See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/12252018

Neonatal Pneumonia: Comparison of 4 vs 7 Days of Antibiotic Therapy in

Term and Near-Term Infants

Article  in  Journal of Perinatology · November 2000

DOI: 10.1038/sj.jp.7200416 · Source: PubMed

CITATIONS READS

23 708

8 authors, including:

William D Engle Gregory L Jackson

University of Texas Southwestern Medical Center University of Texas Southwestern Medical Center
73 PUBLICATIONS   1,498 CITATIONS    56 PUBLICATIONS   1,523 CITATIONS   

SEE PROFILE SEE PROFILE

Dorothy M Sendelbach Marcia Pritchard

University of Texas Southwestern Medical Center University of Texas Southwestern Medical Center
23 PUBLICATIONS   382 CITATIONS    6 PUBLICATIONS   65 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Gregory L Jackson on 06 January 2016.

The user has requested enhancement of the downloaded file.

Original
Article
& & & & & & & & & &
Neonatal Pneumonia: Comparison of 4 vs 7 Days of Antibiotic
Therapy in Term and Near-Term Infants
William D. Engle, MD
Gregory L. Jackson, MD
Dorothy Sendelbach, MD
Diane Ford, PNP
Barbara Olesen, PNP
Kathleen M. Burton, PNP
Marcia A. Pritchard, MD
William H. Frawley, PhD
OBJECTIVE:
To compare a 4 - day course of antibiotic therapy to a 7 - day course in
selected term and near - term neonates with pneumonia.
METHODS:
The diagnosis of pneumonia was made in neonates admitted to the
normal Newborn Nursery ( NBN ) who later had signs of respiratory distress
and whose chest radiographs were consistent with pneumonia. Infants were
excluded if any of the following was present: moderate or thick meconium -
stained amniotic fluid, prior antibiotic therapy > 24 hours, or need for
supplemental oxygen > 8 hours. Infants who were asymptomatic after 48
hours of antibiotic therapy were prospectively randomized to a 4 - day group
( n = 35 ) or a 7 - day group ( n = 38 ) . Infants in the 4 - day group were
observed in the hospital for 24 hours following cessation of antibiotics and
were seen in follow up within several days of discharge.
RESULTS:
The groups were comparable with regard to demographic factors, duration
of rupture of membranes, and incidence of maternal chorioamnionitis.
Median postnatal age at the time of identification of respiratory distress
symptomatology was 19 hours ( range 0.5 to 55 hours ) in the 4 - day group
and 12 hours ( range 1 to 72 hours ) in the 7 - day group. No study infants
had a positive blood culture. Mean reduction in length of hospitalization
was 2.1 days, with estimated savings of greater than US$700 per shortened
hospitalization. Two infants in the 4 - day group developed tachypnea
during the 24 - hour observation period. However, no infants were
Departments of Pediatrics, Radiology and Academic Computing, University of Texas
Southwestern Medical Center at Dallas, Dallas, TX.
Address correspondence and reprint requests to William D. Engle, MD, Department of
Pediatrics, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines
Boulevard, Dallas, TX 75390 - 9063.
Journal of Perinatology 2000; 20:421 ± 426
# 2000 Nature America Inc. All rights reserved. 0743-8346/00 $15
www.nature.com/jp
& & & &
rehospitalized for sepsis or pneumonia following discharge. With 95%
confidence, the true rate of success for the 4 - day group was at least 92%.
CONCLUSION:
Four days of antibiotic therapy plus a 24 - hour period of observation for
selected cases of neonatal pneumonia appears to be comparable to 7 days of
therapy. It is important to note that newborns in our institution receive a
single dose of penicillin soon after birth as part of our group B streptococcal
sepsis prophylaxis program, and all infants in this study received prophylaxis
prior to the onset of respiratory symptoms. Furthermore, only infants who
were asymptomatic after 48 hours of antibiotic therapy were included in this
study, and a 24 - hour observation period at the end of the 4 - day course was
required. These qualifications should be taken into account before use of this
approach is considered, and additional studies are necessary to further
establish its safety and benefits.
Journal of Perinatology 2000; 20:421 ± 426.
INTRODUCTION
Pneumonia is the most common form of neonatal infection and one
of the most important causes of perinatal death.1 The diagnosis of
pneumonia in the term or near-term neonate in the normal
Newborn Nursery (NBN) is based on clinical and radiographic
findings, with blood cultures rarely yielding an etiologic agent;
failure to obtain an organism may be secondary to previous
exposure to antibiotics, or because the infection is localized or
caused by a viral agent. Although respiratory distress presenting at
the time of birth has been well described with neonatal
pneumonia,1,2 in our experience, neonates with pneumonia generally
are asymptomatic for 8 to 24 hours after birth. Furthermore, while
the initial chest radiograph may be consistent with either retained
lung fluid or pneumonia, failure to clear or a worsening picture on
subsequent films suggests the diagnosis of pneumonia.3
A review of our database revealed that 141 infants were diagnosed
with pneumonia in 1997; this group represented 1.1% of the 12,526
admissions to the NBN during that year. In general, neonates with
the diagnosis of pneumonia in the NBN received broad spectrum
intravenous antibiotics, usually ampicillin and gentamicin, for a
total of 7 days.4 Determining the duration of antibiotic therapy for
neonatal pneumonia poses a dilemma for the clinician; the infection
must be adequately treated, while avoiding needlessly prolonged
hospitalization and exposure to antibiotics. In our experience, many
421
 Engle et al.
infants with radiographically and clinically diagnosed pneumonia
responded promptly to therapy, and continuing to administer
antibiotic therapy for 7 days seemed excessive. Therefore, we
hypothesized that for selected neonates diagnosed with pneumonia, 4
days of antibiotic therapy would be comparable to 7 days.
METHODS
Infants admitted to the NBN at Parkland Memorial Hospital must
have an estimated gestational age greater than or equal to 35
weeks and a birth weight greater than or equal to 2100 gm;
those with significant transitional issues, such as birth depression
or respiratory distress apparent in the delivery room, are admitted
directly to the Neonatal Intensive Care Unit (NICU). However,
otherwise stable infants not requiring cardiorespiratory monitoring
or respiratory support may be transferred to an observation
nursery located within the NBN, in which the nurse:patient ratio
is 1:4, vital signs are recorded every 4 hours, and infants may
receive intravenous therapy, including antibiotics; infants in the
observation nursery are managed by experienced Pediatric Nurse
Practitioners in conjunction with one of the authors (G. J. or D.
S.). Infants eligible for this prospective, randomized trial were
those admitted to the NBN from April 1998 through March 1999
who were subsequently diagnosed with pneumonia.
A combined obstetrical±neonatal approach to prevent group B
streptococcal infection was implemented at our institution in January
1995, and was in effect during this study.5 Women with preterm
labor ( <37 completed weeks by last menstrual period, or clinical
and ultrasound examination), prolonged rupture of membranes
(18 hours), or a previous infant with group B streptococcal
infection received intravenous ampicillin (2 gm every 6 hours). All
infants received aqueous penicillin G 50,000 U intramuscularly
within 1 hour of delivery. Exceptions were those infants whose
mothers had chorioamnionitis treated with ampicillin and
gentamicin during labor; asymptomatic neonates with negative
blood cultures in the latter group received ampicillin and gentamicin
intramuscularly for 48 hours following delivery.
Prior to initiation of antibiotic therapy for pneumonia, a single
blood culture and complete blood count with differential cell count
were obtained. Neutrophil values were analyzed according to the
reference ranges for absolute total neutrophils (ATN), absolute total
immature neutrophils (ATI), and the proportion of immature:total
neutrophils (I:T) as reported by Manroe et al.6 The radiographic
diagnosis of pneumonia in study patients was based on the finding of
persistent perihilar streakiness and/or peripheral fluffy infiltrates.3
Sequential radiographs were obtained in most cases to reduce the
likelihood of misdiagnosing retained lung fluid, and radiographic
findings were confirmed by Pediatric Radiology staff, as well as by at
least two attending physicians in the NBN. Infants received
intravenous ampicillin 50 mg/kg every 8 hours and gentamicin 2.5
mg/kg every 12 hours.
422
Neonatal Pneumonia
Only neonates with the diagnosis of pneumonia who were
asymptomatic (respiratory rate <70 breaths/minute, absence of
grunting, flaring and retractions, no need for supplemental
oxygen, and feeding orally) after 48 hours of antibiotic therapy
were eligible for randomization by selection of a sealed, opaque
envelope to either the 4-day (2 additional days of antibiotic
therapy) or the 7-day (5 additional days of antibiotic therapy)
group. Infants were excluded from the study prior to
randomization if: (a) there was moderate or thick meconium
staining of the amniotic fluid; (b) they had received postnatal
antibiotics for another reason, such as maternal chorioamnionitis
(diagnosed by maternal temperature 388C during labor), for
>24 hours at the time of diagnosis of pneumonia; (c)
respiratory symptoms (respiratory rate 70 breaths/minute,
grunting, flaring, and/or retractions) persisted >48 hours after
initiation of antibiotics for pneumonia; (d) the chest radiograph
was not consistent with pneumonia; (e) sustained oxygen
supplementation for greater than 8 hours (in which case infants
were transferred to the NICU and not included in the study); or
(f) failure to obtain parental consent. One infant was excluded
prior to randomization when the diagnosis of congenital viral
syndrome with pneumonia became apparent.
In both study groups, parents were contacted by phone within 1 to
2 days following their infant's discharge from the hospital. In
addition, since infants in the 4-day group received less therapy, they
were: (a) observed in the nursery for 24 hours following
discontinuation of antibiotics and prior to discharge and (b) re-
examined as an outpatient within 2 to 3 days following discharge.
Therapy was considered to be successful if the infant fed well with no
apparent respiratory distress following discharge, as ascertained by
phone contact and/or direct observation, and if there was no
subsequent need for rehospitalization for sepsis and/or pneumonia.
This study was approved by the Institutional Review Boards at The
University of Texas Southwestern Medical Center at Dallas and
Parkland Memorial Hospital, and informed parental consent was
obtained for all infants.
Sample Size
To our knowledge, there has been no formal attempt to estimate
the probability of a successful outcome with the 7-day treatment.
Our experience with this length of therapy has been favorable,
and we would estimate that the probability of a successful
outcome is at least 0.98. Whatever its value, establishing
equivalence of the two treatments (4-day and 7-day) would
require a sample size ranging anywhere from 850 to 8400,
depending on relatively minor variations in the assumptions that
were made.7
We elected to proceed with a sample size sufficient to assess
the feasibility of the new therapy and, for each treatment, to
yield a reasonably high 95% lower confidence bound on the
probability of a successful outcome (defined as infants doing
well, as ascertained by phone contact or direct observation, and
Journal of Perinatology 2000; 20:421 ± 426
Neonatal Pneumonia Engle et al.

Table 1 Infants who Received Antibiotic Therapy for Pneumonia

During Study Period*
Randomized ( n = 73 )
4 - day study group 35
7 - day study group 38

Exclusions ( n = 139 )
Exposure to moderate or thick meconium 37
Prior antibiotic therapy for > 24 hours 6
Symptoms persisting > 48 hours after initiation of 15
antibiotic therapy for pneumonia
Chest radiograph not consistent with pneumoniay 51
Transferred to NICU 5
Congenital viral syndrome 1
Inability to obtain consent 24
Total number of infants with treated pneumonia 212
*April 1998 ± March 1999.
y
Figure 1. Illustration of the 95% lower confidence bound assuming Infants who received antibiotic therapy for presumed pneumonia; however, radiographic
findings did not allow inclusion in this study.
successful outcomes. The dotted line illustrates that for the number of
subjects in the 4-day group (35 ) , the 95% lower confidence bound is
approximately 92%. the criteria for randomization: 35 were randomized to the 4-day
group and 38 to the 7-day group.
Characteristics of the neonates in the 4-day and 7-day groups are
shown in Table 2. Birth weight, gestational age, ethnicity, route of
no rehospitalization for sepsis or pneumonia). It was considered delivery, Apgar scores, cord pH, age of onset of respiratory symptoms,
likely that for a moderate number of subjects, there would be and age at initiation of antibiotic therapy were similar between the
few, if any, unfavorable outcomes for either treatment. For groups. Twenty-one neonates in each group (67% and 59% of the
planning purposes, we plotted the 95% lower confidence bound neonates in the 4- and 7-day groups, respectively) were greater than
versus sample size, assuming no unfavorable outcomes. As 8 hours of age when they developed respiratory distress. The
sample size increases, the curve quickly rises and then ``flattens
out'' (Figure 1). We concluded that at least 35 in each
Table 2 Characterization of the Study Population
treatment group would be adequate both to appraise the viability
of the new treatment and to establish 95% confidence bounds for 4 days 7 days
( n = 35 ) ( n = 38 )
both therapies that exceeded the 90% level.8
Birth weight ( g ) * 3575 ‹ 743 3446 ‹ 589
Statistical Analysis Gestational age ( weeks ) y 39 ‹ 2 39 ‹ 2
An intention-to-treat analysis was performed. Continuous data Gender ( M:F ) 27:8 20:18z
Ethnicity
were evaluated using Student's t-test or the Wilcoxon rank sum Hispanicx 25 ( 72 ) 29 ( 76 )
test; categorical variables were analyzed using Fisher's exact test African ± American 4 ( 11 ) 5 ( 13 )
or chi square contingency table analysis.8,9 A significance level of Caucasian 6 ( 17 ) 4 ( 11 )
0.05 was used. Cesarean section 9 ( 26 ) 9 ( 24 )
Apgar scores
1 minute 8‹1 8‹1
RESULTS 5 minutes 9‹1 9‹1
Cord pH 7.23 ‹ 0.07 7.23 ‹ 0.10
Two hundred twelve infants with presumed pneumonia based on Age at onset of respiratory 19 [ 0.5 ± 55 ] 12 [ 1 ± 75 ]
clinical and radiographic findings received antibiotic therapy during symptoms ( hours ) (,{
Age at initiation of antibiotics 20 [ 1 ± 74 ] 23 [ 1 ± 75 ]
the study period; this group represents 1.7% of the infants admitted to ( hours )
the NBN during this period. One hundred thirty-nine infants who Duration of respiratory symptoms 13 [ 1 ± 44 ] 24z [ 2 ± 65 ]
received antibiotics for pneumonia were excluded from the study ( hours )
prior to randomization (Table 1). This group included 51 neonates *Values are presented as means ‹ SD.
y
with clinical pneumonia in whom chest X-rays were not diagnostic z
As estimated by clinical examination.18
p = 0.05.
of pneumonia; demographic variables in these neonates were x
Numbers in parentheses are percentages.
comparable to those in the two study groups. These infants responded (
{
Numbers are medians.
Numbers in brackets are minimum and maximum values.
well to 5 to 7 days of antibiotic therapy. Seventy-three neonates met
Journal of Perinatology 2000;20:421 ± 426 423
 Engle et al.
male:female proportion was higher in the 4-day group (p=0.05),
and duration of respiratory distress was longer in the 7-day group
(p=0.05). One infant who was randomized to the 7-day group
received only 4 days of treatment because the mother of the baby
signed out against medical advice. Because of the intention-to-treat
design of the study, this infant's data were analyzed as part of the 7-
day group.
Sepsis risk factors and prior exposure of study infants to
antibiotics are shown in Table 3. The mothers of four babies in each
group had rupture of membranes 18 hours. The number of babies
whose mothers received intrapartum antibiotics for chorioamnionitis
or preterm labor <37 weeks, and the number of infants who received
postnatal antibiotics prior to the onset of respiratory distress for
chorioamnionitis exposure were similar between the groups. The
remainder of the study infants received a single dose of aqueous
penicillin G 50,000 U intramuscularly within 1 hour of delivery as
part of our group B streptococcal prophylaxis protocol.
The majority of infants (60% and 58% in the 4- and 7-day
groups, respectively) presented with tachypnea alone. An additional
20% and 29% of the 4- and 7-day infants, respectively, manifested
tachypnea with grunting, retractions, and/or nasal flaring
(p=0.50). Chest radiographic findings in study patients were as
described above. More than one chest X-ray was obtained in 94% and
89% of the 4- and 7-day neonates, respectively. In a blinded,
retrospective review of chest radiographs obtained in study patients
and patients with other diagnoses, performed by one of the
investigators (M. A. P.), the concordance with the original diagnosis
was 83%. Only one CBC was obtained in 26% and 22% of the 4- and
7-day groups, respectively, and two or more CBCs were obtained in
the remaining study infants. Neutrophil values were abnormal in
66% of infants in the 4-day group and 76% of those in the 7-day
group (p=0.46; Table 4). None of the blood cultures obtained in
any of the study infants was positive.
As anticipated by the study design, there was a 2-day difference in
length of hospitalization between the 4-day (6.0‹1.3 days, range 5
Table 3 Sepsis Risk Factors and Prior Exposure to Antibiotic Therapy
4 days 7 days
( n = 35 ) ( n = 38 )
Mother; intrapartum antibiotics
given for
Rupture of membranes 18 hours* 4 ( 11 ) 4 ( 11 )
Chorioamnionitis 2 (6) 2 (5)
Preterm labory 4 ( 11 ) 6 ( 16 )
Infant; postnatal antibiotics
given for
Chorioamnionitis exposure 2 (6) 2 (5)
GBS prophylaxisz 33 ( 94 ) 36 ( 95 )
*Numbers in parentheses are percentages.
y
z
Labor and / or anticipated delivery < 37 weeks.
Single dose of penicillin G.
424
Neonatal Pneumonia
Table 4 Neutrophil Values in Study Patients
4 days ( n = 35 ) 7 days ( n = 38 )
Any neutrophil abnormality* 23 ( 66 ) 29 ( 76 )
2 CBCs obtained 26 ( 74 ) 26 ( 78 )
Total number of CBCs 69 77
Abnormal ATNy 28 ( 41 ) 40 ( 52 )
Abnormal ATI 23 ( 33 ) 30 ( 39 )
Abnormal I:T 13 ( 19 ) 21 ( 27 )
y
ATN = absolute total neutrophil count; ATI = absolute total immature neutrophil count;
I:T = proportion of immature to total neutrophil count. ( Ref.6 ).
*Values in parentheses are percentages.
to 11) and the 7-day (8.1‹1.4 days, range 4 to 14) groups
(p<0.0001). This reduction in length of stay in our institution
would result in greater than US$700 in decreased patient charges per
hospitalization. High rates of phone contact in both groups (97%
and 100% in the 4- and 7-day groups, respectively) and for return
for follow-up in the 4-day group (100%) were observed. All infants
were doing well as ascertained by phone contact or direct observation,
and no study infants were rehospitalized for sepsis or pneumonia.
Two infants in the 4-day group developed tachypnea during the 24-
hour observation period, while being considered asymptomatic after
48 hours of antibiotic therapy and at the completion of the 4-day
course of antibiotics. In one instance, the symptoms resolved and the
infant was discharged 2 days later. In the second case, tachypnea
persisted, work up for other causes of respiratory distress was
negative, chest radiograph showed persistent infiltrates, and
antibiotic therapy was restarted; this infant was treated for an
additional 5 days and then discharged without further problems.
Based on our study protocol, which included a 24-hour observation
period, both of these infants were successfully treated. At the time
enrollment into the study was discontinued, the point estimate of
success for 4 days of therapy plus observation was 100%.
Furthermore, with 95% confidence, the true rate of success was at
least 92% for both the 4- and 7-day groups.
DISCUSSION
The optimal length of antibiotic therapy for neonatal pneumonia is
unclear, which may be a reflection of the diverse clinical
manifestations of this disorder. Recommendations vary from 7 days4
to as long as 3 to 6 weeks2 for disease caused by S. aureus. Treatment
must be adequate to prevent relapse, while therapy should not be
excessive, particularly in the age of early discharge. Therefore, it
would be helpful to more clearly define subpopulations of infants
with pneumonia who may benefit from a shorter course of antibiotic
therapy. Experience in our NBN indicated that, while 7 days of
therapy was adequate in the vast majority of neonates, it appeared
that a selected group of patients, i.e., those whose respiratory signs
resolved within 48 hours of the initiation of therapy, might respond
to a shorter course. The decision to evaluate a 4-day course, and to
Journal of Perinatology 2000; 20:421 ± 426
Neonatal Pneumonia
compare this period to a 7-day course, was made because this length
of therapy would allow 2 additional days of antibiotic therapy
following symptom resolution but, if shown to be successful, would
substantially reduce duration of hospitalization.
The diagnosis of pneumonia in the term and near-term neonate is
challenging and is based on clinical assessment of the baby in
conjunction with supporting information, primarily the interpretation
of sequential chest radiographs.3 Although previous studies have
shown considerable variability among radiologists' interpretations of
chest radiographs10 a retrospective review by one of us (M. A. P.),
performed without benefit of corresponding clinical information,
yielded a concordance of 83% with the original radiographic
diagnosis. Other methods to establish the diagnosis and etiology of
neonatal pneumonia have included gastric aspirates, tracheal
aspirates, and in some instances, lung puncture.11 ± 13 Others have noted
the usefulness of acute phase reactants, such as C-reactive protein, or
sequential neutrophil values in this clinical setting.14,15 Although the
majority of the infants in this study had abnormal neutrophil values,
this finding was not a requirement for entry into the study, and acute
phase reactants were not measured systematically.
Philip4 has noted that respiratory distress developing after the first
12 hours postnatal is strongly indicative of the diagnosis of
pneumonia. In the current study, the median postnatal age at which
signs of respiratory distress appeared was 19 and 12 hours in the 4-
and 7-day groups, respectively. Furthermore, the majority of infants
in each of the treatment groups were well until at least 8 hours after
birth. This delay in presentation makes it much less likely that
infants with retained lung fluid or an aspiration syndrome were
included in the study groups.16,17 It should be noted that 90 infants
who did not have pneumonia but were diagnosed with retained lung
fluid (based on sequential chest radiographic findings and clinical
course) were ill enough to be admitted to the Observation Nursery
during the study.
As shown in Table 1, we excluded a significant number of babies
who received treatment for presumed pneumonia during the study
period. Although the 37 infants exposed to moderate or thick
meconium-stained amniotic fluid were felt to have a course more
consistent with pneumonia than with meconium aspiration, they
were excluded since this distinction often is difficult. In 51 neonates,
the clinical course strongly suggested pneumonia, although the chest
radiographs were not considered diagnostic. The finding of
pneumonia at autopsy with a normal chest radiograph prior to death
has been reported.3 Although we elected to include only infants in
whom the chest radiographs were consistent with pneumonia in this
randomized trial, it is plausible that our conclusions, with further
study, could be generalized to those with chest radiographs not
diagnostic of pneumonia as well as to neonates with meconium-
stained amniotic fluid. In addition, while sequential chest X-rays
appear to be helpful in establishing the presence of pneumonia, and
the absence of retained lung fluid, it is possible that the diagnosis of
pneumonia could be made with fewer radiographs, with a greater
interval between exams, than in the current study.
Journal of Perinatology 2000;20:421 ± 426
Engle et al.
There were marginal statistical differences in gender and duration
of respiratory symptoms between the 4- and 7-day groups. Since
these differences occurred as part of the randomization process and
were not related to duration of therapy, we do not feel that they are
clinically significant. The longer interval between median values for
age at onset of respiratory symptoms and age at initiation of
antibiotics (11 vs 1 hour in the 7- and 4-day groups, respectively)
suggests that the apparent need for immediate antibiotic therapy was
more common in the 4-day group; however, analysis of respiratory
symptomatology suggests that the groups are comparable.
True hospital costs are difficult to define; using hospital charges
as an estimate, we project that a reduction in length of stay of 2 to 3
days for an infant in our observation nursery would result in savings
of greater than US$700. In our institution, with 100 to 200 term and
near-term infants diagnosed with pneumonia per year, these savings
would be substantial. Other potential benefits which might result
from a shorter course of antibiotic therapy, such as enhanced
parent±infant bonding, improved lactation, and reduced antibiotic
resistance were not examined in this study.
In selected term and near-term neonates with pneumonia, 4 days
of antibiotic therapy appears to be a feasible alternative to 7 days of
therapy. It should be emphasized that this conclusion applies only to
those infants who were free of symptoms after 48 hours of antibiotic
therapy with no need for supplemental oxygen and in whom no
bacterial pathogen was isolated. These infants should be
distinguished from those begun on antibiotics for ``rule out sepsis''
scenarios who, after 48 hours, are clinically well, diagnostic tests
have failed to yield a diagnosis, and antibiotics are discontinued.
Based on their presentation and radiographic findings, we considered
that the infants enrolled in the current study were infected. It also
should be emphasized that all of the study infants received either a
single dose of intramuscular penicillin within 1 hour of delivery or,
in the case of four infants, were begun on intramuscular ampicillin
and gentamicin soon after delivery because of exposure to maternal
chorioamnionitis treated with antibiotics during labor. Our practice
of early universal antibiotic administration may have influenced
both the time of onset of respiratory symptoms as well as the
duration of antibiotic therapy.
Based on our finding of two infants who developed tachypnea
during the 24-hour observation period, we suggest that a similar
interval of closely monitored follow up is advisable if this program
were to be implemented. This suggestion is analogous to the
American Academy of Pediatrics recommendation that all neonates
who participate in an early discharge program should be seen in
follow up within 2 to 3 days of discharge from the hospital. Our
decision to consider these two infants in the ``successfully treated''
category was based not just on 4 days of antibiotic therapy but on 4
days of therapy in conjunction with a period of observation. It
appears that this approach would allow many neonates to benefit
from a shorter course of antibiotic therapy, while capturing the few
infants who might require additional hospitalization and/or
treatment. Additional randomized trials are required to further
425
 Engle et al.
establish the safety and benefits of an abbreviated course of
antibiotics for neonatal pneumonia.
Acknowledgement
The authors acknowledge the excellent secretarial assistance of Marilyn Dixon in the
preparation of this manuscript.
References
1. Hickey SM, McCracken GH Jr. Postnatal bacterial infections. In: Fanaroff AA,
Martin RJ, editors. Neonatal ± Perinatal Medicine: Diseases of the Fetus and
Infant. St. Louis: Mosby; 1997.
2. Marks MI, Klein JO. Bacterial infections of the respiratory tract. In:
Remington JS, Klein JO, editors. Infectious Diseases of the Fetus and Newborn
Infant. 4th ed. Philadelphia: WB Saunders; 1995.
3. Haney PJ, Bohlman M, Sun CJ. Radiographic findings in neonatal
pneumonia. Am J Roentgenol 1984;143:23 ± 26.
4. Philip, AGS. Neonatology: A Practical Guide. Philadelphia: WB Saunders; 1996.
5. Siegel JD. Prophylaxis for neonatal group B streptococcus infections. Sem
Perinatol 1998;22:33 ± 49.
6. Manroe BL, Weinberg AG, Rosenfeld CR, Browne R. The neonatal blood
count in health and disease: I. Reference values for neutrophilic cells. J
Pediatr 1979;95:89 ± 98.
7. Farrington CP, Manning G. Test statistics and sample size formulae for
comparative binomial trials with null hypothesis of non - zero risk difference
or non - unity relative risk. Stat Med 1990;9:1447 ± 1454.
426
View publication stats
Neonatal Pneumonia
8. Mood AM, Graybill FA. Introduction to the Theory of Statistics. New York:
McGraw - Hill; 1963.
9. Hollander M, Wolfe DA. Nonparametric Statistical Methods. New York: John
Wiley and Sons; 1999.
10. Davies HD, Wang EE, Manson D, Babyn P, Shuckett B. Reliability of the chest
radiograph in the diagnosis of lower respiratory infections in young children.
Pediatr Infect Dis J 1996;15:600 ± 604.
11. Yeung CY, Tam ASY. Gastric aspirate findings in neonatal pneumonia. Arch
Dis Child 1972;47:735 ± 740.
12. Sherman MP, Goetzman BW, Ahlfors CE, Wennberg RP. Tracheal aspiration
and its clinical correlates in the diagnosis of congenital pneumonia.
Pediatrics 1980;65:258 ± 263.
13. Klein JO. Diagnostic lung puncture in the pneumonias of infants and
children. Pediatrics 1969;44:486 ± 492.
14. Philip AGS. Response of C - reactive protein in neonatal group B streptococcal
infection. Pediatr Infect Dis J 1985;4:145 ± 148.
15. Manroe BL, Rosenfeld CR, Weinberg AG, Browne R. The differential leukocyte
count in the assessment and outcome of early - onset neonatal group B
streptococcal disease. J Pediatr 1977;91:632 ± 637.
16. Miller MJ, Fanaroff AA, Martin RJ. Respiratory disorders in preterm and term
infants. In: Fanaroff AA, Martin RJ, op cit.
17. Potter EL, Craig JM. Pathology of the Fetus and the Infant. 3rd ed. Chicago,
IL: Year Book Medical Publishers; 1975.
18. Ballard JL, Khoury JC, Wedig K, Wang L, Eilers - Walsman BL, Lipp R. New
Ballard score, expanded to include extremely premature infants. J Pediatr
1991;119:417 ± 423.
Journal of Perinatology 2000; 20:421 ± 426