtaining a urine specimen from 60% to almost 97%(Buys
A practical positive "sepsis screen" takes into account
two or more positive tests as given below:
1. Leukopenia (TLC <5000/cmm)
2. Neutropenia (ANC <1800/cmm)
3. Immature neutrophil to total neutrophil (I/T) ratio
4. Micro ESR (> 15mm 1st hour)
5. CRP +ve
Isolation of microorganisms from blood, CSF, urine,
pleural fluid or pus is diagnostic.
There are a variety of tests which are helpful for
screening of neonates with sepsis.
The most useful and widely used is the white blood
cell count and differential count.
An absolute neutrophil count of < 1800 per cmm is an
indicator of infection.
Neutropenia is more predictive of neonatal sepsis
But it may be present in
and periventricular hemorrhage.
Immature neutrophils (Band cells + myelocytes +
metamyelocytes) to total neutrophils ratio (l/T) > 0.20
means that immature neutrophils are over 20 percent
of the total neutrophils because bone marrow pushes
even the premature cells into circulation, to fight
Platelet count of less than 100,000 per cmm,
Toxic granules on peripheral smear,
Gastric aspirate smears showing more than 5
leucocytes per high power field are also useful
indirect evidences of infection.
The micro-ESR may be elevated with sepsis and fall of
> 15 mm during first hour indicates infection.
Acute phase reactants are also frequently used in
predicting neonatal sepsis.
The most widely used is C-reactive protein (CRP) which
has a high degree of sensitivity for neonatal sepsis.
The CRP can be affected by
prolonged rupture of membranes.
There are a variety of other tests which can be used to
predict sepsis but it may be difficult to perform them
at all places and hence the clinical acumen remains
If possible, lumbar puncture should be done in all
cases of late onset (>72 hours) and symptomatic early
onset sepsis because 10-15 percent of them may have
At a small hospital, one may only depend on the CSF
The implications of detecting meningitis in the
setting of septicemia include:
the need for using antibiotics with a high CSF
and provision of antibiotic treatment for at least 3
weeks, administered parenterally throughout.
No investigation is required as a prerequisite to
start treatment in a clinically obvious case.
Early treatment is crucial.
Institution of prompt treatment is essential for
ensuring optimum outcome of neonates with sepsis
who often reach the health care facilities late and in a
Supportive care and antibiotics are two equally
important components of the treatment.
It should be realized that antibiotics take at least 12 to
24 hours to show any effect and it is the supportive
care that makes the difference between life and death
early in the hospital course.
Supportive care care is to
The purpose of supportive
normalize the temperature,
stabilize the cardiopulmonary status,
prevent bleeding tendency.
There is no role of intravenous immunoglobulin
therapy in neonatal sepsis.
Antibiotic therapy – empirical,
upgradation and modification.
There is generally no distinction in the choice of
empirical antibiotics, be it suspected EOS or LOS as
the bacterial and sensitivity profile in India seems to
be is similar in both situations(Sundaram V et al, Zaidi AK et
al, NNPD network.).
Starting empirical antibiotics.
As the profile of organisms is similar for EOS and
LOS, the following policies can be used irrespective of
whether it is EOS or LOS.
Policy for community acquired sepsis (Paul VK,et al)
Ampicillin + Gentamicin/Amikacin (empirical)
If evidence of staphylococcus : Cloxacillin +
If evidence of meningitis: Add Cefotaxime.
Policy for nosocomial sepsis:
It is not possible to suggest a single antibiotic policy
for use in all newborn units.
Every newborn unit must have its own antibiotic
policy based on the local sensitivity patterns and the
profile of pathogens.
Preferably choose Penicillin plus an Aminoglycoside
Cephalosporins rapidly induce the production of
extended spectrum β-lactamases (ESBL),
cephalosporinases and fungal colonization.
Antibiotic therapy should cover the common
causative bacteria, namely,
Staphylococcus aureus and
A combination of ampicillin and gentamicin is
recommended for treatment of sepsis and
In cases of suspected meningitis, cefotaxime should
be used along with an aminoglycoside.
In late-onset sepsis to cover nosocomial
staphylococcal infection, first line of antibiotics may
comprise of cloxacillin 100 mg per kg per day and an
aminoglycoside (gentamicin or amikacin).
In nosocomial sepsis, antibiotic sensitivity pattern of
organisms responsible for nursery infection should be
known and the antibiotic therapy should be started
Usually staphylococci and Gram negative bacilli
(Pseudomonas, Klebsiella) should be covered using
aminoglycoside (gentamicin or amikacin) and a third
generation cephalosporin (cefotaxime).
For resistant staphylococcal infection, vancomycin
(30 mg per kg per day) should be used.
On confirmation of sensitivity pattern, appropriate
antibiotics are used singly or in combination.
In a baby in whom the antibiotics were started on low
suspicion, these may be stopped after 3 days, if baby is
clinically well and the culture is negative.
However, if a baby appears ill even though the
cultures are negative, antibiotic therapy should be
continued for 7 to 10 days as bacterial infection can
occur with negative cultures.
The duration of antibiotic therapy in sepsis depends
upon the pathogen, site of infection and the clinical
response of the baby.
7-10 days therapy is required for soft tissue infections
Deep-seated infections (osteomyelitis) and
meningitis may require therapy for 3-6 weeks.
Upgradation of empirical
Empirical upgradation must be done if the expected
clinical improvement with the ongoing line of
antibiotics does not occur.
At least 48-72 hours period of observation should be
allowed before declaring the particular line as having
If any new sign appears and/or the existing signs fail
to begin remitting, it would be considered that the
expected clinical improvement has not occurred.
Current evidence does not support the use of serial
quantitative CRP as a guide for deciding whether or
not antibiotics should be upgraded empirically.
In case the neonate is extremely sick or deteriorating
very rapidly and the treating team feels that the
neonate may not able to survive 48 hours in the
absence of appropriate antibiotics, a decision may be
taken to bypass the first line of antibiotics and start
with the second-line of antibiotics.
Antibiotic therapy once culture
report is available.
It must first be assessed whether the positive blood
culture is a contaminant. The following are suggestive
growth in only one bottle (if two had been sent),
growth of a known non-pathogen: eg. aerobic spore
mixed growth of doubtful significance and
onset of growth beyond 96 hours in the absence of a
history of prior exposure of antibiotics in the 72 hours
before sending the blood culture.
If the growth is a non-contaminant, the antibiotic
sensitivity must be assessed to decide whether
antibiotics need to be changed or not.
The following guidelines would allow a rationale use
If the organism is sensitive to an antibiotic with a
narrower spectrum or lower cost, therapy must be
changed to such an antibiotic, even if the neonate was
improving with the empirical antibiotics and/or the
empirical antibiotics are reported sensitive.
If possible, a single sensitive antibiotic must be used,
the exception being Pseudomonas for which 2 sensitive
antibiotics must be used.
If the empirical antibiotics are reported sensitive, but
the neonate has worsened on these antibiotics, it may
be a case of in vitro resistance.
Antibiotics may be changed to an alternate sensitive
antibiotic with the narrowest spectrum and lowest cost.
If the empirical antibiotics are reported resistant but
the neonate has improved clinically, it may or may not
be a case of in-vivo sensitivity.
In such cases are careful assessment must be made
before deciding on continuing with the empirical
One must not continue with antibiotics with in vitro
resistance in case of Pseudomonas, Klebsiella and
MRSA; and in cases of CNS infections and deep-seated
If no antibiotic has been reported sensitive, but one or
more has been reported ‘moderately sensitive’, therapy
must be changed to such antibiotics at the highest
permissible dose. Use a combination, in such cases.
Duration of antibiotics.
Culture positive sepsis: Give sensitive antibiotics for
total duration of 10-14 days.
There is no definitive published literature regarding
the optimum duration of antibiotics for neonatal
neonates infected by Staphylococcus aureus require
14 days of antibiotics.
infected by non-Staphylococcus aureus organisms,
without meningitis or deep-seated infections, and
who become completely asymptomatic by day five,
one may consider a shorter duration of antibiotics.
Culture negative sepsis:
Asymptomatic neonate at risk of EOS: stop
Suspected EOS or LOS and the neonate becomes
completely asymptomatic over time: stop antibiotics.
Suspected EOS or LOS and the neonate improves but
does not become asymptomatic: repeat a CRP assay
(Ehl S et al, Engle WD et al, Saini SS et al)
If CRP + ve: continue antibiotics
If CRP –ve: stop antibiotics
Suspected EOS or LOS and the neonate have not
improved or have worsened: upgrade antibiotics as
per the empiric antibiotic policy. Simultaneously,
alternative explanations for the clinical signs must be
Gram stain-proven meningitis or meningitis
suspected on CSF examination:
Give total of 21-day course of parenteral antibiotics
that cross uninflamed meninges.
Anti-meningitic doses must be used throughout the
course and use only antibiotics with a proven in vitro
Monitoring protocol following
diagnosis of meningitis:
At least twice weekly head circumference monitoring,
Input/Output monitoring, daily weight monitoring (for
Daily neurological examination (focal neurological
Hearing screen after 4-8 weeks,
Ultrasound head in the first week and at the end of
antibiotic therapy (look for ventricular size, ventricular
wall enhancement, midline shift, intraventricular debris).
Ventriculitis may require 6 weeks of antibiotics.
CECT head may be required in case of rapidly rising OFC
with suspicious USG, focal seizures, focal neurological
deficits or infection with Citrobacter koseri and
May be treated for 7-14 days. This duration has no
evidence to back it.
Start empirical treatment with
Cefotaxime/Ceftriaxone plus Amikacin, and modify as
per culture report.
Nalidixic acid or nitrofurantoin should not be used to
treat UTI since they do not achieve therapeutic
concentrations in the renal parenchyma and blood
UTI occurring in the setting of generalized septicemia
may not be associated with VUR or malformations.
However, an isolated UTI could be associated with
Hence, after treatment of isolated UTI, all cases must
be started on Amoxycillin 10 mg/kg once a day oral
prophylaxis, till such time that a renal ultrasound,
MCU and DMSA scan are performed to exclude VUR
Proven bone or joint infections:
Must be treated for at least 6 weeks.
Start empirical treatment with Cloxacillin or
Vancomycin (plus an aminoglycoside for first 1-2
weeks) and modify as per culture report.
Of this, at least 4 weeks must constitute parenterally
administered antibiotics. The rest of the course may
be enterally administered.
a) Intravenous immunoglobulins (IVIG): The
currently available evidence does not support the use
of IVIG. The currently undergoing International
Neonatal Immunotherapy Study is expected to be
provide some important and definitive information in
this aspect(The INIS Study).
b) Colony stimulating factors: There is currently
no evidence to support the use of colony stimulating
factors either as a treatment modality or as a
c) Blood Exchange Transfusion (BET): BET may be
performed in a case of deteriorating sepsis with
sclerema provided the general condition of the baby
allows the procedure .
Superficial infections with local
Superficial infections can be treated
application of antimicrobial agents.
Pustules can be punctured with sterile needles and
cleaned with spirit or betadine.
Purulent conjunctivitis can be treated with neosporin
or chloramphenicol ophthalmic drops.
Oral thrush responds to local application of
clotrimazole or nystatin (200,000 units per ml) and
Superficial infections must be adequately managed; if
neglected they can lead to sepsis or even an epidemic.
Prevention of infections
A good antenatal care goes a long way in decreasing
the incidence, morbidity and mortality from neonatal
All mothers should be immunized against tetanus.
All types of infections should be diagnosed early and
treated vigorously in pregnant mothers.
Babies should be fed early and exclusively with
expressed breast milk (or breastfed) without any
Cord should be kept clean and dry. Unnecessary
interventions should be avoided.
This is the simplest and the most effective method for
control of infection in the hospital.
All persons taking care of the baby should strictly
follow hand washing policies before touching any
The sleeves should be rolled above the elbows. Rings,
watches and jewellery should be removed.
Wash hands up to elbows with a thorough scrub for 2
minutes with soap and water taking care to cover all
areas including the under surface of well trimmed
Rinse thoroughly with running water. Dry hands with
sterile hand towel/paper towel.
Wash hands up to the wrist for 20 seconds in
Hands should be rewashed after touching
contaminated material like one’s face, hair, papers
It is preferable to use bar soaps rather than liquid
soaps as the latter tend to harbor organisms after
In emergency situations bactericidal and virucidal
solutions like Sterillium can be used to clean hands
before touching babies.
Surgical, elbow- operated taps should be used in the
hospitals for hand washing.
Prevention of infection in hospital
The nursery environment should be clean and dry
with 24 hour water supply and electricity.
There should be adequate ventilation and lighting.
The nursery temperature should be maintained
Overcrowding should be avoided.
All procedures should be performed after wearing
mask and gloves.
Unnecessary invasive interventions such as needle
pricks and setting up of intravenous lines should be
kept to the barest minimum.
There should be no compromise in the use of
disposables. Stock solutions for rinsing should be
Every baby must have separate thermometer and
stethoscope and all barrier nursing measures must be
Strict house-keeping routines for washing,
disinfection, cleaning of cots and incubators should
be ensured and these policy guidelines should be
available in the form of a manual in the nursery.
The use of prophylactic antibiotics for prevention of
nosocomial infections is strongly condemned.
They are not only useless but also dangerous because
of the potential risk of emergence of resistant strains
Control of outbreak
General measures for the control of an outbreak
detailed epidemiological investigations,
increased emphasis on hand washing,
review of protocols,
procedures and techniques,
disinfection and sterilization of nursery
assessment of the need for additional measures.
The nursery may be fumigated using formalin 40%
and potassium permanganate (70 gms of KMNO4
with 170 ml of formalin for 1000 cubic feet area).
Alternatively, bacillocid spray for 1-2 hours may be
used. Linen and cotton should be washed thoroughly,
dried and autoclaved.
Use of disposable items for invasive and non- invasive
interventions (catheters, probes, cannulae, chest
tubes etc.) though costly, reduces the risk of
Depending upon the pathogen and type of outbreak,
culture surveys of susceptible patients, cohorting of
infants in nursery and a review of antibiotic policy
may be necessary.
Most of the times a scrupulous reinforcement of
general control measures may be sufficient to stop the
In conclusion, manifestations of neonatal sepsis are
A high index of suspicion with or without lab
evidences of infection is the key for early diagnosis.
Prompt institution of antibiotic therapy and
supportive care will save most of the cases of neonatal
Take home messages.
In India, both early and late onset sepsis are caused
by similar organisms with similar antibiotic
Clinical features of neonatal sepsis are non-specific
and any unexplained clinical deterioration should be
investigated for sepsis.
There is no role of performing sepsis screen in early
onset neonatal sepsis.
Lumbar puncture (LP) for CSF examination must be
performed in all symptomatic neonates being
initiated on antibiotics, with the exception of
premature neonates presenting with respiratory
The traditional cut-offs for interpretation of
cerebrospinal fluid values are based on relatively old
studies with methodological problems.
A new set of guidelines for interpretation is proposed.
Routine urine culture in all neonates with nonspecific symptoms is not recommended.
Every newborn unit must have its own antibiotic
policy based on the local sensitivity patterns and the
profile of pathogens.
Apart from appropriate antibiotics, the survival of a
sick septic newborn often depends upon aggressive
There may be a potential role of intravenous
immunoglobulins in treatment of neonatal sepsis but
larger studies are required.
There is currently no role for the use of colony