Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
1
ORIGINAL STUDIESStudy of bacterial flora in the initial oropharyngeal aspirate and
blood of potentially septic neonates
Pauline Alonday-Pugay, MD, Anna Lisa Ong-Lim, MD...................1
Retrospective analysis of the clinical profile and immediate outcome
of pediatric patients aged three months to 17 years, diagnosed with
pleural effusion or empyema thoracis, admitted to a tertiary
government hospital during january 1998 to december 2000
Dexter D. Cheng, MD, Alexander O. Tuazon, MD.........................6
A randomized placebo-controlled trial on the use of probiotics in
the prevention of nosocomial infection in pediatric patients with
hematologic and oncologic diseases
Arlene Grace L. Simon, MD, Jossie Rogacion, MD.......................10
A five-year retrospective study on the common microbial isolates
and sensitivity pattern on blood culture of pediatric cancer patients
admitted at the philippine general hospital for febrile neutropenia
Ailyn T. Isais-Agdeppa, MD, Lulu Bravo, MD................................17
Clinical correlation of neonatal and maternal hematological
parameters as predictors of neonatal sepsis
Willa Antoniette B. Mayuga, MD, Pura Flor D. Isleta, MD............21
Serum concentration of pyrazinamide suspension in children with
tuberculosis: A therapeutic drug monitoring
Daisy O. Sanchez, MD, Cecilia C. Maramba-Untalan, MD*..........42
The analysis of clinical and social profile of congenital rubella
syndrome seen among up-pgh patients from the years 1993 -
2002 (a 10 year prevalence review)
Coralee Lianko Agnas , MD.........................................................49
PIDSP JOURNAL
Vol. 9 No. 2
July - December 2005
PEDIATRIC INFECTIOUSDISEASE SOCIETY OF
THE PHILIPPINES
*Department of Pediatrics, UP-PGH
Keywords: oropharyngeal aspirate, blood culture, potentially septic neonates
STUDY OF BACTERIAL FLORA IN THE INITIAL OROPHARYNGEAL
ASPIRATE AND BLOOD OF POTENTIALLY SEPTIC NEONATES
Pauline Alonday-Pugay, MD*, Anna Lisa Ong-Lim, MD*
ABSTRACT
There are a considerable number of admissions
of neonates because of vertical transmission of infec-
tion. The sterility of amniotic fluid is questionable in cases
of premature rupture of membrane (PROM), spontane-
ous preterm labor, and infection. A Prospective Risk
Analytical Study was conducted to evaluate the use of
oropharyngeal aspirate in the diagnosis of sepsis in po-
tentially septic neonates born at OB Admitting Section,
Philippine General Hospital from March to August 2004.
Included were 107 neonates, of these, 15 (14.02%) have
maternal risk factor of PROM; 34 (31.78%) have ma-
ternal risk factor of spontaneous preterm labor alone; 38
(35.51%) have maternal risk factor of UTI; 15 (14.02%)
have maternal risk factor of respiratory tract infection;
and 2 (4.6%) have maternal risk factor of fever alone.
Immediately after birth, fluid from the oropharyngeal are
and blood was sent for culture. The neonates were fol-
lowed up within 72 hours, and it was noted that 50
(46.73%) developed jaundice, 17 (15.89%) had pneu-
monia, 7 (6.54%) had necrotizing enterocolitis (stage IA
to IV) and 3 (2.80%) had decreased activity, tempera-
ture instability and hypoglycemia, collectively grouped
under sepsis. Of the 107 subjects, 33.6% have oropha-
ryngeal aspirate growth and 8.4% have blood culture
growth. Among the subjects who have oropharyngeal
aspirate growth, 25% have a concomitant blood CS
growth, irregardless of the organism isolated. Of the
subjects who had growth on their blood culture, 77%
have the same organism isolated in their oropharyngeal
aspirate. PROM and UTI highly correlated with posi-
tive growth in the oropharyngeal aspirate. Acinetobacter
baumanii and Alkaligenes faecalis isolated from
oropharyngeal aspirate has 100% correlation with blood
culture and E. coli isolated from the oropharyngeal aspi-
rate correlated 83.33% with blood culture. The sensitiv-
ity of oropharyngeal aspirate culture growth irregardless
of the organism isolated is 100% and the specificity is
75%. In conclusion, oropharyngeal aspirate when posi-
tive for growth may not reflect growth in blood culture,
and when negative for growth highly correlates with a
blood culture that has no growth.
INTRODUCTION
One of the biggest goals of doctors in every
institution is to decrease the incidence of neonatal
septicemia. In the PGH alone, the statistics points
unfavorably to the high number of neonatal deaths due to
infection. Early detection by surveillance is a must to
eliminate high mortality rates caused by infection.
Neonatal mortality due to neonatal infection
brought about by maternal infection is being faced hand
in hand by pediatricians and obstetricians. Subclinical
ascending infections through the lower female genital tract
are predominant worldwide. The importance of prenatal
check-up has been emphasized, as well as proper
hygiene. But still, there are a considerable number of
admissions of neonates in the Neonatal Intensive Care
Unit because of vertical transmission of infection. Usual
source of bacterial infection are the colonizers. The
bacteria load may increase and thus, may predispose the
immunologically immature newborn to sepsis. The overall
mortality rate of neonates with congenital neonatal sepsis
ranges between 25 to 90%.
The amniotic fluid is normally sterile, but its
sterility is questionable in cases of premature rupture of
membrane, spontaneous preterm labor, and infection,
especially genitourinary infection and respiratory tract
infection.
Preterm labor occurs in 20% of pregnancies in
low-income countries whereas prelabor rupture of
membranes and septicemia may occur in 5-10% of such
settings. A wide variety of bacteria present in the normal
vaginal flora of pregnant women such as anaerobes and
Escherichia coli can also cause ascending infections,
usually after rupture of membranes, resulting in
intraamniotic infection. Chorioamnionitis resulting from
such infections can lead to preterm labor and fetal
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
4
morbidity and in the newborn, were important causes of
septicemia and meningitis.
Considerable evidence derived from animal
studies shows administration of bacteria in pregnant
animals can induce preterm labor and subsequent
neonatal morbidity or mortality. Systemic maternal
infections have also been associated with preterm labor
and neonatal complications. The rate of preterm delivery
associated with pneumonia ranges from 15-48%. The
Gold standard for diagnosing intrauterine infection is a
positive culture of amniotic fluid collected aseptically via
transabdominal amniocentesis which that is invasive.
Microorganism gain access to the amniotic cavity by
ascending from vagina and cervix, hematogenous
dissemination (transplacental), retrograde seeding from
peritoneal cavity through the fallopian tube, and accidental
introduction at the time of invasive procedure. Of these
mentioned, the most common pathway is the ascending
route of infection. Studies have shown that in virtually
all cases of congenital pneumonia, inflammation of
chorioamniotic membrane is present. Other studies also
show that bacteria identified in cases of congenital
infection are similar to the bacteria isolated from the lower
genital tract.
Intraamniotic infection is difficult to diagnose.
Clinical criteria includes maternal fever, maternal or fetal
tachycardia. The infection may be polymicrobial, but
collecting amniotic fluid samples without contamination
of normal vaginal flora requires invasive procedure. A
number of studies have shown a correlation between
vaginal infections and preterm birth, The most advanced
and serious stage of ascending infection is fetal infection.
The nose and the external ear canal in the
newborn infants have bacterial colonizers and in one study
done among infants < 7 days old, they noted that the
newborn’s nose has this pattern of bacterial colonization:
normal flora (39%); potentially pathologic gram-positive
microorganisms (23%); gram-negative enteric rods
(16%); and sterile cultures (22%). The most frequent
bacterial isolates from nostrils were Staphylococcus
epidermidis (39%); Staphylococcus. aureus (11%) and
Escherichia coli (8%). The cultures from the external
ear canal showed normal flora (37%); potentially
pathologic gram-positive microorganisms (5%); gram-
negative enteric rods (24%); and sterile cultures (34%).
The most frequent bacteria cultured from the ear canal
were: Staphylococcus. epidermidis (37%); E. coli
(8%); and Klebsiella pneumoniae (7.5%).1
In another study, Gram positive septicemia the
commonest site for colonization of the same organism in
newborn was found to be umbilicus followed by nose,
throat and external auditory canal, while in Gram negative
septicemia rectum and umbilicus were the commonest
sites.2
Oropharyngeal fluid aspirates have not yet been
explored yet. In utero, the fetus swallows amniotic fluid.
The aspirate that is present in the newborns oropharyngeal
area may represent the environmental milieu that the
newborn is exposed to while in utero. Swallowing
potentially septic amniotic fluid may predispose the
newborn to sepsis. Thus, will an initial oropharyngeal fluid
aspirate culture reflect the blood culture of potentially
septic neonates?
DEFINITION OF TERMS
Potentially septic neonates included in the study
are those that have the maternal risk factors of infection
that occurred in the last trimester of pregnancy. This
includes maternal fever alone, maternal urinary tract
infection (dysuria, vaginal discharge), maternal respiratory
tract infection, prelabor rupture of membranes, and
spontaneous preterm labor. Maternal fever as a risk factor
is defined in this paper as the presence of temperature
of more than 37.5° C with no other symptoms and signs
pertaining to a focus of infection. Prelabor rupture of
membranes denotes spontaneous rupture of fetal
membranes before the onset of labor, whether term of
preterm. Pregnancies with intact fetal membranes and
spontaneous preterm labor must be distinguished for
clinical as well as for research purposes from those in
which there has been rupture of membranes. Preterm
labor is precisely defined as spontaneous onset of labor
occurring before the 37 completed weeks or 259
completed days of gestation, without any other symptoms
and with intact fetal membranes. If the pregnant mother
was having symptoms of dysuria with fever, and had a
midstream, clean-catch urinalysis showing elevated Urine
WBC of more than 5 per high power field and/or urine
culture with more than 100,000 colony-forming units of a
single organism per 1mL of urine then she was considered
as having maternal urinary tract infection. The presence
of cough and/or colds with associated fever was
categorized as having respiratory tract infection.
Neonates were followed up and their morbidities
that occurred within 72 hours were noted. Jaundice for
the purpose of this study was defined as having yellow
skin discoloration with an elevated total bilirubin levels
increasing >5mg/cL per day, >12 mg/dL in a term infant
and >15mg/dL in a preterm infant, direct bilirubinemia
>2mg/dL and the jaundice lasted for more than 1 week.
Infants categorized as having jaundice had jaundice alone
as their main symptom with no other associated
symptoms of disease Those that were labeled as having
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
5
pneumonia were neonates that developed respiratory
distress, rales on lung fields and abnormal temperature
patterns. Necrotizing enterocolitis was defined as
feeding intolerances (residuals, abdominal distention,
vomiting) that were associated with plain abdominal
studies supporting the diagnosis (presence of ileus,
pneumatosis intestinal or pneumoperitoneum). The
neonate was labeled as having sepsis if the neonate had
a blood culture isolate and had clinical signs and symptoms
compatible with sepsis like hypothermia, hyperthermia,
poor suck, lethargic.
The use of initial oropharyngeal fluid aspirate as
an alternate means to reflect the possible isolate in the
neonate is thought of because the fluid in that anatomical
body part reflects the intrauterine environment that the
neonate was exposed to. This aspirate was the first
isolate that we usually obtain in the newborn because
suctioning of the oropharyngeal area was part of initial
resuscitation. And because the collection of the specimen
is done within minutes from birth, it is not much affected
by external, internal and temporal factors.
Only symptoms that developed within 72 hours
of life were noted since symptoms after 72 hours of life
were more of acquired than congenital. Presence of
growth in the blood culture was the gold standard in
determining infection.
OBJECTIVES
This study aims to identify bacterial isolates in
the oropharyngeal fluid of potentially septic neonates
delivered from mothers with infection risks. The initial
blood culture will be correlated with the oropharyngeal
fluid isolate. This study aims to identify if the initial
oropharyngeal fluid aspirate correlates with the causative
organism causing sepsis in the neonate. If this has a
relationship then, the use of a more non-invasive and less
expensive laboratory examination as an alternate is an
option.
This study aims to see if development of
symptoms of infection within 72 hours of life correlates
with the presence of growth in the oropharyngeal fluid
aspirate. Symptoms developing immediately within 72
hours of birth and the result of oropharyngeal fluid aspirate
may predict the result of the blood culture.
METHODOLOGY
Neonates admitted via vaginal delivery in the
OBAS, PGH were be selected. Patients that were
excluded were those who had no maternal risk for
infection and those born via caesarian section. Those
included in the study were patients whose mothers had
a high risk of amniotic fluid contamination like maternal
infection in the last trimester, prelabor rupture of
membrane or preterm labor.
The sample size was computed assuming
maximum variability with a margin of error of 0.1. The
sample size needed for the study is 96 subjects.
A protocol was submitted to the adviser and a
copy was sent to Research Information and Development
Office for ethical review and for registration.
Informed consent was obtained prior to the
delivery of the newborn subject. The fluid in the
oropharyngeal area was obtained observing the aseptic
technique. A syringe was attached to a suction tip to
aspirate the oropharyngeal fluid before the newborn was
suctioned using high negative pressure wall suction. The
specimens were placed in a sterile vial with 1cc of sterile
NSS and was sent to the Microbiology Laboratory for
bacterial studies. Concommitantly, a baseline blood
culture was extracted, using only 1 site and 1 bottle. The
specimen was sent to the Microbiology Laboratory for
culture.
Factors that were considered and noted during
the neonate’s hospital stay during the first 72 hours of
life were the development of illness such as jaundice,
necrotizing enterocolitis, pneumonia and sepsis.
The study design is a prospective risk analytical
study. The difference between the presence or absence
of morbidities and maternal risk factors were compared
using independent samples t-test.
RESULTS
A total of 107 subjects fit the inclusion criteria.
Table I below shows the maternal risk factors of babies
who were included in the study
The neonates were followed up within 72 hours
and the table II shows the morbidities that occurred.
Table I- Maternal Risk Factors
Maternal risk Frequency Percent
Factors
Prelabor Rupture of Membranes 15 14.02
Spontaneous Preterm Labor 34 31.78
Urinary Tract Infection 38 35.51
Respiratory Tract Infection 15 14.02
Fever alone 2 4.67
Total 107 100.00
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
6
Table II- Neonatal Morbidities
Neonatal Morbidities Frequency Percent Cum
None 30 28.04 28.04
Jaundice 50 46.73 74.77
Pneumonia 17 15.89 90.65
Necrotizing enterocolitis 7 6.54 97.20
Sepsis 3 2.80 100.00
Total 107 100.00
Of the 107 subjects, 33.6% had oropharyngeal
aspirate growth and 8.4% had blood culture growth.
Among the subjects who have (+) oropharyngeal aspirate
growth, 25% have a concomitant blood CS growth,
irregardless of the organism isolated. Of the subjects
who had growth on their blood culture, 77% had the same
organism isolated in their oropharyngeal aspirate.
Table III- Oropharyngeal aspirate growth
Sensitivity: 1.0000 (0.6637, 0.9910)
Specificity: 0.7245 (0.6237, 0.8074)
Predictive Value Positive: 0.2500
Predictive Value Negative: 1.0000
Lower bound Upper bound
Likelihood
Ratio (+): 3.6296 2.1628 4.8900
Likelihood
Ratio (-): 0.0000 0.0080 0.5160
There was a significant difference among the
maternal risk factors with regards to presence of
oropharyngeal aspirate growth. Of all the maternal risk
factors, prelabor rupture of membrane has the highest
positive to negative ratio of oropharyngeal aspirate
growth.
Oropharyngeal aspirate Frequency Percent
None 71 66.36
Bacillus sp 2 1.87
Candida 2 1.87
Escherichia coli 6 5.61
Klebsiella pneumonia 2 1.87
Klebsiella rhinoscleromatis 2 1.87
Staphylococcus aureus 2 1.87
Streptococcus epidermidis 12 11.21
Streptococcus viridans 3 2.8
Acinetobacter baumanii 2 1.87
Alkaligenes faecalis 2 1.87
Enterobacter claocae 1 0.93
Total 107 100.00
Table IV- Blood culture growth
Blood Culture Frequency Percent
None 98 91.59
Escherichia coli 5 4.67
Streptococcus epidermidis 1 0.93
Acinetobacter baumanii 1 0.93
Alkaligenes faecalis 2 1.87
Total 107 100.00
Table V- Sensitivity and specificity of Oropharyngeal
aspirate culture as plotted against Blood Culture
Test Blood Blood Total
CS (+) CS (-)Oropharyngeal aspirate
culture positive 9 27 36
Oropharyngeal aspirate 0 71 71
culture negative
Total 9 98 107
Table VI- Maternal Risk Factor Contributing to Presence of
Oropharyngeal aspirate Growth
15
34
38
15
5
107
Fisher’s exact = 0.001
Maternal Risk Factor Oropharyngeal Total
aspirate
No With
growth(%) growth
Prelabor Rupture of
Membranes
Preterm Labor
Urinary Tract Infection
Respiratory Tract Infection
Fever
Total
3 12
20% 80%
27 7
79.41% 20.59%
26 12
68.42% 31.58%
12 3
80% 20%
3 2
60% 40%
71 36
66.36% 33.64%
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
7
Sensitivity : 1.0000 (0.5407,0.9864)
Specificity : 0.7273 (0.3952,0.9256)
Predictive value positive : 0.6667
Predictive value negative : 1.0000
Table VII- Neonatal jaundice
Table VIII- Neonatal pneumonia
Test Blood Blood Total
CS (-) CS (+)
Oropharyngeal fluid aspirate (-) 8 0 8
Oropharyngeal fluid aspirate (+) 3 6 9
Total 11 6 17
Test Blood Blood Total
CS (-) CS (+)
Oropharyngeal fluid aspirate (-) 33 0 33
Oropharyngeal fluid aspirate (+) 17 0 17
Total 50 0 50
Test Blood Blood Total
CS (-) CS (+)
Oropharyngeal fluid aspirate (-) 26 0 26
Oropharyngeal fluid aspirate (+) 4 0 4
Total 30 0 30
Lower Upper
Bound Bound
Likelihood Ratio (+): 3.6667 1.3918 9.6124
Likelihood Ratio (-): 0.0000 0.0119 0.7335
Table IX- No symptoms
DISCUSSION
The noted limitation of this study is a small sample
size with a wide margin of error, making wide confidence
interval and thus compromising the precision. This study
has some biases, as not all mothers with risk factors,
who have potentially septic neonates, were recruited. The
recruitment of subjects and collection of specimen is
according to the time that is convenient to the investigator.
Confounding factor identified in the study is the
administration of antibiotics to the mother. To delimit,
larger population must be included, ideally all who have
risk factors. However, the confounding factor of antibiotic
use cannot be controlled as it would be unethical to defer
administration of antibiotics if the clinical suspicion of
infection is likely.
The most common maternal risk factor of
neonates noted is urinary tract infection, however,
majority of them have no oropharyngeal fluid bacterial
growth. There are only fifteen (15) neonates with
maternal risk factor of prelabor rupture of membrane,
but it can be noted that 80% of them have oropharyngeal
bacterial growth. The highest blood:oropharyngeal
aspirate bacterial growth ratio is the Escherichia coli.
Cross-examination of variables in Tables VII to IX will
show that among neonates who developed jaundice had
no bacterial growth in their blood however 34% of them
have oropharyngeal fluid aspirate growth. It is important
to note that in table VIII, where neonates who had
pneumonia, all of the neonates who had bacterial growth
in the oropharyngeal fluid aspirate also has bacterial
growth in their blood. Those neonates who had no
symptoms within the first 72 hours of life had no bacterial
growth in their blood; however 4 of them have an isolate
from the oropharynx. It is noteworthy that the bacterium
isolated from their oropharynx is Staphylococcus
epidermidis, which can be attributed as collection error
probably because of aseptic technique.
CONCLUSION
The sensitivity of oropharyngeal fluid aspirate
culture irregardless of the organism isolated is 100%, the
sensitivity is 72.35%. However this is not as reflective
of the true sensitivity and specificity as there is a wide
margin of error. Prelabor rupture of membrane has the
highest risk of having bacterial growth in the neonates’
oropharyngeal fluid. Patients with neonatal pneumonia
who have bacterial growth in the oropharyngeal fluid have
a high percentage of having bacteremia.
RECOMMENDATION
A further study with a larger sample size to have
a narrower margin of error is recommended.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
8
REFERENCES
1. Bergstrom,S. Infection-related morbidities in the mother,
fetus and neonate. The American Society for Nutritional
Sciences. 2003, May; 133: 1656-1660.
2. Purwar,M., Ughade, S., Bhagat, B., Agarwal, V. & Kulkarni,
H. (2001) Bacterial vaginosis in early pregnancy and
adverse pregnancy outcome. J. Obstet. Gynaecol. Res.
27: 175–181.
3. Blanco, J. D. (1998) Clinical intra amniotic infection. In:
principles and practice of medical therapy in pregnancy
(Gleicher, N. ed., Gall, S. A., section ed.), pp. 853–857. 3rd
edition. Appleton and Lange Stamford, CT.
4. Segal, OE, et al. Bacterial colonization of the nose and
external ear canal in newborn infants. Israel Journal of
Medical Science. 1983 Dec; 19(12).
5. Bhatia, BD, et al. Bacterial flora in mothers and babies
with reference to causative agent in neonatal septicemia.
Indian Pediatrics. 1989 May; 26(5).
6. Romero R, et al. The role of infection in preterm labor and
delivery. . Paediatric and Perinatal Epidemiology 2001,
14 (Suppl 2), 41-56
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
9
RETROSPECTIVE ANALYSIS OF THE CLINICAL PROFILE AND
IMMEDIATE OUTCOME OF PEDIATRIC PATIENTS DIAGNOSED WITH
PLEURAL EFFUSION OR EMPYEMA THORACIS, ADMITTED TO A
TERTIARY GOVERNMENT HOSPITAL DURING JANUARY 1998 TO
DECEMBER 2000
ABSTRACT
Objective: To review the clinical characteristicsors of
patients who developed complicated and non-complicated
effussions
Design: This study was a case review over a 3-year
period (Jan. 1, 1998 to Dec. 31, 2000) of medical records
of children aged 3 months to 17yrs with a final diagnosis
of pleural effusion or empyema thoracis
Setting: A tertiary government hospital
Results: Sixteen (16) case records diagnosed with pleural
effusion or empyema thoracis from a primary non-
tuberculous respiratory tract infection were reviewed as
to demographics, duration of illness, presenting symptoms,
coexisting illness, clinical and laboratory results, and
outcome. Of the 16 cases, 6 had non-complicated
resolution of parapneumonic pleural effusion while 9 had
chest tube thoracostomy inserted. The duration of
confinement for complicated effusions was significantly
longer (p-value 0.019) than those of non-complicated
parapneumonic effusions.
Conclusion: The presenting signs and symptoms of
patients with pleural effusion are febrile, cough, productive
cough, dyspnea, and chest pain. The development of
complicated parapneumonic effusion is correlated to a
prolonged duration of confinement (p-value 0.019).
INTRODUCTION
Pleural effusion is the accumulation of pleural
fluid within the pleural space beyond which the lymphatic
system can remove. This may be due to many causes
including thoracic diseases, trauma, and iatrogenic injury,
but the most common of which is infection. It has been
said that as many as 40% of hospitalized patients with
bacterial pneumonia have an accompanying pleural
effusion.1 However, only about 10% of these cases need
operative intervention for their resolution. It is in these
Dexter D. Cheng, MD*, Alexander O. Tuazon, MD*
conditions that the morbidity and mortality rates are
increased.2
Empyema thoracis by definition is pus in the
pleural space. Light describes this as pleural effusions
with thick, purulent appearing pleural fluid.1 However,
he goes on to elaborate parapneumonic effusions as any
pleural fluid associated with bacterial pneumonia, lung
abscess, or bronchiectasis. In contrast, complicated
parapneumonic effusions, to which empyema thoracis
belongs, refer to those effusions that do not resolve without
tube thoracostomy.iii
Empyema thoracis is a problem that has been
recognized for centuries.1 Again, there are many causes,
but it is an infected parapneumonic effusion that is the
most common cause.2 Early diagnosis and aggressive
management is necessary to prevent decreased pulmonary
function and the local and systemic sequelae of active
infection. Delays in diagnosis will allow the empyema to
become organized, at which point, percutaneous drainage
techniques and systemic antibiotics are ineffective.
It is the objective of this eview to determine
whether there are clinical factors that may predispose to
the development of complicated parapneumonic effusions
and empyema thoracis. If such factors do exist, then early
surgical managment may be instituted.
GENERAL OBJECTIVES
1. To determine the clinical characteristics of patients
that will correlate with a diagnosis of complicated
effusion.
2. To determine and compare non-complicated and
complicated effusion in relation to the following
outcomes:
a. Length of hospital stay
b. Complications
c. Mortality
*Department of Pediatrics, UP-PGH
Keywords: pleural effusion, empyema thoraces
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
10
Specific Objectives
1.To determine whether patient factors (age, sex, birth
order, social class, presenting illness, duration of illness,
etc, will correlate with the development of complicated
effusion
2. To determine and compare the duration of
confinement in days of patients with non-complicated
pleural effusion and complicated effusions.
3.To determine and compare the complications of
patients with non-complicated pleural effusion and
complicated effusions.
4. To determine and compare the mortality rate of
patients with non-complicated pleural effusion and
complicated effusions.
METHODOLOGY
A retrospective review of medical records of
children aged 3 months to 17 years admitted to a tertiary
medical center with a final diagnosis of pleural effusion
or empyema thoracis from January 1998 to December
2000 was included in the study.
DEFINITIONS
A. Non-Complicated pleural effusion is the accumulation
of pleural fluid within the pleural space
B. Complicated effusions are those effusions that satisfy
the criteria defined by Light1 as follows:
1. pH of pleural fluid less than 7.0
2. pleural fluid glucose less than 60 mg%
C. Empyema thoracis is effusion with thick, purulent
appearing pleural fluid
EXCLUSION CRITERIA
a. Patients receiving chemotherapy or radiotherapy.
b. Patients diagnosed with tuberculous effusions.
c. Patients who are admitted within 10 days from a
previous hospitalization.
d. Patients who are transferred from another hospital.
A total of 82 cases were identified. However, a
total of 56 were excluded because 50 were of tuberculous
etiology, 5 were nosocomial pneumonias from other
hospitals, and 1 was traumatic in origin. Data collected
from the remaining 16 cases included demographics,
presenting symptoms and duration of illness, coexisting
illness, clinical and laboratory results, and outcome.
RESULTS
Sixteen (16) case records diagnosed with pleural
effusion or empyema thoracis from a primary non-
tuberculous respiratory tract infection were reviewed.
They ranged in age from 6 months to 202 months (mean
86 months). Seven (44%) were male. All cases were
confined at the charity ward. 11 cases (69%) were the
3rd born or younger. The duration of illness ranged from
4 to 60 days (mean 26 days). Common presenting
symptoms were pyrexia, cough, productive cough,
dyspnea, and chest pain (Table 1). Nine (9) cases (57%)
had intake of antibiotics either through self-medication
or through a physician prior to consult at the tertiary
hospital. One patient had acute gastroenteritis while the
other 15 did not have any concomitant illnesses at the
time of consult.
Of the 16 cases, 6 had non-complicated resolu-
tion of parapneumonic pleural effusion while 9 had chest
tube thoracostomy inserted. Of the latter, 5 were dis-
charged with open tube thoracostomy in place (Table 2).
The duration of confinement, as shown in table
2, for complicated effusions ranged from 15 to 45 days
(mean 29 days), while those with non-complicated
effusions ranged from 4 to 22 days (mean 10 days). There
was 1 case of nosocomial pneumonia in a patient with
complicated effusion. There were no cases of
pneumothorax nor were there any mortalities.
Table 2. Outcome of 16 Cases of Pleural Effusion
Type of
Effusion
Complicated
Non-
complicated
Number
of Cases
10
6
Average
Duration of
Confinement
(Days)
29
10
Number
Discharged
with
OpenTube
Thoracostomy
5
N/A
Table 1. Presenting Symptoms in 16 Patients Diagnosed
With Pleural Effusion or Empyema Thoracis
Symptom
Febrile
Cough
Productive cough
Dyspnea
Chest Pain
Weakness
Number (%)
15 (94)
15 (94)
11 (69)
11 (69)
7 (44)
1 (7%)
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
11
DISCUSSION
Pleural effusions and empyema thoracis are
relatively common and are associated with significant
morbidity and mortality. Treatment is directed at
controlling the infection and maximizing pulmonary
function.1 Delays in the diagnosis will significantly
increase morbidity and will result in a prolonged clinical
course.vi Furthermore, valuable medical resources are
consumed that may not necessarily influence patient
recovery. These assets are important considerations
most especially in this setting of constrained medial care.
This study focused on non-tuberculous
parapneumonic pleural effusions and empyema thoracis.
The demographics, duration of illness, presenting
symptoms, coexisting illness, clinical and laboratory
results, and outcome were reviewed. However, there
was no statistical significance in the results that correlates
to the development of complicated effusions. The
presenting symptoms noted correlate with those in other
studies.1
The diagnosis of effusion was based on the history
and physical examination, together with chest
roentgenogram and pleural fluid analysis. The criteria
were as described by Light.vi The outcome of pleural
effusion is dependent on the stage of the disease.
Complicated effusions that become organized will no
longer respond to percutaneous drainage and systemic
antibiotics alone. vi Many factors are believed to
contribute to its development. These include prolonged
duration of illness, prior antibiotic intake, and delay in
proper diagnosis. These three factors, however, were
not shown to be statistically significant with the
development of complicated effusions in this review.
The duration of confinement for patients with
complicated effusions was shown to be significantly
longer in this review (p-value 0.019) of 16 cases. The
small sample failed to show any difference in the
incidence of morbidities and mortalities.
In our setting, most cases of both complicated
and non-complicated effusions are secondary to or are
complicated by pulmonary tuberculosis. In this 3-year
review, tuberculous effusions accounted for 68% of all
cases while parapneumonic effusions accounted for 20%.
CONCLUSION
In the review of cases over the three years from
January 1998 to December 2000, only 16 cases of
bacterial parapneumonic effusion were seen. The
presenting signs and symptoms of patients with pleural
effusion are febrile, cough, productive cough, dyspnea,
and chest pain.
Due to the small sample size, the statistical power
is weak and differences are not readily observed. The
only significant result was the correlation of prolonged
hospitalization with the presence of complicated pleural
effusions (p-value 0.019). In order to further elicit any
probable relationships, a larger sample size is needed. It
is suggested that the review include several other tertiary
hospitals and to increase the duration of the review.
REFERENCES
1. Light RW, Girard WM, Jenkinson SG, George RB:
Parapneumonic effusions. Am J Med 1980; 69: 507-511.
2. Brewin A, Arango L, Hadley WK, Murray JF: High-dose
penicillin therapy and pneumococcal pneumonia. JAMA
1974;230:409-413.
3. Light, RW: Pleural Diseases; 3rd Edition. William and
Wilkins 1995: 130.
4. Hippocrates. Genuine works of Hippocrates. Translated
by F. anderer. London (UK): Syndenham Society; 1847.
5. LeMense GP, Strange C. Sahn SA: Empyema thoracis:
therapeutic management and outcome. Chest 1995;
107:1532-7.
6. Light RW: Pleural diseases, 3rd Edition. William and
Wilkins 1995: 138-142.
7. Shaknar KR, Kenny SE, Okoye BO, Carty HM, Lloyd
DA, Losty PD: Evolving experience in the management
of empyema thoracis. Acta Paediatr 200 Apr; 89(4): 417-
20.
8. Chu M, Dewar L, Burgess J, Busse E: Empyema thoracis:
lack of awareness results in a prolonged clinical course.
J Canadien de Chirurgie 2001; 44(4): 284-288.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
12
A RANDOMIZED PLACEBO-CONTROLLED TRIAL ON THE USE OF
PROBIOTICS IN THE PREVENTION OF NOSOCOMIAL INFECTION IN
PEDIATRIC PATIENTS WITH HEMATOLOGIC AND ONCOLOGIC
DISEASES
Arlene Grace L. Simon, MD*, Jossie Rogacion, MD*
ABSTRACT
Background: Despite measures to decrease its
incidence, nosocomial infection remains to be a major
problem in most tertiary hospitals. The use of prophylactic
antibiotics leads to an increase in bacterial resistance,
rather than decreasing the nosocomial infection rate.
Other measures have been proposed to combat antibiotic
resistance and the development of nosocomial infection,
such as the use of probiotics.
Objective: The efficacy of the use of probiotics in the
prevention of nosocomial infection and effect on duration
of hospitalization was tested in patients with hematologic
and oncologic diseases.
Study design: Randomized placebo-controlled double-
blind trial using placebo and probiotics.
Intervention: 50 patients out of 76 predicted sample size,
with hematologic and oncologic diseases who had a
minimum predicted duration of hospitalization of 3 days
were enrolled in a double-blind trial and randomly assigned
on admission to take either 1 capsule of probiotics (n=25)
containing at least 1.5 billion cells of Lactobacillus
acidophilus, Bifidobacterium bifidus and
Bifidobacterium longus or a comparable placebo (n=
25) once daily during their whole hospital stay. The
patients were monitored and assessed daily for the
development of nosocomial infection and for compliance
to administration of the capsules. The patients were also
assessed for any possible adverse effects. The counting
of the duration of hospital stay started from admission to
the emergency room till discharge. Patients were asked
to follow up 1 week after discharge.
Results: This paper presents the preliminary results of
the study involving 50 patients. Baseline characteristics
on admission were found to be similar. The use of
probiotics prevented the occurrence of nosocomial
infection in patients with hematologic and oncologic
diseases as compared to those receiving placebo, 0% vs.
20%, with a relative risk of 0.2 (fisher’s exact test= 0.025,
p< 0.05). However, there was no significant statistical
difference between the duration of stay of the 2 groups
(placebo = 9.4 days compared to probiotics = 8.6 days).
The difference in readmission was analyzed using z-test
and was also found to be not statistically significant. No
adverse effects were reported in both groups.
Conclusion: The use of probiotics in patients with
hematologic and oncologic diseases may play a role in
the reduction of the occurrence of nosocomial infection,
but may not necessarily cause a reduction in the duration
of hospital stay. Further confirmation of the results is
recommended upon completion of the study.
INTRODUCTION
Measures have been proposed to prevent the
occurrence of nosocomial infection in most tertiary
hospitals, such as strict hand-washing, the use of gowns,
masks and gloves and the segregation of wards into
infectious and non-infectious cases.1 However,
nosocomial infection still remains to be a major problem.
It prolongs the stay of the patient in the hospital and is a
cause of additional financial burden to the family of the
patient. Patients afflicted with chronic illnesses, such as
those with hematologic and oncologic diseases are said
to be particularly susceptible to nosocomial infection due
to the secondary impairment of their immune response.
The use of chemotherapy, in particular, contributes to the
impairment of the immune response of oncologic patients.1
Prophylactic antibiotics may be given to prevent the
occurrence of opportunistic infections in oncologic
patients. However, the use of prophylactic antibiotics may
lead to the emergence of resistance to these antibacterial
agents and may even lead to the depletion of good
microorganisms. It is this growing emergence of
resistance to antibacterial agents which gave birth to
bacteriotherapy. Bacteriotherapy is the use of harmless
bacteria to displace pathogenic organisms and is said to
be an alternative and promising way of combating
infection.2
Keywords: probiotics, nosocomial infection, hematologic, oncologic, duration of hospitalization.
*Department of Pediatrics, UP-PGH
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
13
Probiotics are live microorganisms which colonize
the intestine, modifying the intestinal microflora and their
metabolic activities, leading to presumed beneficial
effects to the host system.3 Probiotics are said to be
non-pathogenic because they naturally inhabit the human
intestine. The use of probiotics in diarrhea have been
proven by numerous studies. These studies have shown
that probiotics prevent the occurrence of diarrhea, as well
as lessen it severity by promoting the growth of these
beneficial live microorganisms and preventing the
overgrowth of harmful microorganisms by means of their
direct effect on the intestinal mucosal system.4 It is said
to promote healing of the intestinal mucosa by reducing
gut permeability and by enhancing the local intestinal
immune response.3 A study on the prevention of
nosocomial diarrhea on infants was undertaken in Poland
by Szajewska et. al. Lactobacillus GG was given to
hospitalized infants and the results showed that probiotic
supplementation caused a decrease in the occurrence of
nosocomial diarrhea (6.7% vs. 33.3%), with a resulting
decrease in the length of stay of the patients given
Lactobacillus GG.5
Recently, it has been postulated that probiotics
may be beneficial not only in the gastrointestinal tract
which they normally inhabit, but also in other mucosal
surfaces such as the respiratory tract and urinary tract
systems. It has been postulated that probiotics produce a
beneficial effect on the over-all host system because of
its’ effect on both the humoral and cellular immune
systems. A study in Finland by Hattaka et. al. explored
the use of milk cultured with probiotics in the prevention
of respiratory and gastrointestinal infections in healthy
children attending day care centers.6 The study showed
a significant decrease in the number of absences due to
respiratory and gastrointestinal illnesses in the children
supplemented with probiotics compared to the placebo
group (16% difference). There have been no studies,
however, regarding the use of probiotics in children
afflicted with chronic illnesses.
This study was conducted to determine if
probiotics can be used in the prevention of the occurrence
of nosocomial infection in children with hematologic and
oncologic diseases.
OBJECTIVES
General Objectives
To determine if supplementation with probiotics
will improve the outcome of hospitalized pediatric patients
with hematologic or oncologic diseases.
Specific Objectives
To determine if hospitalized pediatric patients
with hematologic or oncologic diseases given probiotic
supplements will have a lesser frequency of nosocomial
infections compared with hospitalized patients given
placebo.
To determine if patients with probiotic
supplementation will have a shorter duration of
hospitalization as compared with patients given placebo.
Definition of Terms
Probiotics – nutritional supplements containing live
bacterial or yeast cultures which commonly inhabit the
intestinal system.3
Hematologic diseases – disorders that produce either
quantitative or qualitative defects involving the cellular
elements of the blood and blood tissues and those affecting
hemostasis, such as in aplastic anemia and hemophilia.7
Oncologic diseases – diseases which cause proliferation
of abnormal cells called oncogenes producing cancer and
proliferation of tumor suppressor genes, such as in
leukemia and retinoblastoma.7
Nosocomial infection - infections appearing in hospitalized
patients not present nor incubating at time of admission
from the hospital. The onset is beyond 72 hours from
admission to the hospital, starting from admission to the
emergency room.
METHODOLOGY
Setting
This randomized, double blind, placebo-controlled
clinical study was conducted at the Philippine General
Hospital Pediatrics wards and Emergency Room from
July to October, 2002.
Patient Recruitment
Children aged 1 to 16 years of age with
hematologic or oncologic diseases, with a predicted length
of hospital stay of at least 3 days were recruited in the
study consecutively upon admission at the Pediatric
Emergency Room. The conduct of the research was
explained to the parents by the investigator and written,
informed consent was secured prior to enrollment in the
study. Children who were placed on nothing per orem,
with known allergy to cow’s milk, other concomitant
disorders, those who were in severe respiratory distress
needing oxygen and ventilatory support, as well as those
with neurologic deficits such as signs of increased
intracranial pressure due to intracranial bleed and patients
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
14
with assessment of nosocomial infection on admission
were excluded from the study. All the patients whose
parents gave written, informed consent were randomized
for the trial.
Sample Size Computation
Sample size was calculated based on the
assumption that the use of probiotics will result in a 20%
reduction in the occurrence of nosocomial infection. This
assumption was based on an earlier study by Szajewska
et. al.,5 which reported a 26.6% decrease in the
occurrence of nosocomial diarrhea (33.3% in the control
group) in normal children. It was estimated that with a
power of 90% and at alpha level of 0.10, we needed 38
children per group to show a 20% difference.
Randomization
The patients were recruited as they arrived at
the Pediatric Emergency Room. The patients were
randomized into the 2 groups by means of a computer
generated randomization table.
Random Allocation/Blinding
The GNC Kyo-dophilus® (probiotics) capsules
were repackaged by the Industrial Pharmacy laboratory
of the University of the Philippines, Manila into yellow
size 1 capsules. The same laboratory prepared the
placebo capsules and packaged the cornstarch placebo
into the same yellow size 1 capsules. This repackaging
was necessary to ensure that the physicians and the
patients will be blinded during the study. A person not
directly involved in care of the subjects was tasked to
encode the capsules into 2 groups (1 and 2). The capsules
were packaged in air-tight envelopes (7 capsules per
pack) labeled 1 or 2 with the corresponding patient number
and patient name. Both the physician in charge and the
patients were unaware of which capsules contain
probiotics or placebo. The code was revealed during the
data analysis.
Intervention
On admission to the study, complete history and
physical examination were performed. The patients
included in the study were given either 1 capsule of
probiotics or placebo everyday from day 1 of ER
admission until the last hospital day. In the event that the
patients get admitted to the wards, intervention and
monitoring were continued until discharge. One capsule
of the GNC Kyo-dophilus® probiotics contains a minimum
of 1.5 billion live cells of Lactobacillus acidophilus,
Bifidobacterium bifidum and Bifidobacterium longum. The
probiotic capsules were lyophilized, ensuring stability in
room temperature.
The mothers were given a set of 7 capsules on
admission. The mothers were instructed to give their
children 1 capsule once a day after the morning meal.
For patients less than 7 years of age or those who had
difficulty in swallowing the capsules, the mothers were
instructed to mix the powdered contents in 1 tsp of milk
to be given after the morning meal. For patients who
vomited the capsules or the powdered contents, the
mothers were given instructions to give another dose 1
hour after the vomiting. The mothers were instructed to
list down difficulties in giving the capsules.
The supplementation was continued until the
patient was discharged from the hospital. Patients who
developed nosocomial infection were instructed to
continue supplementation until discharge. Monitoring for
these patients continued until the patients were discharged
from the hospital. Parents were instructed not to give
their children other products with probiotic organisms.
Infection Surveillance
Baseline laboratories were ordered on the first
hospital day, which included a baseline complete blood
count, chest x-ray, urinalysis, stool exam and blood
culture. The patients were visited daily and were
monitored for the occurrence of nosocomial infection and
compliance to the administration of the drug. The parents
were also instructed to write down problems in
administration and noted adverse effects on
supplementation. Laboratories were repeated depending
on the clinical suspicion of nosocomial infection. The
patients were also followed up 7 days after discharge
and assessed for the development of nosocomial
infection. A diagnosis of nosocomial infection was made
if the criteria for nosocomial infection as drafted by the
Infectious Disease Section of the Department of
Pediatrics of the Philippine General Hospital was fulfilled.
Classification of Nosocomial Infection:
A. Respiratory infection
1. Clinical signs and symptoms of lower respiratory
tract infection (LRTI)
(cough, fever, increase in purulence of secretions,
new auscultatory findings) + new infiltrates on
chest x-ray
1.1 with + culture (ETA/TTA/LT)
1.2 with (-) blood culture
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
15
2. Clinical signs and symptoms of lower respiratory
tract infection (LRTI) + new infiltrates on chest
x-ray
2.1 with (+) blood culture
2.2 with (-) blood culture
B. Urinary tract infection
1. Colony counts of > 100,000/ml of appropriately
collected specimen or visible organism on gram
stain in a patient with previously (-) urine culture
or normal urinalysis
1.1 with no clinical signs and symptoms (fever,
dysuria, costovertebral angle tenderness,
suprapubic tenderness)
2. Colony counts > 10,000 colonies/ml. Of
appropriately collected specimen of a new
pathogen with a previously (+) urine culture:
2.1 with clinical signs/symptoms
2.2 without clinical signs/symptoms
C. Gastrointestinal tract infection
1. Diarrhea with a patient with no diarrhea on
admission
1.1 with (+) stool culture
1.2 with (-) stool culture
2. Any other signs of GI disturbance (nausea,
vomiting, abdominal tenderness) not previously
present on admission (peritonitis, NEC, intra-
abdominal abscess)
2.1 with (+) stool culture
2.2 with (-) stool culture
D. Bacteremia
1. Presence of a (+) blood culture in a patient with
no definite focus of infection
2. Bacteremia with Infective endocarditis
E. Sepsis
1. Signs and symptoms of sepsis (poor suck/activity,
jaundice, acidosis, hypotension, hyperglycemia,
etc.) with definite focus of infection
1.1 with (+) blood culture
1.2 with (-) blood culture
Checking of Compliance
The number of remaining capsules and empty
shells were checked daily. Patients whose length of stay
exceeded 7 days were given a weekly supply of 7
capsules. All remaining capsules were returned on the
last hospital day, prior to discharge of the patient.
Outcome Measures
The primary outcome measures were the
absence or occurrence of nosocomial infection as defined
by the criteria for nosocomial infection by the Section of
Infectious Diseases and the number of days the patient
was hospitalized starting from the day of admission at
the emergency room. A secondary outcome was the
number of children with readmissions due to nosocomial
infection within 7 days from discharge.
Withdrawals/Deviations From The Study
Withdrawal is defined as those patients who are
taken out from the study by their parents, those who are
brought home against advice by their parents disrupting
therapy and those who are placed on nothing per orem
for more than 2 days during the duration of the study.
Deviations from the study include those patients who are
placed on nothing per orem for less than 48 hours and
those who vomited the capsules during the initial
administration. The mothers of the patients who were
placed on nothing per orem for less than 48 hours were
instructed to resume administration once feeding, while
the mothers of the patients who vomited during the initial
administration were instructed to re-administer the
capsules 1 hour after the vomiting.
STATISTICAL ANALYSIS
An intention to treat analysis was performed.
Data were analyzed using Statistica software. T-test was
done comparing the difference between the means of 2
independent samples. In comparing the distribution of
discrete variables between two groups, chi-square test
or Fisher’s exact test when appropriate were performed.
The z test was used to determine the significance of
readmissions between the two groups. All tests were
performed at a significance level at p < 0.05.
RESULTS
A total of 56 patients out of 76 predicted sample
size, with hematologic and oncologic diseases were
recruited for the study and randomized using a computer
generated blocked randomization table. Of the 56 patients
recruited, 6 patients were still hospitalized at time of
analysis and were not included in the partial analysis of
this study. This paper presents the preliminary data
analysis of 50 patients included in the study.
Of the 50 patients discharged, there were 2
withdrawals: 1 from each group and were due to the
patient being brought home per request of the parents.
The patient from the treatment group who was brought
home against advice developed intracranial bleed due to
hyperleukocytosis on the 3rd ward day and was brought
home per request of the parents. This patient did not
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
16
develop nosocomial infection during the whole hospital
stay. The patient from the control group who dropped
out of the study developed nosocomial pneumonia on the
4th hospital day and was brought home per request
because the parents did not want to push through with
chemotherapy. Since this is an intention to treat type of
study, the data up to the day that these patients dropped
out of the study were included in the analysis.
DESCRIPTION OF STUDY POPULATION
Data gathered on admission (Table 1) revealed
that 74% of the patients were male (n=37) and in the 2-
5 years age group (n=21; 42%). Of the 50 patients
included in the data analysis, 26 (52%) had oncologic
diseases while 24 (48%) had hematologic diseases.
Majority of the patients were hospitalized due to fever
(n=20; 40%), followed by pallor (n=15; 30%) as the most
common presenting symptom.
Table 1. Characteristics of patients included in the study
(N = 50)
CHARACTERISTIC
AGE Distribution
2-5 years
6-9 years
10 – 13 years
14 –16 years
Presenting complaint
Fever
Pallor
Bleeding
Diarrhea
Disease Type
Hematologic
Aplastic anemia
Thalassemia
Hemophilia
Others
Disease Type
Oncologic
ALL
AML
Lymphoma
Others
NUMBER
21
11
11
7
20
15
12
3
24
9
6
5
4
26
11
9
2
4
%
42%
22%
22%
14%
40%
30%
24%
6%
48%
37.5%
25%
20.8%
16.7%
52%
42.3%
34.6%
7.6%
15.4%
Comparison of Baseline Characteristics of Placebo
and Probiotic Groups
Table 2 lists the comparison of baseline
characteristics of the 2 groups, such as age, disease type
and presenting complaints. There was no significant
difference in mean age between the placebo group (mean
age = 7.92 years) and the probiotics group (mean age =
7.48 years). With regards to the disease type, it was found
that there are more hematologic patients (n= 14) in the
control group as compared to the treatment group (n =
10). In contrast, the treatment group had more oncologic
(n = 15) patients than the control group (n = 11). This
difference, however, was found to be not statistically
significant, using the Fisher’s exact test. Fever was the
most common presenting complaint in both groups, with
the treatment group having more patients presenting with
fever (11 vs. 9). Again, the difference in presenting
complaints was not statistically significant (x2= 2.3636,
DF =3).
1: t = 0.34, p > 0.05
2: x2 test = 2.3, df = 3, p > 0.05
3: Fisher’s exact test = 0.4, df = 2
DATA ANALYSIS
Frequency of Nosocomial Infection
None of the patients in the treatment group
developed nosocomial infection compared to 20% in the
placebo group (5 out of 25), showing a relative risk
reduction of 100% and an absolute risk reduction of 20%,
with number needed to treat = 5 patients. The difference
was statistically significant using the Fisher’s exact test
(0.025, p< 0.05).
Of the nosocomial infection detected, 1 patient
developed nosocomial candiduria, while 4 patients
developed nosocomial pneumonia. It is noteworthy that
all the patients who developed nosocomial infection are
leukemia patients.
Table 2. Comparison of baseline characteristics of placebo
and probiotic groups
Characteristic
Age (mean) (1)
Presenting
complaint (2)
Fever
Disease type (3)
Hematologic
Oncologic
Control
(N = 25)
7.92 years
9
14
11
Treatment
(N= 25)
7.48 years
11
10
15
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
17
Table 3. Frequency of patients who developed nosocomial
infection
Duration of Hospitalization
The patients in the placebo group had a mean
duration of hospitalization of 9.4 days (sd = 7.0) as
compared to the probiotics group with a mean duration
of hospitalization of 8.6 days (sd = 6.7; t value = 0.43, p>
0.05).
Secondary Outcome: Readmissions
Of the 25 patients in the probiotics group, 2
patients were readmitted within 1 week due to continuation
of chemotherapy, while in the placebo group, 1 patient
was readmitted for infection, while 2 patients were
readmitted for bleeding and another patient was
readmitted for chelation therapy. None of the 50 patients
had readmissions for nosocomial infection. The data for
re-admissions was analyzed via the z test and was found
to be not statistically significant (z value = - 0.879, α =
0.05; significant z value if < -1.65).
There were no mortalities among the patients
recruited. Of the 50 patients, only 2 patients were
discharged with no improvement. These 2 patients were
the 2 drop-outs who were brought home against advice
by their parents.
All the patients who developed nosocomial
infection received appropriate antibiotic therapy and were
discharged improved.
Surveillence for Adverse Effects
No adverse effects were noted during
supplementation with either the placebo or probiotic
capsules in any of the 50 patients included in the study.
DISCUSSION
This randomized, double blind, placebo-controlled
study is the first of its’ kind to explore the prophylactic
effects of probiotics on nosocomial infection, particularly
in the respiratory, gastrointestinal and urinary tract
systems, in patients with hematologic and oncologic
diseases. The study was undertaken to be able to find a
safe alternative to giving prophylactic antibiotics, which
in the end leads to an increase in antibiotic resistance.
Oncologic and hematologic patients were chosen as the
main subjects in the study because they were found to
be particularly susceptible to developing nosocomial
infection because of their need for frequent admissions
and their exposure to immunosuppressive therapy.
Hematologic and oncologic patients need frequent
admissions due to blood transfusions or chemotherapy.
Baseline characteristics of both groups were
similar and statistical analysis showed that group
differences were not significant. Results revealed that
the treatment group had 0% infection rate as compared
to 20% in the control group. Absolute risk reduction was
100% and relative risk was 0.20, with number needed to
treat = 5. This means that we need to treat only 5 patients
to prevent 1 nosocomial infection. All the patients who
developed nosocomial infection were oncologic patients.
Oncologic patients were found to be more susceptible to
acquiring nosocomial infection due to their impaired
neutrophil counts which decreases the body’s ability to
fight off infection .1 Furthermore, hematologic patients
were found to have a lesser length of stay as compared
to oncologic patients, because most of them get admitted
for blood transfusions only and not for infection or
chemotherapy although the difference did not reach
statistical significance.
The lack of statistical significance with regards
to duration of hospitalization can be explained by the fact
that despite the placebo group having more episodes of
nosocomial infection than the probiotics group, which was
supposed to result in a longer duration of hospitalization,
there were more patients in the placebo group with
hematologic disorders than oncologic disorders as
compared to the probiotics group. Other factors that may
contribute to the lengthening of the duration of stay include
the need for chemotherapy, the procurement of funds
for chemotherapy, as well as the completion of antibiotic
therapy. Despite the lack of difference in the duration of
hospitalization, the decrease in occurrence of nosocomial
infection in the probiotics group implies a better outcome,
since they can proceed to definitive chemotherapy earlier
and there is lesser need for funds to acquire antibiotics
needed to treat nosocomial infection.
Proposed Mode of action:
Probiotics work by being able to replenish the
depleted intestinal microflora which is often caused by
the use of antibiotics.8 It is presumed to promote healing
Placebo Probiotics Total
group group
Nosocomial infection 5 0 5
No nosocomial infection 20 25 45
Total 25 25 50
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
18
of the intestinal mucosa by reducing gut permeability and
by enhancing the local intestinal immune response,
particularly that of IgA. It is also considered safe and
has no known adverse effects.9
The study in Finland by Hattaka et. al. is the first
to examine if probiotics can also be effective in other
mucosal systems. They hypothesized that probiotics
prevent the occurrence of infection by means of
stimulating non-specific immunity or by enhancing
humoral and cellular immunity.5 Probiotics presumably
prevent respiratory tract infections, even though it is not
in direct contact with it because of its’ stimulatory effect
on secretory IgAat all mucosal surfaces.10
An article published in the Nutrition Hospital
Journal in 2001 promotes the use of probiotics in immuno-
compromised patients, particularly ICU patients. Listed
effects of probiotics in the immune system aside from
stimulating IgA response include the inhibition of IgE
production, stimulation of the macrophage function,
stimulation of the NK activity, stimulation of apoptosis,
promotion of growth and regeneration, reduction of
endotoxin production and the production of anti-
oxidants.11 Further studies are needed to demonstrate
these effects.
CONCLUSION
Probiotics may play a role in the prevention of
the occurrence of nosocomial infection in patients with
hematologic and oncologic diseases, but it may not
necessarily cause a reduction in the duration of hospital
stay.
RECOMMENDATIONS/LIMITATIONS
Since this paper presents only the preliminary
results of the study, further confirmation of the results is
recommended upon completion of the study.
A more objective way of testing compliance
should have been done in the form of stool cultures to
detect the presence of probiotic organisms. Unfortunately,
this was not done due to the unavailability of the test in
the country.
REFERENCES
1. Dix, Kathy. Infection Control in the ICU: The Final Frontier.
Infection Control Today. 2002, June.
2. Huoniven P. Bacteriotherapy: the time has come. British
Medical Journal Aug. 2001;323:353-354.
3. Katelaris P. Probiotics control of diarrheal disease. Asia
Pacific Journal of Clinical Nutrition1996;5:39-43.
4. Agarwal K., Bhasin S., Faridi M., Mathur M., Gupta S.
Lactobacillus cassei in the control of Acute Diarrhea: A
Pilot Study. Indian Pediatrics 2001; 38: 905 – 910.
5. Szajewska H, Kotowska M, Mrukowicz JZ, Armanska M,
Mikolajczyk W. Efficacy of Lactobacillus GG in prevention
of Nosocomial Diarrhea in infants. Journal of Pediatrics
2001, March; 138(3): 361-365.
6. Hattaka K. et. al. Effect of long term consumption of
probiotic milk on infections in children attending day care
centers: double blind, randomized trial. British Medical
Journal 2001; 322:1327-1329.
7. Behrman Richard, Kliegman Robert, Jenson Hal. Nelson
Textbook of Pediatrics. 2000; 16: 1456, 1533.
8. Vanderhoof JA. Lactobacillus GG reduced diarrhea
incidence in antibiotic treated children. Pediatrics 1999,
Nov.; 135:564-568.
9. Wanke C. Do probiotics prevent childhood illnesses?
British Medical Journal 2001; 322: 1318-1319
10. Friedrich M.J., A Bit of Culture for children: Probiotics
may improve health and fight disease. JAMA 2000, Sept.;
284(11):
11. Bengmark S., Garcia De Lorenzo A., Culebras J.M. Use of
Pro, Pre and Synbiotics in the ICU. Nutrition Hospital 2001;
16 (6): 239-256.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
19
A FIVE-YEAR RETROSPECTIVE STUDY ON THE COMMON
MICROBIAL ISOLATES AND SENSITIVITY PATTERN ON BLOOD
CULTURE OF PEDIATRIC CANCER PATIENTS ADMITTED AT THE
PHILIPPINE GENERAL HOSPITAL FOR FEBRILE NEUTROPENIA
Ailyn T. Isais-Agdeppa, MD*, Lulu Bravo, MD*
ABSTRACT
Rationale: Febrile neutropenia is a common clinical
problem among pediatric cancer patients. This often
necessitates hospitalization and immediate empiric broad
spectrum antimicrobial therapy. In selecting the initial
antibiotic regimen, the type, frequency and antibiotic
susceptibilities of the microbial isolates should be
considered.
Objective: To determine the bacterial blood isolates and
antimicrobial sensitivity pattern of pediatric cancer patients
admitted at the Philippine General Hospital for febrile
neutropenia.
Methods: Retrospective study of all pediatric patients
diagnosed with a hematologic or oncologic malignancy
admitted at the Philippine General Hospital from January
1999 to December 2003 for febrile neutropenia.
Results: A total of 90 patients were included in the study.
Sixty two per cent (62%) of the subjects were diagnosed
to have hematologic malignancy while 38% are oncologic
cases. Only 7% of the patients had growth in the blood
culture while 93% had no growth in the blood culture.
Causative organisms isolated were Streptococcus
viridans, Gram negative bacilli, Staphylococcus
epidermidis, Candida sp., Salmonella sp. and Haffnea
alveii. All patients received Ceftazidime as empiric
antimicrobial therapy in combination with Netilmycin for
patients admitted in 1999 or Amikacin for patients
admitted from January 2000 to December 2003. S.
viridans, G(-) bacilli, Salmonella sp. and H. alveii
showed sensitivity to Ceftazidime. All patients given
Ceftazidime with Netilmycin/Amikacin became afebrile
on the 3rd day of antibiotics with increasing absolute
neutropenic count except for patients who developed
nosocomial infections causing delay in recovery and
prompting shift of antibiotics.
Conclusion: Majority of pediatric cancer patients with
febrile neutropenia yielded no growth in blood culture.
However, most cases responded well with the empiric
antibiotic combination of Ceftazidime and Amikacin. Few
cases with isolates were also sensitive to the current
antibiotic regimen. The microbiological laboratory should
also act in consultation with the clinician to determine
the optimal approach for blood cultures in the
immunocompromised patients.
INTRODUCTION
Infection and its complications were the main
causes of morbidity and mortality among pediatric patients
receiving chemotherapy for malignant diseases.9
Granulocytopenia carries the risk of bacterial infection.
If prolonged, this can lead to fungal infection. As the
neutrophil count decreases below 500 cells/cu.mm, the
risk of infection increases proportionately.3 Hence,
antimicrobials play a major role in the management of
patient with febrile neutropenia.
International trend shows that gram-positive
bacteria account for 60-70% of microbiologically
documented infections. Some of these are methicillin
resistant. However, gram-negative bacilli especially P.
aeruginosa, E. coli and Klebsiella sp. remain prominent
causes of infection. Fungal infections on the other hand,
are usually superinfections. In some cases, Candida
sp. or other fungi can cause primary infection.6
It is therefore important in the selection of the
initial antibiotic regimen for febrile neutropenia to consider
the type, frequency of occurrence and antibiotic
susceptibility of bacterial isolates recovered from other
patients at the same hospital. Current recommendation
in the empiric antibiotic management of febrile
neutropenia at the Philippine General Hospital is based
on the 1997 Guidelines for the Use of Antimicrobial
Agents in Neutropenic Patients with Cancer prepared
by the Infectious Diseases Society of America (IDSA)
Fever and Neutropenia Guidelines Panel. This
recommends the use of Vancomycin plus Ceftazidime
(or Cefepime) or a carbapenem or combination therapy
with an aminoglycoside plus an anti-pseudomonal beta-Keywords: febrile neutropenia, malignancy, granulocytopenia, absolute neutrophil count, blood
culture
*Department of Pediatrics, UP-PGH
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
20
lactam as initial intravenous therapy.7 This is based on
international trend of microbiologic growth and sensitivity
pattern.
This study was conceived to determine whether
bacterial blood isolates among pediatric patients with
febrile neutropenia at the Philippine General Hospital is
comparable with the international trend. It aims to provide
local data on the bacterial blood isolates and antimicrobial
sensitivity patterns. This study also hopes to provide
valuable information that will aid in the decision-making
and improvement in the care of pediatric patients with
febrile neutropenia especially in the selection of initial
antibiotic regimen.
DEFINITION OF TERMS
Febrile neutropenia – a single temperature
measurement of >/= 38.3 C in the absence of obvious
environmental causes or temperature of >/= 38.0 C for>/
= 1 hour and granulocytopenia by routine complete blood
count determination Granulocytopenia – absolute
neutrophil count of </= 500 cells/cu.mm.
REVIEW OF RELATED LITERATURE
Children with cancer can be severely
compromised. This can be due to immunodeficiency
secondary to malignancy, the therapy or both. Most
chemotherapeutic agents also inhibit the inflammatory
response to invading microbes. In these patients, the
physical signs and symptoms are not reliable in predicting
infection.3 The neutrophil count on the other hand, is
important in predicting risk and response to infection. As
the absolute neutrophil count goes down, the patient’s
immunological status becomes compromised and the risk
of infection is greater. In such patients with impaired
inflammatory response coupled with granulocytopenia,
fever most of the time is the only sign of infection. This
demands urgent empirical antibiotic therapy with broad
spectrum coverage. This is to decrease the risk of septic
shock, ARDS, hypotension, renal and other organ
dysfunction and death.2
For febrile neutropenia cases, the Infectious
Diseases Society of America (IDSA) recommends a
thorough physical examination, complete blood count,
blood urea nitrogen, creatinine and transaminase level
determination and blood cultures as part of initial
management. Sample for blood culture is drawn from a
peripheral line including a set from central venous access
device. Urinalysis and urine culture is warranted if the
patient has symptoms of urinary tract infection or has a
chronic urinary catheter. Chest radiograph is indicated
when the patient has respiratory symptoms. High
resolution chest computed tomography is requested in
the presence of respiratory symptoms and normal chest
radiographic findings.6
According to the 2002 Guidelines for the Use of
Antimicrobial Agents in Neutropenic Patients with Cancer
published by IDSA, the bacterial causes of febrile
episodes in neutropenic patients are as follows:
Gram-positive cocci and bacilli
Staphylococcus species *
Coagulase-positive (Staphylococcus aureus)
Coagulase-negative(Staphylococcus
epidermidis and others)
Streptococcus species *
Streptococcus pneumoniae
Streptococcus pyogenes
Viridans group
Enterococcus faecalis/faecium *
Corynebacterium species *
Gram-negative bacilli and cocci
Escherichia coli *
Klebsiella species *
Pseudomonas aeruginosa
Enterobacter species
Anaeorobic cocci and bacilli
Bacteroides species
Clostridium species
Fusobacterium species
Propionibacterium species
Peptococcus species
Veillonella species
Peptostreptococcus species
The 2002 Guidelines for the Use of Antimicrobial
Agents in Neutropenic Patients with Cancer published
by IDSA provided the following algorithm in the
management of cancer patients with febrile neutropenia:
* the most common cause of bacteremia
Factors that favor a low risk for severe infection
among patients with neutropenia are absolute neutrophil
count >/= 100 cells/cu.mm, absolute monocyte count of
.>/= 100 cells/cu.mm, normal findings on chest radiograph,
nearly normal results of hepatic and renal function tests,
duration of neutropenia of less than 7 days, resolution of
neutropenia expected in less than 10 days, no intravenous
catheter-site infection, early evidence of bone marrow
recovery, malignancy in remission, peak temperature of
less than 39 C, no neurological and mental changes, no
appearance of illness, no abdominal pain and no
comorbidity complications.6
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
21
Study Population
The study involved chart review of all pediatric
cancer patients admitted at the Philippine General
Hospital from January 1999 to December 2003 selected
on the basis of the following inclusion criteria: Pediatric
patients aged 1-19 years old diagnosed with oncologic or
hematologic malignancy documented to have
granulocytopenia (neutrophil count </= 500 cells/cu.mm)
by routine complete blood count determination. A single
temperature measurement of >/=38.3 C in the absence
of obvious environmental causes or temperature of >/=
38.0 C for >/= 1 hour no overt site or focus of infection
at the time of admission
DATA COLLECTION
Charts of patients who fulfill the criteria for the
study were retrieved and reviewed. The final blood
culture results prior to the start of any antibiotic regimen
was recorded including the antimicrobial sensitivity
patterns.
OUTCOMES MEASURED
The outcome of interest was the determination
of the common bacterial blood isolates and antimicrobial
Fever (temperature >/=38.3C) + Neutropenia (<500 neutrophils/cu.mm.)
Low Risk High Risk
Oral IV Vancomycin not needed Vancomycin needed
Ciprofloxacin
+
Amoxicillin-
clavulanate
(adults only)
Monotherapy
Cefepime
Ceftazidime
or
Carbapenem
Two Drugs
Aminoglycoside +
Antiseudomonal
penicillin
Cefepime
Ceftazidime or
Carbapenem
Vancomycin +
Vanomycin +
Cefepime
Ceftazidime
or
Carbapenem +/-
Aminoglycoside
Reassess after 3-5 days
OBJECTIVES
General objectives:
To determine the bacterial blood isolates and
antimicrobial sensitivity pattern of pediatric cancer patients
admitted at the Philippine General Hospital for febrile
neutropenia.
Specific objectives:
1. To determine the bacterial blood isolates of pediatric
cancer patients admitted at the Philippine General
Hospital for febrile neutropenia.
2. To determine the antimicrobial sensitivity patterns
of bacterial blood isolates of pediatric cancer patients
admitted at the Philippine General Hospital for febrile
neutropenia.
3. To determine whether the current empiric antibiotics
given to pediatric cancer patients admitted at the
Philippine General Hospital for febrile neutropenia
are effective.
METHODOLOGY
Study Design
Retrospective, descriptive study
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
22
sensitivity pattern of pediatric cancer patients with febrile
neutropenia at the Philippine General Hospital.
ANALYSIS
Descriptive analysis was done on the data
collected.
RESULTS
From January 1999 to December 2003, there
were a total of 90 febrile neutropenia cases. 56% of the
subjects were male while 44% were female. 56% of the
subjects belong to the 0-6 years of age.
Table 2. Gender Distribution of Study Population
Gender N %
Male 50 62
Female 40 38
Total 90 100
Table 3. Frequency Distribution of Study Population
According to Nature of Malignancy
Nature of Malignancy N %
Hematologic 56 62
Oncologic 34 38
Total 90 100
Table 4. Frequency Distribution of Study Population According
to Blood Bacterial Isolates
Blood Bacterial Isolates N %
Streptococcus viridans 1 1
Gram-negative bacilli 1 1
Staphylococcus epidermidis 1 1
Candida sp. 1 1
Salmonella sp. 1 1
Haffnea alveii 1 1
Table 1. Age Distribution of Study Population
Age Group (Years)
0-3
4-6
7-9
10-12
13-15
16-18
>18
Total
N
25
25
3
18
16
3
0
90
%
28
28
3
20
18
3
0
100
Among the 90 subjects, 62% had hematologic
malignancy i.e. acute lymphocytic leukemia, acute
myelogenous leukemia. 38% were oncologic cases
comprising of osteosarcoma, embryonal carcinoma,
neuroectodermal sinus tumor, nasopharyngeal carcinoma,
Wilm’s tumor, Ewing’s sarcoma, rhabdomyosarcoma,
malignant mixed germ cell tumor and malignant peripheral
nerve sheath tumor.
Only 7% of the patients had growth in the blood
culture. 93% had no growth. Table 4 and 5 indicates the
blood microbial isolates and antimicrobial sensitivity
patterns of these isolates. Among the patients who had
growth in the blood culture, 50% grew Gram-negative
bacteria, 33% had Gram-positive bacteria and 17% had
fungal isolate.
All patients received Ceftazidime as empiric
antimicrobial therapy in combination with Netilmycin for
patients admitted in 1999 or Amikacin for patients
admitted from January 2000 to December 2003. Majority
patients given Ceftazidime with Netilmycin/Amikacin had
lower temperature and less febrile episodes on the third
day of antibiotics with increasing absolute neutrophil count
except for patients who developed nosocomial infections
causing delay in recovery and prompting shift of
antibiotics.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
23
DISCUSSION
Bacteremia is detected in only a handful of
pediatric cancer patients with febrile neutropenia. Studies
done by Mahmud, Latiff, Celiken, Ariffin and Muller had
growth in blood culture in only 25%, 24%, 27% 35.4%
and 24% of their subjects respectively.10,8,4,2,11 This is
comparable with the blood culture results of pediatric
cancer patients with febrile neutropenia at the Philippine
General Hospital. In terms of major etiologic agents, the
trend varies per institution. Although Gram-positive
bacteria accounts for 60-70% of microbiologically
documented infections based from the IDSA data, this is
not so in other institutions. In a study done by Mahmud
among pediatric cancer patients with febrile neutropenia
in Riwalpindi 51.7% of his subjects grew Gram-positive
bacteria in their blood culture.10 This is similar to the
study done by Celkan in Istanbul, Turkey where 60% of
the isolated bacteria were coagulase-negative
Staphylococcus sp. and Staphylococcus aureus.4
63.4% of the subjects in a study done by Kern grew
coagulase-negative Staphylococcus sp.7. Gram-negative
bacteria on the other hand, was the predominating isolates
in the studies done by Latiff and another one by Ariffin
with yield of 64% and 62% respectively.8,2 While the
international trend shows that gram-positive bacteria is
the predominant isolate, the 2002 Guidelines for the Use
of Antimicrobial Agents in Neutropenic Patients with
Cancer published by IDSA still recommends the use of
broad-spectrum antibiotics that would cover for both
gram-positive and gram-negative bacteria. Hence, most
patients respond well to the empiric antibiotics given.10,
8,4,2,11
CONCLUSION
Majority of pediatric cancer patients with febrile
neutropenia yielded no growth in blood culture. However,
most cases responded well with the empiric antibiotic
combination of Ceftazidime and Amikacin. The
predominant pathogen isolated were Gram-negative
bacteria which is in contrast with the international trend.
These are also sensitive to the current antibiotic regimen.
The empiric antibiotic therapy are effective and well
tolerated in febrile neutropenia.
RECOMMENDATION
The types of microbial isolates and sensitivity
patterns should be continuously monitored to identify
trends and to gauge the suitability of antibiotics chosen
for empirical therapy. The microbiological laboratory
should also act in consultation with the clinician to
determine the optimal approach for blood cultures in the
immunocompromised patients.REFERENCES
1. Ammann RA, Hirt A, Luthy AR, et. al. Predicting bacteremia
in children with fever and chemotherapy-induced
neutropenia. Pediatric Infectious Disease Journal. 2004
January;23(1):61-7.
2. Ariffin H, Navaratnam P and Lin HP. Surveillance study of
bacteremic episodes in febrile neutropenic children.
International Journal of Clinical Practice. 2002
May;56(4):237-40.
3. Behrman RE, Kliegman RM, Jenson HB. Nelson Textbook
of Pediatrics 16th edition 2000:785-787.
S I R S I R S I R S I R S I R S I R
Ceftazidime +
Amikacin
Meropenem
Piperacillin-
Tazobactam
Ciproflo-
xacin
Netilmycin
Co-amoxy-
clav
Table 5. Sensitivity Pattern of Blood Microbial Isolates
Anti- S. G(-) S. Candida Salmo- H.
biotic viridans bacilli epider sp. sp. nella alveii
midis
+
+
+
+
+
+
+
+
4. Celkan T, Ozkan A, Apak H, et. al. Bacteremia in childhood
Cancer. Journal of Tropical Pediatrics. 2002
December;48(6):373-7.
5. Dubey AP, Singhal D and Prakash SK. Febrile episodes in
childhood malignancies. Indian Pediatrics. 2002
October;39(10):952-7.
6. Hughes WT, Armstrong D, Bodey GP, et al. 2002
Guidelines for the use of antimicrobial agents in
neutropenic patients with cancer. CID 2002;34:730-751.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
24
47. Kern WV. Risk assessment and risk-based therapeutic
strategies in febrile neutropenia. Infectious Disease
2001;14:415-422.
8. Latiff Z, Zulkifli SZ and Jamal R. Risk Assessment and
microbiological profile of infections in pediatric cancer
patients with febrile neutropenia. Malays J Pathol. 2002
December;24(2):83-9.
9. Muller J., Schmidt M, Koos R, et. al. Management of
fever in immunocompromised children. Original Abstracts-
2001: abstract nr 35, 6th International Symposium on Febrile
Neutropenia. 2001.
10. Mahmud S, Ghafoor T, Badsha S, et. al.. Bacterial infections
in pediatric patients with chemotherapy induced
neutropenia. J Pak Med Assoc. 2004 May;54(5):237-43.
11. Muller J, Schimdt, Koos R, et. al. Management of fever in
immunocompromised children. Original Abstracts-2001:
abstract nr 35, 6th International Symposium on Febrile
Neutropenia. 2001
12. Rebedea I, Radu I and Davila C. Empirical therapy for
bacterial infections in neutropenic patients. Original
Abstracts-2001: abstract nr 35, 6th International Symposium
on Febrile Neutropenia. 2001
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
25
ABSTRACT
Objective: To determine the in vitro antimicrobial activity
of probiotic containing milk against community acquired
bacterial pathogens in the pediatric age group.
Design: Experimental Study
Methods: In vitro analysis of the antimicrobial activity
of probiotic milk in comparison to antibiotics against
control strains based on the Kirby Bauer Method of disc
susceptibility testing. Kruskall Wallis Test was used to
analyze the difference between the mean zones of
inhibition of the different control groups. Chi square and
Fisher Exact Test was used to analyze % susceptibility.
Results: Zones of inhibition of probiotic containing milk
were observed among the three bacterial pathogens
tested. The mean zones of inhibition produced by the
probiotic milk showed smaller means for Streptococcus
pneumoniae ( 9.04 +/-5.26) and H. influenzae (9.76 +/
- 7.65) as compared to their respective antibiotics:
Penicillin G (30.18+/-0.70) and Chloramphenicol (30.96
+/- 1.86) (p value <0.000001). A larger mean zone of
inhibition was produced for Staphyloccocus aureus
(16.96+/- 5.30) compared to Oxacillin (12.92+/- 0.65).
Comparison of the % susceptibility showed higher
susceptibility of Streptococcus pneumoniae to Penicillin
G and H. influenzae to Chloramphenicol when compared
to the probiotic containing milk. Staphyloccocus aureus
however, showed a better susceptibility to the probiotic
containing milk (88%) vs. Oxacillin (64%) with a P value
of 0.04.
Conclusion: Probiotic milk containing Lactobacillus and
Bifidobacteria was observed to have in vitro
antimicrobial activity against Streptococcus pneumoniae,
H. influenzae, and Staphyloccocus aureus . It has a
better antimicrobial activity against S. aureus as
demonstrated by a larger zone of inhibition and increased
proportion of disc susceptibility than Oxacillin.
Simone Rosario C. Cid, MD*, Ma. Carina Cruz, MD*, Vincent Faustino, MD*, Alexander O.
Tuazon, MD*
IN VITRO STUDY ON THE ANTIMICROBIAL ACTIVITY OF PROBIOTIC
MILK AGAINST COMMON PEDIATRIC COMMUNITY ACQUIRED RES-
PIRATORY PATHOGENS
INTRODUCTION
There has been an increasing scientific and
commercial interest in the use of beneficial
microorganisms, or probiotics for the prevention and
treatment of disease. For many years, probiotics such as
Lactobacillus and Bifidobacteria in food products have
been touted for their reputed health benefits. However,
scarcity of supporting evidence for these health effects
was, in previous years, largely anecdotal. Until recently,
evidence has started to accumulate, as studied in literature,
from good quality clinical control trials with randomized,
placebo controlled design; and deductions from well-
founded in vitro studies.
The use of probiotics to control certain infections
has, likewise, started to gain acceptance. The alarming
rise of inappropriate antibiotic use, and antimicrobial
resistance, along with renewed interest in ecological
methods to prevent infections, makes probiotics a very
interesting field for research. Latest studies to date on
their potential use in infectious disease include an in vitro
study showing comparable antimicrobial activity of
probiotic containing milk with that of breastmilk against
common bacterial isolates in a hospital setting.1 Another
double blind, randomized controlled trial in daycare centers
studied the effect of probiotic milk on diarrhea and
respiratory infections in children, and reported for the
first time reductions in respiratory infections and their
severity.2 The extent to which probiotic containing milk
is able to exert activity against common bacterial
pathogens, however, has not been studied. Because of
this new benefit, an in vitro study to investigate the
antimicrobial activity of probiotic- containing milk against
common community acquired bacterial pathogens in
children is being undertaken. A comparison of the
antimicrobial activity of the commonly used antibiotics
against these pathogens in vitro also needs to be
determined.
Keywords: probiotics, antimicrobial activity, community, acquired pathogens
*Department of Pediatrics, UP-PGH
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
26
OBJECTIVES
General Objectives
This study was undertaken to determine the in
vitro antimicrobial activity of probiotic- containing milk
against community acquired bacterial pathogens in the
pediatric age group.
Specific Objectives
1. To determine the in vitro antimicrobial activity of
probiotic-containing milk against pediatric community
acquired bacterial pathogens, namely: H. influenza,
S. pneumoniae, and S. aureus
2. Compare the in vitro antimicrobial activity of
probiotic-containing milk versus Chloramphenicol,
Penicillin G, and Oxacillin based on zones of inhibition
and on known susceptibility patterns of H. influenza,
S. pneumoniae, and S. aureus to these antimicrobials
MATERIALS AND METHODS
Samples
One commercial milk formula containing
probiotics Lactobacillus and Bifidobacteria (Gain Plus,
ABBOT Pharmaceuticals) was prepared as instructed
by its manufacturer using sterile water and containers.
Using sterile syringes, two ml of prepared milk formula
was transferred into sterile plastic petridishes ready for
use.
The pathogens used as reference bacterial
strains are control srains as follows: Haemophilus
influenzae ATCC 49247, Steptococcus pneumoniae
ATCC 41619, and Staphylococcus aureus ATCC 25922
obtained from the Microbiology and Infectious Disease
Center in Alabang
Controls
Sterile water was used as negative control. The
antibiotics Chloramphenicol, Penicillin G and Oxacillin
were used as positive control.
Preparation of the Culture Media
Standardization of the culture media and
preparation of the control strains through broth cultures
was done by a single senior medical technologist in
accordance with the National Committee for Clinical
Laboratory Standards ( NCCLS 2000). Probiotic-
containing milk and Antibiotic Susceptibility Testing
For testing susceptibility of the bacterial
pathogens, the Kirby Bauer Disk Diffusion method was
employed. 75 sterile blank 6 mm disks were dipped into
the probiotic -containing milk preparations then placed
and equally distributed on the prepared MHA culture
media streaked with the three respiratory pathogens
alongside antibiotic disks specific for a particular bacteria;
Penicillin G for S. pneumoniae, Chloramphenicol for H.
influenza , and Oxacillin for S. aureus . The same was
done for sterile water which was used as negative control.
All disks were incubated for 18-24 hours at a temperature
of 37 0C. Measurement of the zones of inhibition produced
by the probiotic-containing milk, antibiotics (positive
control), and sterile water (negative control) were done
on the 24th hour using a standard caliper of 0.05 mm.
Results were recorded and tabulated accordingly.
Outcome Measurement and Data Analysis
In vitro antimicrobial activity of probiotic
containing milk against common community acquired
bacterial pathogens was measured by zones of inhibition
1. Measurement of the mean zones of inhibition of
milk containing Lactobacilli and Bifidobacteria
2. Measurement of the mean zones of inhibition of
the negative and positive controls
3. Comparison of the mean zones of inhibition of
probiotic-containing milk and known zones of
inhibition based on standard susceptibility patterns
of antibiotics: Chloramphenicol, Pen G, and Oxacillin
against H. influenza, S. pneumoniae, and S.
aureus, respectively, was done using the Kruskal
Wallis Test and Mann Whitney U test. A p value of
<0.05 was considered statistically significant.
4. Measurement of the percentage susceptibility
defined as proportion of disks showing susceptibility
of test organisms to antibiotics an probiotic
containing milk, based on standard susceptibility
patterns of the antibiotics use for each organism
was compared using Chi-square test and Fisher
Exact Test
RESULTS
This study included one commercial milk formula
containing the probiotics Lactobacillus salivarius and
Bifidobacteria; and control strains namely, S.
pneumoniae, H. influenzae and S. aureus
The zones of inhibition produced by the probiotic-
containing milk , Penicillin G, Chloramphenicol, and
Oxacillin against the these pathogens were compared.
The zone of inhibition produced by the milk is a measure
of its activity to the reference bacteria, with a larger value
interpreted as representing a better in vitro antimicrobial
activity.
Table 1 and 2 showed a significant difference
in the mean zones of inhibition between the probiotic-
containing milk and Penicillin G (p value <0.000001).
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
27
For Hemophilus influenzae, the result likewise,
showed a significant difference in the mean zones of
inhibition. When compared to Chloramphenicol, the
probiotic-containing milk was noted to have a weak in
vitro antimicrobial activity against this pathogen. (Tables
3 & 4)
When the proportion of the % susceptibility was
compared between the 3 groups using Chi- square ,
significant differences were noted. Streptococcus
pneumoniae showed a 100% susceptibility to Penicillin
G, with only 8% to the probiotic containing milk. H.
influenzae showed an 87.5% susceptibility to
Chloramphenicol as against 8% with the milk tested.
Staphylococcus aureus, on the other hand, showed a
Using established standard susceptibility patterns,
zones of inhibition produced by the milk and Penicillin G
for Streptococcus were labeled as susceptible if it had a
zone of > 20mm, 29 mm for Chloramphenicol against H.
influenza, and 13 mm for Oxacillin against
Staphylococcus aureus.
For Staphylococcus aureus, comparison was
significantly different. Compared to Oxacillin, the mean
zone of inhibition produced by the probiotic containing
milk was significantly larger, hence, better antimicrobial
activity. (Table 5 & 6)
Table 3. Comparison of the Mean Zone of Inhibition for
Hemophilus influenzae
Mean +/- SD 9.76 + 7.65 6.0 + 0 30.96+ 1.86
Median 6 6 30.5
Probiotic- Sterile Chloram- P value
milk water phenicol
<0.000001 (S)
Kruskall
Wallis Test
Table 2. P Values for the Comparison of Two Groups
Variables p value
Probiotic- containing Milk
Vs
Sterile Water
Probiotic- containing Milk
Vs
Penicillin G
Penicillin G
Vs
Sterile Water
0.005
0.000001
0.000001
Table 1. Comparison of the Mean Zone of Inhibition for
Streptococcus pneumoniae
Mean +/- SD 9.04 + 5.26 6.0 + 0 30.18+ 0.70
Median 6 6 30
Probiotic- Sterile Penicillin P Value
containing Water
Milk
<0.000001 (S)
Kruskall
Wallis Test
Table 4. P Values for the Comparison of Two Groups
Probiotic- containing Milk
Vs
Sterile Water
Probiotic- containing Milk
Vs
Chloramphenicol
Chloramphenicol
Vs
Sterile Water
Mann Whitney U Test
Variables P Value
0.009
0.000001
0.000001
Table 5. Comparison of the Mean Zone of Inhibition for
Staphylococcus aureus
Mean +/- SD 16.96 + 5.30 6.0 + 0 12.92+ 0.65
Median 15 6 30
Probiotic- Sterile Oxacillin P Value
containing Water
Milk
<0.000001 (S)
Kruskall
Wallis Test
Probiotic- containing Milk
Vs Sterile Water
Probiotic- containing Milk
Vs Oxacillin
Oxacillin
Vs Sterile Water
Table 6. P Values for the Comparison of Two Groups
Variables P Value
0.00004
0.0004
0.000001
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
28
In ancient times, the benefit and health potential
of foods containing live bacteria have been recognized.
During the beginning of the 20th Century, Elie
Metchnikoff proposed a scientific rationale for the
beneficial effects of bacteria in yogurt and attributed long
life to intake of yogurt containing Lactobacillus species.
Since then, multiple antimicrobial properties have been
suggested.
The word ‘probiotic’ is derived fron the Greek,
‘for life’.5 Today, they are defined as live microbial food
supplements with a demonstrated effect on human health.
There are several commercially available supplements
contaning probiotics, either as fermented food products
or dairy based foods. Recently, with the advancement in
research and infant nutrition, it has been introduced in
some milk formulas and powdered milk supplements.
The main mechanisms whereby probiotics exert
protective or therapeutic effects are not fully elucidated,
but multiple mechanisms have been postulated including
lactose digestion,6 production of antimicrobial agents,7,8
competition for space nutrients, and immunomodulation:
adjuvant like effects on intestinal and systemic immunity9
and non-humoral immunity.10
In a review of recent studies, probiotic therapy
showed substantial evidence of clinical benefit in pediatric
patients with viral gastroenteritis and prevention of
nosocomial diarrhea;11 in the treatment of Clostridium
defficile diarrhea.12 The use of Lactobacillus GG was
also found to be effective in the prevention of early atopic
disease in children with high risk.13 These clinical studies
have attempted to establish the value of probiotics in the
prophylaxis and treatment of childhood diseases. Fewer
studies have been done to evaluate their role in the
prevention of community acquired infections. In one such
study, long term consumption of probiotic milk was found
to be beneficial in reducing incidence of diarrhea and
respiratory infections.
To evaluate further the probiotics’ potential in
infections, the in vitro antimicrobial activity of a probiotic
milk containing Lactobacillus and Bifidobacteria
against common bacterial respiratory pathogen was
determined. Zones of inhibition were observed in all the
three organisms tested. However, among these 3
organisms, only Staphylococcus aureus was found to be
susceptible to the probiotic containing milk with a more
superior antimicrobial activity over Oxacillin. In a similar
study comparing the same probiotic containing milk and
breastmilk, zones of inhibition were likewise larger for S.
aureus , and the observed antimicrobial activity was
comparable to breastmilk.
Table 8. P Values for the Comparison of the % Susceptibility
Between Two Groups
DISCUSSION
Increased risk of disease in the pediatric age
group has obvious public health and economic
consequences, such as direct medical costs as well as
indirect costs of parents having to take time off from
work to look after sick children.3 The successful
prevention of infections could be extremely useful for
families and for society in general. Since prevention would
obviate the need for treatment, the use of probiotic
bacteria to prevent common childhood infections have
been proposed.4
better susceptibility pattern to the probiotic containing milk
compared to Oxacillin: 88% vs. 64%. (Table 7)
Comparison of the % susceptibility between two
groups using the Fisher Exact Test showed similar results.
Moreover, the results further show that the %
susceptibility of S. aureus to probiotic containing milk is
significantly higher than Oxacillin. (p value 0.04).
Table 7. Comparison of the % Susceptibility
Probiotic Sterile Anbtibiotic p value
Containing Water
Milk
Streptococcus 2 (8.0%) 0 (0%) 25 (100%) 0.000001
pnueumoniae (Penicillin G) (Significant)
H. influenzae 2 (8.0%) 0 (0%) 7 (87.5%) 0.000001
(n=8) (Significant)
(Chloram-
phenicol)
Staphylococcus 22 (88.0%) 0 (0%) 16 (64.0%) 0.000001
Aureus (Oxacillin) (Significant)
Probiotic-
containing
Milk
Vs
Sterile Water
Probiotic-
containing
Milk
Vs
Antibiotic
Antibiotic
Vs
Sterile Water
0.49
(Not
Significant)
0.000001
(Significant)
0.000001
(Significant)
0.49
(Not
Significant)
0.00006
(Significant)
0.000001
(Significant)
0.000001
(Not
Significant)
0.04
(Significant)
0.000001
(Significant)
Streptococcus
PneumoniaeHemophilus
influenzaeStaphylococcus
aureus
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
29
REFERENCES
1. Torralba, R. In vitro study of the antimicrobial effect of
probiotic containing milk vs. breastmilk against 5 common
pediatric bacterial isolates at MDH. July 2001.
2. Hatakka, K, et. Al. Effect of long term consumption of a
probiotic milk on the infections in children attending day
care centers: A double blind randomized trial. BMJ
2001;322:1327-9.
3. Bell DM, Gleiter, Merces, AA. Et al. Illness associated with
child day care: A study of incidence and cost. Am J. of
Public Health. 1989; 79: 479-84.
4. Wanke, C. Do probiotics prevent childhood illnesses. BMJ
June 2001; 322: 1318-1319.
5. Alvarez- Olmos MI, Oberhelman RA. Probiotic agents and
infectious disease: A modern perspective on traditional
therapy. CID. 2001.June 1;32 (11) 1567-76.
6. Salminen, S., Bouley, MC. Functional food science and
gastrointestinal Physiology and function. BT J. Nutrition.
1998;80 ( Suppl 1) 147-71.
7. Mcfarlane G, Cummings JH. Probiotics and prebiotics: can
regulating the intestinal bacteria benefit health. BMJ
1999;318:999-1003.
8. SilvaM, Jacoubs NV, Deneke C. Gorbarch SL. Antimicrobial
substance from a human lactobacillus strain. Antimicrobial
agents Chemother 1987; 31: 1231-3.
9. Koop- Hoolihan. Prophylactic and Therapeutic Uses of
Probiotics: A Review: J Am Diet Assoc. 2001 Feb;
101(2):229-38
10. Sheih YH, Chiang BL, Wang LH, Liao, CK, Gill, HS.
Systemic Immunity- enhancing Effects in Healthy subjects
Following Dietary Consumption of Lactic acid Bacterium
Lactobacillus rhamnosus HN001. J. Am Coll Nutr.2001
Apr;20 (2 Suppl): 149-56.
11. Szajewska et al. Efficacy of Lactobacillus GG in Prevention
of Nosocomial Diarrhea in Infants. J Pediatr.200.Mar;
138(3):361-5.
12. Marteau, PR, et al.Protection from Gastrointestinal
Diseases With the use of Probiotics Am.J Clin Nutr 2001
Feb; 73(2 suppl): 430S- 436S.
13. Kolliomaki M, et al. Probiotics in the Primary Prevention of
Atopic Diseases: A randomized Placebo Controlled Trial.
Lancet 2001 Apr 7;357(9269): 1076-9.
14. Kimura S. et al. Multiple antibiotic resistant lactic acid
bacteria preparation Eliminated MRSA from the decubitus
of a bed ridden elderly patient. Clin Med J. 1997;110:157-9.
The significant antimicrobial activity of the
probiotic containing milk observed in this experimental
study may shed light on an interesting case report in
1997 which describes successful treatment of a decubitus
edcer colonized by MRSA with the use of a Lactobacillus
preparation.14 Studies of this potential use may have
profound impact in the coming years.
CONCLUSION
In conclusion, milk containing Lactobacillus and
Bifidobacteria has in vitro antimicrobial activity against
S. pneumonia and H. influenzae but less than Penicillin
amd Chloramphenicol, respectiveley. It has better
antimicrobial activity against S. aureus as demonstrated
by larger zones of inhibition and increased proportion of
disc susceptibility to than Oxacillin.
RECOMMENDATION
This experimental study has presented data
showing that probiotic containing milk has a significant
antimicrobial activity against Staphylococcus aureus.
Further research is required to delineate this potential. It
is therefore recommended that clinical in vivo studies be
done to measure this in vitro advantage, in order to
translate it to clinically relevant outcomes.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
30
THE EFFECT OF TOPICAL APPLICATION OF MUPIROCIN IN
INTRAVENOUS CATHETER SITE IN THE INCIDENCE OF SUPERFICIAL
PHLEBITIS
Ronald Allan N. Geraldez, MD*, Ma. Liza M. Gonzales, MD*
ABSTRACT
Background: Superficial phlebitis is a common
complication of venoclysis although its incidence
especially in pediatric hospital setting is not often known
and evaluated. A standard aseptic technique in IV line
insertion is observed to decrease its incidence but the
use of topical antibiotic is rarely used.
Study objectives: Our objective was to determine if
topical application of antibiotic mupirocin will affect the
incidence of superficial phlebitis as compared to using
alcohol alone in the preparation of the IV insertion site.
Setting: Pediatric department wards and emergency room
of the University of the Philippines-Philippine General
Hospital.
Methodology: In a randomized control study, 69 pediatric
patients for intravenous catheter insertion were evaluated.
Thirty-six patients were assigned in the control group
whose IV insertion site were prepared with alcohol alone
while 33 patients in the case group received topical
mupirocin after application of alcohol in the IV insertion
site. The IV insertion site were than evaluated daily by
the investigator for the development of superficial phlebitis
until the IV cannula were removed.
Results: Eight out of the 36 patients (22%) in the control
group while 4 out of the 33 patients (12%) in the case
group developed phlebitis.
Conclusion: The use of topical mupirocin in the IV
insertion site prior to cannulation can decrease the
incidence of superficial phlebitis.
Background
Phlebitis in insertion site is not an uncommon
complication of peripheral intravenous catheterization with
cases reported to range from a low 2.3% to as high as
31%.1.2,3,4 It can manifest as an inflammation in the
insertion site to cellulitis and suppuration in the contiguous
areas to a more severe catheter related sepsis.
Key Words: intravenous cannulation, mupirocin, pediatrics, superficial phlebitis, topical antimicrobial
In UP-PGH pediatric wards and emergency
room, a proper and successful IV catheter insertion is
one of the most basic skill that interns and residents should
learn. This can be made difficult by the fact that
peripheral veins of pediatric patients are small and often
difficult to locate visually and by palpation. Formal IV
therapy training for hospital personnel has been shown
to decrease leakage, phlebitis and infiltration
complications of IV cannulation.5 In PGH, however, most
medical interns learn the technique by practice while
serving their rotation in the pediatrics department.
A proper and effective way of doing IV
catheterization entails an aseptic technique that every
health personnel performing the procedure should
observe. However, the incidence of phlebitis may still
occur in the insertion site requiring the removal of the IV
cannula and reinsertion of new cannula in another site.
Several factors may predispose to the
development of superficial phlebitis in IV cannulation site
foremost of which is the length of time the cannula is in
place or the dwell time. The Center for Disease Control
guidelines recommends replacement of IV catheter every
48 to 72 hours for adults but no such recommendation
for pediatric patients exists.6 Studies however has shown
that there is no significant differences between phlebitis
rate of cannula with dwell time of 72 hour and 96 hour.7,1
Thus, considering the difficulty of successful cannulation,
the limited number of skin sites, and the cost of the devise,
the cannula may be left in place for longer than 48 hours.9
Among neonates, catheter life is on the average lasts for
only 30 hours. Extravasation, erythema, accidental
displacement, and worse, phlebitis may require the
removal and reinsertion into another site of the cannula.
Without these complications however, catheter can be
safely maintained with adequate monitoring for up to 144
hours.3
It is advisable that IV cannula should be used
only once per attempt but this is not always the case in
PGH where the same cannula can be used several times*Department of Pediatrics, UP-PGH
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
31
in an attempt at successful cannulation. A transparent
polyurethane cannula (Insyte) is the most commonly type
of used which compared to butterfly steel cannula is
said to decrease the risk of phlebitis.10, 11, 12
A significant factor in the development of
phlebitis is the infusate or the type and frequency of
medication or fluid infused and pushed. Drug irritation is
the most reliable predictor of phlebitis.13 Total parenteral
nutrition, blood products, potassium and sodium
bicarbonate drips are just some of the infusate commonly
implicated in phlebitis.
There are commercially available IV cannula
dressings available in the market such as transparent
dressing and sterile gauze but these are more expensive
and thus has not been popularly used. The use of gauze
versus transparent dressing shows no relationship with
IV complications such as bacterial colonization and
phlebitis although a study has shown less evidence of
phlebitis with adhesive bandage compared with gauze.14,3
Bacterial colonization or the presence of positive culture
in the cannula tip is widely believed to be not correlated
with the development of phlebitis.15 There has been no
reported study comparing these commercially available
dressing with adhesive tape (Leukoplast) which is
generally what is being used in PGH.
The most practical way of preventing phlebitis is
the employment of aseptic technique in the performance
of the peripheral IV cannulation procedure. The usual
practice in this institution is to topically clean the area in
the skin with alcohol and the secure the site with
leukoplast. Application of a topical antibiotic
prophylactically in the insertion site such as mupirocin
ointment is not usually done except although occasionally,
povidone iodine, a topical antiseptic is used. Some studies,
however has shown that the use of antimicrobial ointment
has resulted in higher proportion of phlebitis.16
Mupirocin, a topical antibiotic available in
ointment form has been used in the treatment of
secondarily infected wounds. It is likewise often used in
the care of indwelling central lines to prophylactically
prevent phlebitis or even treat secondarily infected central
line site. It is effective against gram positive and gram
negative organisms including methicillin resistant
Staphylococcus aureus. As to date, there is no local
study investigating the efficacy of the application of
mupirocin in the prevention of the development of
superficial phlebitis in peripheral cannulation site.
OBJECTIVES
The primary objective is to assess whether single
topical application of mupirocin ointment in the peripheral
IV cannulation site administered prior to IV cannulation
compared to applying alcohol alone decreases the
incidence of phlebitis.
Secondary objectives include:
1. to determine and assess the factors that predisposes
to the development of superficial phlebitis namely:
a. the length of time the IV cannula is in place
b. the type of IV medications infused
c. the number of attempts before the cannula has been
inserted
d. the type of personnel (i.e. intern, resident) who
performed the cannulation
Type of Study
A randomized controlled trial in which mupirocin
ointment topically applied prior to IV cannulation is
compared with using alcohol alone.
Participants
Patients in the pediatric wards and emergency
room requiring indwelling peripheral IV cannulation for
more than 24 hours.
Exclusion Criteria
Patients whom IV cannula was removed within
24 hours from insertion.
Outcome Measures
The presence of signs of phlebitis in the area
where IV cannulation has been placed as assessed by
the investigator.
METHODOLOGY
All patients in the emergency room and wards
for intravenous catheter insertion that was referred for
inclusion in the study was randomly assigned to case
group or control group. For every patient enrolled in the
study a card was drawn from a set of cards randomly
marked with “B” and unmarked cards. Those assigned
with unmarked cards was put in the control group while
those with marked “B” was placed in the case group
The choice of IV site for each patient was at
the discretion of the physician and or intern as is the choice
of the IV catheter to be used. The insertion technique
was done percutaneously without prior skin incision. The
skin was prepared with alcohol. All patients in the case
group will have topical mupirocin applied to the area
covering at least 0.25 cm prior to IV insertion. The
insertion site would then covered with an adhesive tape.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
32
The catheters were subsequently handled ac-
cording to the normal practice of the attending medical
and nursing staff. Each patient was seen daily by the
investigator and the patient was questioned about pain in
the insertion site, or/and the IV site was inspected and
palpated. The presence of phlebitis was defined as the
presence of a palpable cord or the presence of at least 2
of the following physical changes along the course of the
vein: warmth, erythema, tenderness and induration.
RESULTS
Of the 92 patients enrolled in the study, only 36
in the control group and 33 in the case group were
included in the evaluation. Twelve patients in the control
group out of 48 (25%) and 11 out of the 44 (25%)
patients in the case group were excluded because their
IV cannula were removed in less than 24 hours or the
patients were discharged before the investigator was able
to assess the IV site. Subject characteristics for both
groups were similar with respect to sex although there in
terms of age, there were more subjects in the 1 to 12
months age group among the control. For both control
and case group, the locations of the IV insertions had
almost similar distribution and most were done in the hand.
Final result had shown that 8 out of the 36 (22%) subjects
in the control group developed phlebitis as compared to
the case group in whom 4 out of the 33 (12%) subjects
had phlebitis.
A. Dwell time
The average dwell time for subjects in the control
group is 62.9 hours (SD 30.3 hours) as compared to the
case group with average dwell time of 62.8 hours (SD
30.4). If subjects whom IV cannula were electively
removed or those whom cannulation where removed
because they were no longer need for were excluded,
the average dwell time became 62.7 hours (SD 29.9) for
the control and 65 hours (SD 27.6) for the case group.
Of the total of 12 patients who had phlebitis, 5 had their
cannula in place for 24-48 hours, 3 of them for 49-72
hours, 2 for 73-96 hours while 2 had their cannula in place
for 145-168 hours.
B. Type of Infusate
Of the 12 patients in the control group who
developed phlebitis, 8 were given IV fluids while 4 had
heparin lock. Eight were given IV antibiotics while 3
were given blood products.
Table 2. Subjects in the control group who developed phlebitis
and the type of infusate given
1 D5 0.3 NaCl Mannitol,
Dexamethasone
2 D5 IMB Ampicillin, Famotidine
Metronidazole
3 D5 IMB Penicillin G, Vitamin K,
Amikacin Paracetamol
4 Meropenem
5 Piperacillin- Midazolam
Tazobactam
6 pNSS pRBC
7 D5IMB + Meropenem Famotidine
K2 + Ca200
8 D5IMB Ampicillin, Vitamin K
Amikacin
Subjects
with
phlebitis IV fluids Antibiotic Others IV meds
Type of infusate
Control Group
Control
I. Sex:
Male 19
Female 12
II. Age Group
0-<1mo 3
1-12mos 19
1-5yrs 6
6-12yrs 7
13-18yrs 1
III. IV Cannulation Site
Hand:
Right 15
Left 8
Forearm:
Right 0
Left 3
Foot:
Right 5
Left 4
Leg:
Right 0
Left 1
Scalp: 0
Mupirocin
I. Sex:
Male 20
Female 13
II. Age Group
0-<1mo 3
1-12mos 11
1-5yrs 11
6-12yrs 6
13-18yrs 2
III. IV Cannulation Site
Hand:
Right 13
Left 9
Forearm:
Right 0
Left 4
Foot:
Right 4
Left 2
Leg:
Right 0
Left 0
Scalp: 1
Subject Characteristics:
Dwell Time
24-48 hrs
49-72 hrs
73-96 hrs
97-120 hrs
121-144 hrs
145-168 hrs
# of
Subjects
16
10
4
4
1
1
+
Phlebitis
5
2
1
0
0
0
# of
Subjects
14
10
5
1
2
1
+
Phlebitis
0
1
1
0
2
0
Control Mupirocin
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
33
Table 2. Subjects in theMupirocin group who developed
phlebitis and the type of infusate given
C. Number of times the IV cannula were used
There could be several attempts in IV insertion
in a subject and a cannula could be used several times by
the intern or the residents before a successful cannulation.
Majority of IV insertion fortunately was successful on
first attempt. Seven out of 51 of those who developed
phlebitis had the IV cannula used only once in an attempt.
There were 2 subjects whose cannula was used 4 times
but did not developed phlebitis.
1 D5 0.3 Nacl Furosemide
2 Penicillin G
Choramphenicol
3 pNSS Cefepime
Metronidazole plt conc
4 D5 0.3 Nacl plt conc
Subjects
with
phlebitis IV fluids Antibiotic Others IV meds
Type of infusate
Mupirocin
Table 4. No of times the cannulla has been used and the inci-
dence of phlebitis
No. of
Times the
IV Cathwas
used
1
2
3
4
30
3
1
2
+ Phlebitis
5
1
1
1
# of
Subjects
21
10
2
0
2
2
0
0
Mupirocin Control
D. Type of Personnel Performing the cannulation.
In PGH pediatric wards and emergency room,
the interns were the first in line who should perform the
IV insertion, thus 70% (36/53) of the IV insertion were
performed by the interns while the rest were done by the
residents. Seven out of the 36 insertions done by the
interns developed phlebitis while 4 out of the 18 insertions
done by the residents had phlebitis.
Table 5. The type of personnel performing the IV cannulation
and the incidence of phlebitis
Type of
Personnel
Interns
Residents
# of
Subjects
24
13
+ Phlebitis
4
4
# of
Subjects+ Phlebitis
3
1
Control Mupirocin
20
13
DISCUSSION
Phlebitis is the most common complication of IV
therapy and several factors has been implicated in its
pathogenesis. Chemical factors such as irritant drugs
and physical factors such as duration of cannulation are
just few of the identified risk factors for the development
of phlebitis.
Among the 69 subjects both in the control and
mupirocin group evaluated in this study, 12 developed
phlebitis thereby giving an incidence rate of 17%. This
is within range of the incidence of phlebitis reported in
other studies which is from a low of 2.3% to a high
31%.1,2,3,4 For those in the control group, 22% (8/36)
developed phlebitis, a rate that is higher than in the case
group with 12% (4/33). With using topical topical
mupirocin the absolute risk reduction is 10% and the
relative risk reduction of 46%. The results therefore
suggests that applying topical mupirocin (Bactroban) in
the IV insertion site may the decrease the risk of phlebitis.
However, using the statistical analysis Fisher’s exact test
this finding is not statistically significant with a 2-tail p
value of 0.34781
The rationale for the use of topical antimicrobial
in the preparation of the skin for IV insertion can decrease
the bacterial load of the skin thus decreasing the
colonization in the point of entry. Topical antimicrobial is
often employed in the care of central venous catheter
but its use in percutaneous IV cannulation is not by
standards observed. Some studies even discourage the
use of antimicrobial ointment because it can result in
higher incidence of phlebitis.
The length of time that the IV cannula is in place
is traditionally believed to be directly correlated with the
incidence of phlebitis. Two subjects however whose
dwell time exceeds 144 hours did not developed phlebitis
although 2 subjects with dwell time of 121-144 hours did.
Five subjects in the control group with dwell time less
than 48 hours had phlebitis. The per day risk of phlebitis
is not evident in this study and this supports the no longer
acceptable practice of replacing IV cannula every three
days.
In this study, the IV insertion performed by
interns showed 16% (7/44) rate of phlebitis as compared
with residents who had 19% (5/26). In PGH, it is usually
the interns who were first in line to do the IV insertion
and the procedure would be referred to the residents in
cases of difficult insertion. The higher rate of phlebitis
among those done by residents can be probably explained
by their more difficult tasks of IV insertion especially
among chronic patients.
# of
Subjects
+
Phlebitis
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
34
Since IV cannula is sometimes used several times
in an insertion, the number of times the cannula has been
used before a successful cannulation has been recorded
and analyzed as a risk factor of phlebitis. Although
majority of the insertion has been successful on 1st
attempt, 14% (7/51) has developed phlebitis while 23%,
(3/13), 50% (1/2) and 50%(1/2) has developed phlebitis
when the cannula has been reinserted twice, thrice and
four times respectively. It is therefore recommended
that the IV cannula be used only once and be discarded
if attempt is unsuccessful. The cost of IV cannula will
however make this suggestion impractical.
These results lead to the question of whether
there is a direct relationship between difficulty of insertion
as evaluated subjectively by those performing the IV
cannulation and by the number of attempts before a
successful cannulation has been made. A similar study
with such an objective is therefore recommended.
The type of infusate such as irritant IV medication
is a significant factor in the development of phlebitis.
Among subjects who developed phlebitis, some of the
REFERENCES
1. Lai. Safety of prolonging peripheral cannula and i.v. tubing
use from 72 hours to 96 hours. Am J Infect Control 1998
Feb; 26 (1): 66-70
2. Shimandle, Johnson, Baker, Stotland, Karrison, Arnow.
Safety of peripheral intravenous catheters in children.
Infect Control Hosp Epidemiol 1999 Nov; 20(11): 736-40
3. Stephen, Loewenthal, Wong, Benn. Complications of
intravenous therapy. Med J Aust 1976 Oct 9; 2(15): 557-9
4. Tager, Ginsberg, Ellis, Walsh, Dupont, Simchen, Faich.
An epidemiologic study of the risks associated with
peripheral intravenous catheters. Am J Epidemiol 1983
Dec;118 (6): 839-51
5. Palefski, Stoddard. The infusion nurse and patient
complication rates of peripheral-short catheters. A
prospective evaluation. J Intraven Nurs 2001 Mar-Apr;
24(2): 113-23
6. Oishi. The necessity of routinely replacing peripheral
intravenous catheters in hospitalized children. A review
of literature. J Intraven Nurs 2001 May-Jun; 24 (3); 174-9
7. Homer, Holmes. Risks associated with 72- and 96-hour
peripheral intravenous catheter dwell times. J Intraven
Nurs 1998 Sep-Oct; 21(5): 301-5
8. Tobin. The Teflon intravenous catheter: incidence of
phlebitis and duration of catheter life in the neonatal
patient. J Obstet Gynecol Neonatal Nurs 1988 Jan-Feb;
17 (1): 35-42
drugs that has been infused and recognized to be often
implicated in the development of phlebitis has been IV
fluids with KCl and Ca gluconate incorporation and blood
products packe RBC and platelet concentrate. There is
however particular classification of infusate as to their
abiltity and degree to irritate the veins and cause phlebitis.
The application of topical mupirocin may not decrease
the incidence of chemical phlebitis or those caused by
irritant infusate in contrast to those with infectious etiology
although both factors may co-exist in the development
of phlebitis.
The potential of topical mupirocin in decreasing
the incidence of phlebitis therefore needs further
confirmation. The amount of ointment used per
cannulation is so small that it will not greatly increase the
cost of IV therapy. The only problem however is the
more tedious way of inserting the cannula since the skin
site can have more glare from the shiny ointment once
applied as reported by the residents and interns who
participated in the study. However, if proven by further
studies and by evidenced based medicine, the decreased
cost of reinsertion and decreased pain for the patient can
be beneficial.
9. Garland, Dunne, Havens, Hintermeyer, Bozzette, Wincek,
Bromberger, Seavers. Peripheral intravenous catheter
complications in critically ill children: A prospective study
Pediatrics 1992 Jun; 89 (6 Pt 2): 1145-50
10. McKee, Shell, Warren, Campbell. Complications of
intravenous therapy: a randomized prospective study –
Vialon vs. Teflon. J Intraven Nurs 1989 Sep-Oct; 12(5):
288-95
11. Karadag, Gorgulu. Effect of two different short peripheral
materials on phlebitis development.J Intraven Nurs. 1998
Sep-Oct; 1 (5):301-5
12. Maki, Ringer. Risk factors for infusion-related phlebitis
with small peripheral venous catheters. Ann Inter Med
1991 May 15; 114 (10): 845-54
13. Catney, Hillis, Wakefield, Simpson, Domino, Keller,
Connely, White, Price, Wagner. Relationship between
peripheral intravenous catheter Dwell time and
development of phlebitis and infiltration. J Infus Nurs
2001 Sep-Oct; 24(5): 332-41
14. Pettit, Kraus. The use of gauze versus transparent
dressings for peripheral intravenous catheter sites. Nurs
Clin North Am 1995 Sep; 30 (3): 495-506
15. Righter, Bischop, Hill. Infection and peripheral venous
catheterization. Diagn Microbiol Infect Dis 1983 Jun; 1
(2); 89-93.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
35
16. Vost, Longstaff.. Infection control and related issues in
intravascular therapy. Br J Nurs 1997 Aug 14-Sep 10; 6
(15); 846-8, 850, 852 passim
17. Campbell, Carrington. Peripheral i.v. cannula dressings:
advantages and disadvantages. Br J Nurs 1999 Nov 25-
Dec 8 ; 8(21): 1420-2
18. Garland, Nelson, Cheah, Hennes, Johnson. Infectious
complications during peripheral intravenous therapy with
Teflon catheters: a prospective study. Pediatr Infect Dis
J 1987 Oct; 6 (10): 918-21
19. Harris, Alford, Dan, Savage. Bacteremia related to IV
cannulation: variability of underlying venous infection.
South Med J 1980 Jun; 73(6):719-22
20. Hershey, Tomford, McLaren, Porter, Cohen. The natural
history of intravenous catheter-associated phlebitis. Arch
Intern Med 1984 Jul; 144(7): 1373-5
21. Hoffmann, Weber, Samsa, Rutala. Transparent
polyurethane film as an intravenous catheter dressing. A
meta-analysis of the infection risks. JAMA 1992 Apr 15;
267(15):2072-6
22. Hoffmann, Western, Kaiser, Wenzel, Groschel. Bacterial
colonization and phlebitis-associated risk with transparent
polyurethane film for peripheral intravenous site dressings.
Am J Infect Control 1988 Jun; 16(3): 101-6
23. McKee, Shell, Warren, Campbell. Complications of
intravenous therapy: a randomized prospective study –
Vialon vs. Teflon. J Intraven Nurs 1989 Sep-Oct; 12(5):
288-95
24. Meylan. Increased risk of bacterial colonization of
intravenous catheters covered with transparent adhesive
polyurethane bandages, compared to classical gauze
bandages. Schweiz Med Wocheenschr 1987 Dec 12; 117
25. Monreal, Quilez, Rey-Joly, Rodriguez, Sopena, Neira, Roca.
Infusion phlebitis in patients with acute pneumonia. Chest
1999; 115: 1576-80
26. Nelson, Garland. The natural history of Teflon catheter-
associated phlebitis in children. Am J Dis Child 1987 Oct;
141 (10): 1090-2
27. Parras, Ena, Bouza,Guerrero, Moreno, Galvez, Cercenado.
Impact of an educational program for the prevention of
colonization of intravascular catheters. Infect Control
Hosp Epidemiol 1994 Apr; 15 (4 Pt 1): 239-42.
28. Tomford, Hershey, McLaren, Porte, Cohen. Intravenous
therapy team and peripheral venous catheterassociate
complications. A prospective controlled study. Arch
Intern Med 1984 Jun; 1444(6): 1191-4
29. VandenBosch, Cooch, Treston-Aurand. Research
utilization: adhesive bandage dressing regimen for
peripheral venous catheters.Am J Infect Control 1997
Dec;25(6): 513-9
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
36
ABSTRACT
Objective: To evaluate the neonatal and maternal clinical
manifestations and their hematological parameters,
individually and in combination, as parameters which can
be used to formulate a scoring system in determining
neonatal sepsis.
Design: A cross-sectional study conducted at the
Neonatal Intensive Care Unit of a tertiary care teaching
hospital.
Methods: The study consisted of 100 neonates admitted
at Neonatal Intensive Care Unit at the UP-PGH Medical
Center who were clinically suspected of sepsis at birth
and within 24 hours of life. A perinatal history, clinical
profile, symptoms and laboratory data were recorded in
each case. The neonatal hematological parameters
included were total leukocyte count, total neutrophil count,
lymphocytes, immature cells, immature to total leukocyte
ratio, immature to mature cells ratio, nucleated red blood
cells, lymphocytes, absolute neutrophil count, platelet
count, and toxic granules. The maternal hematological
parameters consisted of total leukocyte count, total
neutrophil count, lymphocytes and platelet count. These
parameters were evaluated based on the standard
reference values. A blood culture was the standard
indicator for proven sepsis.
Results: There were 17 out of 100 neonates (17%)
who had culture proven sepsis and they were
predominantly preterm. Among the different parameters,
the preterm infants, neonatal platelet count and maternal
total leukocyte count were significantly associated with
neonatal sepsis with p value of 0.047, 0.02, and 0.006
respectively. Based on these factors, a scoring system
was devised to predict the probability of sepsis. A score
of 3 had a 100% sensitivity and 91.3% specificity.
Conclusion: A scoring system for predicting neonatal
sepsis could be obtained by correlating the clinical
manifestations of the neonate and the mother together
with their hematological parameters.
CLINICAL CORRELATION OF NEONATAL AND MATERNAL
HEMATOLOGICAL PARAMETERS AS PREDICTORS OF NEONATAL
SEPSIS
*Department of Pediatrics, UP-PGHKeywords: Neonatal sepsis, hematological parameters, scoring system, new-
born, perinatal infection
Willa Antoniette B. Mayuga, MD*, Pura Flor D. Isleta, MD*
INTRODUCTION
Sepsis neonatorum is used to describe the
systemic response to infection in newborn infants. It
continues to be the major cause of morbidity and mortality
in the newborn.1 Neonatal sepsis occurs in 1 to 8 cases
of all live births.2 In the Philippines, the incidence is
estimated between 4 to 9 cases per 1000 live births.3 In
the Neonatal Intensive Care Unit of the University of
the Philippines – Philippine General Hospital, it is
estimated between 2 to 7 cases per 1000 live births with
an average sepsis rate of 7%.4
Neonatal sepsis is categorized as early or late
onset. Eighty-five percent of newborns with early onset
of infection present within 24 hours, 5% present at 24-
48 hours, and a smaller percentage of patients present
between 48 hours and 6 days of life.5 The susceptibility
of the newborn is related to immaturity of both the cellular
and humoral immune systems at birth. This feature is
particularly evident in preterm neonate. Early-onset
sepsis syndrome is also associated with acquisition of
microorganisms from the mother through blood-borne
transplacental infection of the fetus, ascending infection,
and infection upon passage through an infected birth canal
or exposure to infected blood at delivery.6 Late-onset
sepsis syndrome occurs at 7-90 days of life and is
acquired from the care-giving environment.
The early signs of sepsis in the newborn are
nonspecific. Therefore, many newborns undergo
diagnostic studies and the initiation of treatment before
the diagnosis has been determined. The definitive
diagnosis of septicemia is made by a positive blood
culture.1 The incidence of culture proven sepsis is
approximately 2 in 1000 live births. Of the 7-13% of
neonates who are evaluated for sepsis, only 3-8% have
culture proven sepsis. The mortality rate of untreated
sepsis can be as high as 50%.5 Thus, most clinicians
believe that the hazard of untreated sepsis is too great to
wait for confirmation by positive cultures. They initiate
treatment while awaiting culture results.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
37
However, due to the high cost of antibiotics,
inavailability of blood cultures in some community
hospitals, and the time it takes for the blood culture result
to come out, several studies have examined the laboratory
findings associated with sepsis. There is a lack of
consensus on the essential test that would identify
neonates with acute infection. In a systematic review to
determine the value of diagnostic tests for bacterial
infection in early life, it was reported that the accuracy
of tests varies enormously and the tests are of limited
value in the diagnosis of infection.7 In another study, a
combination of hematological and biochemical tests (eg.
acridine orange leukocyte cytospin test, nitroblue
tetrazolium and C-reactive protein) may provide a more
rapid and accurate diagnosis of bacteremia than
conventional microbiological methods.8 In recent years,
various investigators have evaluated some highly sensitive
and specific inflammatory markers (eg. C-reactive protein,
interleukin-6, interleukin-8, plasma elastase) to diagnose
neonatal sepsis and shock. Although these markers are
sensitive and specific, they require sophisticated and
expensive kits and are therefore impractical for routine
clinical work-up in a community health delivery systems,
particularly in developing countries.1
The use of hematological parameters for
determining sepsis was evaluated in different studies.
There was significant heterogenicity across these studies.9
The possible sources were population, age, subjects,
methodological quality, different leukocyte indices,
different cut-offs and interpretation of test results by
different laboratory observers. However, these
parameters remain to be rapid, economical, feasible,
practically possible in all laboratories and most especially,
these hematological parameters can be used as a tool in
screening neonates with sepsis1 which in turn may
decrease the antibiotic usage.10
GENERAL OBJECTIVE
This study is designed to evaluate the neonatal
and maternal clinical manifestations and their
hematological parameters, individually and in combination,
as parameters which can be used to formulate a scoring
system in predicting the probability of neonatal sepsis.
SPECIFIC OBJECTIVES
1. To provide a rapid identification of sepsis based on
complete blood count and peripheral blood smear in
correlation with clinical symptoms
2. To compare neonates who are more prone to
infection based on gestational age, weight, sex and
manner of delivery
3. To determine whether the newborn and maternal
symptoms correlate well with neonatal sepsis
4. To determine which of the newborn hematological
parameters, namely: the white blood count (WBC)
or total leukocyte count (TLC), total neutrophil count
(TNC), lymphocytes, immature cells, immature to
total neutrophil cells (I/T) ratio, immature to mature
cells (I/M) ratio, absolute neutrophil count, nucleated
red blood cells (NRBC), platelet count and toxic
granulation, are significant in predicting sepsis
4. To determine whether the maternal white blood count,
different count, and platelet count are also significant
in predicting sepsis
STUDY DESIGN
This is a cross-sectional study conducted at the
Neonatal Intensive Care Unit of a tertiary care teaching
hospital.
METHODS
Subjects
The study consisted of 100 neonates admitted at
Neonatal Intensive Care Unit (NICU) at the Philippine
General Hospital from July to September 2003 who were
clinically suspected of sepsis at birth and within 24 hours
of life or had maternal history of infection.
Inclusion Criteria:
Neonates with respiratory distress syndrome,
cyanosis, apnea, transient tachypnea, meconium
aspiration syndrome, pneumonia, low Apgar score, birth
asphyxia, lethargy, temperature instability, and
hypoglycemia.
Neonates with maternal history of infection such
as upper respiratory tract infection, pneumonia, urinary
tract infection, vaginitis, premature rupture of membrane,
chorioamnionitis, with or without antibiotic intake during
pregnancy.
Newborn with gestational age of 30 weeks and
above by pediatric aging and with a weight of more than
or equal to 1000 grams.
Study Procedure
Each neonate was examined by a pediatric
resident rotating in NICU or neonatology fellow who
recorded the signs and symptoms of the neonate,
predisposing perinatal factors and the clinical assessment
of the neonate.
Initial tests performed were complete blood count,
peripheral smear and blood culture. Blood samples (2
ml) were collected from the umbilical cord, peripheral
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
38
venous or arterial puncture within 24 hours of admission
before initiation of antibiotic therapy.
A 0.5-1 ml of blood sample was anticoagulated
with ethylene diamine tetra acetic acid. The total
leukocyte count and platelet count were measured on a
Coulter STKS. White blood cells were corrected for
nucleated red blood cells. Peripheral blood smears were
drawn on clean slides and stained by Wright’s stain. A
differential leukocyte count was done to obtain the total
neutrophil count (TNC), immature neutrophil count (IM),
including bands and stabs; and mature neutrophil count
(M). Neutrophils were classified as band forms when
there were no nuclear segmentation or when the width
of the nucleus at any constriction was not less than one
third the width at its widest portion. Band forms together
with less mature cell form were classified as immature
polymorphonuclear (PMN) leukocytes. Using these
values, I/M and I/T ratios were computed. One hundred
neutrophils were further examined for degenerative
changes such as toxic granulation, Dohle bodies, and
vacuolization. Toxic granulation was graded as 0 or (-)
which indicated normal granulation or no toxic granules
seen, (+) slight, (++) approximately 50% of neutrophils
contained dark granules, (+++) very high granulation in
most cells, and (++++) gross toxic granulation with the
nucleus obscured by toxic granules.11
One milliliter of blood was inoculated aseptically
into 20ml of brain heart infusion broth for culture and
sensitivity. Newborn infants with positive blood cultures
were considered to have proven sepsis while the others
were still considered as clinically suspected of infection.
The clinical manifestations and hematological
parameters were compared, individually and in
combination, with the blood culture result.
Statistical analysis
Data were analyzed by using T-test to compare
two groups with numerical data, Chi-square test to
compare or associate nominal data and Fisher Exact test
when the expected frequencies are less than 5. A level
of 0.05 was considered statistically significant. The
reference values of the neonatal hematological
parameters of Manroe, et al were used as the standard
values.12 The maternal reference values used were taken
from the values for pregnancy.13-14 The results that were
statistically significant in this study were used to design a
hematologic scoring system that will predict the probability
of sepsis.
RESULTS
Neonatal profile and neonatal sepsis
There were 17 neonates who had culture proven
sepsis which had a prevalence of 17%. Eleven of the 17
neonates were preterm (64.7%) and 6 were full term
neonates (35.3%). There was a significant correlation
between preterm and positive blood culture with p value
of 0.047. The neonatal profile showed that 12 males
(70.6%), 11 appropriate for gestational age (64.7%) with
mean birth weight of 2000 grams and 11 infants delivered
via caesarian section (64.7%) had culture proven sepsis.
However, these data were not significant (Table 1).
Table 1: Association of neonatal profile with neonatal sepsis
Patient’s
Profile
Pre-term
Full-term
Sex
Male
Female
Weight(gm)
Mean +/-
SD
Range
Weight for
pediatric
age
AGA
LGA
SGA
Manner of
Delivery
LSCS
OFE
SVD
Blood CS (+)
(n=17)
11 (64.7%)
6 (35.3%)
12 (70.6%)
5 (29.4%)
2012+/-920.7
1000 – 3800
11 (64.7%)
2 (11.8%)
4 (23.5%)
11 (64.7%)
0 (0%)
6 (35.3%)
Blood CS (-)
(n=83)
32 (38.6%)
51 (61.4%)
50 (60.2%)
33 (39.8%)
2289+/-768.6
1000 – 3900
63 (75.9%)
1 (1.2%)
19 (22.9%)
41 (49.4%)
7 ( 8.4%)
35 (42.2%)
Total
43
57
62
38
74
3
23
52
7
41
p value
0.047 (S)
(chi-square test)
> 0.05 (NS)
(chi-square test)
> 0.05 (NS)
(t-test)
> 0.05 (NS)
> 0.05 (NS)
> 0.05 (NS)
(Fisher test)
> 0.05 (NS)
(chi-square test)
The bacterial species isolated showed that 14 of the 17
(82.3%) blood culture isolates were Alkaligenes faecalis
followed by Acinetobacter (n=1), Diphteroides (n=1) and
Staphylococcus epidermidis (n=1).
Clinical profile and neonatal sepsis
Majority of the clinical manifestations of the
newborns who were suspected with sepsis had
concomitant respiratory diseases (n=77) and only 23
patients had primary impression of sepsis clinically or
based on the maternal history of infection. Among the
neonates with culture proven sepsis, there were 12
neonates who had respiratory problems (70.6%).
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
39
There were 13 out of the 67 mothers who had
illnesses during pregnancy with culture proven sepsis as
shown in Table 2. The maternal illnesses were upper
respiratory tract infection, urinary tract infection, and
premature rupture of membrane.
It was also noted that there were no significant
difference in the Apgar scores, neonatal symptoms
and maternal illnesses.
Table 2: Association of the neonatal clinical symptoms and
maternal symptoms with Neonatal Sepsis
Neonatal hematological parameters and neonatal
sepsis
The details of the neonatal hematological
parameters are shown in Table 3. The mean total
leukocyte count of the neonates with sepsis was
significantly lower than those without sepsis (p=0.03).
However, when compared to the reference values, there
was no significant difference. It could be noted that there
were 14 neonates with sepsis (82.4%) whose total
leukocyte count were within the normal range. The
nucleated red blood cells, total neutrophil count or
segmenters, lymphocytes, immature cells, I/T ratio, I/M
ratio, absolute neutrophil count and toxic granules were
not statistically significant. It was only the platelet count
which was significant (p=0.02) when compared to the
reference value.
Clinical profile
Apgar score at
1-min
1 – 3
4 – 6
7 – 9
Mean
Apgar score at
5-mins
3
4 – 6
7 – 9
10
Mean
Neonatal
Symptoms
Rule out Sepsis
HMD
TTN
Pneumonia
MAS
Maternal
Symptoms
(+)
(-)
4 (23.5%)
4 (23.5%)
9 (52.9%)
6
1 (5.9%)
4 (23.5%)
12 (70.6%)
0
8
5 (29.4%)
6 (35.3%)
1 (5.9%)
3 (17.6%)
2 (11.8%)
13 (76.5%)
4 (23.5%)
Blood CS (+)
(n=17)
14 (16.9%)
18 (21.7%)
51 (61.4)%6
1 (1.2%)
9 (10.8%)
71 (85.5%)
2 (2.4%)
8
18
(21.7%)14
(16.9%)20
(24.1%)17
(20.4%)14
(16.9%)
54 (65.0%)29
(35.0%)
Blood CS (-)
(n=83)
18
22
60
2
13
83
2
23
20
21
20
16
67
33
Total
> 0.05 (NS)
> 0.05 (NS)
> 0.05 (NS)
(chi-square
test)
> 0.05 (NS)
(chi-square
test)
p value
TLC (x109//L)
< 5 or >25 3 (17.6%) 13 (15.7%) 16 >0.05 (NS)
Normal 4.50-41.90 70 (8 4.03 84 (Fisher test)
14(82.4%) 15.6
TLC (x109//L)
Mean +/- SD 6.29 +/- 7.54 9.17 +/ 0.05 (NS)
-6.34 (t-test)
Range 1.02-32.85 0.48-28.56
Median 3.50 8.10
TNC(x109/L) 16 (94.1%) 81 (97.6%) 97 >0.05 (NS)
<0.78 or 1.45
Normal 1 (5.9%) 2 (2.4%) 3 (Fisher test)
Lymphocytes
(x109/L)
Mean+/-SD 4.17+/-2.75 5.60+/-3.17
Range 1.08-12.32 0.18-16.71 0.05 (NS)
Median 4.04 4.85 (test)
Immature<0.05
or >1.45
Normal16 (94.1%)
1 (5.9%)
73 (88%)
10 (12%)
89
11
> 0.05
(NS)
(Fisher
test)
Immature
Mean +/-
SD
Range
Median
0.64 +/
- 1.26
0 – 4.45
0
0.52 +/
- 1.22
0 – 6.97
0
> 0.05
(NS)
(t-test)
Table 3: Association of the Neonatal Hematological
Parameters with Neonatal Sepsis
Blood CS (+)
(n=17)
Blood CS (-)
(n=83)
Total p valueHematological
parameters
NRBC
Mean +/- SD
Range
Median
25.24 +/
- 44.81
0 – 146
2
5.46 +/
- 13.54
0 – 85
2
0.05 (NS)
(t-test)
NRBC
0 7(41.2%) 39 (47.0%) 46 >0.05 (NS)
>0 10 (58.8%) 44 (53.0%) 54 (chi-square
TLC (x109//L)
Mean +/- SD 12.14 +/- 16.66 +/ 0.03 (S)
8.68 -7.78
Range 4.50-41.90 1.70 – 41.00
Median 10.2 15.6
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
40
Table 3: Association of the Neonatal Hematological
Parameters with Neonatal Sepsis (cont’d)
Maternal hematological parameters and neonatal
sepsis
The obtained results of the maternal total
leukocyte count showed no significant difference but
when compared to the reference value for pregnant
women12, the results were significant. There were
94.1% mothers who had leukocytosis in the confirmed
sepsis group with p value of 0.006. The other
hematological parameters namely segmenter, lymphocyte
and platelet count had no statistical significance (Table
4).
Table 4: Association of the Maternal Hematological
parameters with Neonatal Sepsis
WBC (x109/L)
Mean +/- SD
Range
Median
WBC (x109/L)
>12
≤12
Segmenter
(x109/L)
Mean +/- SD
Range
Median
Segmenter
(x109/L)
1.8 – 7
<1.8 – >7
Lymphocyte
(x109/L)
Mean +/- SD
Range
Median
Lymphocyte
(x109/L)
1 – 4.8
< 1 - > 4.8
Platelet count
(x109/L)
Mean +/- SD
Range
Median
Platelet count
(x109/L)
150 - 400
< 150 - > 400
Maternal Blood CS(+) Blood CS(-) Total p-value
parameters (n=17) (n=83)
> 0.05 (NS)
(t-test)
0.006 (S)
(chi-square
test)
> 0.05 (NS)
(t-test)
> 0.05 (NS)
(Fisher-
test)
> 0.05 (NS)
(t-test)
> 0.05 (NS)
(Fisher-
test)
> 0.05 (NS)
(t-test)
> 0.05 (NS)
(Fisher-
test)
14.29 +/- 5.18
5.9 – 30.4
13.20
49 (59.0%)
34 (41.0%)
11.01 +/-4.76
7.87 – 25.84
10.19
1 (1.2%)
82 (98.8%)
2.18 +/- 1.13
2.33 – 7.44
1.97
75 (90.4%)
8 (9.6%)
292.5 +/- 101.5
69 – 677
277
1 (1.2%)
82 (98.8%)
15.47+/-3.61
9.3–23.19
15.30
16 (94.1%)
1(5.9%)
12.53+/- 3.89
6.70–21.24
12.71
1 (5.9%)
16 (94.1%)
1.85 +/- 0.9
0.88–3.84
1.71
16 (94.1%)
1 (5.9%)
335.7+/- 97.3
135 – 467
336
0 (0%)
17 (100%)
65
35
2
98
1
99
I/M
> 0.3 3 (17.6%) 7 (8.4%) 10 >0.05 (NS)
< 0.3 14 (82.4%) 76 (91.6%) 90 (Fisher test)
ANC
Mean +/- SD 7164 +/ 9788 +/ 0.05 (NS)
-7291 -6745 (t-test)
Range 1620-32500 582 -34850
Median 5044 8959
Platelet count
(x109/L)
Mean+/-SD 200.5+/ 243.4 +/
-108.8 -85.4
Range 54-375 55-456 0.05 (NS)
Median 203.50 243.00 (test)
Platelet count
(x109/L)
Mean+/-SD 6 (35.3%) 9 (10.8%) 15 0.02 (S)
< 150 6 (35.3%) 55-456 85 (Fisher
> 150 11 (64.7%) 74 (89.2%) test)
Toxic granules
(+) 6(35.3% 30 (36.1%) 36 >0.05 (NS)
(-) 11 (64.7%) 53 (63.9%) 64 (chi-square
test)
Toxic
granules
> 3 + < 3
Normal
0 (0%)
17 (100%)
0 (0%)
83 (100%)
Blood CS (+)
(n=17)
Blood CS (-)
(n=83)
Total p valueHematological
parameters
I/T
Mean +/- SD
Range
Median
0.108 +/
- 0.197
0 – 0.640
0
0.063 +/
-0.131
0 –0.784
2
0.05 (NS)
(t-test)
I/T
>0.16 4(23.5%) 13 (15.7%) 17 >0.05 (NS)
<0.16 10 (58.8%) 44 (53.0%) 83 (Fisher-
square test)
I/M
Mean +/- SD 0.216+/ 0.074 +/ 0.05 (NS)
-0.473 -0.159 (t-test)
Range 0-1.790 0 – 0.750
Median 0 0
0
100 NA
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
41
The factors which showed statistical significance
were preterm neonates, neonatal platelet count and ma-
ternal total leukocyte count. In table 5, the odds ratio of
having a positive blood culture result was 3 times (OR =
2.92) higher for preterm as compared to full term neo-
nates. The odds ratio for neonates with low platelet count
was 5 times (OR = 4.74) higher than the neonates whose
platelet count was within normal range. Lastly, for neo-
nates whose maternal total leukocyte count (TLC) was
>12, the odds ratio was 11 times higher for neonatal sep-
sis as compared to patients whose maternal TLC was
≤12. Among these factors, the maternal total leukocyte
count had the highest sensitivity of 94.1% and negative
predictive values of 97.1% (Table 6).
A scoring system was devised based on the
significant factors that were obtained in this study. There
is a significant difference in the median scores of the
patients with positive blood culture and negative blood
culture. Higher median scores were noted among the
neonates with positive blood culture (Table 7).
Since the median scores were significantly
different, each score was then computed for its individual
accuracy of determining sepsis. A score of 3 was both
highly sensitive and specific for neonatal sepsis. The
chance of getting a positive blood culture given all the 3
factors was 100%. While, if the factor were absent, the
chance of a negative blood culture was 91.3% (Table 8).
Table 8: Sensitivity, Specificity of Score Associated with
Neonatal Sepsis
Scores Sensitivity Specificity PPV NPV
1 100 35.0 11.4 100
2 100 50.0 30.0 100
3 100 91.3 60.0 100
DISCUSSION
A high index of suspicion is important in the
diagnosis and treatment of neonatal infection because it
is hampered by vague, nonspecific or nonexistent clinical
manifestation. Thus, it is difficult to establish a diagnosis
based on clinical picture alone. However, it is imperative
that treatment is instituted early because of the high
mortality associated with the neonatal infection.
In this study, there was 17% culture proven
neonates with sepsis which were predominantly preterm
(64.7%) and males (70.6%). This was possibly due to
impaired defense mechanisms and low immunoglobulin
G levels in males and low birth weight neonates.15 In
addition, newborns particularly the preterm, have less
effective phagocytosis and chemotactic activity.
Therefore, rapid invasion of offending organism occur
very fast. They also have relative immunoglobulin M
deficiency rendering them more vulnerable to gram
negative infections.16
Infections occurring at less than 72 hours of age
usually are caused by bacteria acquired in utero or during
delivery, whereas infection after that time most likely have
been acquired after birth.13 Thus, it is essential to know
the maternal illnesses which can predispose to neonatal
sepsis. These are prolonged rupture of membranes, foul
smelling amniotic fluid, maternal fever or other symptoms
suggestive of infection, unexplained fetal distress and
previous septic infant. It was noted that there were 76.5%
neonates with culture proven sepsis who had maternal
history of infection. However, the result showed no
significant difference which could be attributed to prior
antibiotic intake of the mothers during the time of illness.
The most common presenting symptom in the
early onset of sepsis is respiratory distress. It is
Scores Bld CS (+)(n=17) Bld CS (-)(n=83) Total
0 0 (0%) 21 (100%) 21
1 5 (11%) 39 (89%) 44
2 9 (30%) 21 (70%) 30
3 3 (60%) 2 (40%) 5
Table 7: Scoring
p <0.001 (S)
Table 6: Sensitivity, Specificity of Factors Associated with
Neonatal Sepsis
Factors Sensitivity Specificity PPV NPV
Preterm 64.7 60.2 25.0 89.3
Neonatal Platelet
count (<150) 35.3 89.2 40.0 87.1
Maternal
TLC (>12) 94.1 41.0 24.6 97.1
Correlation of neonatal and maternal factors with
neonatl sepsis
Factors OR 95% CI p value
Preterm 2.92 0.88 – 9.98 < 0.05
Neonatal
Platelet count 4.74 1.16 – 19.65 < 0.05
(<150)
Maternal TLC (>12) 11.1 1.42 – 87.73 < 0.01
Table 5: Factors Associated with Neonatal Sepsis
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
42
manifested most commonly on the first day of life, with
majority of cases at less than 12 hours.6 This was evident
in this study wherein there were 77 out of the 100
neonates (77%) presented with respiratory problems.
Other clinical signs of bacteremia include unexplained
low Apgar scores, poor perfusion, hypotension,
bradycardia, and unstable temperature.
Because of the low positivity of blood culture, its
inavailability in some peripheral health centers and the
time allotted for the result to be obtained, the need for
other tests in diagnosing neonatal septicemia is
warranted.1
The complete blood count with differential is
widely used, either singly or in conjunction with other
test or clinical findings, as a diagnostic tool for neonatal
sepsis. 17
The criteria of Manroe with 2 of 3 indices (total
PMN count, immature PMN count, and I/T ratio)
abnormal were the most reliable of the published criteria
evaluated and would have identified all infants with sepsis
and all infants with probable sepsis.12,17
In this study, there were more hematological
parameters of the neonates studied. These were total
leukocyte count, total neutrophil count, immature cells,
immature to total neutrophil cells (I/T) ratio, immature to
mature cells (I/M) ratio, platelet count and toxic
granulation. Nucleated red blood cells, lymphocytes and
absolute neutrophil count were also included because
there were no studies done yet in determining its
association with sepsis. Moreover, maternal infection
was noted to be one of the major risk factors in early
neonatal sepsis which could be documented by a complete
blood count. Thus, the maternal hematological
parameters, consisting of total leukocyte count, total
neutrophilic count, lymphocytes and platelet count, were
also used as indices in predicting neonatal sepsis.
In the neonatal hematological parameter, only the platelet
count was significant with p value of 0.02.
Thrombocytopenia was seen frequently in sepsis
proven group. This could result from increased platelet
destruction, sequestration secondary to infections, failure
in platelet production due to decreased number of
megarkaryocytes or damaging effects of endotoxin on
the platelets.11
Total leukocyte count, total neutrophilic count and
immature cells showed no significant association with
sepsis. In a study by Akenzua, it was stated that total
neutrophil count was of limited value for the diagnosis of
infection since elevation is often late and inconsistent. In
addition, newborn infants with proven bacterial infection
had normal neutrophil count but the bands increased
beyond the normal range.18 In another study, although
neutropenia in the newborn is most often secondary to
infection, there are many causes of neutropenia including
isoimmune neutropenia, congenital neutropenia, and
neutropenia due to inborn error of metabolism.19 Lastly,
neutrophilia in the absence of an increase in band may
occur in patients with no evidence of infection,
presumably the result stress or other non specific causes.
Therefore, neutrophilia itself is not a reliable or sensitive
test of infection.11
Neutrophil ratios were often abnormal during
neonatal sepsis. However, this was not evident in this
study possibly because of the variation in interpretation
of peripheral smears by different observers.
Toxic granules also showed no significant
difference in this study. The presence of toxic granules
represents the production of unusual neutrophils during
the stress leucopoiesis and infection. It is invariably
present during sepsis, a change never seen in healthy
newborn infants but are not always increased in infection. 11
The other parameters, namely: nucleated red blood cells,
lymphocytes and absolute neutrophilic count, were not
statistically significant.
The maternal hematological parameter which
showed significant correlation between sepsis was total
leukocyte count with a negative predictive value of
97.1%. The leukocyte count usually ranges from 5,000
to 12,000 x 109/L. During labor and the early puerperium,
it may be markedly elevated. The cause is not known
but probably represents the reappearance in the
circulation of leukocytes previously shunted out of the
active circulation. During pregnancy, there is neutrophilia
that consists of mature forms.13
In this study, there were 3 parameters, namely:
preterm neonates, neonatal platelet count of <150 x 109/
L, and maternal total leukocyte count of > 12,000 x 109/
L that were statistically significant. Based on these
factors, a scoring system was formulated which could
be used in determining the presence of sepsis. The
presence of one factor corresponded to a score of 1 which
indicated a positive predictive value of 11% and a score
of 2 indicated 30% positive predictive value. The
probability of getting a positive blood culture increases
with an increasing score. The highest score was 3 with
100% sensitivity and 91.3% specificity for predicting
neonatal sepsis.
It could be noted that in this study, there were
different parameters used for predicting sepsis compared
to other hematological scoring sytem.12 This was because
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
43
the said parameters were the ones found to be statistically
significant. The difference in the sensitivity of
hematological values to other studies could be due to wide
range of the subjects and some degree of observer
variability in reporting the peripheral smear.
CONCLUSION
Neonatal sepsis, especially in its early stages, may
be difficult to diagnose because of its nonspecific clinical
symptoms. Because the prognosis for sepsis largely
depends on early identification and treatment, these
neonates are subjected to extensive diagnostic evaluation
and empiric treatment.
The usefulness of a scoring system based on the
clinical manifestations of the neonate and mother
supported by their hematological parameters can provide
information in determining the probability of sepsis in
Pediatric Diagnosis and Treatment, 14th edition. Hay WW,
Hayward AR, Levin MJ, Sondheimer JM: editors. New
Jersey: Appleton and Lange, 1999.
7. Fowlie PW, Schimdt B. Diagnostic tests for bacterial
infection from birth to 90 days- a systematic review.
Archives of the Diseases in Child Fetal Neonatal Edition.
1998; 79: 92-98.
8. Kite P, Millar MR, Gorham P, Congdon P. Comparison of
five tests used in diagnosis of neonatal bacteremia.
Archives of Disease in Childhood. 1988, 63: 639-643.
9. Da Silva O, Ahisson A, Kenyon C. Accuracy of leukocyte
indices and C-reactive protein for the diagnosis of neonatal
sepsis: a critical review. Pediatric Infectious Diseases
Journal. 1995; 14: 362-366.
10. Philip AG. Decreased use of antibiotics using a neonatal
sepsis screening technique. The Journal of Pediatrics.
1981, 98(5): 795-799.
11. Zipursky A, Palko RT, Milner MIS, and Akenzua MB.
The Hematology of Bacterial Infections in Premature
Infants. Pediatrics. 1976, 57(6): 839-853.
12. Manroe BL, Weinberg AG, Rosenfeld CR, Browne R. The
neonatal blood count in health and disease. The Journal
of Pediatrics. 1979, 95(1): 89-98.
13. Cunningham FG, Grant N. Maternal Adaptations to
Pregnancy. Williams Obstetrics, 21st edition. New York:
Mc Graw Hill Medical Publishing Division. 2001.
14. Nelson DA, Morris MW. Basic Examination of Blood.
Clinical Diagnosis and Management, 18th edition. Henry
JB: editors Philadelphia: W.B. Saunders Company. 1991.
15. Anwer S, Mustafa S. Rapid Identification of Neonatal
Sepsis. Journal of Pakistan Medical Association. 2000;
50(3).
16. Limjoco-Sarte L, Gonzales RM. Septicemia in Filipino
Infants and Children: Analysis of 724 Bacteriologicallly
Proven Cases in an Urban Hospital. Philippine Journal of
Pediatrics. 1981; 30(6): 190-198.
17. Rodwell RL, Tudehope DA. Early diagnosis of neonatal
sepsis using a hematologic scoring sytem. The Journal
of Pediatrics. 1988; 112(5): 761-767.
18. Akenzua GI, Hui YT, Milner R, Zipursky A. Neutrophil
and Band counts in the Diagnosis of Neonatal Infections.
Pediatrics. 1974; 54(1): 38-42.
19. Engle WD, Rosenfeld CR. Neutropenia in high risk
neonates. The Journal of Pediatrics. 1984; 105(6): 982-
986.
neonates. Since this scoring system is highly sensitive
and specific for neonatal sepsis, it could also serve as
the basis for a more rational approach to antibiotic use.
A significant decrease in the use of antibiotics may prevent
the emergence of resistant organisms, decrease the
chance of side effects and minimize cost.
RECOMMENDATIONS
It is recommended that more subjects will be
included in future studies wherein there will be a control
group composed of healthy, asymptomatic neonates and
a test group composed of neonates with probable sepsis
or proven sepsis. The group can also be divided into full
term or preterm neonates to determine differences in
their hematological characteristics. It also suggested that
one interpreter of the laboratory results be assigned to
decrease the observer variability.
REFERENCES
1. Manucha V, Rusia U, Sikka M, Faridi MMA, Madan N.
Utility of hematological parameters and C-reactive protein
in the detection of neonatal sepsis. Journal of Pediatrics
and Child Health. 2002; 38: 459-464.
2. Baley J, Goldfarb J, Neonatal Infections. Care of the high
risk neonate, 4th edition. MH. MH Klaus and AA Fanaroff:
editors. Philadelphia: WB Saunders Company, 1993.
3. David E., Tan CT, Que SP., Cruz NGS. Determining the
criteris for early discontinuation of antibiotic therapy in
suspected neonatal sepsis. Philippine Journal of
Microbial Infectious Diseases. 1980; 9 (2): 145-155.
4. Annual Reports of the Section of Neonatology of the
Department of Pediatrics UP-PGH Medical Center. 2000-2002.
5. Bellig L, Ohning B. Neonatal Sepsis. EMedicine Journal
Pediatrics/Neonatology. 2003; 4(1). 6. Thilo EH,
Rosenberg AA. Infections of the Newborn. Current
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
44
SERUM CONCENTRATION OF PYRAZINAMIDE SUSPENSION IN
CHILDREN WITH TUBERCULOSIS: A THERAPEUTIC DRUG
MONITORING
Daisy O. Sanchez, MD*, Cecilia C. Maramba, MD*
ABSTRACT
Rationale. Therapeutic drug monitoring (TDM) is a
process of adjusting drug dosages on the basis of serum
drug concentrations for the purpose of optimizing drug
therapy. This study introduces the use of TDM in the
management of mycobacterial infections. Pyrazinamide
(PZA) has been marketed as tablet in other countries.
It is only in the Philippines wherein pyrazinamide is
available both in tablet and suspension forms. No study
has been done on pyrazinamide suspension use for the
treatment of tuberculosis to this date.
Objectives. To examine the serum concentration of
pyrazinamide suspension in children with tuberculosis
Design. Descriptive study
Methods. Thirty pediatric patients who were taking
pyrazinamide suspension for at least 1 week as part of
chemotherapy for tuberculosis were included in this study.
Blood was taken prior to the dose then 2, 4, 8 hours after
administration of PZA suspensiongfor thefirst 4 patients.
Specimens were submitted to the Pharmacology
Department Laboratory of the University of the
Philippines – College of Medicine and were analyzed
using High Performance Liquid Chromatography
technique. The samples from the first 4 patients were
used to determine the time when the drug reaches its
maximum concentration (Tmax). For subsequent
patients, 2 determinations were taken at the time when
the drug reaches its maximum concentration and trough
level.
Results. At at a Tmax of 2 hours, the mean serum
concentration of PZA suspension is at 34.6+ 11.86 ug/
ml. The mean serum trough level is 4.55+ 4.63 ug/ml.
There were no significant differences in serum
concentration of PZA suspension among 3 brand names
of PZA (p-value: 0.506).
Keywords. Pyrazinamide, antituberculosis, therapeutic drug monitoring, pharmacokinetics,
tuberculosis
*Department of Pediatrics, UP-PGH
Conclusion. Mean serum concentration of PZA
suspension falls within the established therapeutic range
for pyrazinamide. But 2 subjects failed to reach the
therapeutic levels. No subject reached toxic levels
INTRODUCTION
In the past 4 years, tuberculosis ranks 6th in the
leading causes of morbidity in the Philippines. It is also
the 5th leading cause of mortality from 1989-1993.
Children under 15 years of age represent 1.3 million cases
per year and 450,000 deaths per year. Despite the
widespread use of BCG vaccine and the availability of
effective drugs, TB remains a major health problem. The
eradication of TB has proven to be an elusive goal for
clinicians and policy-makers. Failure of TB control is
not a new phenomenon in our country. Several factors,
including irrational antibiotic use, collapse of public health
infrastructures, the HIV epidemic, war, famine, increasing
inequality and poverty, and prohibitive cost of medicines,
have all contributed to the increasing incidence of TB all
over the world. In certain situations, drugs provide
suboptimal serum concentrations and these are associated
with worse treatment outcomes. Recurrence occurs in
2.4-5.5% of cases even when the patient receives directly
observed treatment. It is important to maintain high
standards of quality assurance, as low quality drugs often
penetrate emerging markets, resulting in low cure rates
for patients and increased resistance.
Definition of Terms
1. Therapeutic drug monitoring (TDM) – is the process
of using serum drug concentrations to optimize drug
therapy. TDM is useful when serum concentrations
show a better correlation with the therapeutic effects
or the incidence of adverse effects than does the
size alone. TDM requires the accurate timing of
doses and blood collection and the avoidance of assay
interferences.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
45
2. Tmax – time when drug reaches maximum/peak
serum concentration
3. Trough – serum concentration of the drug at time 0
4. Cmax - peak serum concentration
Review of literature
Pyrazinamide is one of the most frequently
administered drugs for the treatment of TB. It was
synthesized in the 1940s by Hall and Spoerri and formerly
used only as salvage therapy. Convincing results of
experimental studies have elevated it to a central role in
tuberculosis chemotherapy as an essential addition to
isoniazid and rifampicin which make it possible to shorten
the treatment to 6 months. The major contribution of
pyrazinamide is to increase the sterilizing power of an
antituberculosis regimen, measured by prevention of
relapse after initially successful therapy. Pyrazinamide
is administered at 15-30 mg per kg with a maximum daily
dose of 2 grams. When administered twice weekly, the
dose of pyrazinamide is 50-70 mg/kg/dose with a
maximum dose of 4 grams. A randomized, four-period,
crossover study, a single oral dose of 30 mg/kg of
pyrazinamide tablet in a fasting state results in a mean
peak serum concentration of 53.4 ug/ml at 1.43 hours. 11
The mean serum concentration of pyrazinamide tablet
was slightly increased by antacid and modestly decreased
by food (p=0.0138). The optimum sampling time was 1
hour after the dose, which is the closest to Cmax.
Similarly, a prospective study of C. Lacroix regarding
pyrazinamide kinetics in 9 healthy subjects showed a rapid
absorption in the fasting state (tmax </= 1 hour).12 In
Peloquin’s two-way, randomized, crossover study of
isoniazid, rifampicin, and pyrazinamide in 1997, the
calculated steady-state range for pyrazinamide (dose of
20mg/kg) was 27.03 to 53.12 ug/ml.16
Current knowledge on PZA came from results
described mostly in healthy volunteers. However, the
pharmacokinetics of pyrazinamide in children with
tuberculosis differ from that found in healthy volunteers.
In the prospective, multiple-dose population
pharmacokinetic study of Min Zhu has shown these
differences.17
Pyrazinamide has been marketed as tablet (500
mg) in other countries. In the Philippines, it is available
in tablet and suspension form. The suspension form
(250mg/5 ml) is very unstable. It crystallizes when it is
allowed to stand in room temperature and especially when
it is refrigerated. Much of the studies published to this
date used the tablet form of the drug. Bioavailability
studies on PZA were based of the tablet preparation.
Locally, according to the Bureau of Food and Drug
Administrative Order 67 series of 1989, all drug
manufacturers, traders and distributors are required to
submit bioavailability tests on the products sought to be
registered. At present, there are 5 existing brands of
pyrazinamide in the market. The purpose of this study is
to determine the serum level of pyrazinamide suspension
in order to optimize drug therapy and verify that they
achieve the therapeutic level. The National Jewish Center
for Immunology and Respiratory Medicine proposed a
therapeutic range for PZA at 2-hour which is at 20-60
ug/ml. Generally speaking, drug concentration at the site
of drug action are difficult to determine and typically not
obtainable. Therapeutic drug monitoring assumes that
there is a better correlation between serum concentration
and drug effects than between the dose prescribed and
the drug effects.
Peloquin of the Infectious Disease
Pharmacokinetics Laboratory in Denver suggested an
approach to the use of antibiotic serum concentration.15
It is clear that some drug concentration is required to
achieve a therapeutic effect. The target ranges should
be designed with a margin of safety with respect to their
efficacy. Once a decision is made to use a given
antibiotic, a goal should be set for the desired serum
concentration. Monitoring antimycobacterial drug levels
allows clinician to identify those patients who are not
absorbing, metabolizing or eliminating their drugs normally.
OBJECTIVES
General objective
To examine the serum concentration of PZA suspension
in children with tuberculosis
Specific objectives:
1. To present the demographic characteristics of
subjects in terms of the following: age, sex, duration
of therapy, organ of affectation
2. To compute the Tmax in 4 patients in order to
determine the correct timing of peak level in
subsequent patients
3. To compare the peak and trough of PZA suspension
in children with tuberculosis with the established
therapeutic range and duration of therapy
4. To compare the peak and trough levels of 3 most
commonly used brands of PZA suspension with each
other
METHODOLOGY
Study design. Descriptive study
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
46
Study population
Thirty pediatric patients who were taking PZA
suspension for at least 1 week as part of chemotherapy
for tuberculosis were included in this study. The subjects
were recruited from the TB registry of the INTROP
section of the Department of Pediatrics and from admitted
patients at Ward 11.
Inclusion criteria:
1. Thirty pediatrics patients diagnosed to have
tuberculosis of any organ and are taking PZA
suspension as part of antituberculosis therapy for at
least 1 week
2. normal CBC, serum creatinine, AST, ALT, uric acid
3. informed consent from parent or guardian
Exclusion criteria:
1. poor compliance with the medications
2. abnormal CBC, serum creatinine, AST, ALT, uric
acid
3. failure to give consent.
4. liver or renal insufficiency
5. adverse reactions to anti-TB medications
DATA COLLECTION
Last intake of pyrazinamide suspension of the
thirty patients is 24 hours prior to the procedure. The
investigator observed the patient take the pyrazinamide
suspension and recorded the exact date and time. Three
to five ml of blood was drawn from the subjects, who
were at least on their 2nd week of pyrazinamide
suspension therapy, via direct venipuncture. An indwelling
IV catheter attached to a heplock was inserted in order
to minimize punctures. Blood was taken prior to the dose
then 2, 4, 8 hours after administration of pyrazinamide
suspension. The specimens were placed in glass test
tubes labeled with the patient’s name, date and time of
collection, and the drug to be assayed. It was then
allowed to clot and then stored in an icebox at –20
degrees centigrade and was submitted to the
Pharmacology Laboratory of the UP-College of
Medicine. Blood specimens were centrifuged for 5
minutes. The plasma was analyzed using High
Performance Liquid Chromatography technique. This
method can measure the amount of pyrazinamide in the
blood to as low as 0.1 ug/ml. Samples for the first four
patients were used to determine Tmax. For subsequent
patients, 2 determinations were taken at Tmax and trough.
Outcome measured.
Serum concentration of pyrazinamide suspension in
children with tuberculosis
STATISTICAL ANALYSIS
Frequency of distribution included measures of
central tendency for age, duration of treatment, serum
concentrations at hour 0 and 2. Differences among
subjects and groups were determined by analysis of
variance (ANOVA) model. Paired samples were
compared. Sample size calculations were based on an
estimated mean difference of 20 and a standard deviation
of the difference of 15. The test of equality of means
was carried out at 0.017 level of significance (overall
level of significance for pair wise comparisons of
pyrazinamide brand names is 0.05). A sample size of 9
pairs per brand name gives a probability of 0.815 of
rejecting the null hypothesis of equal means if the
alternative holds.
RESULTS
The characteristics of subjects who participated
in the study are described in Table 1 and 2. Thirty
subjects (16 males; 14 females) were included. The mean
age is 9.46 +/-4.65 (range: 1-18 years). The median
duration of therapy is 9 days (range: 7-35 days). The
subjects received 20-25 mg/kg/day of pyrazinamide
suspension in keeping with the standard clinical practice
of the institution. Twenty-six subjects have concomittant
intake (intravenously and orally) of other drugs on the
day of blood sampling such as oxacillin, cefepime,
penicillin G, cefuroxime, amikacin, furosemide,
acetazolamide, dexamethasone, prednisone,
phenobarbital, propranolol, lanoxin, kalium durule, vit B
complex, dibencozide and multivitamins. All of these
drugs were assayed to check for interference in
pyrazinamide level. Only acetazolamide was found to
interfere with the assay of pyrazinamide. All patients
received pyrazinamide in combination with other
antituberculosis drugs including isoniazid, rifampicin,
ethambutol and streptomycin. Among the study
population, 40% has pulmonary tuberculosis. This was
followed by tuberculous meningitis and Potts disease at
23% and 13%, respectively.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
47
Table 1. Mean age and duration of therapy of patient population
Variable Mean Median Standard Range
Deviation
Age (years) 9.4667 10.5 4.6589 1.00-18.00
Duration of 12.1 9.0 8.1213 7.00-35.00
therapy
(days)
Table 2. Patient Population Demographics
Variable N Percentage
Male 16 53.3%
Female 14 46.6%
Organ of affectation
Lungs 12 40%
CNS 7 23.3%
Bone 4 13.3%
Disseminated 3 10%
Heart 2 6.6%
Endobronchial 1 3.3%
Kidney 1 3.3%
Table 3 shows that Tmax was determined using
the first four subjects. Based on the computation, Tmax
of 2 hours was used for the subsequent subjects (mean
of 2.98+/-1.1).
Table 3. Computed Tmax in first 4 patients
Patient Hour 0 Hour 2 Hour 4 Hour 8 Tmax
#1 4.24 20.25 8.43 4.55 2.28
#2 0.26 7.14 10.55 8.26 1.98
#3 12.17 24.06 23.09 15.85 4.44
#4 9.23 28.62 20.37 13.32 3.22
Table 4. Suspention trough and Tmax serum concentration
of pyrazinamide
Patient no. Duration Serum concentration (ug/ml)
of therapy 0 hour 2 4 8
(Through) (Tmax)
3 16 days 9.23 28.62 20.37 13.32
13 7 4.76 25.98
14 9 0** 58.64
17 7 2.94 45.80
22 10 0** 44.48
23 7 0** 42.02
24 10 0** 41.94
25 8 0** 21.48
28 7 8.02 38.62
29 10 13.97 52.58
1 30 days 0.26* 7.14 10.55 8.26
2 14 4.24 20.25 8.43 4.55
4 9 12.17 24.06 23.09 15.85
8 11 0.88* 34.36
9 9 4.19 27.05
10 30 14.49 43.53
18 9 0.85* 45.79
19 7 2.64 47.42
21 8 7.12 37.43
20 8 0.93* 24.30
PZA-CIBA by Novartis
6 35 days 0** 40.03
7 7 0.52 32.20
11 7 0** 16.55*
12 30 4.9 18.38*
15 7 3.28 22.35
16 7 3.83 40.45
26 9 9.28 38.81
27 9 10.10 34.62
28 8 11.62 44.95
5 18 6.49 37.32
ZINAPLEX by Pediatrica
*failed to reach therapeutic level (20-60ug/ml)
**trough levels <1ug/ml
ZCURE by Natrapharm
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
48
Table 5. Mean serum level of pyrazinamide suspension in
population grouped into brand names
Zinaplex n Mean Standard
ug/ml Deviation
Hour 0 0 3.8920 4.9944
Hour 2 0 40.0160 11.7294
Duration of therapy 0 9.1 2.7669
Zcure
Hour 0 0 4.7770 5.0055
Hour 2 0 31.1330 12.87226
Duration of therapy 0 13.5 8.9100
PZA-Ciba
Hour 0 10 5.0020 4.2789
Hour 2 10 32.6660 9.9970
Duration of therapy 10 13.7 10.5098
Repeated Measures Analysis of Variance was
used to analyze the data. Hour (0 hour, 2 hour) was
considered as within-subject effect; group (Zinaplex,
Zcure, PZA-Ciba) as between–subject effect and
duration as a covariate. A significant difference between
0 hour and 2 hours was noted (p-value: 0.000). However
there were no significant differences in serum
concentration of pyrazinamide suspension among three
brand names of pyrazinamide (p-value: 0.506).
Within-subject effects
Effect/Source p-value
Hour 0.000*
Hour•duration 0.378
Hour•group 0.242
Between-subject effects
Duration 0.550
Group 0.506
Table 6.
DISCUSSION
Subjects were grouped according to brand names
of pyrazinamide suspension. Ten subjects were included
for each brand name. The serum concentration of
pyrazinamide suspension and duration of therapy is
reported in Table 4. In the zinaplex group, half of the
subjects had a trough level of 0 ug/ml. On two occasions,
PZA-Ciba also had the same serum trough level, while 4
out of 10 subjects belonging to the Zcure group reported
a serum trough concentration of <1 ug/ml. Generally
speaking, a low to absent serum drug level prior to the
next dose indicates that the drug did not sustain the desired
serum concentration necessary for it to exert bactericidal
activity most probably because of its short half-life.
Pharmacodynamic parameters of efficacy can
either be concentration-dependent or time-dependent. In
concentration-dependent killing agents, the higher the drug
concentration, the greater the extent of bactericidal
activity (Cmax:MIC ratio). On the other hand, time-
dependent agents kill bacteria only when the concentration
at the site is higher than the minimum inhibitory
concentration or MIC (time of the concentration above
MIC). Thus, the extent of killing is dependent on the
time of exposure. Drug concentration at the site of drug
action or tissue sample is difficult to determine and that
is the why serum drug concentration represent the next
best alternative.
Previous studies have shown that the mean
serum concentration of pyrazinamide tablet in healthy
volunteers is at 5-7 ug/ml at the 23rd hour following
administration of 1.5 gram dose of pyrazinamide tablet
once daily (C. Peloquin, 1998). In addition, Min Zhu et
al in 2002 determined the population parameter of
pyrazinamide tablet in children and adults with
tuberculosis. The absorption of pyrazinamide tablet in
children was 32% slower than in adults. The median
Cmax was 21.1 ug/ml, which was almost 50% lower than
the adult value of 41.1 ug/ml. The volume of distribution
was 71% lower in children and the half-life was 43%
shorter for children. According to this data, children
appeared to absorb pyrazinamide tablet more slowly but
eliminated it more quickly than adults. Another study
done in India showed that slow absorption was also found
in 10 patients with tuberculosis, aged 6-12 years old after
a single oral dose of pyrazinamide tablet. It is not clear if
these differences were due to chance, formulation, brand
name, race, concurrent TB infection, or a combination of
these factors. Whether or not the trough level 0-1 ug/ml
is clinically significant, a full-scale pharmacokinetics and
pharmacodynamics on pyrazinamide suspension on
Filipino children needs to be done.
Table 3 presents the computed Tmax from the
first 4 subjects. There is a perceptible wide variability of
time for drug to reach maximum concentration. Again,
this could be due to several circumstances as mentioned
earlier. Additional research is needed to explore how
the derived parameters can be used to optimize
antituberculous drug therapy.
As regards to the Cmax, 93% of subjects were
able to achieve a serum concentration level which was
within the expected therapeutic range (mean 34.6+/-11.86
*Significant at the 0.05 level
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
49
ug/ml). Two out of the 30 patient population reported a
serum concentration of <20 ug/ml, both belonging to the
PZA-Ciba group. Zinaplex group had the highest mean
serum concentration of 40 +/- 11.72 ug/ml.
In terms of duration of therapy, the present study
shows that there is no significant difference in serum
concentration of pyrazinamide suspension as presented
in Table 6.
The therapeutic range of pyrazinamide has been
reported to be at 20-60 ug/ml. In our study, at Tmax of 2
hours, the mean serum concentration of pyrazinamide
suspension is 34.6+/-11.86 ug/ml. Thus, pyrazinamide
suspension provides sufficient serum drug level at the
second hour to elicit the desired therapeutic response.
The result of the present study justifies that the suspension
form achieves the desired serum concentration at Tmax
but this study poses many questions such as, what would
be the role of race, formulation, brand name in the serum
level of pyrazinamide; what is a better measure or
parameter of efficacy, Cmax:MIC or time>MIC? In this
context, the results of the present study might constitute
a useful baseline or reference for future bioavailability
studies on pyrazinamide suspension.
CONCLUSION
Serum concentration of pyrazinamide suspension
in 30 pediatrics patients aged 1-18 years old, who are on
this formulation for 7-35 days were analyzed in this study.
Forty percent of the study population had pulmonary
tuberculosis. The Tmax of 2 hours was used based on
the computed value from the first four subjects. The
mean serum trough level of pyrazinamide suspension is
4.557+/-4.63 ug/ml. Seven out of 30 subjects accounted
for the lowest serum trough level of 0 ug/ml, 5 of whom
belong to the zinaplex group. The mean serum
concentration of pyrazinamide suspension at Tmax is
within the therapeutic level for pyrazinamide and is
notably being achieved also by the suspension form.
However, there were two patients who failed to reach
therapeutic levels and both of them belong to the PZA-
Ciba group.
Zinaplex has the highest mean serum
concentration at 2 hour followed by PZA-Ciba and Zcure.
However, it also has the lowest mean serum trough level
and has the most numbered of subjects (5) that recorded
the lowest value of 0 ug/ml, but these differences did not
reach statistical significance.
This study also demonstrates that duration of
therapy does not significantly affect serum drug
concentration. Neither the duration of therapy nor the
brand name of the formulation statistically significantly
affect the serum drug level of pyrazinamide suspension
of Tmax.
RECOMMENDATION
A full-blown bioavailability study is needed to
confirm the kinetics of pyrazinamide suspension
particularly on Filipino children. The role of race,
formulation, burden of TB disease on serum drug level is
yet to be determined.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
50
REFERENCES
1. Philippine Health Statistics 2001, DOH. www.doh.gov.ph
2. The Philippine Pediatric Society. Tuberculosis in Infancy
and Childhood, 2003, p.3
3. Rajesh Gupta, Jim Y. Kim, et al. Responding to Market
Failures in Tuberculosis Control. Science. August
2001;(293):1049.
4. Weis SE, Slocum PC, Plais FX, et al. The Effect of Daily
Observed Therapy on the Rates of Drug Resistance and
Relapse in Tuberculosis, New England Journal of Medicine,
1994;(330):1179-84.
5. Theodore Brody et al, Emma D. Underdown. Human
Pharmacology. Second Edition. St. Louis, Missouri:
Mosby-Year Book, Inc.,1994. p.6.
6. Hall SA, Spoerri PE. Synthesis in the Pyrazine Series II.
Preparation and Properties of Amino Pyrazine, Journal of
American Chemical Society 1940;(62):664-5.
7. Clini V, Grasi C. The Action of New Antituberculous Drugs
on Intracellular Tubercle Bacilli. Antibiotic Chemotherapy,
1970;(16):20-6.
8. Mitchison DA. Basic Mechanism of Chemotherapy. Chest
1979;(76):771-81.
9. Lawrence J. Geiter. Contribution of Pyrazinamide to
Antituberculosis Chemotherapy. The Journal of Infectious
Diseases, 1991;(164):610.
10. Pyrazinamide Drugdex Drug Evaluations. Micromedex
Healthcare series. 2003. (116).
11. Charles A. Peloquin, et al. Pharmacokinetics of
Pyrazinamide under Fasting Conditions, with Food, and
with Antacids. Pharmacotherapy 1998;(18):1205-1211.
12. C. Lacroix, T. Phan Hoang, J. Nouveau, C. Guyonnaud, G.
Laine, O. Lafont. Pharmacokinetics of Pyrazinamide and
its Metabolites in Healthy Subjects. European Journal of
Clinical Pharmacology 1989,(36):395-400.
13. Q.L. Kintanar. Bureau Circular 1997. Bureau of Food and
Drugs, Department of Health. www.doh.gov.ph
14. Charles A. Peloquin. Therapeutic Drug Monitoring:
Principles and Application in Mycobacterial Infections.
Drug Therapy, July 1992:31-54.
15. Charles A. Peloquin. Tuberculosis Drug Serum Levels.
Clinical Infectious Disease 2001, (33):584-585.
16. Charles A. Peloquin, et al. Population Pharmacokinetic
Modeling of Isoniazid, Rifampin, and Pyrazinamide.
Antimicrobial Agents and Chemotherapy, December
1997;41(12):2670-2679.
17. Min Zhu et al. Population Pharmacokinetic Modeling of
Pyrazinamide in Children and Adults with Tuberculosis.
Pharmacotherapy 2002;22(6):686-695.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
51
THE ANALYSIS OF CLINICAL AND SOCIAL PROFILE OF
CONGENITAL RUBELLA SYNDROME SEEN AMONG UP-PGH
PATIENTS FROM THE YEARS 1993 - 2002 (A 10 YEAR PREVALENCE
REVIEW)
Coralee Lianko Agnas , MD*
Objectives :
1. To describe the socio-demographic profile of patients
with Congenital Rubella Syndrome (CRS) s e e n
in UP-PGH for the last 10 years.
2. To compare the clinical profile of patients with
Congenital Rubella Syndrome (CRS) seen in UP-PGH
to that of foreign published data.
3. To describe the frequency distribution of occurrence
of clinical findings of patients with Congenital
Rubella Syndrome (CRS).
4. To describe maternal factors such as antenatal
maternal viral infection and MMR vaccination
history among siblings with Congenital Rubella
Syndrome (CRS).
Design: A descriptive study design.
Methodology: There were total of 58 cases of
Congenital Rubella Syndrome (CRS) classified and
retrieved in the Biostatistics and Epidemiology Section
of the UP-PGH, OPD Division. The 58 case numbers
of identified diagnosis in the charts were tracked down
and analyzed if they fit in to the inclusion criteria of
the study.
Results: In the wide spectrum of clinical manifestations
of Congenital Rubella Syndrome (CRS), the youngest
patient brought in for consult is within the1st month
with the mean age of 8 months, with slight male
preponderance.
Baseline anthropometrics showed 30% of
infants classified as severely wasted although 20% of
study population had normal weight . Majority (69%)
had no stunting. Congenital catarract ranked the most
common clinical presentation of patients with CRS,
accounting for 24 cases (49%) followed by patent
ductus arteriosus of 15 cases (31%). Most clinical
findings that co-exists were: ocular, cranial, growth
retardation, cardiac and sensorineural findngs. There is
a high occurrence of antenatal maternal viral infection
in this review – accounting to about 37 cases (71%)
of study population.
Conclusions: This research study documented the
continued existence of CRS and majority of the
patients were delivered primarily by an unvaccinated
women. The disease continued to be dynamically
prevalent despite the illness is a vaccine preventable
occurrence.
INTRODUCTION:
Congenital Rubella Syndrome is a serious multi-
systemic disease with a wide spectrum of clinical
expression and sequelae. This is comprised of
constellation of birth defects which may occur when
the mother acquires the infection during pregnancy ,
especially on her 1st trimester . In Congenital Rubella
Syndrome (CRS) virtually every organ system maybe
involved , singly , multiply , transiently or progressively
and permanently.
The risk of congenital defects of the disease
is greatest with the primary maternal infection during
the first trimester which may lead to abortion , stillbirth,,
or a child with anatomic defects. The risk is about
90% if the infection occurs before 11th week of
pregnancy, and this decreases by 10-20% at the end
of the 1st trimester, with an overall risk of 70% and
maternal infection after the 16th week of pregnancy
poses low risk of congenital defects although fetal
infection may still occur.
From the records of Philippine General Hospital
from the years 1993-2002, there are still 56 cases of
congenital rubella syndrome (CRS) despite the
introduction of MMR vaccination by our DOH since
1983. What are the factors that may have contributed
to the Congenital Rubella Syndrome profile that is
still prevalent in our community at present times?
In the study of Elias, Erik on Rubella (August
2002) states that after licensing the live attenuated
vaccine in the US in 1969, a dropped has occurred in
the number of cases of Rubella and Congenital Rubella
Syndrome (CRS). As shown in the statistics of 1969
(from the total of 57,686 cases of rubella and 62
cases of CRS ). From 1992-1998, the cases of rubellaKeywords : congenital , epidemiology , socio-demographic , immunization , vaccination .
*Department of PEdiatrics, UP-PGH
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
52
recorded annually decreased by the range of 128-
364 cases and CRS decreased by 2-9 cases per
year, although 15 cases of isolated outbreaks were
still reported. The median cases per outbreak was 21
and most recent cases occurred in New York (1997-
1998), Kansas(1998) and Nebraska (1999).
Likewise , CRS continuous to occur despite 25
years of immunization program in Canada and still
requires further measures to eliminate this preventable
disease. Between the years 1986-1995, a mean of 3
cases of congenital rubella syndrome (CRS) per year
were reported to the Notifiable Diseases Reporting
System in Canada. (Infectious Disease and Immunization
Committee; Canadian Pediatric Society).
In Kuala Lumpur, Malaysiathe incidence of CRS
from 1993 - 1998 varies from 19-93 per 100,000
deliveries.13
In Canada, the epidemiologic pattern of rubella
underscored the importance of the collection and
analysis of information on demographic, vaccination
history, source of infection in relation to transmission
and outbreaks.2
In a local study done, one of the variables
considered was the educational attainment of parents
or guardians on how they perceive the disease entity
itself and the benefits of vaccination.17
In a GENEVA conference , held November 28,
2001, the UNICEF (United Nations Children’s Fund)
hailed the massive immunization of individual between
15 months to 25 years old , to significantly reduce
the cause of birth defects in children which is now
estimated to be about 110, 000 cases of CRS per
year worldwide.
The intent of this paper is to have a wide
panoramic view of the demographic and psychosocial
profile of Congenital Rubella Syndrome among Filipino
children to guide us in early diagnosis and subsequently
pave way for the policy makers to formulate a timely
medical intervention and prevention programs to abort
further spread of virus and hinder specific disabilities
for progressive deterioration.
GENERAL OBJECTIVE
To describe the clinical and social profile of
patients with Congenital Rubella Syndrome (CRS)
seen in UP-PGH from the years 1993-2002.
SPECIFIC OBJECTIVES:
• To describe the socio-demographic profile
of patients with Congenital Rubella Syndrome (CRS)
seen in UP-PGH for the last 10 years .
• To compare the clinical profile of patients in terms
of type and number of clinical signs of Congenital
Rubella Syndrome (CRS) seen in UP-PGH versus
foreign published data.
• To describe the frequency distribution of
occurrence of clinical findings of patients with
CRS seen in UP-PGH.
• To describe maternal factors such as frequency of
antenatal maternal viral infection and MMR
vaccination history among siblings with Congenital
Rubella Syndrome.
METHODOLOGY
Study design : Descriptive study design
Sample size/Population:
Population considered are all charity pediatric
patients seen and/or admitted at Pedia ER, OPD,
Wards 9 and 11, PICU, and NICU of UP-Philippine
General Hospital from years 1993-2002. Excluded in
the study were those admitted at paywards and other
departments who were not referred nor co-managed
by pediatric residents.
Inclusion criteria:
1. patients aged 1 month to 18 years old.
2. patients consulted and / or admitted at UP-PGH:
PER, OPD, Wards 9 and 11 and special areas like
PICU and NICU.
3. patients admitted in other wards but seen or co-
managed by pediatric residents.
4. patients whose charts met the clinical case definition
and classification of Congenital Rubella Syndrome
(CRS) in APPENDIX E and with pertinent data
needed in the study.
Exclusion criteria:
1. patients seen at paywards. or other departments
not referred or co-managed by pediatric residents.
2. patients more than 18 years old at the time of
consult.
Actual Methodology:
There was total of 58 cases of Congenital
Rubella Syndrome (CRS) classified and retrieved in
the Biostatistics and Epidemiology Section of the
Record Division, 5th floor OPD, UP-PGH which were
tracked down and analyzed if they fit in to the
inclusion criteria.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
53
However, only a total of 52 charts were
reviewed. The rest were missing due to the typhoon
that has embarked Metro Manila in the past years
destroying considerable amount of old documents in
the years 1993-1994 in the attic of the record section.
STATISTICAL ANALYSIS
In this research study , the test of 2 proportion
was used and the frequency distribution for descriptive
data was likewise adapted .
RESULTS and ANALYSIS
Eight-year incidence of CRS cases and
sociodemographic profile
A total of 52 patient charts from 1995 to 2002
were reviewed. The frequency distribution of congenital
rubella syndrome cases from years 1993-2002 is
summarized as figure 1-1. A gradual increasing trend
was seen in 1998 up to the last two previous years.
YEAR
YEAR
20022001200019991998199719961995
Fre
quency
16
14
12
10
8
6
4
2
0
8
14
7
4
2
3
77
Figure 1. Frequency distribution of confirmed and probable
CRS cases from 1993-2002, UP-PGH Data
The socio-demographic characteristics are
summarized in Table 1. The mean age was 8 months.
The youngest patient on the time of consultation was 1
month old while the oldest was 60 months old. There
was a slight male preponderance 29 (56%) as opposed
to females. Mean weight on consult was 5 kg while the
mean height was 65 cm. Thirty percent (30%) of these
children are classified as severely wasted while
20 % had normal weights. Majority (69%) had no
stunting.
Much of our registered patients come from the
greater Manila and Southern Tagalog area (65% & 30%
respectively), There were 9 patients identified living in
Paranaque City, 6 patients in Las Pinas City, while
Mandaluyong, Taguig and Quezon City had 4 each, and
Pasay, Caloocan, Makati and Metro Manila had 1 patient
each. Majority of the parents of congenital rubllea
syndrome (CRS) finished high school (26%), as compared
to those who finished collegiate level (19%), and no
formal education (13%).
Antenatal maternal infection and vaccination
The number of women who had a history of
rubella (confirmed by history alone) is summarized in
Figure 2. There is a high occurrence (37 out of 52) of
antenatal maternal rubella in this review, while 12 mother’s
charts did not indicate the presence or absence of rubella
exposure. Two siblings underwent MMR vaccination and
this was seen in the years 2001 and 2002 respectively.
YEAR
20022001200019991998199719961995
Count
10
8
6
4
2
0
RUBELLA
not inidcated
yes
5
8
5
4
1
3
5
6
3
6
2
1
2
1
Figure-2. Antenatal history of maternal rubella in CRS cases
from 1995-2002 , UP-PGH Data
Table-1. Socio-demographic characteristics of patients with
Congenital Rubella Syndrome seen from 1995-2002 , UP-PGH.
UP-PGH Patients with CRS Percentage
Characteristics N= 52 %Age in months 8 + 10
(mean + SD)Sex
Males 29 56
Females 23 44Baseline Anthropometrics
Length (mean + SD) 65.3 + 16 cmWeight (mean + SD) 5.7 + 3 kg
Waterlow ClassificationWasting
Normal 15 28.8
Mild 11 21.2Moderate 10 19.2
Severe 16 30.8Stunting
Normal 36 69.2
Mild 5 9.6
Moderate 7 13.5
Severe 4 7.7
Geographic DistributionCentral Luzon 1 19
Southern Tagalog 16 30.7Western Visayas 1 1.9NCR* 34 65.3
Educational Background of Parents•
No formal education• 7 13.5Elementary• 6 11.5High school• 14 26.9
College• 10 19.2Vocational• 2 3.8
Not stated 13 25*National Capital Region
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
54
The common clinical findings of CRS is tabulated in
table-2. Only 5 clinical findings were mentioned in the 52
charts, namely cataract, PDA, hepatitis, jaundice,
pulmonary artery stenosis, extrauterine growth
retardation, glaucoma and hemolytic anemia. Congenital
cataract ranked as the most common syndrome finding
accounting for 24 (49%) of cases, followed by PDA in
15 cases (31%). Barring the differences survey design
and geography, we roughly compared each proportion
of findings with that of an 11- year US- based
epidemiologic data previously published by Schluter et.al.
Among all the mentioned CRS findings, PDA, hepatic
derangement and pulmonary artery stenosis was
statistically significantly higher in the US. This could be
explained by the relatively early detection of
abnormalities. (p-value <.05). Other socio-economic
variables, such as cost for diagnostic work-up , short
hospital stay and even parent preference may possibly
explain the low sign/symptom detection rate of the above
3 findings in our setting.
Table 2. Clinical Findings of CRSClinical UP-PGH Foreign p-value
Findings *N=122(%) Census
(ranked)*
Cataract 23(44) 55(45.1) .628
Patent Ductus 16(31) 62(50.8) .22^
Arteriosus (PDA)
Hepatomegaly/ 5(10) 43(35.2) .0009^
Hepatitis
Jaundice 5(10) 18(14.8) .3955
Pulmonary Artery 3(6) 22(18) .04^
Stenosis
Extrauterine Growth 2(4) 15(12.3) .94
Retardation
Glaucoma 1(2) 4(3.3)
Hemolytic Anemia 1(2) rare NA***
Clinical Profile of Congenital Rubella Syndrome (CRS) cases
*Only 5 clinical signs were identified in this study.
Several signs co-exist
**Data from the Schluter et.al., Changing Epidemiology of CRS in the US,1985-
1996
^ p-value significant at <.05
***Not applicable
Syndrome Profile of CRS
In UP-PGH, we identified the frequency of
combinations of organ systems affected. (See figure-3).
From the years 1995 and onwards, the presence of
cataract was a consistent syndrome finding. Cardiac
findings, PDA, PAS and ASD were not that common.
Jaundice and hepatosplenomegaly had been shown to be
consistent, although less common.
The combinations of findings were not accurately
determined in this review due to several reasons. Each
patient came in at different periods of the follow-up, the
presenting symptoms might have not incited a high index
of suspicion so as to subject the patient to intensive work-
up. So far, in this study , ocular plus cranial plus growth
retardation plus sensorinueral findings and cardiac
findings is still the most common combination.
Figure-3 Syndrome Profile of Congenital Rubella from 1995-
2002, UP-PGH Data
5
1
3
2
Cataract
sensorineural deafness
PDA
PAS
2002
10
1
6
4
11
1
Cataract
sensorineural deafness
PDA
PAS
microcephalymeningoencephalitis
Hepatitis/ jaundice
2001
7
1
2
4
1
1
1
Cataract
Microphthalmia
PDA
Microcephaly
Full anterior fontanelle
Blue Berry muffin
syndrome
Thombocytopenic purpura
2000
3
2
1
1
1
1
1
11
Cataract
EUGR
Sensorineural deafness
PDA
PAS
ASD
Microcephaly
Mental retardation Hepatosplenomegaly
1999
1
1
1
Cataract
EUGR
Hepatitis/Jaundice
1998
21
2
Cataract EUGR
PDA
1997
4
1
3
2
1
3
2
1
2
Cataract
EUGR
PDA
PAS
VSD
Microcephaly
TTP
Blueberry Muffin Syndrome
Hepatitis/Jaundice
1996
44
2
2
2
2
Cataract alonePDA
Microcephaly
hepatitis/jaundice
hepatosplenomegaly
generalized
lymphadenopathy
1995
Fre-
quency
Findings
n=52
Year
5
1
3
2
Cataract
sensorineural deafness
PDA
PAS
2002
10
1
6
4
11
1
Cataract
sensorineural deafness
PDA
PAS
microcephalymeningoencephalitis
Hepatitis/ jaundice
2001
7
1
2
4
1
1
1
Cataract
Microphthalmia
PDA
Microcephaly
Full anterior fontanelle
Blue Berry muffin
syndrome
Thombocytopenic purpura
2000
3
2
1
1
1
1
1
11
Cataract
EUGR
Sensorineural deafness
PDA
PAS
ASD
Microcephaly
Mental retardation Hepatosplenomegaly
1999
1
1
1
Cataract
EUGR
Hepatitis/Jaundice
1998
21
2
Cataract EUGR
PDA
1997
4
1
3
2
1
3
2
1
2
Cataract
EUGR
PDA
PAS
VSD
Microcephaly
TTP
Blueberry Muffin Syndrome
Hepatitis/Jaundice
1996
44
2
2
2
2
Cataract alonePDA
Microcephaly
hepatitis/jaundice
hepatosplenomegaly
generalized
lymphadenopathy
1995
Fre-
quency
Findings
n=52
Year
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
55
DISCUSSION :
Since the introduction of MMR (measles,
mumps, rubella) vaccine in the Philippines since the
year 1983, there remains sporadic cases of CRS that
occurs in the country and eradication of the disease
seemed to be an elusive dream.
In this study there is not much appreciable
difference in the infection rates by sex distribution, a
difference of 4 cases (8%) of slight male
preponderance was noted. Only for the fact that
most of the disease was primarily noted during
“infancy period” where clinical features are clearly
seen if not without complications, except for 3 isolated
cases where patients presented beyond 3 years of
life and were noted to reside outside Greater Manila
Area (GMA), and sought consult at UP-PGH, for
definitive surgical correction of either congenital
catarract or cardiac anomaly.
The bulk of 34 cases (65%) out of 52
population were sporadically distributed along the
metropolis and 16 cases (32%) in region IV and
approximately (2%) in each of regions III and VI. A
large number of cases were seen in Metro Manila
where subjects are residing in the nearby vicinity of
the studied institution (UP-PGH) and could avail the
charity rates of medical consult and could take
advantage of the laboratories and medical manpower
the hospital could offer.
There is poor correlation of educational
attainment of parents to having a child infected with
congenital rubella syndrome (CRS). The study reflects
that the disease affects equally different people of
varying degree of literacy and levels of social strata,
which should alert the public health officials and mass
media to formulate infection control programs and
strategies to heightened public awareness of the spread
of infection since the disease targets the entire
population.
In the study of Schluter in the most frequent
clinical presentation of CRS is the sensorineural
deafness which is about 80-90% of the total population
followed by congenital catarract or glaucoma which
is about 35% and cardiac anomaly of roughly 30%.18
In our institution (UP-PGH) most of our
CRS cases sought consult due to ophthalmologic
problem; about 23 cases (44%) of the study group
population. Opacification of the lens from catarract
are more common (22 (42%) than glaucoma 1(2%))
followed by congenital heart defects of 19 cases
(36%) may which patent ductus arteriosus to be more
common than pulmonary stenosis accounting for 16
cases (84%) and 3 cases (16%) respectively.
According to written by Baja-Panlilio, et all.
The table of birth attendance , since 1990 documented
that most of our mothers; (40%) delivered at home
and attended by traditional birth attendant, and (30%)
by midwives, (1%) by nurse or others and only (27%)
by physician.19
And in most cases of non-institutionalized
delivery of high risks mothers , there is a missed
chance of the infant being evaluated by a trained
health care providers and corollary be worked up
and be seen by tertiary hospital subspecialties.
So, it follows that our congenital rubella
syndrome cases would present at OPD/PER with the
chief complaint of opacification of the eyes noted
during infancy or in blatant cardiac failure which is
probably due to inadequate anticipatory guidance
from our traditional birth attendants (TBA) and/or
midwives and which resulted to less preparatory
movement on the part of the parents.
In rich countries, where most mothers
delivered in a hospital, newborns have the benefit of
being seen by a pediatrician or neonatologist prior to
discharge and would present at ambulatory clinics
later life for either a child with neurodevelopmental
delay and/or for definitive management of hearing
impairment.
CONCLUSION:
This research study documented the continued
existence of CRS and majority of these born to
mothers, who had history of antenatal maternal viral
infection during pregnancy accounting for 37 cases
(71%) of the 52 study population . The disease
continued to be dynamically prevalent despite the
illness is a vaccine preventable occurrence .
It is therefore recommended for public health
officials to have high index of suspicion to the wide
spectrum of clinical manifestations of CRS and
subsequently send suspended infants for laboratory
work-ups, thereby improving the country’s surveillance
system and in turn improve the Filipinos’ quality of
future generation.
RECOMMENDATIONS:
Serologic testing of unvaccinated person for
rubella immunity before vaccination probably is not
necessary because this would only postpone the
opportunity for the vaccine to be administered.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
56
Unless contraindicated, it is suggested that vaccine be
offered to: primary health care clinics, STD clinics,
travel clinics postpartum clinics, health care personnel,
and to all unvaccinated person who lack the evidence
of rubella immunity especially in child-bearing age
and foreign born individuals.
The state health department along with the
mass media should formulate programs and strategies
to heighten public awareness and consequently
educate its citizenry to bring patients to hospital for
diagnostics and early intervention of the impact of
associated disabilities and advice significant others
on ways and means to deter the further shedding of
virus in the community.
REFERENCES:
1. Behrman , Richard et.al., (2000) Nelson Textbook of
Pediatrics (16th edition) W.B. Saunders Company:pp 951-
953.
2. Reef , Susan , Coronado ,Victor: Congenital Rubella
Syndrome , Archive Page, www.deafandblind.com. Canada
Communicable Disease . Report: surveillance of CRS and
other rubella associated adverse pregnancy outcomes
, March 1,1996, Vol.22-05.
3. Canadian Pediatric Society, Prevention of Congenital
Rubella Syndrome; Ontario, Canada. Pediatrics and
Child Health ; 1999 pp 155-157.
4. Lambert , Scott.,Tychslu, Lawrence: Advances in Surgical
Management: Congenital/juvenile Cataract, November
2002.
5. Sharp, Merck and Dohme et all. General Information of
the Disease and the Vaccine, Asia: September 2002
6. Mitz, Sanchez, Scandas: Does Rubella Vaccination
prevent CRS; Vaccination Home Page, July 26, 2002.
7. Elias Ezike et al. August 21 , 2002. Rubella. Department
of Pediatrics, Division of Pediatric Infectious Disease;
Children’s Hospital: Michigan.
8. Mehlita M, Nehlish and Thomas Roslyn. Antenatal
Screening for Rubella – infection or immunity? NICU,
Northwick Park Hospital, Harrow, HAI 3UJ, British
Medical Journal, 2002.
9. Dominique, Marie MD. Screening and vaccinating
adolescents and adults to prevent congenital rubella
syndrome. Canadian Task Force on Preventive Health
Care. January 1994.
10. The Forsight Foundation . Referrence Site for Congenital
Rubella Syndrome , 2002.
11. Save the Children Foundation . A slender Thread: The
Vulnerable Newborn. USA, May 8, 2003.
12. UNICEF. Campaign to prevent Birth Defects in Central
Asia. UNICEF, Media, New York 1996 - 2003.
13. Kaw Bing, Chua and Sai Kit, Lam. Congenital Rubella –
Possible Threat to the New Millennium. Department of
Microbiology, University of Malaya , Kuala Lumpur, April
2003 .
11. Center for Disease Control. Rubella and Congenital
Rubella Syndrome – United States, 1994—1997. USA,
1997.
12. Center for Disease Control. Control and Prevention of
Rubella: Evaluation and Management of Suspected
Outbreaks, Rubella in pregnant women , Surveillance for
CRS. July 11, 2003.
13. Whitemeyer, Sharon MD. Rubella (German Measles).
University of New Mexico .Health Sciences Center , May
2003.
14. Ala , Romeo, MD; Mantos , Donna Belle, MD; Rodriquez,
Jose, MD . Factor Affecting Maternal Compliance to
Expanded Program of Immunization. Department of
Medicine, Cebu City, 1986.
PIDSP Journal, 2005 Vol. 9 No. 2Copyright © 2005
57