Bacteremia in Private Pediatric Practice

(1)

Bacteremia

in Private

Pediatric

Practice

Michael A. Baron, MD, and Howard D. Fink, MD

From the Departments of Pediatrics, Yale-New Haven Hospital and Yale University

School of Medicine, New Haven, Connecticut, and Milford Hospital, Mi/ford, Connecticut

ABSTRACT. The capifiary white blood cell count and

differential, a test easily done in office practice, was used

in unexplained febrile illness of infants and toddlers to

help distinguish those babies most likely to have

bacter-emia who would benefit from blood culture and early

bacteriologic diagnosis. Four criteria were used as indi.

cators suggestive of bacterial infection: white blood cell count 15,000/cu mm, total segmented neutrophils 10,000/cu mm, total band cells 500/cu mm, and total

polymorphonuclear leukocytes (segmented neutrophils

plus band cells) 10,500/cu mm. These measurements

were found helpful in separatinga small group of

bacter-emic babies at high risk of complications from a large group of babies who recovered without antibiotic

treat-ment and without complications. There were 146 febrile illnesses recorded in babies 3 to 24 months of age;

bacter-emia was proven in eight of these. Three or four blood

cell count criteria were fulfilled in seven of the eight

bacteremic babies and in only ten (7.2%) ofthe remaining

138 febrile ifinesses (P < .001 by test). Pediatrics 66:

171-175, 1980; bacteremia, private practice, infants and toddlers, white blood cell counts, neutrophil counts, blood culture.

The febrile infant or toddler whose physical

ex-amination does not reveal a specific focus of

infec-tion presents a dilemma to the clinician. Such a

baby may have bacteremia, a febrile state in which

a major pathogen circulates in the blood for hours

to days in temporary balance with the body’s

im-mune defenses. The causal organism is

Streptococ-cuspneumoniae in most cases, Haemophilus

influ-enzae group B in fewer, and other major pathogens

in a stifi smaller group.’ Entry is often by the

respiratory route and outcome is one of either

spon-taneous clearing or localization in meninges, lungs,

skin, or other major site.’ Low-dose intravenous

Received for publication Sept 17, 1979; accepted Nov 26, 1979. Reprint requests to (M.A.B.) 75 New Haven Aye, Milford, CT 06460.

or outpatient antibiotic treatment does not affect

the incidence of meningitis complicating

bactere-rnia.5’6

Since babies with bacterernia are at serious risk

of major pyogenic complications, early diagnosis is

important.’ It is now well established that the

history and physical examination are unreliable as

indicators of bacteremia.’ In one study, trivial

illness was the clinical diagnosis in 52% of infants

with culture-proven bacteremia.2 There have been

many studies from hospital outpatient departments

demonstrating that bacteremia is most closely

as-sociated with: age under 24 months, especially

un-der 12 months; fever exceeding 102 F (38.9 C); white

blood cell count more than 15,000/cu mm, and ESR

more than 30 mm/hr’’78 It is also well established

that the incidence of bacteremia increases almost

arithmetically with increasing degrees of fever.’7

To our knowledge, however, there have been no

studies of bacteremia reported from middle-class

private pediatric practice. We wish to report on a

three-year experience with bacteremia in a

two-man partnership practice, and the use of the

capil-lary white blood cell count and differential as

indi-cators of this condition.

METHODS

Our patients reside in the city of Milford, CT,

and its surrounding towns. The population of

Mil-ford is 52,000, 97.8% ofwhom are white, 1.1% black,

and 1.1% Hispanic. In June 1979, the

unemploy-ment rate was 4.3%. Each year, 10,000 office

exam-inations, including follow-up visits, are performed.

Our patients are for the most part middle class;

14.8% receive state aid. Of the remaining 85.2%,

65% have medical insurance covering physician

fees, and 80% have insurance covering hospital

charges.

Total WBC counts and differentials were

per-formed by standard methods with capillary blood.

(2)

RESULTS

172 BACTEREMIA

by multiplying the WBC count in cells/cu mm by

the percentage of segmented neutrophils. The total

band count was determined by multiplying the

WBC count by the percentage of band neutrophils.

The total polymorphonuclear leukocyte count was

derived by adding the total segmented neutrophils

and total band cells.

A search for children with bacteremia was

con-ducted in our practice from April 1, 1976 through

March 31, 1979. During this time, a WBC count and

differential was performed on every febrile baby

aged 3 to 24 months whose history and physical

examination did not reveal a specific focus of

infec-tion. If the infant developed evidence of

recogrnz-able bacterial or viral disease during the course of

an illness, he was excluded from the study. Patients

with pneumonia, otitis media, or roseola were

there-fore not studied. The highest temperature recorded

in the history, physical examination, or clinical

course was used for evaluation purposes in the

study. This was done whether or not antipyretics

were administered.

Following the initial evaluation, the patient

waited 20 minutes in the laboratory for the test

results. The WBC count and differential were then

scanned for evidence of bacterial infection and a

decision was made whether to take a blood culture.

No rigid rule was used in making this decision, nor

was any single variable always used in set fashion.

The entire picture including the baby’s age,

behav-ior in the examining room, and height of the fever

were taken into account. The possibility that

bac-terial infection would occur increased as the number

of these four criteria that were met increased: WBC

count 15,000/cu mm; total segmented neutrophils

10,000/cu mm; total band cells 500/cu mm; and

total polymorphonuclear leukocytes 10,500/cu

mm. As the value of one or more of these counts

increased, the likelihood that bacterial infection

would occur also increased.

After the evaluation of the blood cell count and

in some cases the performance of a blood culture,

follow-up was carefully assured and the patient was

sent home. In 15% of cases a different procedure

was used: the laboratory drew blood for the WBC

count, differential, and blood culture

simultane-ously before releasing the patient.

TABLE I. Incidence of Bacteremia

During the study period 146 episodes of fever

were recorded in babies aged 3 to 24 months, in

whom the histories, physical examinations, and

clinical courses did not indicate a cause for the

fever. Temperature ranged from 100.6 to 106 F (38.1

to 41.1 C). WBC counts and differentials were done

in all 146 illnesses; blood cultures were performed

in 42 of them (29%). One such illness was seen on

an average of every 7.5 days. Eight of the 42 blood

cultures were positive, seven for Spneumoniae and

1 for H influenzae group B. The incidence of

bac-teremia was 5.5%.

Temperature was 103 F (39.5 C) in 76 ifinesses.

All eight cases of bacteremia were in this group,

and the incidence of bacteremia in patients with a

temperature of 39.5 C was 10.5% (Table 1). In this

group of 76 ifinesses, one of which was seen every

14.4 days, 32 blood cultures were taken.

Details of the eight infants with bacteremia are shown in Table 2, and these cases are compared to

the 138 additional febrile illnesses in Table 3. These

138 illnesses were treated symptomatically only; antibiotics were not administered. All 138 infants

recovered without complications. All but one

re-turned for follow-up examination after the febrile episode.

The entire study group of 146 illnesses was

di-vided into three groups depending on the number

of blood cell count criteria that were fulfilled. There

were 99 babies whose blood cell counts met none of

the criteria that suggested bacterial infection. In

this group, 14 blood cultures were drawn

simulta-neously with the blood drawn for the blood cell

counts. All of these blood cultures were negative.

There were 30 patients who had counts fulfilling

either one or two of the criteria; blood cultures were

done in 13 of these patients and one was positive.

There were 17 patients with blood cell counts

ful-filling either three or four of the criteria; blood cultures performed in 15 of the 17 were positive in

seven (Table 4). Three or four blood cell count

criteria were fulfilled in seven of eight bacteremic

babies but in only ten (7.2%) of the remaining 138 babies with febrile illnesses (P < .001 by

x2

test).

A total of 42 babies had blood cultures taken

Tempera tore Range No. of

ill-nesses

Blood Cultures Incidence of

Bacteremia (%) No. Per- No. Posi- Positivity

F C

formed tive Rate

Group 1 100.6-102.9 38.1-39.4 70 10 0 . . . 0

Group 2 103.0-106.0 39.5-41.1 76 32 8 1/4 10.5

Total study 100.8-106.0 38.1-41.1 146 42 8 1/5 5.5

at Viet Nam:AAP Sponsored on September 7, 2020 www.aappublications.org/news

(3)

.‘

ili

... =) io’SIE4 ‘‘ ,‘‘

= 0 I

. o.I

-I’ )

.l 0 .4 -

-2i

E- Al

.

cii,

a...,

E oI

.5 000

I

EEE EEE Al

. ‘ )))

.

,S I N 0

‘ ci:ci: riDc,:

I

Al c’

-Lu

-I

n’: ‘‘

Lc

n,

I

4

l

#{149}

u I

.s cl) .‘I 0

N

I-

I

ol c e

I

E’I

I

, ,, C,, ‘ C’ C’ - t.- V

H

_, I CN-

-. .

I

u 0

‘ cI o© LOOLO ,0

I

C’- CCCC1C’N cot’-

4

*u -u

- c’i .6LO© eq

i

+ - - - - - - c

-0

E

-r 0 I t,.. 0

, E-I c’i c

‘-r

o r’.

f-i ,4 ci ,

c4 V

,

I

c1 o eic oo 0

I

I 4 .0

C,, C’ t’.. t- 5)

.

I

Cl) c0c “ ., I

i:.: xc LdLtC

,

, I ,

0 -- -ei

E-

.2

4 C,, ol LO’

0 - E-41 Cl ‘-I

I

i

- 4 .‘

I

c, o

-

I

,

.* I

. 5) I c ‘

4 I c,

V I

t-.‘ I a

0

I

: cit-o ©L(

‘,

I

o LOLO t’.t’-E- .

I

)

Cl)

Q? ?I o_

* ..

gi-I

.

5)

z

c6cc6 i6 ‘

,

I

‘ cq

cI ci - ci - c c’

-.E

CCLO

-?

?

#{176} .‘ Os 00 00 i,..© 0

c

co S

r4 LOc’’ cLO . 0

Qo

.

E- I - - - - -

,

w LO iL

0 Al

.‘

#{231}x LO c’ ©

. ! - - cq cq

c1 .

.0 ILl

r4 L4 .0 -J₄

(4)

174 BACTEREMIA

TABLE 4. Blood Cultures Grouped by Number of

Blood Cell Count Criteria Fulfilled

No. of Criteria Fulfilled No. of Patients Blood Cultures No. Performed No. Positive 0 1-2 3-4 99 30 17 14 13 15 0 1 7

Totals 146 42 8

during the study. Of the 15 such babies whose blood

cell counts met three or four criteria, seven had

positive cultures. Of the 27 babies whose counts

met fewer than three criteria, only one had a

posi-tive culture (P < .001 by Fisher’s exact test). During the same time period, blood cell counts

and cultures were occasionally done in patients

older than 24 months. It is worth mentioning that

two additional cases of bacteremia in older patients

were diagnosed during this time. One, a

30-month-old girl, cleared spontaneously before her initial

blood culture grew S pneumoniae five days after

being drawn. The other, a 36-month-old boy, had

pneumococcal bacteremia which cleared on a

ten-day course of oral penicillin. The blood cell counts

of both patients fulfilled all four criteria suggestive

of bacterial infection.

All ten cases of bacteremia occurred during the

colder months of the year: November, 3 cases;

De-cember, 2 cases; January, 1 case; March, 2 cases;

April, 2 cases. Data were also compiled on the

socioeconomic status of these ten patients. Seven

of the ten lived in single-family dwellings owned by

their parents. These homes ranged in size from five

to nine rooms with three to five people living in

them Only one of the ten was a single-parent

family. Eight of the ten families were covered by

health insurance and nine of the ten families had at

least one parent either an owner of a business or

otherwise employed full-time.

OUTCOME

One case of persistent bacteremia was identified.

Because of an error in reporting from the

labora-tory, a 5-month-old boy had three consecutive blood

cultures over 4#{189}days before antibiotics were

started. These cultures all grew Spneumoniae with

increasing rapidity: the first in three days, the

sec-ond in one day, and the third in eight hours. He

received intravenous penicillin for ten days and

recovered without sequelae.

Two complications were noted. The first occurred

in a 13-month-old boy whose WBC count was

23,300/cu mm and total polymorphonuclear

leuko-cytes were 14,912/cu mm. Blood culture was taken

and follow-up assured before he was sent home.

When the blood culture was reported positive for H

influenzae group B 24 hours later, he returned with

cellulitis of the cheek.

The second complication occurred in a

20-month-old boy whose WBC count was 16,300, total

seg-mented neutrophils 14,018, total band cells 652, and

total polymorphonuclear leukocytes 14,670/cu mm.

Blood culture was drawn, follow-up assured, and he

was allowed to go home. When the blood culture

was reported positive for S pneumoniae 24 hours

later, he returned with periorbital cellulitis.

The remaining five babies had pneumococcal

bacteremia and ranged in age from 5 to 21 months;

none had a specific focus ofinfection. Blood cultures

were repeated and high-dose intravenous penicillin

was administered until the babies were afebrile and

asymptomatic for 48 hours. Intravenous therapy

was given for two to six days, and oral penicillin was

given for a total of ten days. All five patients

re-covered without sequelae.

DISCUSSION

The WBC count and the ESR have been used

extensively to provide early identification of those

infants most likely to have bacteremia.’’#{176} In one study of children aged 3 to 24 months, a

tempera-ture of 104 F (40 C), WBC 15,000/cu mm, and!

or ESR 30 mm/hr distinguished 75% of children

with bacteremia; bacteremia was five times as likely

if these tests were positive thai if they were not.8

In 1974 Todd9 reported the value of total

neutro-phil counts in screening for bacterial infection. He

found that a total segmented neutrophil count 10,000/cu mm or a total band cell count 500/cu

mm is associated with an 80% chance of bacterial

infection. The present study confirms his results.

When three or four white blood cell count criteria

were present, bacteremia was 12 times as likely

than when they were not.

The WBC count and differential are easily

per-formed on capillary blood and the results are

avail-able in minutes. Although false-negative and

false-positive values obviously occur, the correlation of

these tests with bacteremia is remarkably high. The

WBC count and differential can therefore be used

to select patients for blood culture, especially under

conditions in which counts are obtained more easily

or inexpensively than are blood cultures. In the

present study, the use of neutrophil counts

permit-ted such a reduction in the number ofblood cultures

performed that the rate of culture positivity was

higher than that ofany other test done in or through

our office. When the blood culture, WBC count,

and differential are performed simultaneously,

neu-trophil counts provide immediate identification of

most patients at greatest risk of bacteremia.

(5)

fore, it is advantageous to do this test even when

blood cultures are performed on every febrile infant.

The good economic condition of the bacteremic

patients is another striking aspect of this study. In

the August 1979 issue of Pediatrics in Review,

Smith’#{176} stresses the need for data on bacteremia

from private practice, since some studies “ . . .

sug-gest it is the poor and the crowded who suffer the

disease,” and that without such data it appears that

“this illness . . . may be unique to urban emergency

rooms.” The present study indicates unequivocally

that bacteremia and its complications occur in

ad-vantaged populations with incidences very similar

to those reported from hospital outpatient

depart-ments.

Much remains to be learned about optimal

man-agement of bacteremia and about means of

distin-guishing the potentially serious bacteremias from

those more likely to clear spontaneously or with

outpatient therapy. In any case, early diagnosis can

only lead to better management.

ACKNOWLEDGMENTS

The authors thank Drs Paul L. McCarthy and Howard

A. Pearson for their comments, and Dr Sydney Z. Spiesel

for the statistical analysis. Thanks are due Angela

Kaza-novicz for preparation of the tables, and Ruth Jankovsky and Georgiann Caxnbras for secretarial expertise.

REFERENCES

1. McGowan JE, Bratton L Klein JO, et vi: Bacteremia in febrile children seen in a “walk-in” pediatric clinic. N Engi

JMed 288:1309, 1973

2. Teele DW, Pelton SI, Grant MJA, et a!: Bacteremia in febrile children under 2 years of age: Results of cultures of blood of 600 consecutive febrile children seen in a “walk-in” clinic. J Pediatr 87:227, 1975

3. McCarthy PL, Grundy GW, Spiesel SZ, et a!: Bacteremia in children: An outpatient clinical review. Pediatrics 57:861, 1976

4. McCarthy PL, Jekel JF, Dolan TF: Temperature greater than or equal to 40 C in children less than 24 months of age: A prospective study. Pediatrics 59:663, 1977

5. Myers MG, Wright PF, Smith AL, et al: Complications of occult pneumococcal bacteremia in children. J Pediatr 84: 656, 1974

6. Bratton L, Teele DW, Klein JO: Outcome of unsuspected pneumococcemia in children not initially admitted to the hospital. J Pediatr 90:703, 1977

7. McCarthy PL, Dolan TF: Hyperpyrexia in children. Am J

Dis Child 180:849, 1976

8. McCarthy PL: Controversies in pediatrics: What tests are indicated for the child under 2 with fever. Pediatrics 64:PIR 51, 1979

9. Todd JK: ChildhOOd infections: Diagnostic value of periph-eral white blood cell and differential cell counts Am J Dis

Child 127:810, 1974

10. Smith AL: Commentary: The febrile infant. Pediatr Rev 1: 35, 1979

THERAPEUTICS

Therapeutics is the pouring of drugs of which one knows nothing into a

patient of whom one knows less.

(6)

1980;66;171

Pediatrics

Michael A. Baron and Howard D. Fink